Abstract
Host control of Mycobacterium tuberculosis is dependent on the activation of CD4+ T cells secreting IFN-γ and their recruitment to the site of infection. The development of more efficient vaccines against tuberculosis requires detailed understanding of the induction and maintenance of T cell immunity. Cytokines important for the development of cell-mediated immunity include IL-12 and IL-23, which share the p40 subunit and the IL-12Rβ1 signaling chain. To explore the differential effect of IL-12 and IL-23 during M. tuberculosis infection, we used plasmids expressing IL-23 (p2AIL-23) or IL-12 (p2AIL-12) alone in dendritic cells or macrophages from IL-12p40−/− mice. In the absence of the IL-12/IL-23 axis, immunization with a DNA vaccine expressing the M. tuberculosis Ag85B induced a limited Ag-specific T cell response and no control of M. tuberculosis infection. Codelivery of p2AIL-23 or p2AIL-12 with DNA85B induced strong proliferative and IFN-γ-secreting T cell responses equivalent to those observed in wild-type mice immunized with DNA85B. This response resulted in partial protection against aerosol M. tuberculosis; however, the protective effect was less than in wild-type mice owing to the requirement for IL-12 or IL-23 for the optimal expansion of IFN-γ-secreting T cells. Interestingly, bacillus Calmette-Guérin immune T cells generated in the absence of IL-12 or IL-23 were deficient in IFN-γ production, but exhibited a robust IL-17 secretion associated with a degree of protection against pulmonary infection. Therefore, exogenous IL-23 can complement IL-12 deficiency for the initial expansion of Ag-specific T cells and is not essential for the development of potentially protective IL-17-secreting T cells.
Protective immunity against Mycobacterium tuberculosis requires the orchestration and integration of innate and adaptive immune responses to generate a robust and long-lived memory T cell response. Tuberculosis represents an enormous burden to global health, with ∼9 million cases and 2 million deaths annually (1). More effective antituberculosis vaccines are required to control this transmission of infection. The development of improved vaccines will require detailed understanding of the induction and then expression of T cell responses in infected lungs. Innate immunity to M. tuberculosis is triggered by the activation of pattern recognition receptors on macrophages and dendritic cells (DC)3, including TLR2 and TLR9 (2, 3), which activate the maturation and migration of DC to draining lymph nodes. M. tuberculosis-activated DC stimulate the differentiation of Ag-specific CD4+ T cells and CD8+ T cells, both of which are critical for protective response to M. tuberculosis infection (4, 5). The recruitment of Ag-specific IFN-γ-secreting CD4+ T cells to the lung is a critical component of immunity to tuberculosis, because it leads to the activation of infected macrophages to inhibit the replication of M. tuberculosis (6).
The regulatory cytokines IL-12 and IL-23 are important for the activation and clonal expansion of Ag-specific CD4+ T cells in the draining lymph nodes of M. tuberculosis-infected lungs (7, 8, 9). Both IL-12 and IL-23 are heterodimeric cytokines, which share a common p40 subunit along with a unique p35 and p19 subunit, respectively. Both cytokines signal through a common IL-12Rβ1 chain in their receptors, and this chain is essential for biological responsiveness to IL-12 and IL-23 and regulation of optimal IFN-γ responses (10, 11). The IL-23p19 chain signals through the IL-23R binding chain, which is expressed on CD4+ T cells (11, 12, 13). Whereas IL-12 drives Th1-like CD4+ T cell response characterized by IFN-γ secretion, IL-23 also promotes the differentiation of T cells into IL-17 secretion (14, 15). These IL-17-secreting CD4+ T cells have a dominant role in the induction of chronic autoimmune inflammation of the CNS and joints (16, 17). Furthermore, IL-12 and IFN-γ can suppress the IL-17-producing T cells, and this is consistent with the increased incidence of autoimmune diseases observed in mice deficient in IL-12, IL-12Rβ2, and IFN-γ (18, 19, 20).
The relative roles of IL-12 and IL-23 in M. tuberculosis infection have been examined in mice deficient of the common p40 chain (IL-12p40−/−), which lack both IL-12 and IL-23. These mice are markedly susceptible to M. tuberculosis infection; however, the absence of IL-12 alone in IL-12p35−/− mice causes only partial compromise of protective immunity to M. tuberculosis. This suggests a role for other IL-12p40 binding molecules, such as IL-23, in the protective response (21). Surprisingly, IL-23p19-deficient mice effectively control M. tuberculosis replication, and neither the number of IFN-γ-specific CD4+ T cells nor the local IFN-γ mRNA expression were reduced at the site of infection (22). This suggested that IL-23 was not essential for the development of resistance during the acute stage of infection. By contrast, during immunization with a DNA vaccine expressing Ag 85B (DNA85B), plasmid-expressing IL-23 was as effective as plasmid IL-12 as an adjuvant to increase the frequency of IFN-γ-secreting T cells and protection against aerosol M. tuberculosis infection (23).
