Brucella abortus is a facultative intracellular bacterium that infects humans and domestic animals. The enhanced susceptibility to virulent B. abortus observed in MyD88 knockout (KO) mice led us to investigate the mechanisms involved in MyD88-dependent immune responses. First, we defined the role of MyD88 in dendritic cell (DC) maturation. In vitro as well as in vivo, B. abortus-exposed MyD88 KO DCs displayed a significant impairment on maturation as observed by expression of CD40, CD86, and MHC class II on CD11c+ cells. In addition, IL-12 and TNF-α production was totally abrogated in MyD88 KO DCs and macrophages. Furthermore, B. abortus-induced IL-12 production was found to be dependent on TLR2 in DC, but independent on TLR2 and TLR4 in macrophages. Additionally, we investigated the role of exogenous IL-12 and TNF-α administration on MyD88 KO control of B. abortus infection. Importantly, IL-12, but not TNF-α, was able to partially rescue host susceptibility in MyD88 KO-infected animals. Furthermore, we demonstrated the role played by TLR9 during virulent B. abortus infection. TLR9 KO-infected mice showed 1 log Brucella CFU higher than wild-type mice. Macrophages and DC from TLR9 KO mice showed reduced IL-12 and unaltered TNF-α production when these cells were stimulated with Brucella. Together, these results suggest that susceptibility of MyD88 KO mice to B. abortus is due to impaired DC maturation and lack of IL-12 synthesis. Additionally, DC activation during Brucella infection plays an important regulatory role by stimulating and programming T cells to produce IFN-γ.
Brucella is a facultative intracellular bacterium that infects humans and domestic animals (1). After entering the host, Brucella is taken up by dendritic cells (DC)4and macrophages (2). The recognition of microbes by cells of innate immunity involves host cell membrane receptors termed TLR (3). TLRs, upon activation by microbial products, transduce signals via a common adaptor molecule, MyD88, consequently leading to host cell activation (4). MyD88 is critical for TLR-mediated activation of the transcription factor NF-κB and the induction of proinflammatory cytokines (5, 6). Although the available evidence suggests that multiple rather than single TLRs are required for innate defense against most pathogens, it is not clear how signals from different TLRs are orchestrated in generating protective responses against microbial pathogens (7, 8).
In previous studies, we and others showed that Brucella abortus signals through TLR2 and TLR4 (9, 10, 11) but TLR2 plays no role in controlling the infection. Additionally, Huang et al. (12) demonstrated that heat-killed B. abortus (HKBA) activates DC and macrophages to secrete TNF-α and IL-12 by two distinct TLR pathways. TNF-α was TLR2 dependent whereas IL-12 secretion was TLR2 independent but MyD88 dependent. Furthermore, the same group showed that IL-12 synthesis by DC is dependent on TLR9 and that TLR9 knockout (KO) mice failed to mount an effective Th1 response (13). Regarding in vivo infection, Weiss et al. (14) demonstrated that MyD88 is a critical molecule during B. abortus infection because MyD88 KO mice did not control the bacteria as efficiently as wild-type, TLR4 KO, TLR2 KO, or TLR2/4 KO mice. However, the mechanisms involved in MyD88-deficient mice susceptibility to brucellosis are not known.
In this study, we confirmed that the MyD88 molecule plays a critical role in host control of Brucella infection in vivo, and we defined the immune mechanisms related to this enhanced susceptibility. First, bone marrow (BM)-derived and splenic DCs from MyD88 KO mice do not mature as efficient as cells from wild-type mice when stimulated with HKBA. Second, DCs and macrophages from MyD88 KO mice do not produce significant amounts of TNF-α and IL-12 compared with wild-type mice. Furthermore, IL-12 was shown to be critical to MyD88 KO resistance because Brucella CFU were reduced in spleens of MyD88 KO mice treated with rIL-12. Additionally, MyD88 KO macrophages produce low levels of the inflammatory chemokines RANTES, MCP-1 and MIP-1α, and NO. Third, lack of IL-12 synthesis by DC has a direct impact on IFN-γ production by lymph node cells as observed in MyD88 KO mice. Finally, we observed that TLR9 is an important molecule during Brucella infection; however, the lack of MyD88 induces a more profound effect in host susceptibility. In conclusion, impaired DC maturation and reduced IL-12 production are major factors affecting host susceptibility to B. abortus infection in MyD88 KO mice.
