Trypanosoma Infection Favors Brucella Elimination via IL-12/IFNγ-Dependent Pathways

This study develops an original co-infection model in mice using Brucella melitensis, the most frequent cause of human brucellosis, and Trypanosoma brucei, the agent of African trypanosomiasis. Although the immunosuppressive effects of T. brucei in natural hosts and mice models are well established, we observed that the injection of T. brucei in mice chronically infected with B. melitensis induces a drastic reduction in the number of B. melitensis in the spleen, the main reservoir of the infection. Similar results are obtained with Brucella abortus- and Brucella suis-infected mice and B. melitensis-infected mice co-infected with Trypanosoma cruzi, demonstrating that this phenomenon is not due to antigenic cross-reactivity. Comparison of co-infected wild-type and genetically deficient mice showed that Brucella elimination required functional IL-12p35/IFNγ signaling pathways and the presence of CD4+ T cells. However, the impact of wild type and an attenuated mutant of T. brucei on B. melitensis were similar, suggesting that a chronic intense inflammatory reaction is not required to eliminate B. melitensis. Finally, we also tested the impact of T. brucei infection on the course of Mycobacterium tuberculosis infection. Although T. brucei strongly increases the frequency of IFNγ+CD4+ T cells, it does not ameliorate the control of M. tuberculosis infection, suggesting that it is not controlled by the same effector mechanisms as Brucella. Thus, whereas T. brucei infections are commonly viewed as immunosuppressive and pathogenic, our data suggest that these parasites can specifically affect the immune control of Brucella infection, with benefits for the host.

This study develops an original co-infection model in mice using Brucella melitensis, the most frequent cause of human brucellosis, and Trypanosoma brucei, the agent of African trypanosomiasis. Although the immunosuppressive effects of T. brucei in natural hosts and mice models are well established, we observed that the injection of T. brucei in mice chronically infected with B. melitensis induces a drastic reduction in the number of B. melitensis in the spleen, the main reservoir of the infection. Similar results are obtained with Brucella abortus-and Brucella suis-infected mice and B. melitensis-infected mice co-infected with Trypanosoma cruzi, demonstrating that this phenomenon is not due to antigenic cross-reactivity. Comparison of co-infected wild-type and genetically deficient mice showed that Brucella elimination required functional IL-12p35/IFNγ signaling pathways and the presence of CD4 + T cells. However, the impact of wild type and an attenuated mutant of T. brucei on B. melitensis were similar, suggesting that a chronic intense inflammatory reaction is not required to eliminate B. melitensis. Finally, we also tested the impact of T. brucei infection on the course of Mycobacterium tuberculosis infection. Although T. brucei strongly increases the frequency of IFNγ + CD4 + T cells, it does not ameliorate the control of M. tuberculosis infection, suggesting that it is not controlled by the same effector mechanisms as Brucella. Thus, whereas T. brucei infections are commonly viewed as immunosuppressive and pathogenic, our data suggest that these parasites can specifically affect the immune control of Brucella infection, with benefits for the host.
Keywords: infection control, Trypanosoma cruzi, Trypanosoma brucei brucei, Brucella melitensis, Brucella abortus, brucellosis, Mycobacterium tuberculosis inTrODUcTiOn In natural populations, individual resistance to infection is remarkably diverse. This has been linked to many factors [reviewed in Ref. (1)]. Among them, persistent infection can enhance the ability to control unrelated pathogens, a phenomenon termed the "Mackaness effect" in reference to the seminal work of Mackaness (2,3) demonstrating cross-protection between L. monocytogenes,

MaTerials anD MeThODs ethics statement
The procedures used in this study and the handling of the mice complied with current European legislation (directive 86/609/ EEC) and the corresponding Belgian law "Arrêté royal relatif à la protection des animaux d' expérience du 6 avril 2010 publié le 14 mai 2010. " The Animal Welfare Committee of the Université de Namur (UNamur, Belgium) reviewed and approved the complete protocol (Permit Number: 12-188).