To address the relative contribution of IL-12 and IL-23 in the development of antimycobacterial T cells and protective immunity to M. tuberculosis infection, we have examined the effects of IL-12 and IL-23 as coimmunogens with an antituberculosis DNA vaccine in IL-12p40-deficient mice. Plasmid IL-23 was as effective as plasmid IL-12 in complementing the deficiency of IL-12p40 in the induction of T cell responses and partially restoring the protective efficacy of the vaccine against M. tuberculosis.
Materials and Methods
Bacterial growth conditions
M. tuberculosis H37Rv (ATCC 27294) was grown in Proskauer and Beck liquid medium for 14 days for aerosol infection and Mycobacterium bovis bacillus Calmette-Guérin (BCG, Pasteur strain) was grown in Middlebrook 7H9 broth supplemented with albumin, dextrose, and catalase (BD Biosciences) for 14 days at 37°C. The bacteria were enumerated on OADC-enriched Middlebrook 7H11 agar and stored in 15% glycerol/PBS at −70°C. For construction of IL-12p40-expressing plasmid, Escherichia coli MC1061 was grown in Luria-Bertani broth or on Luria-Bertani agar, and Circlegrow broth (BIO 101) was used for large-scale preparations. Ampicillin (100 μg/ml) was supplemented as necessary.
Preparation of vaccines
The construction of DNA vaccines encoding M. tuberculosis Ag85B (DNA85B), murine IL-12 (p2AIL-12) and murine IL-23 (p2AIL-23) has been described previously (23, 24). The plasmid expressing murine IL-12p40 as a monomer was constructed by excising the IL-12p35 subunit and the 2A fragment from p2AIL-12 with the use of restriction enzymes Mlu and Apa, resulting in a 5′ overhang that was end filled using T4 polymerase. The blunt ends were self-ligated to yield IL-12p40 and the sequence integrity confirmed. The bioactivity of plasmids expressing IL-12 or IL-23 was confirmed previously (23, 25). The DNA used in immunizations was prepared by equilibrium centrifugation in a continuous CsCl-ethidium bromide gradient. Endotoxin levels in the purified DNA were <6 pg of endotoxin per microgram of plasmid DNA, as determined by a Chromogenic end-point Limulus amoebocyte lysate (QCL-1000) assay (Cambrex).
Immunization
Female C57BL/6 mice (6–8 wk old), which were obtained from Animal Resources Centre, Perth, Australia, and the C57BL/6 IL-12p40−/− (IL-12p40−/−) mice (The Jackson Laboratory) were maintained in specific pathogen-free conditions. Mice were immunized with a total of 100 μg of DNA85B mixed with either 100 μg of p2AIL-12, p2AIL-23, or control vector by i.m. injection into both tibialis anterior muscles. Control mice were either immunized by i.m. injection with 200 μg of parental control vector (pcDNA3) or s.c. with BCG (5 × 105 CFU) >100 days before challenge.
Bone marrow-derived DC (BMDC) preparation
BMDC were generated as previously described (26) with the following modification. In brief, cells were incubated with a mixture of hybridoma supernatants from M5.114 (anti-MHC class II), RA3-6B2 (anti-B220), 53-6.7 (anti-CD8), GK1.5 (anti-CD4), and RB6-8C5 (anti-Ly-6G). Cells bound to Abs were eliminated using goat anti-rat IgG microbeads (MACS; Miltenyi Biotec). Progenitor cells were cultured in complete RPMI 1640 medium containing 10% FCS, 50 mM 2-ME, and 2 mM glutamine with recombinant murine GM-CSF (10 ng/ml) and recombinant murine IL-4 (5 ng/ml) (PeproTech) for 5 days at 37°C in 5% CO2.
Transfection of BMDC
Following a 5-day culture, differentiated BMDC were transfected with control, p40 (monomer), p2AIL-12, or p2AIL-23 vector using a Mouse Macrophage Nucleofector kit (Amaxa Biosystems) according to the manufacturer’s instructions. Forty-eight hours post-transfection, supernatants were collected and the levels of IL-12p40 and IL-12p70 were determined by a cytokine ELISA. Secreted IL-23 was measured by an IL-23 bioassay.
IL-23 bioassay
The induction of IL-17 from OT-II TCR transgenic CD4+ T cells (provided by B. Heath, Walter and Eliza Hall Institute, Melbourne, Australia) was used to assay for the presence of biologically active IL-23. CD4+ T cells were purified from lymph nodes of OT-II transgenic C57BL/6 mice using magnetic bead sorting (MACS; Miltenyi Biotech). Purified OT-II CD4+ T cells were cultured in 96-well plates (2 × 105 cells/well) with syngeneic, irradiated wild-type APCs (2 × 105 cells/well). The OT-II CD4+ T cells were stimulated with varying doses of the H-2b-restricted OVA peptide323–339 (ISQAVHAAHAEINEAGR), and supernatants from cultures of BMDC transfected with IL-12p40 monomer, p2AIL-12, or p2AIL-23. The data shown are with a peptide concentration of 0.01 μg/ml. Induction of IL-17 was measured by ELISA.