Materials and Methods
MyD88, TLR2, TLR4, and TLR9 KO mice were provided by S. Akira (Osaka University, Osaka, Japan). The wild-type strain C57BL/6 mice were provided by Federal University of Minas Gerais (UFMG; Belo Horizonte, Brazil). Genetically deficient and control mice were maintained at UFMG and used at 6–8 wk of age. All animal experiments were preapproved by the Institutional Animal Care and Use Committee of the UFMG.
B. abortus virulent strain 2308 was obtained from our own laboratory collection (15). They were grown in Brucella broth medium (BD Biosciences) for 3 days at 37°C.
Infection and Brucella counts in spleens
Mice were infected i.p. with 106 CFU of B. abortus strain 2308. To count residual Brucella CFU in the spleens of mice, five animals from each group were examined at each sampling period. Spleens from individual animals were homogenized in PBS, 10-fold serially diluted, and plated on Brucella broth agar (Difco). Plates were incubated at 37°C and the number of CFU was counted after 3 days as previously described (15).
Cytokine production in macrophage and lymph node cells
For macrophages, MyD88 KO, TLR9 KO, and C57BL/6 mice were inoculated i.p. with 2 ml of 3% thioglycolate (Difco) and, 4 days later, the elicited peritoneal exudate cells were harvested in cold serum-free RPMI 1640. The medium used in macrophages cultures consisted of RPMI 1640 (Invitrogen Life Technologies) supplemented with 2 mM l-glutamine, 25 mM HEPES, 5% heat-inactivated FBS (Sigma-Aldrich), penicillin G sodium (100 U/ml), and streptomycin sulfate (100 μg/ml; supplemented RPMI 1640). The macrophages were washed and resuspended in supplemented RPMI 1640 at 106/ml and dispensed into the wells of a 96-well plate. Cells were allowed to adhere at 37°C and 5% CO2 for 3 h, washed once with serum-free RPMI 1640, and 200 μl of supplemented RPMI 1640 was added to each well, in the presence or absence of 20 U of IFN-γ/ml. The cultures were stimulated by addition of 109 heat-killed bacteria (heat inactivation was performed at 80°C for 2 h)/ml, 1 μg/ml Escherichia coli LPS, zymosan (50 μg/ml), or CpG (1 μg/ml) in a total volume of 200 μl of medium/well. For lymph node cells, freshly removed lymph nodes (inguinal, mesenteric, axillary, and brachial) of mice infected with B. abortus S2308 were placed in petri dishes containing 5 ml of PBS and passed through steel mesh to obtain single-cell suspensions. The lymph node cell suspension was then transferred to a sterile 15-ml tube. Cells were then washed twice with sterile PBS containing penicillin G sodium (100 U/ml), streptomycin sulfate (100 μg/ml). Afterward, lymph node cells were cultured in supplemented RPMI 1640 and placed at 106 cells/well in 96-well tissue-culture plates. The murine lymph node cells were stimulated with 1 μg/ml E. coli LPS (Sigma-Aldrich), zymosan (50 μg/ml), or 109 heat-inactivated B. abortus/well. Unstimulated cells were used as a negative control, and cells stimulated with Con A (5 μg/ml; Sigma-Aldrich) were a T cell-activating control. These preparations were incubated at 37°C and 5% CO2. Levels of TNF-α, IFN-γ, IL-12 (p40), IL-4, RANTES, MIP-1α, and MCP-1 in the supernatants were measured by a commercially available ELISA Duoset kit (R&D Systems).
