NFATc1 Mediates Toll-Like Receptor-Independent Innate Immune Responses during Trypanosoma cruzi Infection

Host defense against the intracellular protozoan parasite Trypanosoma cruzi depends on Toll-like receptor (TLR)-dependent innate immune responses. Recent studies also suggest the presence of TLR-independent responses to several microorganisms, such as viruses, bacteria, and fungi. However, the TLR-independent responses to protozoa remain unclear. Here, we demonstrate a novel TLR-independent innate response pathway to T. cruzi. Myd88 −/− Trif −/− mice lacking TLR signaling showed normal T. cruzi-induced Th1 responses and maturation of dendritic cells (DCs), despite high sensitivity to the infection. IFN-γ was normally induced in T. cruzi-infected Myd88 −/− Trif −/− innate immune cells, and further was responsible for the TLR-independent Th1 responses and DC maturation after T. cruzi infection. T. cruzi infection induced elevation of the intracellular Ca2+ level. Furthermore, T. cruzi-induced IFN-γ expression was blocked by inhibition of Ca2+ signaling. NFATc1, which plays a pivotal role in Ca2+ signaling in lymphocytes, was activated in T. cruzi-infected Myd88−/−Trif−/− innate immune cells. T. cruzi-infected Nfatc1 −/− fetal liver DCs were impaired in IFN-γ production and DC maturation. These results demonstrate that NFATc1 mediates TLR-independent innate immune responses in T. cruzi infection.


Introduction
The host defense against invasion of intracellular pathogens relies on Th1 cell-derived IFN-c that activates macrophages to kill the engulfed pathogens [1].Toll-like receptor (TLR)-mediated recognition of pathogens has been established to induce activation of innate immune cells such as dendritic cells (DCs) and subsequent development of Th1 cells [2,3].However, recent evidence also indicates the presence of TLR-independent mechanisms for the recognition of microorganisms such as bacteria, viruses, and fungi [4,5].Accordingly, TLR-independent mechanisms for Th1 development have been demonstrated in several infectious models such as fungal and bacterial infections [6,7].However, TLR-independent recognition of protozoa remains unknown.
Trypanosoma cruzi is an intracellular protozoan parasite that causes Chagas' disease, a chronic disorder characterized by cardiomyopathy and malformation of the intestine [8].Several components of T. cruzi have been shown to induce TLR-dependent activation of innate immunity and subsequent development of Th1 cells [9][10][11][12][13][14].The absence of TLR-dependent activation of innate immunity results in high susceptibility to T. cruzi infection [15,16] due to defective type I interferon (IFN)-mediated induction of the GTPase IFN-inducible p47 (IRG47) [17].Invasion of infective metacyclic trypomastigotes of T. cruzi into host cells induces a close interaction between the parasites and the host, because T. cruzi utilize several host-derived factors in order to establish the infection.These include activation of Ca 2+ signaling pathways and phosphatidylinositol-3 kinases [18][19][20].However, it remains unclear how T. cruzi-mediated activation of host cytoplasmic signaling pathways is regulated and whether it is TLRdependent or -independent.
In T cells, the nuclear factor of activated T cells (NFAT) family of transcription factors has been shown to mediate production of cytokines including IFN-c [21,22].The NFAT family of proteins comprises four closely related members (NFATc1, NFATc2, NFATc3, and NFATc4) that are activated by Ca 2+ signaling, and NFAT5 that is regulated by osmotic stress.The role of NFAT proteins in T cells has been well characterized [21,22].However, little is known about the role of NFAT proteins in innate immune responses, although some of the NFAT members are highly expressed and can modulate gene induction in macrophages (Mw) [23,24].
Here, we analyzed the mechanisms of TLR-independent activation of innate immunity during T. cruzi infection using T. cruzi-infected Myd88 2/2 Trif 2/2 mice.Our results demonstrate that NFATc1 mediates TLR-independent induction of IFN-c in innate immune cells, leading to effective Th1 responses during T. cruzi infection.