Ab measurement
To detect serum Ag85-specific Ab levels, plates were first coated with Ag85 (2 μg/ml). Five-fold dilutions of sera were incubated for 1 h before washing and the addition of goat anti-mouse IgG, goat anti-mouse IgG1, or goat anti-mouse IgG2a Abs conjugated to alkaline phosphate (Sigma-Aldrich). Following washing, n-nitrophenyl-phosphate (1 mg/ml) (Sigma-Aldrich) was added and absorbance was measured at 405 nm. The mean absorbance plus three SDs of normal mouse sera, diluted at 1/100, was adopted as the cutoff absorbance for determining Ab titers.
Lymphocyte proliferation
Two weeks following the last immunization, single-cell suspensions of splenocytes were prepared. Pooled samples were enriched for T cells using a nylon wool column and cultured with gamma-irradiated syngeneic splenocytes (2 × 105 cells/well) in 96-well plates with Ag85 (3 mg/ml), Con A (3 mg/ml), or medium alone for 72 h and subsequently pulsed with 1 μCi of [3H]thymidine for 6 h. Results were calculated by subtracting the mean cpm in control wells from test samples.
Cytokine assays
Ag-specific IFN-γ secreting cells were measured by ELISPOT or the secretion of IFN-γ was determined by ELISA as previously described (25). Briefly, ELISPOT plates were coated with monoclonal anti-IFN-γ AN18 Ab and blocked with 2% FCS/PBS buffer for 1 h, followed by washing. Single-cell splenocyte suspensions (2 × 105/well) were cultured with Ag85 (3 mg/ml), Con A (3 mg/ml), or medium alone for 16 h at 37°C in an atmosphere of 5% CO2. Cells were removed, and the plate wells were washed before the addition of biotinylated anti-IFN-γ mAb (clone XMG 1.2) to each well for 2 h. The number of IFN-γ-secreting cells was enumerated by color development following incubation with avidin-alkaline phosphatase conjugate. For cytokine ELISA, plates were coated with anti-IFN-γ (clone AN18), anti-IL-17, anti-IL-12p70, or IL-12p40 Abs (all obtained from R&D Biosciences) and the supernatant samples and cytokine standards added. Following a 1-h incubation period, XMG 1.2 biotin (IFN-γ detection antibody), anti-IL-17 biotin, or anti-IL-12p40 biotin (both from R&D Biosciences) was added followed by streptavidin-HRP. Substrate solution was allowed to develop and absorbance measured at a dual wavelength of 405/492 nm.
Draining lymph node cultures and intracellular cytokine flow cytometry
Wild-type and IL-12p40−/− mice were s.c. injected with M. bovis BCG (5 × 105 CFU/mouse) in the base of the tail and footpad. Draining para-aortic, inguinal and popliteal lymph nodes were removed at 1, 2, 3, 4, and 6 wk. Single-cell suspensions were plated at 3 × 106/well and cultured with complete RPMI 1640 medium containing BCG sonicate (10 μg/ml) overnight at 37°C in an atmosphere of 5% CO2. Brefeldin A (10 μg/ml; Sigma Aldrich) was added to the cultures 4 h before surface staining to block the export of proteins from the Golgi. Cells were incubated with FcγRII/III (CD16/CD32) Ab (BD Pharmingen) to prevent nonspecific binding, followed by surface staining with anti-CD4 Alexa 700 (BD Pharmingen). Cells were fixed and permeabilized with 1× Cytofix/Cytoperm buffer (BD Pharmingen) for 20 min and washed extensively. Cytoplasmic IFN-γ and IL-17 were detected with anti-IFN-γ FITC (Caltag Laboratories) and anti-IL-17 PE (BD Pharmingen), respectively, and analyzed on a BD LSR-II flow cytometer.
M. tuberculosis challenge
Six weeks after the last boost with DNA vaccine, mice were challenged with aerosol M. tuberculosis H37Rv using a Middlebrook airborne infection apparatus (Glas-Col) with an infective dose of ∼100 viable bacilli per lung. Four weeks following M. tuberculosis challenge, homogenized organs were plated in 10-fold dilutions on Middlebrook 7H11 Bacto agar supplemented with OADC Middlebrook enrichment containing 10% oleic acid, and bacterial colonies were counted at 3 wk.
Statistical analysis
Statistical analysis of the results from immunological assays and log-transformed bacterial counts were conducted using ANOVA. Differences with p < 0.05 were considered significant.
Results
Plasmid IL-23 complements IL-12p40−/− DC for secretion of bioactive IL-23
To examine the complementation of IL-12p40 deficiency in vitro, BMDC were transfected with plasmids expressing IL-12p40 alone, IL-12, IL-23 or control plasmid. Forty-eight hours later the supernatants were collected and analyzed for presence of IL-12p40, IL-12p70 and IL-23. Significant levels of IL-12p40 were detected in cultures transfected with IL-12p40, p2AIL-12 and p2AIL-23 (Fig. 1⇓A). The secretion of IL-12p70 was identified only in DC transfected with p2AIL-12 (Fig. 1⇓B). Similarly, transfection with p2AIL-23 resulted in the exclusive production of IL-23, which was detected by IL-23-mediated stimulation of IL-17 production in a bioassay (Fig. 1⇓C). BM macrophages transfected with the same plasmids produced a similar pattern of cytokine release (data not shown). Therefore, the expression of IL-12p40 alone was not sufficient to restore the secretion of IL-12 or IL-23, and the coordinate expression of p40 and p35 or p40 and p19 is necessary for the production of IL-12 and IL-23, respectively. Furthermore, transfection with p2AIL-23 did not lead to IL-12 production.