To asses the amount of NO produced, macrophage culture supernatants from MyD88 KO, TLR9 KO, or C57BL/6 mice were assayed for accumulation of the stable end product of NO, NO⨪2 which was determined by the Griess reaction. Briefly, culture supernatants (50 μl) were mixed with 50 μl of Griess reagent (1% sulfanilamide, 0.1% naphthylethyline diamine dihydrochloride, and 2.5% phosphoric acid) in triplicate in 96-well plates at room temperature for 5 min. The OD at 550 nm was then measured. NO⨪2 was quantified by comparison with Na (NO2; Sigma-Aldrich) as a standard.
Generation of BM-derived and splenic DCs
DCs were generated from BM mononuclear cells from wild-type C57BL/6, TLR9 KO, or MyD88 KO mice, in medium containing rGM-CSF as previously described (16). Briefly, femurs and tibiae were collected from mice with 4–12 wk old. After removing bone adjacent muscles, marrow cells were extracted by flushing RPMI 1640 medium through the bone interior. Unwanted fragments were removed by filtering the marrow suspension through 70-μm cell strainers (Falcon; BD Biosciences). Bone marrow cells were then resuspended on DC culture medium (RPMI 1640 medium, 10% heat-inactivated fetal/bovine serum, 100 U/ml penicillin/streptomycin, 50 mM 2-ME, 200 U GM-CSF), and plated on 100 mm petri dishes (1 × 107cells/10 ml DC culture medium). On days 2 and 4, the cells were refed. On day 5, cells were removed from petri dishes, plated on 6-well plates (3 × 106 cells/well) in 3 ml of fresh DC culture medium. After overnight culture, on day 6, cells demonstrated differentiated morphology and expressed DC markers (CD11c+, MHC class II (MHCII)low and CD40low; data not shown), confirming to be immature DCs. For splenic DC, C57BL/6, TLR9 KO, TLR2 KO, or MyD88 KO mice were inoculated with 109 HKBA and the splenocytes were removed 24 h later. Then, spleen cell suspensions were treated with appropriate Abs for FACS analysis as described below.
FACS analysis and anti-TLR-blocking assay
On day 6, BMDC were stimulated for 24 h with E. coli LPS (1 μg/ml), HKBA (103 bacteria/cell), zymosan (50 μg/ml), or CpG (1 μg/ml). Following in vitro stimulation for BMDCs or in vivo HKBA inoculation for splenic DC (24 h), cells were stained with fluorescein (FITC)-coupled HL3 (anti-CD11c), PE-coupled AF6–120.1 (anti-MHCII) and, biotin-conjugated GL1 (anti-CD40) and 3/23 (anti-CD86) from BD Biosciences. Samples were then fixed and analyzed on a FACScan flow cytometer (BD Biosciences). Acquired data were analyzed by using FlowJo software (Tree Star). Additionally, 2 × 105 differentiated BMDCs or macrophages were plated on 96-well plates and incubated with blocking Abs anti-TLR2 at 5 μg/ml (Cell Sciences), and/or anti-TLR4 at 20 μg/ml (Imgenex) for an hour followed by Ag stimulation as indicated above. After BMDC and macrophage culturing, the cell-free supernatant was collected and stored until assayed for IL-12 production.
MyD88 KO mice treatment with recombinant cytokines
MyD88 KO mice were i.p. injected daily with mouse rTNF-α (0.5 μg; PeproTech) in 0.2 ml of PBS, starting at 3 h before infection. Control mice were injected with PBS only. Additionally, MyD88 KO mice were treated 36 h prior infection with 1 × 108 PFU of a recombinant nonreplicating adenovirus expressing murine IL-12 (Ad5mIL-12) diluted in RPMI 1640 plus 1% of normal mouse serum in a total volume of 100 μl. This Ad5-based recombinant virus was gifted by Dr. F. L. Graham (McMaster University, Ontario, Canada) and it contains the p35 subunit cDNA of murine IL-12 in early region 1 of adenovirus type 5 and the cDNA for p40 in early region 3 as previously described by Bramson et al. (17). This IL-12-expressing recombinant virus has also protected mice against lethal Klebsiella pneumoniae (18). Then, five mice per group were infected i.p. with 106 CFU of B. abortus strain S2308 on day 0 and residual Brucella CFU in the spleens were determined 1 wk after treatment.