Brucella infection
We used wild-type B. melitensis 16M and strains stably expressing a rapidly maturing variant of the red fluorescent protein DsRed (mCherry-Br) (24), the mCherry protein (mCherry-Br), under the control of the strong Brucella spp. promoter, PsojA. Construction of the mCherry-Br strains has been described previously in detail (25). We also used B. abortus 2308 and Brucella suis 1330. All Brucella strains were grown in biosafety level III laboratory facilities. Cultures were grown overnight with shaking at 37°C in 2YT medium (Luria-Bertani broth with double quantity of yeast extract) and were washed twice in RPMI 1640 (Gibco Laboratories) (3,500 × g, 10 min) before inoculation of the mice. The mice were anesthetized with a cocktail of xylazine (9 mg/kg) and ketamine (36 mg/kg) in PBS before being inoculated intranasally (i.n.) with 2 × 10 4 CFU of B. melitensis, B. abortus, and B. suis in 30 µl of PBS [described in Ref. (25)]. For the i.p. infection with B. melitensis, the mice received 2 × 10 4 CFU/500 μl of PBS. Control animals were inoculated with the same volume of PBS. The infectious doses were validated by plating serial dilutions of the inocula. The mice were sacrificed at the selected time after infection by cervical dislocation. Immediately after sacrifice, spleen cells were collected for bacterial count and flow cytometry analyses.  (27) were used in this study. T. brucei infection is characterized by multi-wave parasitemic development, in which every wave represents a parasite population of different antigenic type. DNac mutant parasitemia is considerably lower and displays no peak of infection (27). The mice were infected by i.p. injection of 5,000 parasites/mouse. Every 2-3 days, the number of parasites present in the blood was estimated using a counting chamber and a light microscope. For T. brucei elimination, the mice were treated by i.p. injection of Berenil (diminazene aceturate, 14 mg/kg, Sigma-Aldrich) in 200 µl of distillated water (28). A parasite lysate was obtained by three freeze/thaw cycles as described in Ref. (29).

M. tuberculosis infection
For Trypanosoma cruzi, we used the Tulahuen strain (genotype TcVI). The mice were infected by i.p. injection of 1,000 blood trypomastigotes as previously described (30). Blood parasitemia was evaluated regularly by microscopic examination.

antibiotic Treatment
Antibiotic treatment was administered to both immunized and control mice for 2 weeks. The oral treatment was a combination of rifampicin (12 mg/kg) and streptomycin (450 mg/kg) [adapted from Ref. (31)] prepared fresh daily and given in the drinking water. To ensure that the antibiotic treatment was effective, some mice from each group were sacrificed 1 week prior to the challenge, and the colony-forming unit counts were evaluated in the spleen.

Brucella counting
Spleens were crushed and transferred to PBS/0.1% X-100 Triton (Sigma-Aldrich). We performed successive serial dilutions in RPMI to obtain the most accurate bacterial count and plated them onto 2YT medium. The colony-forming units were counted after 5 days of culture at 37°C.

enzyme-linked immunosorbent assay (elisa)
Brucella-specific murine IgM and IgG isotypes were determined by ELISA. Polystyrene plates (Nunc 269620) were coated with heat-killed B. melitensis (10 7 CFU/ml). After incubation overnight at 4°C, the plates were blocked for 2 h at room temperature (RT) with 200 µl of PBS-3.65% casein. The plates were then incubated overnight at 4°C with 50 µl of serial dilutions of the serum in PBS-3.5% casein. The sera of unimmunized mice were used as the negative control. After four washes with PBS, isotypespecific goat anti-mouse horseradish peroxidase conjugates were added (50 μl/well) at appropriate dilutions (hIgM from Sigma; LO-MG2a-9 HRPO). The plates were incubated for 2 h at RT and washed four times in PBS before adding 100 µl of substrate solution (BD OptEiA) to each well. After 10 min of incubation at RT in the dark, the enzyme reaction was stopped by adding 25 μl/well of 2 N H2SO4, and absorbance was measured at 450 nm.