Transfection of IL-12p40−/− DC with p2AIL-23 restores the secretion of IL-23 exclusively. IL-12p40−/− BMDC cultures were transfected with control (Cont), IL-12p40 (p40), p2AIL-12 (IL-12), or p2AIL-23 (IL-23) vectors. Supernatants were collected after 72 h and tested for the presence of IL-12p40 (A) and IL-12p70 (B) by ELISA and of IL-23 as detected by the IL-23 bioassay (C), which detects the induction of IL-17 from purified OT-II CD4+ T cells. Data are the means of triplicate samples ± SEM and are representative of one of two individual experiments (ND, not detected).
Plasmid IL-23 complements IL-12p40 deficiency in vivo
To determine whether p2AIL-23 would restore differentiation of T cell in vivo, IL-12p40−/− mice were immunized with DNA85B and p2AIL-23, p2AIL-12, or control plasmid. Immunization with DNA85B alone resulted in limited activation of T cells as demonstrated by Ag-specific T cell proliferation and IFN-γ production (Fig. 2⇓, A and B). The IFN-γ response was approximately one-third of the level in wild-type mice. Coimmunization with p2AIL-12 or p2AIL-23 led to the activation of Ag-specific T cells, which proliferated (Fig. 2⇓A) and secreted IFN-γ (Fig. 2⇓B) to equivalent levels observed in wild-type mice immunized with DNA85B alone, however proliferative responses were greater following codelivery of p2AIL-23 compared with p2AIL-12 (Fig. 2⇓A). There were no significant differences between mice coimmunized with p2AIL-12 or p2AIL-23 in the induction of priming IFN-γ-secreting T cell immune responses in IL-12p40-deficient mice.
IL-12p40−/− mice coimmunized with DNA85B and p2AIL-12 or p2AIL-23 generate Ag-specific T cell responses equivalent to wild-type (WT) mice receiving DNA85B. IL-12p40−/− mice were immunized by i.m. injection three times at 2-wk intervals with control (Cont) vector alone (200 μg), DNA85B (85B, 100 μg) with control (100 μg), with p2AIL-12 (IL-12, 100 μg), or p2AIL-23 (IL-23, 100 μg) plasmids. Wild-type C57BL/6 mice received control vector or DNA85B. Two weeks following the final immunization, T cells purified from the spleens of immunized mice were simulated with syngeneic irradiated splenocytes and Ag85 protein. Ag-specific T cell proliferation was measured by the thymidine incorporation 3 days later (A). Splenocytes stimulated with Ag85 protein were enumerated for IFN-γ-secreting cells (SC) by ELISPOT, following 16 h of incubation (B). Data are the SEM for five mice and are representative of one of four experiments. The differences between groups were analyzed by ANOVA (*, p < 0.05; **, p < 0.001; NS, not significant).
Coimmunization with plasmid IL-12, but not plasmid IL-23, reduces the IgG1 Ab response
To characterize the role of p2AIL-23 and p2AIL-12 in complementing humoral immunity in mice deficient in IL-12p40, anti-Ag85B Ab responses were measured. As shown in Fig. 3⇓, IL-12p40-deficient mice mounted Ag85B-specific IgG responses equivalent to those in wild-type mice following DNA85B immunization. Interestingly, IL-12p40−/− mice coimmunized with p2AIL-12, but not those coimmunized with p2AIL-23, demonstrated a reduced anti-Ag85B IgG response (Fig. 3⇓). More detailed analysis revealed that the reduction in the Ag-specific Ab level in mice coimmunized with p2AIL-12 was within the IgG1 compartment (Fig. 4⇓A), and the response of IgG2a was similar in IL-12p40−/− mice receiving DNA85B with p2AIL-12 or p2AIL-23 (Fig. 4⇓B). Therefore, IL-12 and IL-23 have distinct effects on the Ab response to DNA immunization.
Codelivery of p2AIL-12, but not p2AIL-23, reduces Ag-specific IgG response to DNA immunization. Mice were immunized as described in the legend of Fig. 2⇑. Following the final immunization, sera were collected, and the Ab titer for IgG anti-Ag85 Abs was measured by ELISA. The geometric means of the log10 titers of Abs were measured as described in Materials and Methods. Data represent the means ± SEM for five mice and are representative of two separate experiments. The differences between groups were analyzed by ANOVA (*, p < 0.05; **, p < 0.001; †, Ab titer <1; WT, wild-type).
Coimmunization with p2AIL-12, but not p2AIL-23, reduces Ig class switching to IgG1, but not IgG2a. Mice were immunized as described in the legend of Fig. 2⇑. Following the final immunization, sera were collected and the Ab titer measured for Ag-specific IgG1 (A) and IgG2a (B) Abs by ELISA. The geometric means of the log10 titers of anti-Ag85 Abs were measured as described in Materials and Methods. Data are the means ± SEM for five mice and are representative of two separate experiments. The differences between groups were analyzed by ANOVA (*, p < 0.05; **, p < 0.001; †, Ab titer <1; WT, wild type).