Percentage of Ag-specific T cells producing IFN-γ
For intracytoplasmatic cytokine staining, splenocytes from MyD88 KO or C57BL/6 mice 2 wk infected with Brucella (1 × 106 CFU/mice) were adjusted to 5 × 105 cells/well. Splenocytes were maintained in culture at 37°C in 5% CO2 in medium alone or stimulated with recombinant Brucella proteins L7/L12, Omp16, and Omp19 (25 μg/ml, rOmp16, and rOmp19 were provided by Dr. J. Cassataro, University of Buenos Aires, Buenos Aires, Argentina), HKBA S2308 (103 bacteria/cell) or Con A (5 μg/ml). All recombinant proteins were free of LPS by using Sepharose-polymyxin B and contained <0.25 endotoxin U/μg of protein as assessed by Limulus amebocyte assay (Associates of Cape Cod). After 15 h of culture, 1 μl/well of brefeldin A (1-mg/ml stock; Sigma-Aldrich) was added to impair cytokine secretion. After 3 h of incubation, these cells were stained for surface markers and intracellular cytokines. Briefly, cultured cells were incubated for 15 min with Fc-block (FcRII/III; BD Biosciences) in FACS buffer (0.15 M PBS, 2% BSA, 1% NaN3) and stained for surface markers using fluorescein (FITC)-labeled H129.19 (anti-CD4) and Cy5-conjugated 53-6.7 (anti-CD8) mAbs from BD Biosciences by incubation for 15 min with FACS buffer, followed by washes and incubation with 20 μl of streptavidin diluted 1/200 in FACS buffer. After new washes and fixation using 4% formaldehyde solution, these cells were permeabilized and further stained with PE-labeled XMG1.2 (anti-IFN-γ) mAbs from BD Biosciences in a 0.5% saponin solution in PBS. After 30 min, cells were washed with permeabilization solution and resuspended in 4% formaldehyde. A minimum of 30,000 splenocyte-gated events were acquired in list mode and analyzed using a lymphocyte gate determined based on size and granularity profiles. All quadrants were set according to the labeled isotype controls and analyzed using the WinMDI 2.9 software.
MyD88 is pivotal to control murine brucellosis
To determine the contribution of MyD88 in bacterial clearance, numbers of Brucella were monitored in the spleens of MyD88 KO and C57BL/6 mice at 1, 2, 3, and 6 wk following B. abortus infection. The animals were sacrificed weekly and the numbers of CFU were determined. Murine brucellosis was markedly exacerbated in MyD88 KO mice at all intervals studied (Fig. 1). The difference in bacterial numbers in MyD88 KO and C57BL/6 was accentuated at 3 and 6 wk postinfection, 1.9 logs and 1.74 logs, respectively. These in vivo findings reinforce the role of MyD88 in innate immunity against Brucella, functioning as an important molecule in the control of infection.