cytofluorometric analysis
As described previously (16), spleens were harvested, cut into small pieces and incubated for 30 min at 37°C with a mix of DNAse I fraction IX (Sigma-Aldrich) (100 µg/ml) and 1.6 mg/ml of collagenase (400 M and l U/ml). The spleen cells were washed and filtered, and then incubated with saturating doses of puri-

statistical analysis
We used a (Wilcoxon-)Mann-Whitney test provided by the GraphPad Prism software to statistically analyze our results. Each group of deficient mice was compared to the wild-type mice. We also compared each group with each other and displayed the results when required. Values of p < 0.05 were considered to represent a significant difference. *, **, and *** denote p < 0.05, p < 0.01, and p < 0.001, respectively.

B. melitensis
Persists in cD11c + reservoir cells in the spleen and is able to resist to Protective Memory immune response The identification of reservoir cells allowing for the persistence of Brucella in vivo constitutes a crucial step in our understanding of how Brucella escapes the immune system. Following i.n. administration of 2 × 10 4 CFU of B. melitensis, we have shown previously (16) that the bacteria persist in the lungs up to 12 days postinfection and the spleen and liver are colonized starting 5 days postinfection. At 28 days postinfection, B. melitensis is  only detected in the spleen by CFU analysis. At that time, it is not detected in the lungs, liver, brain, ovaries, heart, thigh muscle, and tissues of the tail (data not shown). Microscopic analysis has demonstrated that during the chronic phase of infection of highly susceptible IL-12p40 −/− BALB/c mice, B. melitensis resides in specific reservoir cells expressing a particular phenotype (CD11c + CD11b − CD205 + arginase + ) (32). Due to the low number of B. melitensis persisting in the spleen of resistant mice (10 3 -10 4 CFU/spleen, less than 1 infected cell per 10 4 spleen cells), in situ microscopic analysis of the phenotype of the infected cells is not possible. To solve this problem, we chose to use CD11c-DTR C57BL/6 mice that express the DT receptor under the control of the CD11c promoter. As previously reported (33), injection of DT in DTR mice induced the transitory elimination of CD11c + cells.
To determine if the B. melitensis reservoir cells express CD11c, DTR mice infected for 26 or 48 days with 2 × 10 4 CFU of mCherry-B. melitensis received 500 ng of DT in 500 µl of PBS or PBS alone (control group). Two days later, the mice were sacrificed and the spleens were harvested. Flow cytometry analysis showed that DT induced a depletion of CD11c high cells in control and Brucella-infected DTR mice but not in wild-type mice ( Figure 1A). Elimination of the CD11c + cells in DTR mice led to a drastic reduction in the number of bacteria compared to the mice treated with PBS alone (Figure 1B). Note that, as expected, we observed no significant impact of DT administration on the control of Brucella infection in wild-type C57BL/6 mice ( Figure  S1 in Supplementary Material). These results suggest that Brucella persists in CD11c + reservoir cells in wild-type C57BL/6 mice.
We have previously shown that i.p. (31) or i.n. (16) Brucella infection in wild-type C57BL/6 mice induces a protective memory T cell response able to control and completely eliminate a secondary Brucella infection in a majority of mice. To determine whether  the activation of memory T cells during a secondary infection is able to eliminate Brucella from the first infection persisting in splenic reservoir cells, we compared the CFU levels of nonfluorescent (wild-type) B. melitensis and mCherry-B. melitensis in the spleens of five groups of mice. The precise treatment of each group is described in detail in Figure 2A. As expected, both the i.p. and the i.n. mCherry Brucella challenge was controlled well in mice chronically infected with wild-type Brucella (Figure 2B), thus confirming the presence of a protective memory response in these mice. In striking contrast, the CFU levels of wild-type Brucella were similar in unchallenged mice and in the i.p. and i.n. challenged groups ( Figure 2C). This demonstrates that the memory response is able to eliminate the challenge strain but not the Brucella strain that has settled in the reservoir cells. To the best of our knowledge, this observation has never been reported before and suggests that the reservoir cells protect Brucella against the Brucella-specific memory response.  Figure 3A). At the selected time point following T. brucei infection, the number of parasites was measured in the blood ( Figure 3B). No significant impact of Brucella infection on the course of T. brucei was detected. In the same experiment, at a selected time point following Brucella infection, the mice were sacrificed, the spleens were harvested, and the Brucella CFUs in the spleen were counted by plating (Figures 3C,D). Our results showed that both early (day 7, Figure 3C) and late (day 45, Figure 3D) infection with T. brucei induced a rapid decrease in the CFU count of B. melitensis in the spleen. At 5 days post T. brucei infection (Figure 3C), the CFU level was already decreased by >1 log in the co-infected mice compared to the mice infected with B. melitensis alone. Twenty-one days following early and late T. brucei infection (Figures 3C,D), the CFU level appeared to be reduced by approximately 3 log, and the B. melitensis CFU count was below the detection threshold (10 CFU/spleen) in the majority of the co-infected mice. Similar results were obtained in the i.p. model of B. melitensis infection ( Figure S2 in Supplementary Material). These results suggest that the immune response induced by T. brucei infection is able to eliminate Brucella in its reservoir cells. It is important to remark that this effect is not due to depletion of the entire CD11c + spleen cell population, like in CD11c-DTR mice treated with DT, as the frequency of CD11c + MHCII + cells is not significantly reduced in the spleen of T. brucei co-infected mice compared to the spleen of Brucella-infected mice ( Figure S3 in Supplementary Material).