Both plasmid p2AIL-12 and p2AIL-23 lead to expansion of IFN-γ-secreting T cells after M. tuberculosis infection
We have previously shown that immunization of wild-type mice with DNA85B plasmid and M. tuberculosis infection leads to increased expansion of IFN-γ T cells in the lungs and draining lymph nodes (27). To determine whether T cells stimulated with DNA85B and p2AIL-12 or p2AIL-23 would expand after M. tuberculosis challenge, immunized IL-12p40−/− mice were rested for 6 wk and then exposed to low-dose aerosol infection with M. tuberculosis, and the IFN-γ-secreting T cell responses were measured in the lungs and mediastinal lymph nodes (MLN) 4 wk later. IL-12p40−/− recipients of control vector were unable to mount a T cell response to M. tuberculosis infection, with negligible levels of activated T cells, compared with the response in wild-type mice (Fig. 5⇓, A and B). IL-12p40−/− mice immunized with DNA85B alone also showed no significant reactivation of IFN-γ-secreting T cells. By comparison, IL-12p40−/− mice immunized with DNA85B and plasmid IL-12 or IL-23 showed expansion of Ag-specific T cells secreting IFN-γ in both the infected lung (Fig. 5⇓A) and draining lymph nodes (Fig. 5⇓B). The level of IFN-γ-secreting T cells was markedly less than that observed in lungs (Fig. 5⇓A) of wild-type mice that had received DNA85B. The expansion of IFN-γ-secreting T cells after M. tuberculosis infection was significantly higher in IL-12p40−/− mice coimmunized with p2AIL-12 than those receiving p2AIL-23. In the draining lymph nodes, the response in IL-12p40−/− mice receiving p2AIL-12 and DNA85B was similar to that in wild-type mice receiving DNA85B alone. Immunization of IL-12p40−/− mice with BCG alone did not result in a significant IFN-γ T cell response to M. tuberculosis in either the lungs or MLN. Thus, p2AIL-23 in the absence of IL-12 leads to both the induction and subsequent expansion of IFN-γ-secreting T cells in response to DNA vaccination, albeit at a lower level than in wild-type mice.
IL-12p40−/− mice coimmunized with DNA85B and p2AIL-12 or p2AIL-23 generate Ag-specific IFN-γ T cell responses following M. tuberculosis infection. Mice were immunized as described in the legend of Fig. 2⇑ with the additional group of mice given a s.c. injection of M. bovis BCG (5 × 105 CFU/mouse) ∼100 days before challenge. Six weeks following the final immunization, mice were challenged with M. tuberculosis H37Rv. Four weeks later, Ag-specific T cell responses were evaluated in M. tuberculosis-infected lungs (A) and draining MLN (B) of IL-12p40−/− mice (▪) and infected lungs and MLN of wild-type mice (□), by ELISPOT after ex vivo stimulation with Ag85 protein for 16 h. Data represent the means ± SEM for five mice and are representative of two separate experiments. The differences between groups were analyzed by ANOVA (*, p < 0.05; **, p < 0.001; WT, wild type).
Plasmid IL-23 partially restores protective immune responses to DNA85B in the absence of IL-12p40
To determine whether p2AIL-23 could complement protective immunity in the absence of IL-12, IL-12p40−/− mice were immunized with DNA85B and coimmunized with p2AIL-12 or p2AIL-23, or immunized s.c. with BCG, challenged with M. tuberculosis; the bacterial loads were measured at 4 wk. IL-12p40−/− mice were less able to control M. tuberculosis infection than wild-type mice with ∼1 log10 higher mycobacterial load in the lungs and the spleen (Fig. 6⇓). Immunization of IL-12p40-deficient mice with DNA85B alone had no significant protective effect in either the lung (Fig. 6⇓A) or the spleen (Fig. 6⇓B), compared with the previously reported protective effect of DNA85B in the lungs of wild-type mice (27). The codelivery of p2AIL-12 or p2AIL-23 with DNA85B in IL-12p40−/− mice led to a significant reduction in M. tuberculosis load in the lungs; however, this did not reach the levels observed in DNA85B-immunized wild-type mice (Fig. 6⇓A). The protective effect afforded by BCG in IL-12p40-deficient mice was significantly reduced in the lungs and completely diminished in the spleen. Therefore, IL-12 and IL-23 can partially restore protective efficacy of DNA immunization against pulmonary M. tuberculosis infection.
Coimmunization with p2AIL-12 or p2AIL-23 partially restores the protective effect of DNA85B in IL-12p40−/− mice against M. tuberculosis infection. Mice were immunized as described in the legend of Fig. 2⇑ with the additional group of mice given a s.c. injection of M. bovis BCG (5 × 105 CFU/mouse) ∼100 days before challenge. Six weeks following the final immunization, mice were challenged by aerosol infection with M. tuberculosis H37Rv. Four weeks later, the bacterial loads in the lungs (A) and spleen (B) were determined. Data are the means ± SEM for five mice and are representative of three independent experiments. The differences between groups were analyzed by ANOVA (*, p < 0.05; **, p < 0.001; WT, wild type).