Lack of proinflammatory cytokine production by macrophages and lymph node cells parallels enhanced susceptibility of MyD88 KO mice
IL-12 is an important cytokine driving Th1 cells and therefore, production of IFN-γ. IFN-γ is a pivotal cytokine involved in the control of murine brucellosis (19, 20, 21, 22). TNF-α is also shown to be crucial to destroy live Brucella inside of macrophages (23, 24, 25). First, we determined the levels of IFN-γ, TNF-α, or IL-4 in lymph node cells of MyD88 KO and C57BL/6 mice following 1, 2, 3, and 6 wk postinfection, and after stimulation with HKBA, E. coli LPS, or zymosan. Levels of IFN-γ and TNF-α in the supernatant of HKBA-stimulated MyD88 KO cells were totally abrogated compared with wild-type mouse cells (Fig. 2). However, MyD88 KO cells were able to produce IFN-γ and TNF-α when stimulated with zymosan, a TLR agonist. We also observed a decrease in the synthesis of these cytokines during the course of infection. As for IL-4, low levels were detected; however, no statistical significant difference was observed in MyD88 KO compared with C57BL/6 mice (data not shown). Regarding macrophage production of IL-12 or TNF-α, the same phenomenon was observed. Macrophages of MyD88 KO mice when stimulated with HKBA or E. coli LPS, activated with IFN-γ or not, did not produce significant levels of IL-12 or TNF-α. In contrast, wild-type macrophages produced ∼800 or 2000 pg/ml IL-12 or TNF-α, respectively, when stimulated with HKBA (Fig. 3, A and B). Additionally, the level of IL-12 produced by wild-type mouse macrophages was influenced by the activation with IFN-γ. Furthermore, reduced levels of NO were observed in MyD88 KO macrophages stimulated with HKBA when compared with C57BL/6 macrophages (Fig. 3 C). Regarding inflammatory chemokines, low levels of RANTES, MCP-1, and MIP-1α were detected in MyD88 KO macrophages when stimulated with HKBA compared with C57BL/6 (data not shown).
Impairment of MyD88 KO Ag-specific T cells to produce IFN-γ
To check whether the reduction in Brucella CFU observed in MyD88 KO mice during the course of infection (Fig. 2) is dependent on IFN-γ produced by Ag-specific T cells, we determined the percentage of CD4+ or CD8+ T lymphocytes producing IFN-γ following activation with recombinant B. abortus L7/L12, Omp16, and Omp19 proteins by FACS analysis. These Brucella proteins are described as major immunogenic molecules for this bacterium (26). As demonstrated in Table I, Brucella-primed CD4+ or CD8+ T cells from MyD88 KO mice were not able to produce IFN-γ when stimulated with recombinant proteins. In contrast, CD4+ T cells from C57BL/6 mice produced significant levels of this cytokine to rL/L12, rOmp16, and rOmp19 proteins. This finding reveals that Th1 cell activation to specific Brucella proteins is impaired in MyD88 KO.
|.||Medium .||L7L12+OMP16/19 .||HKBa .||Con A .|
|C57BL/6||1.46 ± 0.35||2.26 ± 0.5ab||3.26 ± 0.5ab||3.36 ± 0.7ab|
|MyD88−/−||0.86 ± 0.05||0.8 ± 0.1||1.7 ± 0.2a||2.0 ± 0.17a|
|C57BL/6||1.06 ± 0.15||1.2 ± 0.26||2.3 ± 0.45ab||3.76 ± 0.49ab|
|MyD88−/−||0.53 ± 0.14||0.7 ± 0.3||0.5 ± 0.26||1.53 ± 0.12a|
|.||Medium .||L7L12+OMP16/19 .||HKBa .||Con A .|
|C57BL/6||1.46 ± 0.35||2.26 ± 0.5ab||3.26 ± 0.5ab||3.36 ± 0.7ab|
|MyD88−/−||0.86 ± 0.05||0.8 ± 0.1||1.7 ± 0.2a||2.0 ± 0.17a|
|C57BL/6||1.06 ± 0.15||1.2 ± 0.26||2.3 ± 0.45ab||3.76 ± 0.49ab|
|MyD88−/−||0.53 ± 0.14||0.7 ± 0.3||0.5 ± 0.26||1.53 ± 0.12a|
Statistically significant compared to nonstimulated cells from the same animal.
Statistically significant compared to MyD88 KO mice.