Results
Normal Th1 response in T. cruzi-infected Myd88 2/2 Trif 2/2 mice Previously, we demonstrated that mice lacking both MyD88 and TRIF, in which TLR-dependent activation was abolished, are highly sensitive to infection with T. cruzi [17].Because TLRs have been shown to control development of Th1 cells, we analyzed Th1 responses in T. cruzi-infected mice.Mice were intraperitoneally (i.p.) infected with T. cruzi trypomastigotes, and at 6 days of infection CD4 + T cells were isolated from the spleen and stimulated with anti-CD3 antibody (Ab) (Figure 1A).In T. cruziinfected wild-type mice, there was considerable production of IFNc compared with that in non-infected control mice, indicating induction of potent Th1 responses.In Myd88 2/2 and Myd88 2/2 Trif 2/2 mice, IFN-c production was similar to that in wild-type mice following T. cruzi infection.Next, we analyzed the antigenspecific Th1 response at 0, 4, 6, and 10 days after T. cruzi infection by stimulating CD4 + T cells with freeze-thawed T. cruzi in the presence of antigen presenting cells (APC) (Figure 1B).This stimulation induced marked production of IFN-c at 6 and 10 days of the infection in wild-type mice.Even in CD4 + T cells derived from T. cruzi-infected Myd88 2/2 and Myd88 2/2 Trif 2/2 mice, antigen-specific production of IFN-c was induced to levels similar to that of wild-type mice.Thus, the antigen-specific Th1 response was not impaired in Myd88 2/2 and Myd88 2/2 Trif 2/2 mice.We also analyzed IFN-c production from CD4 + T cells by intracellular staining (Figure 1C, Figure S1A).The number of IFN-c-producing CD4 + T cells was almost equally elevated in wild-type, Myd88 2/2 and Myd88 2/2 Trif 2/2 mice at 6 days (Figure 1C) as well as at 10 days (Figure S1A) after infection.Consistent with previous studies [25,26], the number of IFN-c producing CD8 + T cells and NK1.1 + natural killer cells was not increased at 10 days after T. cruzi infection (Figure S1B).
Development of Th1 cells is critically controlled by DCs [27,28].In addition, stimulation of TLRs induces maturation of DCs [3].Therefore, we analyzed expression of MHC class II and co-stimulatory molecules on T. cruzi-infected DCs.Bone marrowderived DCs (BMDCs) were infected with T. cruzi trypomastigotes for 6 h, then were washed, cultured for 48 h, and analyzed for expression of MHC class II, CD40, and CD86 by flow cytometry (Figure 1D).T. cruzi infection resulted in enhanced expression of these molecules in wild-type BMDCs.Expression was also increased in BMDCs derived from Myd88 2/2 Trif 2/2 mice after T. cruzi infection, indicating normal maturation of T. cruzi-infected DCs of Myd88 2/2 Trif 2/2 mice.Thus, Th1 cell development and DC maturation were induced during T. cruzi infection even in the absence of TLR-dependent activation of innate immunity.