BCG promotes the expansion of IL-17-producing lymph node cells in IL-12p40−/− mice
Recent studies suggest a role for IL-23 in the promotion of a distinct T cell population expressing IL-17 as an effector cytokine, independent of IFN-γ, the hallmark of Th1-like T cell responses (28). Therefore, we examined the presence of IL-17 in infected lungs and MLN of IL-12p40−/− mice and wild-type mice immunized with DNA85B and coimmunized with plasmid IL-12 or p2AIL-23 and DNA85B or BCG alone. Four weeks following M. tuberculosis infection, Ag-induced secretion of IFN-γ, but not IL-17, was detected in the MLN (Fig. 7⇓A) and lungs (data not shown) of coimmunized IL-12p40−/− and wild-type mice. The IFN-γ T cell response in IL-12p40−/− mice following DNA85B immunization was limited, however this was significantly enhanced and equivalent following codelivery with p2AIL-12 or p2AIL-23. No Ag-specific IFN-γ response was detected in IL-12p40−/− mice immunized with BCG. Interestingly, however, in the absence of the IL-12/IL-23 axis, significant levels of IL-17 were produced by cells from the MLN, but not the lungs, of BCG-immunized IL-12p40−/−, but not by MLN cells from wild-type mice following M. tuberculosis infection (Fig. 7⇓B). Therefore, there was an IL-23-independent production of IL-17 in response to BCG, but not DNA, immunization in IL-12p40−/− mice.
BCG stimulates production of IL-17 following M. tuberculosis infection in the absence of IL-12 and IL-23. Mice were immunized as described in the legend of Fig. 2⇑ with the additional group of mice given a s.c. injection of M. bovis BCG (5 × 105 CFU/mouse) ∼100 days before challenge. Six weeks following the final immunization, mice were challenged by aerosol infection with M. tuberculosis H37Rv. Four weeks later, MLN were stimulated ex vivo with Ag85 protein for 72 h and supernatants were tested for IL-17 (A) and IFN-γ (B). Data represent the means ± SEM for five mice and are representative of two separate experiments. The differences between groups were analyzed by ANOVA (*, p < 0.05; **, p < 0.001; WT, wild type).
To confirm this observation, a direct comparison of the proportion of IFN-γ-secreting cells and IL-17-secreting cells was conducted in BCG-immunized IL-12p40−/− or wild-type mice. No IL-17 was detected in lymphocytes from IL-12p40−/− mice immunized with DNA85B and p2AIL-12 or p2AIL-23 or with DNA85B alone or from wild-type mice immunized with DNA85B or control vector (data not shown). Wild-type mice developed an IFN-γ T cell response as early as 1 wk following BCG vaccination (Table I⇓). This BCG-specific IFN-γ-secreting T cell response in wild-type mice increased over 6 wk with no substantial expansion of IL-17-secreting CD4+ T cells (Fig. 8⇓ and Table I⇓). The opposite effect was seen in IL-12p40−/− mice. From week 2, there was an increase in CD4+ T cells secreting IL-17, which peaked at 4 wk after vaccination with BCG (Fig. 8⇓). Therefore, the absence of IL-12 and IL-23 attenuates the IFN-γ T cell response during the first 6 wk following vaccination, but promotes the development of IL-17-producing CD4+ T cells. In addition, the BCG-responsive CD4+ T cells exhibited increased TNF production in the absence of IL-12 and IL-23 (data not shown). For example, 2 wk following immunization, the proportion of TNF-secreting CD4+ T cells was 0.45% in IL-12p40−/− mice compared with 0.18% in wild-type mice.
BCG stimulates preferentially IL-17 production in draining lymph nodes of IL-12p40−/− mice. Wild-type and IL-12p40−/− mice were immunized s.c. in the base of tail and footpads with M. bovis BCG (5 × 105 CFU/mouse). Draining lymph nodes were isolated after 1, 2, 3, 4, and 6 wk and cells were stimulated with BCG sonicate (10 μg/ml) overnight before intracellular staining for IFN-γ and IL-17. The percentage of IFN-γ- and IL-17-producing CD4+ T cells was determined by gating on CD4+ T cells on duplicate samples. The mean percentages of CD4+ T cells producing IFN-γ or IL-17 4 wk after BCG vaccination are shown. The results at other time points are summarized in Table I⇓.