MyD88 is required for DC maturation following activation with HKBA
To characterize the role of MyD88 in DC maturation, we stimulated BM-derived DC with E. coli LPS or HKBA and analyzed cell surface expression of MHCII, CD40, CD86, on CD11c+ cells. As demonstrated in Fig. 4, DCs from C57BL/6 mice stimulated with E. coli LPS or HKBA showed enhanced expression of MHCII (Fig. 4,A), CD86 (Fig. 4,B), and CD40 (Fig. 4,C), compared with nonstimulated cells. A similar activation profile was observed in MyD88 DCs when stimulated with E. coli LPS. However, no signs of DC maturation were observed when MyD88 DCs were stimulated with HKBA demonstrating that HKBA-induced DC maturation is dependent of MyD88. Because DC populations are heterologous (27, 28) and distinct subsets have been discriminated on the basis of experimental expression of various cell surface markers, we also evaluated the role of MyD88 on maturation of splenic DC in vivo (Fig. 5). This result confirms our finding with BMDC that MyD88 deficiency impaired splenic DC maturation following in vivo HKBA stimulation. Additionally, we observed that splenic DC maturation is independent of TLR2 and TLR9 signaling.
Production of TNF-α and IL-12p40 by DCs after HKBA stimulation is MyD88 dependent
To assess the role of MyD88 on DC activation by HKBA, immature DCs were generated in vitro by culturing BM cells with GM-CSF. Additionally, we measured proinflammatory cytokine production in MyD88 KO and wild-type DC following HKBA stimulation. As shown in Fig. 6, IL-12 (Fig. 6,A) and TNF-α (Fig. 6 B) production by DC were totally abrogated in MyD88 KO mice when compared with C57BL/6 animals. These results confirmed that MyD88 is essential for HKBA-induced proinflammatory cytokine production by BMDC.
Production of IL-12p40 by DCs but not macrophages is TLR2 dependent
To determine whether TLR2 or TLR4 is involved in IL-12 production by DC or macrophages, cells from TLR2 KO, TLR4 KO, or C57BL/6 mice were used in this assay. For DC, cells of TLR2 or TLR4 KO were incubated with E. coli LPS or HKBA. As observed in Fig. 7,A, TLR2 is required for production of IL-12 by DCs stimulated with HKBA. Regarding macrophages, IL-12 production in this cell type is independent of TLR2 and TLR4 (Fig. 7,B). These results were confirmed by blocking TLR2 or TLR4 receptor on DC or macrophages by mAbs (Fig. 7, C and D).
MyD88 KO treatment with rIL-12 enhances resistance to Brucella
To determine whether the greater susceptibility of MyD88 KO mice to brucellosis is due to a lesser ability to produce IL-12 or TNF-α in response to Brucella, these mice were treated with rTNF-α or recombinant adenovirus-expressing murine IL-12 (Ad5mIL-12). This strategy using a nonreplicating adenovirus-expressing murine IL-12 has shown to be more effective than the direct administration of rIL-12 protein (18). After 1 wk of infection, there was a statistically significant reduction in splenic Brucella CFU of MyD88 KO mice treated with rIL-12 but not with rTNF-α (Fig. 8). This result demonstrated that lack of IL-12 production in MyD88 KO is partially responsible for enhanced susceptibility to brucellosis observed in these animals.
TLR9 is required to control Brucella infection in vivo
To determine the role of TLR9 in Brucella infection in vivo, TLR9 KO mice were infected with virulent strain 2308. As demonstrated in Fig. 9, TLR9 KO mice showed enhanced susceptibility to B. abortus infection (log 6.16) compared with C57BL/6 mice (log 5.02) after 2 wk postinfection. However, TLR9 KO mice are more resistant than MyD88 KO animals. This result reflects the involvement of other TLRs that signal through MyD88.
Production of IL-12p40 by macrophage and DCs but not TNF-α is partially dependent on TLR9
To determine whether TLR9 played a role in IL-12 or TNF-α production, we stimulated macrophages or DC with HKBA, E. coli LPS or CpG. IL-12 but not TNF-α synthesis was reduced on either macrophages or DC activated with HKBA (Figs. 10 and 11). In contrast, induction of IL-12 by E. coli LPS, a TLR4 agonist, was not reduced in TLR9 KO mice. Additionally, NO production was intact on TLR9 KO macrophages stimulated with HKBA compared with C57BL/6 (Fig. 10 C). These findings indicate that IL-12 production induced by HKBA on macrophages and DC is partially dependent on TLR9.