T. brucei infection reduces the number of
To determine whether the elimination of B. melitensis observed after T. brucei infection could be the consequence of antigenic cross-reactivity, we tested whether T. brucei can affect the growth of two other Brucella species able to infect humans, B. abortus and B. suis, in the spleen (Figures 4A,B), and whether i.p. T. cruzi infection also reduces the level of B. melitensis ( Figure 4C). T. cruzi is an intracellular parasite displaying different infection kinetics and does not share an antigen with T. brucei. We observed that T. brucei infection leads to a 10-fold reduction of B. abortus and B. suis CFUs in the spleen and that T. cruzi infection also induces a drastic reduction of B. melitensis in the spleen of infected mice. These results strongly suggest that elimination of Brucella following Trypanosoma infection is not due to shared antigens between the pathogens.  compared to wild-type mice (503-fold). These results suggest that the mechanism underlying the elimination of B. melitensis by T. brucei is indeed dependent on the host adaptive immune response and partially requires functional IL-12/IFNγ signaling pathways and CD4 + T cells. Elimination of B. melitensis following co-infection with T. cruzi is also dependent on IL-12 ( Figure  S2 in Supplementary Material). Note that depletion of CD4 + or CD8 + T cells by injection of depleting antibodies to confirm the data obtain with MHCII −/− and TAP1 −/− mice is not feasible in this model. Indeed, CD11c + cells constitute a potential Brucella reservoir cells, and it is well known that some subpopulations of splenic CD11c + dendritic cells express CD4 or CD8 receptors. As nitric oxide produced by iNOS/NOS2 is widely known to be produced in response to IFNγ in B. melitensis (34) and T. brucei (35) models and to negatively affect B. melitensis growth in the spleen (34), we analyzed by flow cytometry the frequency of IFNγ-and iNOS-producing cells in the spleen of wild-type and various deficient mice infected with B. melitensis, T. brucei or coinfected with both pathogens. Our results showed that T. brucei induced a strong increase in the frequency of IFNγ ( Figure 5B) and iNOS-producing cells (Figure 5C) in the spleen. IFNγ is mainly produced by CD3 + CD4 + T cells and iNOS is produced by CD11b + cells (data not shown). As expected, the absence of IL-12p35 significantly reduces the frequency of IFNγ + and iNOS + cells. Interestingly, there was no significant difference between the frequency of IFNγ and iNOS-producing cells in the spleen of TAP1 −/− (deficient in CD8 + T cells) and MHCII −/− (deficient in CD4 + T cells) mice, demonstrating that the absence of T. bruceiinduced B. melitensis control in MHCII −/− mice is indeed due to the absence of CD4 + T cells and not only to the lack of IL-12/IFNγ signaling pathways.