Immunization with BCG stimulates IL-17 production by CD4+ T cells in the absence of IL-12 and IL-23a
Discussion
The differentiation of naive CD4+ T cells into distinct IFN-γ- and IL-4-secreting T cell subsets is regulated during the initiation of T cell responses by the polarizing cytokines, IL-12 and IL-4, respectively. Additional cytokines, including IL-23, IL-18, and IL-27, also influence the development of IFN-γ-secreting T cells during the in vitro activation of T cells (29). In the case of IL-23, in vitro studies suggested that IL-12 plays the initial role in the differentiation of CD4+ T cells into IFN-γ-secreting T cells and up-regulates the IL-23 receptor, rendering the cells responsive to IL-23. More recently, IL-23 has been found to activate IL-17 secreting, proinflammatory CD4+ T cells implicated in autoimmune tissue damage (17). To investigate the relative contributions of IL-12 and IL-23 in the activation of anti-mycobacterial T cell responses, we used plasmids expressing both chains of IL-12 or IL-23. Transfection of IL-12p40−/−-deficient DC or macrophages with either plasmid IL-12 or IL-23 led to the exclusive production of IL-12 or IL-23, respectively, whereas transfection with pIL-12p40 alone failed to produce IL-12 or IL-23, indicating that the coordinate expression of both chains of these cytokines is required to produce functional IL-12 or IL-23 (Fig. 1⇑). Complementation with plasmid IL-12 or plasmid IL-23 during DNA immunization of IL-12p40-deficient mice revealed distinctive properties of IL-23. First, plasmid IL-23 was as effective as plasmid IL-12 in the priming of IFN-γ-secreting T cell responses to the mycobacterial protein encoded by the DNA vaccine (Fig. 2⇑). Second, subsequent exposure to M. tuberculosis led to the re-expansion of IFN-γ-secreting T cells in the MLN of IL-12p40-deficient mice coimmunized with either plasmid IL-23 or IL-12, as measured by ELISPOT (Fig. 5⇑) and IFN-γ release (Fig. 7⇑A). Third, T cells primed by coimmunization with plasmid IL-23 and DNA-85B were as effective as those primed with plasmid IL-12 and DNA-85B in stimulating protective immune responses in the lung to aerosol M. tuberculosis infection. This protective effect, however, was significantly less than that in wild-type mice immunized with DNA-85B alone, indicating that IL-12 and/or IL-23 is required during the expression of immunity to M. tuberculosis in the lung for optimal protective immune responses.
This finding that IL-23 can act independently of IL-12 in complementing IL-12p40 deficiency suggests that IL-23 may contribute significantly to the optimal generation of protective CD4+ T cells. It is consistent with the observations that IL-12p35-deficient mice, which lack IL-12 but retain IL-23, develop partial antimycobacterial IFN-γ T cell responses and control experimental tuberculosis infection more effectively than IL-12p40-deficient mice lacking both IL-12 and IL-23 (21). By contrast, IL-23p19−/− mice, which lack IL-23 but retain IL-12, displayed normal control of aerosol M. tuberculosis infection (22). Therefore, IL-12 can fully compensate for the IL-23 deficiency, whereas IL-23 only partially complements IL-12 deficiency. Nevertheless, IL-23 independent of IL-12 can augment the protective T cell response of wild-type mice following immunization and mycobacterial infections. For example, coadministration of plasmid IL-23 with an antituberculosis DNA vaccine significantly increased the Ag-specific T cell response and led to enhanced protective efficacy against pulmonary M. tuberculosis infection (23). This augmentation of T cell responses by IL-23 was also evident following treatment with the cytokine alone. Delivery of adenovirus-expressing IL-23 to the airways before infection with M. tuberculosis resulted in enhanced antimycobacterial Th1-like responses and increased clearance in the lungs (30).
In addition to its direct effects on T cells, IL-23 may also influence the function of DC and macrophages (31). Targeting DC with a single-chain Ig-IL-23 fusion construct led to the induction of T cell responses to an otherwise tolerogenic peptide, and this effect was mediated by IL-23 triggering the release of IL-12 from the DC (13). This highlights the collaborative effects of IL-23 and IL-12 in generating optimal T cell responses. There are differences, however, in some effects of the two cytokines. For example, coimmunization with plasmid IL-12 reproducibly leads to a reduction in the total IgG (25, 32); however, IL-23 has no effect on B cell responses to DNA vaccines (Figs. 3⇑ and 4⇑).
Although local i.m. delivery of IL-23 or IL-12 restored the T cell responses to DNA immunization in IL-12p40−/− mice to those in normal mice, the recall response and protective efficacy of these Ag-specific T cells following M. tuberculosis infection was less than in immunocompetent mice (Figs. 5⇑ and 7⇑). This suggests that IL-12 and/or IL-23 are required for the restimulation of central memory T cells in the MLN or at the site of infection in the lungs. Treatment of IL-12p40−/− mice with rIL-12 for the first 4 wk of M. tuberculosis infection partially restored the deficient IFN-γ-secreting T cell response, leading to granuloma formation (33). However, withdrawal of the rIL-12 led to progressive infection, indicating that IL-12 is required for the maintenance of T cell immunity through the expansion of memory T cells. The requirement for IL-12 during the restimulation of IFN-γ-secreting T cells and expression of protective immunity by these T cells has been demonstrated for other intracellular pathogens, including Leishmania major (34) and Toxoplasma gondii (35). Dissection of the contribution of IL-23 to the restimulation of effector T cells will require systemic delivery of IL-23 as a protein or virally encoded cytokine.