IFN-c induction in T. cruzi-infected Myd88 2/2 Trif 2/2 DCs and macrophages T. cruzi infection induced maturation of DCs in the absence of TLR signaling.Therefore, we screened genes that were normally induced in T. cruzi-infected DCs of Myd88 2/2 Trif 2/2 mice.BMDCs from wild-type, Myd88 2/2 and Myd88 2/2 Trif 2/2 mice were infected with T. cruzi trypomastigotes for 6 h, then mRNA was extracted and used for DNA microarray analysis.Approximately 80% of genes that were induced in T. cruzi-infected wildtype DCs (about 4000 genes) were MyD88-dependent, as the T. cruzi-mediated induction was reduced in Myd88 2/2 DCs (Figure S2).Some of the genes that were normally induced in Myd88 2/2 DCs, but not induced in Myd88 2/2 Trif 2/2 DCs (MyD88/TRIFdependent genes; 14% of genes that were induced in wild-type DCs) are known to be induced by type I IFNs.In addition, a majority of the small number of genes that were induced even in Myd88 2/2 Trif 2/2 DCs (6%) were IFN-c-inducible genes (Figure S3).In order to corroborate that IFN-c-inducible genes are normally induced in T. cruzi-infected Myd88 2/2 Trif 2/2 DCs, we analyzed mRNA expression of Ifng and IFN-c-inducible genes, including Stat1, and Irgm, by real-time RT-PCR (Figure 2A).T. cruzi infection resulted in robust induction of Ifng, Stat1, and Irgm in wild-type, Myd88 2/2 , Trif 2/2 , and Myd88 2/2 Trif 2/2 DCs.T. cruzi-induced expression of Stat1 and Irgm in wild-type DCs was inhibited by addition of a de novo protein synthesis inhibitor, cycloheximide (CHX) (Figure 2B).In contrast, CHX did not inhibit T. cruzi-induced Ifng expression (Figure 2C).Next, in order to analyze whether the expression of the IFN-c-inducible genes was secondary to induction of Ifng, we used BMDCs derived from Ifngr1 2/2 mice in which the IFN-c-mediated response was abolished (Figure 2D, E).In Ifngr1 2/2 BMDCs, T. cruzi-mediated induction of Stat1 and Irgm was reduced, whereas induction of Ifng was unimpaired.These data indicate that Ifng was induced primarily in response to T. cruzi infection, and Stat1 and Irgm were induced secondary to Ifng induction.In peritoneal Mw, similar patterns of T. cruzi-mediated gene expression were observed (Figure S4).Recently, the CD11c low B220 + NK1.1 + subset of cells was identified as a natural killer (NK) cell subset with a high capacity for IFN-c production in response to IL-12 or a TLR9 ligand [29][30][31].To exclude the possibility of contamination of these cells in preparation of BMDCs or Mw, we purified CD11c high B220 2 NK1.1 2 population from the spleen, and analyzed for IFN-c expression (Figure S5A).CD11c high B220 2 NK1.1 2 cells

Author Summary
Trypanosoma cruzi is an intracellular protozoan parasite that causes Chagas diseases in humans.Invasion of T. cruzi into the host is sensed by Toll-like receptors (TLRs), which recognize microbial components that are present in microbes but not in the host.TLRs are essential for the initiation of immune responses against pathogens.Recent evidence indicates the presence of TLR-independent mechanisms for the recognition of microbes, such as bacteria, viruses, and fungi.However, TLR-independent recognition of protozoa remains unknown.We found that immune responses against T. cruzi were induced even in the absence of TLR signaling.The TLR-independent responses were found to be mediated by IFN-c production in innate immune cells.Furthermore, the TLR-independent IFN-c production was revealed to be mediated by Ca 2+dependent activation of NFATc1, which has been shown to play a pivotal role in cytokine production in T lymphocytes.
Our study provides a novel mechanism for the TLRindependent innate immune response against protozoan parasites.It is also worth noting that the host defense mechanism utilizes a factor (Ca 2+ ) that is a prerequisite for the survival of intracellular protozoan parasites.
showed very low levels of IFN-c expression in response to IL-12/IL-18 stimulation compared with the NK cell subset (Figure S5B).However, these cells from wild-type and Myd88 2/2 Trif 2/2 mice expressed IFN-c in response to T. cruzi infection (Figure 2F).Flow cytometric analysis further demonstrated that T. cruzi-infected CD11c high splenic DCs expressed IFN-c protein (Figure 2G).These findings indicate that T. cruzi infection induces IFN-c production in DCs.