DCs are professional APC critical for bridging innate and adaptive immune responses (29). In the E. coli model, bacterial products can induce functional DC maturation in an MyD88-independent manner (30). These results clearly indicate that the MyD88-independent pathway can lead to DC maturation not only in vitro, but also in vivo. During maturation, DC increase their ability to present Ag and prime T cells by up-regulating their cell surface MHCII and CD40−, CD80−, and CD86− costimulatory molecules (29, 31). In our study, we have demonstrated that B. abortus-mediated DC maturation is dependent on the adaptor molecule MyD88. Additionally, we have confirmed that MyD88 KO mice are highly susceptible to B. abortus infection in vivo and also demonstrated the role of TLR9 in host control of infection. Furthermore, MyD88 KO susceptibility was characterized by lack of DC maturation and production of IL-12 and, therefore, regulating activation of Brucella-specific IFN-γ-producing T cells.
Regarding the ability of host TLRs to recognize Brucella, there are some contradictory data reported in the literature. Our group has reported a role for TLR2 or TLR4 on Brucella signaling, but we only observed the involvement of TLR4 in resistance (9). Others have indicated that neither TLR2 nor TLR4 plays a role in host resistance to B. abortus infection (14). The reason for this discrepancy is not known but may relate to the different B. abortus strains used. However, Weiss et al. (14) demonstrated the critical role of the adaptor molecule MyD88 for Brucella clearance in vivo. Here, we confirmed the predominant role of MyD88 in host resistance to B. abortus infection and further investigated the mechanisms involved in MyD88 KO susceptibility. Initially, inflammatory macrophages from MyD88 KO mice stimulated with HKBA produced no IL-12 and TNF-α. Low levels of these cytokines induced by HKBA in MyD88 KO macrophages were not altered by activation with rIFN-γ. As observed with other intracellular microbial pathogens, IL-12-derived IFN-γ production and TNF-α are critical components of the immune system to control murine brucellosis (19, 20, 21, 22, 23, 24, 25). Additionally, low levels of NO were observed in MyD88 KO macrophages. Previous study has demonstrated the role of NO as an important molecule in the control of Brucella infection (22). Furthermore, low levels of RANTES, MCP-1, and MIP-1α were detected in MyD88 KO macrophages culture stimulated with HKBA compared with wild-type cells. These inflammatory chemokines are potent leukocyte chemoattractants and critical on Th1 cell activation. In contrast to our findings, production of chemokines such as RANTES during Mycobacterium tuberculosis and Pseudomonas aeruginosa infection is MyD88 independent (32, 33). The lack of chemokine production may be another important deficiency in MyD88 KO macrophages when activated by HKBA that could enhance host susceptibility to murine brucellosis.
DC are professional APCs and the major cell type involved in IL-12 production. Because IL-12 is critical for Th1 cell differentiation and IFN-γ production by CD4+ T cells, we investigated the role of MyD88 in DC maturation and activation by B. abortus. First, we confirmed that HKBA induces DC maturation in C57BL/6 mice by increased expression of CD40, CD86, and MHCII molecules on CD11c+ BM or splenic DC. In contrast, DC maturation was totally abrogated on MyD88 KO cells when HKBA was used as stimulus. This result was confirmed when in vivo splenic DC maturation was impaired in MyD88KO mice that received HKBA but not in TLR2 KO or TLR9 KO animals. Conversely, E. coli LPS could induce functional maturation of MyD88-deficient DC as previously reported (34). The inhibition of Brucella-induced DC maturation in MyD88-deficient mice argues in favor of a critical role for MyD88 in the detection of B. abortus by DC. Therefore, these results suggest that recognition of Brucella products by TLRs might be a prerequisite for MyD88-dependent DC maturation. Additionally, we measured IL-12 and TNF-α production in the supernatant of HKBA, E. coli LPS, or zymosan-stimulated DCs. As shown in Fig. 6, production of these proinflammatory cytokines was totally abrogated in MyD88 KO DC either stimulated with HKBA or E. coli LPS but these cells were able to produce IL-12 and TNF-α when stimulated with zymosan.