T. brucei infection Does not affect the course of M. tuberculosis infection in Mice
Brucella melitensis and M. tuberculosis are both facultative intracellular bacteria able to chronically colonize the spleen of infected mice. It is clearly established that IFNγ production (36,37) and activated CD4 + T cells (38) participate in the protective immune response against M. tuberculosis in mice, suggesting that bystander activation of these effector mechanisms by T. brucei could promote the control of M. tuberculosis. Thus, we tested the impact of i.p. T. brucei infection on the course of C57BL/6 mice previously infected by aerosol with M. tuberculosis, as described in Figure 6A. Surprisingly, we observed that, despite the enhanced frequency of IFNγ-producing cells in the spleen of co-infected mice (Figure 6B), T. brucei did not affect M. tuberculosis infection in the spleen either earlier (36 days, Figure 6C) or later (91 days, Figure 6D). This result demonstrates that T. brucei infection specifically impacts B. melitensis infection and that these effects cannot be generalized to all other intracellular bacterial infections.

an attenuated T. brucei Mutant is able to reduce Brucella infection
Our previous experiments showed that T. brucei induces a strong inflammatory response mediated by IFNγ + CD4 + T cells in infected  mice and that this response is indispensable to the reduction of Brucella infection in co-infected mice. To determine whether the intensity and duration of the IFNγ-mediated response are key parameters, we compared a wild-type T. brucei and a DNac T. brucei mutant (27), as described in Figure 7A. Infection with the DNac mutant induced a considerably lower parasitemia and a smaller inflammatory immune response (27). We observed that both wild-type and DNac T. brucei induced similar elimination of B. melitensis in co-infected mice ( Figure 7B). The kinetics of elimination were very similar, except for the early time point of co-infection (day 12). The slower impact of the DNac mutant on Brucella could be due to the fact that it induces IFNγ less rapidly in spleen as demonstrated by the reduced frequency of IFNγproducing cells in the spleen of DNac T. brucei-infected mice early in co-infection compared to wild-type T. brucei-infected mice (Figures 7C,D). Like for wild-type T. brucei, Brucella elimination by DNac T. brucei was impaired in IL-12p35 −/− mice ( Figure S4 in Supplementary Material). On the whole, these results suggest that even lower persistence of parasite associated with lower levels of IFNγ is able to mediate B. melitensis elimination. However, live parasite seems to be required as we failed, in the same experiment, to reduce B. melitensis infection by three repeated injection (once per week) of a lysate of T. brucei ( Figure S5 in Supplementary Material). The lysate has been realized, approximately, from a number of cultivated parasites corresponding to the number of parasites detected in the blood of infected mice at the peak of infection.

Transient T. brucei infection Does not alter Development of the Protective Memory response against B. melitensis
Trypanosoma brucei infection has been described to deeply affect the protective humoral immune memory against unrelated pathogens (18). Thus, we planned to analyze the impact of T. brucei infection on the specific protective immune memory in Figure 8A. Both groups of mice were treated with Berenil, an antiparasitic drug (28). After a resting period, the mice were i.p. challenged with mCherry B. melitensis. A group of naive mice, treated with antibiotic and Berenil, was also i.p. infected with mCherry B. melitensis and was used as the internal control (pri Br cont group). Surprisingly, T. brucei infection did not reduce the ability of B. melitensis immunized mice to control an i.p. Brucella challenge in the blood (Figure 8B) or the spleen (Figure 8C), despite a drastic reduction in the frequency of B cells ( Figure S6A in