BCG immunization of IL-12p40−/− mice revealed an IL-12/IL-23-independent pathway for the activation and killing of M. tuberculosis-infected macrophages mediated by BCG-specific T cells (Fig. 6⇑). Following M. tuberculosis infection of BCG-immunized IL-12p40−/− mice, there was no IFN-γ production by lung or MLN cells; however, there was increased synthesis of IL-17. This was associated with a small, but significant, reduction in M. tuberculosis load in the lungs. IL-17 is predominantly a T cell-derived proinflammatory cytokine, which induces local tissue inflammation through the local production of IL-6, IL-1β, and granulocyte CSF, up-regulation of ICAM-1 expression by endothelium, and recruitment of monocytes and neutrophils (36). IL-17, also known as IL-17A, belongs to a family of related cytokines, including IL-17B–F (37, 38, 39, 40, 41). The receptor for IL-17 is ubiquitously expressed (42) and when engaged by its ligand, induces translocation of NF-κB in a TNFR-associated factor 6-dependent manner (43). The secretion of IL-23 by DC regulates the production of IL-17A and IL-17F by T cells. During pulmonary infection with Klebsiella pneumoniae, IL-17A secretion was dependent on the release of IL-23 by alveolar DC activated through TLR4 (44). Moreover, IL-23 deficiency resulted in reduced IL-17 mRNA expression and IL-17 secretion by pulmonary T cells during M. tuberculosis infection (22). Nevertheless, IL-17 production is not absolutely dependent on IL-23, because BCG immunization stimulated IL-17 production in the absence of IL-23 (Fig. 8⇑). This is consistent with the induction of IL-17-secreting T cells in IL-23p19-deficient mice (45). Although, IL-23 was required for protection against Citrobacter rodentium infection, the development of IL-17-secreting T cells in response to infection can occur independently of IL-23 (46). TGF-β, rather than IL-23, was the key signal for promoting the development of IL-17-secreting T cells during C. rodentium infection, and both TGF-β and IL-6 were required for the IL-17 response to an auto-Ag (47).
IFN-γ may play an important inhibitory role in the development of IL-17-secreting CD4+ T cells (14, 15). IL-17-secreting T cells can be induced to differentiate directly from naive CD4+ T cells, however this effect is antagonized by cytokines controlling the development of Th1 or Th2 cells, namely IFN-γ and IL-4. As a result, in the absence of the IL-12/IL-23 axis, BCG-primed T cells were defective in the IFN-γ response and developed into IL-17-secreting CD4+ T cells (Fig. 8⇑). The peak of this response was between 2 and 4 wk postimmunization, when the frequency of Ag-specific CD4+ T cells secreting IL-17 was 4-fold greater than that in wild-type mice immunized with BCG (Table I⇑). By contrast, T cells from BCG-immunized wild-type mice failed to produce IL-17 in the presence of a strong IFN-γ response (Table I⇑) (48). This is consistent with the recent finding that IFN-γ suppressed the development of mycobacteria-specific IL-17-producing CD4+ T cells (49). Interestingly, priming with a DNA vaccine before BCG, which increases the protective efficacy of BCG (50), was associated with both IL-17 and IFN-γ production in wild-type mice (48). IL-18 may be a contributing factor to the IL-12/IL-23-independent induction of IL-17 production, because it is secreted in response to mycobacterial infection and can stimulate IL-17 production (51). In addition, BCG immunization led to an increased frequency of TNF-secreting T cells in the absence of IL-12 and IL-23. TNF expression and synthesis was up-regulated by IL-17 in a dose-dependent manner in human macrophages (52). T cell-derived TNF activates antimycobacterial mechanisms in infected macrophages (53) and increased TNF-secreting T cells may have contributed to the control of infection observed in the BCG-immunized IL-12p40−/− mice.
In summary, although IL-23 is not essential for protection against M. tuberculosis, IL-23 can complement IL-12p40 deficiency during immunization with DNA vaccines, leading to the induction of IFN-γ-secreting T cells and partial protection in the lungs against M. tuberculosis infection. Furthermore, despite the requirement for IL-23 in the development of inflammatory IL-17-secreting T cell responses in autoimmune diseases, this study provides the first evidence that antimycobacterial IL-17-secreting T cell responses can emerge without IL-23, and the critical factor for the development of this IL-17 response is the absence of IFN-γ.
Acknowledgments
We thank Dr. B. Heath (Walter and Eliza Hall Institute of Medical Research, Melbourne, Australia) for providing the OT-II transgenic mice. The Ag85 protein was obtained through the TB Research Material and Vaccine Testing Contract at Colorado State University (NIAID AI-75320).
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 This work was supported by the National Health and Medical Research Council of Australia and the New South Wales Department of Health through its research infrastructure grant to the Centenary Institute of Cancer Medicine and Cell Biology.
↵2 Address correspondence and reprint requests to Dr. Warwick J Britton, Centenary Institute of Cancer Medicine and Cell Biology, Mycobacterial Research Laboratory, Locked Bag Number 6, Newtown, NSW, Australia. E-mail address: wbritton{at}med.usyd.edu.au
↵3 Abbreviations used in this paper: DC, dendritic cell; BCG, bacillus Calmette-Guérin; BMDC, bone marrow-derived DC; MLN, mediastinal lymph node.
- Received July 20, 2006.
- Accepted September 30, 2006.
- Copyright © 2006 by The American Association of Immunologists