To evaluate whether the lack of IL-12 or TNF-α in MyD88 KO could be responsible for reduced resistance to brucellosis, we treated these mice with rTNF-α or rIL-12. Administration of rIL-12 enhanced resistance of MyD88 KO mice to Brucella while rTNF-α therapy did not influence the outcome of infection. This result demonstrated that lack of IL-12 production in MyD88 KO is partially responsible for enhanced susceptibility to brucellosis observed in these animals.
To determine the TLR requirement for IL-12 production by DC stimulated with Brucella, we first investigated the role of TLR2 and TLR4. Previous studies reported that IL-12 secretion by splenic DC exposed to HKBA is TLR2/TLR4 independent but TLR9 dependent (12, 13). However, in our experiments, we observed that IL-12 production by BMDC is MyD88, TLR2, and TLR9 dependent. Similar results were obtained using anti-TLR treatment on C57BL/6 DC. This contrasting result may be due to different DC subsets used (35, 36, 37). The present study used BMDC, while splenic DC were used by others. Furthermore, other investigators have demonstrated the involvement of TLR2 in TNF-α and IL-6 production by a human monocytic cell line stimulated by HKBA (38). However, when macrophages from MyD88 KO or TLR2 KO were studied after HKBA stimulation, IL-12 production was dependent of MyD88 and independent of TLR2 and TLR4. Therefore, we concluded that TLR signaling to induce IL-12 production by HKBA is regulated differently in BMDC and inflammatory macrophages. A similar finding was observed in BMDC and macrophages stimulated with M. tuberculosis (39). Additionally, we have observed that HKBA secretion of IL-12 by TLR9 KO macrophages or DC was only partially reduced, suggesting a role for other receptors in IL-12 signaling. Also, TNF-α production by macrophages or DC and NO by macrophages from TLR9 KO was not altered when compared with wild-type mice.
Recently, Huang et al. (13) reported the involvement of TLR9 on Brucella-driven Th1 responses. They suggested that DC in T cell areas are stimulated by DNA from HKBA via TLR9 to secrete IL-12 that activates nearby T cells to differentiate as Th1. Our investigation of the role of TLR9 in host control of B. abortus in vivo demonstrated that TLR9 deficiency enhances susceptibility to B. abortus infection. However, in comparison to TLR9, MyD88 deficiency has a more predominant role in host resistance to Brucella which suggests the involvement of other receptors that signals through MyD88 during the course of infection. In summary, the scenario emerging from our study is that susceptibility of MyD88 KO mice to B. abortus infection is due to impaired DC maturation and lack of IL-12 production. Because DC maturation is a critical link between innate and adaptive immunity, MyD88-dependent signaling appears to be required for the development of IFN-γ-producing T cells and efficient control of Brucella infection (40, 41). This was confirmed by lack of IFN-γ production in MyD88 KO lymph node cells of B. abortus-infected mice. Furthermore, we demonstrated by flow cytometry that CD4+ T cells from MyD88 KO mice have a deficient production of IFN-γ when stimulated with Brucella-specific recombinant Ags. Taken together, this study provides new insights into how innate and adaptive immunity are orchestrated to control B. abortus infection.
We thank Dr. S. Akira (Department of Host Defense, Research Institute for Microbial Diseases, Osaka University, Osaka, Japan) for providing the knockout mice.
The authors have no financial conflict of interest.
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.
This work was supported by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), CNPq/Centro Brasil Argentino de Biotecnologia, Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG), and FAPEMIG/Pronex.
Abbreviations used in this paper: DC, dendritic cell; HKBA, heat-killed Brucella abortus; KO, knockout; MHCII, MHC class II; BM, bone marrow.