DiscUssiOn
Immunology arose from the will and the need to ameliorate vaccination. During the 20th century, the molecular biology revolution led immunologists to develop highly reductionist experimental models, far removed from reality. For many years, the most studied immunological models were based on the injection of proteins combined with adjuvants. Much of our understanding of how the immune system works derives from studies of these models. We must not forget that our perception of reality results from the experimental model in which we study it. It may be time to remember that the first successful vaccination actually derives from a cross-reaction between smallpox and the cowpox virus and that multiple unrelated infections in the same host are probably more common than single mono-species infections. The impact of past or chronic unrelated infections on the ability of the host to control infection is well documented [reviewed in Ref. (1)], but the mechanisms remain largely unknown. In this study, we developed an original experimental model of coinfection to study the impact of T. brucei infection on the course of chronic infection with B. melitensis in mice. Over the course of evolution, Brucella has acquired specific stealth strategies that allow it to interfere with its recognition by the immune system and neutralize immune effector mechanisms. For example, after phagocytosis, Brucella controls the intracellular trafficking of its vacuole to avoid degradation by phagolysosomes [reviewed in Ref. (39)]. We have shown previously in a mice model that following i.p. (15) or i.n. infection (16) the spleen is stably and durably colonized by Brucella and constitutes one of the major reservoirs of the infection. In the spleen of highly susceptible IL-12p40 −/− BALB/c mice, Brucella persists within specific CD11c + CD205 + arginase1 + myeloid cells displaying high levels of lipids (32). In this study using CD11c-DTR C57BL/6 mice, we demonstrated that specific elimination of CD11c + cells reduced colony-forming unit counts of Brucella in the spleen, suggesting that Brucella survives in CD11c + reservoir cells in resistant C57BL/6 mice. Our results showed that these reservoir cells constitute a niche that hides Brucella from the primary immune response and even against the protective following B. melitensis infection. We have shown previously that humoral immunity-mediated control of secondary infection with B. melitensis is dependent on the route of infection. In the i.p. model, bacteria disseminate by the blood, and the humoral immune response is indispensable to efficiently control secondary infection (31). In contrast, bacteria disseminate slowly after i.n. infection, are undetectable in the blood, and the absence of B cells does not affect the control of secondary infection (16). We therefore chose to analyze the effects of T. brucei in both i.p. and i.n. B. melitensis infection models.
Mice were i.p. infected with wild-type B. melitensis for 28 days, treated with antibiotic and then infected (Sec Br/T. brucei group) or not (Sec Br cont group) with T. brucei, as described Brucella-specific memory response. Indeed, in mice chronically infected since 45 days with wild-type B. melitensis, a new infection with mCherry B. melitensis challenge strain is completely neutralized. But this specific memory response appears to be unable to eradicate the established wild-type Brucella strain. Thus, during chronic infection phase, Brucella appears to be extremely well equipped to escape the IFNγ-mediated protective immune response (16) and persist in the host. It is therefore really surprising and unexpected, in view of the well documented immunosuppressive effects of many trypanosomes (40), that both T. brucei and T. cruzi infection lead to an almost 2-3 log reduction of the B. melitensis load in the spleen of co-infected mice.
T. brucei is known to suppress both the T-cell (19,20) and B-cell (18)    The fact that two species of Trypanosoma displaying completely different antigens, infectious cycles and dynamics are both able to favor the elimination of three different species of Brucella (melitensis, suis, and abortus) suggests that the underlying mechanism has very little chance of being based on antigenic cross-reactivity. The comparison of various deficient mice has demonstrated that CD4 + T cells, but not CD8 + T cells or γδ + T cells, are indispensable to Brucella elimination, and thus that the immune response is directly implicated in this phenomenon. We also observed that functional IL-12p35/IFNγ signaling pathways are required. Taken together, these results suggest that the strong inflammatory IFNγ-mediated response induced by T. brucei and T. cruzi infection is responsible for the elimination of Brucella. Interestingly, in the absence of CD4 + T cells (MHCII −/− mice), T. brucei induced a strong IFNγ response mediated by CD8 + T cells but was not able to reduce the level of Brucella persistence in the spleen, thus suggesting that IFNγ production is necessary but not sufficient to eliminate Brucella and that CD4 + T cells are key actors in this process. Infection with a strongly attenuated T. brucei mutant, but not repeated injection of a T. brucei lysate, was able to eliminate Brucella in the spleen, suggesting that the level of inflammation is not a limiting factor but that a living parasite is required. B cell deficient mice have been shown to display enhanced control of Brucella infection (43), and B cells have been reported to act as reservoir cells for Brucella in the spleen (44). As T. brucei is well known to induce B cell apoptosis (18), we cannot exclude that Brucella elimination by T. brucei could in part be a consequence of B cell depletion by T. brucei. However, T. brucei infection does not completely eliminate B cells in our model ( Figure S5 in Supplementary Material).
In addition, the colony-forming unit count reduction induced by T. brucei is already observed after 5 days, while no impact on the frequency of B cells in the spleen is detected at the same time [data not shown and Ref. (18)].
The fact that T. brucei infection affects B. melitensis, B. suis, and B. abortus, but not early or late phases of M. tuberculosis infection in the spleen, demonstrates that the mechanism implicated is specific to Brucella species but is not effective against all intracellular bacterial infections. However, the results obtained with M. tuberculosis are unexpected given the strong available evidence that a deficiency of IFNγ responses is associated with increased susceptibility to M. tuberculosis and non-tuberculous mycobacterial infections in animal models and humans (45,46). In addition, Sakai et al. (47) recently nicely confirmed the protective role of IFNγ-producing CD4 + T in the control of extra-pulmonary M. tuberculosis infection. In our experimental model, we observed decreased IFNγ production in vitro in response to known MHC class II epitopes of M. tuberculosis antigens in splenocytes isolated from mice co-infected with M. tuberculosis and T. brucei compared to the responses measured in splenocytes from mice infected with M. tuberculosis (data not shown). This suggests that co-infection with T. brucei could have a negative impact on IFNγ responses by M. tuberculosis specific CD4 + T cells and that direct recognition of infected cells by specific CD4 + T cells rather than the presence of high levels of IFNγ may be essential to restricting the growth of M. tuberculosis. In addition, we cannot exclude that the beneficial impact of IFNγ production may be counterbalanced by some immunosuppressive effects associated to T. brucei infections.
Finally, as T. brucei is especially known for its ability to suppress vaccine-induced protective humoral memory (18), we tested the ability of T. brucei to neutralize Brucella-induced protective memory in i.p. and i.n. infection models. As expected, we observed a significant decrease in the B cell count in the spleen and a reduction of Brucella-specific IgM and IgG2a levels in the serum of co-infected mice compared to control Brucella-infected mice. Interestingly, similar results have been reported in cattle infected with either Trypanosoma congolense or Trypanosoma vivax that display suppressed humoral immune responses to attenuated B. abortus injected subcutaneously (48). Despite this reduction of the Brucella-specific humoral response, we did not observe a reduced ability of the humoral response to neutralize Brucella in the blood or to impair Brucella persistence in the spleen.
On the whole, our results show that both T. brucei and T. cruzi infection are able to positively improve the immune control of Brucella infection in the spleen. This demonstrates that non-antigen specific effector mechanisms could be more efficient at eliminating stealth pathogens like Brucella than antigen specific immune effectors. This finding opens up new perspectives for research on brucellosis as well as for other stealth pathogens. This co-infection model also offers a unique opportunity to identify the effector mechanisms expressed by CD4 + T cells involved in the elimination of Brucella reservoir cells and could be further used to discover new therapeutic strategies for brucellosis without antibiotic treatment. As mice are not a natural host for B. melitensis and T. brucei, extrapolation of these results to natural hosts must be done carefully. However, the geographic distribution of these pathogens is largely overlapping, and all of them infect cattle and humans. Thus, co-infection of the natural host with Brucella and Trypanosoma, although not documented, should not be uncommon, and these effects deserve to be studied in nature.

eThics sTaTeMenT
The procedures used in this study and the handling of the mice complied with current European legislation (directive 86/609/ EEC) and the corresponding Belgian law "Arrêté royal relatif à la protection des animaux d' expérience du 6 avril 2010 publié le 14 mai 2010. " The Animal Welfare Committee of the Université de Namur (UNamur, Belgium) reviewed and approved the complete protocol (Permit Number: 12-188).