Involvement of Ca 2+ signaling in IFN-c induction in T. cruzi-infected Myd88 2/2 Trif 2/2 cells Next, we analyzed the molecular mechanisms for TLRindependent induction of IFN-c after T. cruzi infection.In Myd88 2/2 Trif 2/2 DCs, T. cruzi-induced phosphorylation of MAP kinases such as ERK, p38, and JNK, as well as degradation of IkBa was not observed at all (Figure S8A).In addition, T. cruzi infection did not induce DNA binding activity of NF-kB in Myd88 2/2 Trif 2/2 DCs (Figure S8B).Thus, T. cruzi-mediated activation of NF-kB and MAP kinases was not induced in the absence of TLR signaling.Next, we stimulated DCs with T. cruzi trypomastigotes killed by repeated freeze-thaw steps.Live T. cruzi, but not killed parasites, induced Ifng expression (Figure 4A).Because many studies have demonstrated that T. cruzi utilize the host Ca 2+ signaling to establish the infection [35], we assessed the intracellular Ca 2+ concentration in T. cruzi-infected BMMw using a fluorescent Ca 2+ indicator Fluo-4 AM (Figure 4B, Figure S9A,  B). T. cruzi infection led to rapid increase in intracellular Ca 2+ level in both wild-type and Myd88 2/2 Trif 2/2 Mw, which returned to the basal level after 18 min of the infection.Epimastigotes, which are not able to invade the host cells, did not induce the elevation of Ca 2+ concentration in Mw (Figure S9C).These results prompted us to examine whether Ca 2+ mobilization induced by intracellular invasion of T. cruzi contributed to the TLR-independent Ifng induction.Accordingly, we treated wild-type and Myd88 2/2 Trif 2/2 BMDCs with an intracellular Ca 2+ chelator, bis-(o-aminophenoxy)ethane-N,N,N9,N9-tetraacetic acid tetra (acetoxymethyl) ester (BAPTA-AM), and infected with T. cruzi.In BAPTA-AM pretreated DCs, T. cruzi-induced Ifng expression was severely reduced, although lipopolysaccharide (LPS)-induced response was not impaired (Figure 4C).In this condition, T. cruzi-induced elevation of intracellular Ca 2+ concentration was severely reduced (Figure S10).In addition, stimulation with both phorbol myristate acetate (PMA)/Ca 2+ ionophore or Ca 2+ ionophore alone, which mimics Ca 2+ signaling, induced expression of Ifng in both wild-type and Myd88 2/2 Trif 2/2 Mw (Figure 4D).Taken together, these findings indicate that T. cruzi-dependent intracellular Ca 2+ mobilization mediates TLR-independent Ifng induction.
NFATc1 activation in T. cruzi-infected Myd88 2/2 Trif 2/2 cells In the host cells, especially in T lymphocytes, Ca 2+ mobilization induces activation of cytokine genes via calmodulin/calcineurindependent activation of the transcription factor NFAT. Therefore, we treated Myd88 2/2 Trif 2/2 Mw with FK506 to block calcineurin activation, and infected with T. cruzi.Treatment of FK506 resulted in a marked decrease in T. cruzi-induced expression of Ifng, despite normal LPS-induced response (Figure 5A).Among NFAT members, Nfatc1, Nfatc3, and Nfat5 mRNA were abundantly expressed in BMDCs (Figure S11).A previous study has shown that NFATc1 increased anti-CD3/anti-CD28-induced IFN-c promoter activity in T cells [36].Furthermore, it has been demonstrated that IFN-c production was normal in NFATc3deficient splenocytes [37].In addition, NFAT5 has been shown to be activated by osmotic stress, but not by Ca 2+ signaling [38].Thus, we focused on NFATc1.In wild-type and Myd88 2/2 Trif 2/2 BMMw, T. cruzi trypomastigotes infection induced nuclear translocation of NFATc1 (Figure 5B).T. cruzi-induced nuclear translocation of NFATc1 was blocked by the pre-treatment with BAPTA-AM in wild-type and Myd88 2/2 Trif 2/2 BMMw (Figure S12A, B).These results indicate that NFATc1 is activated in response to T. cruzi infection in a TLR-independent manner.Next, we analyzed whether NFATc1 was involved in the T. cruzi-induced IFN-c production.We obtained RAW264.7 macrophage clones expressing different levels of NFATc1  (A, B) Splenic CD4 + T cells were isolated from wildtype, Myd88 2/2 and Myd88 2/2 Trif 2/2 mice after 6 days (A) or the indicated days (B) of T. cruzi infection or PBS injection, and stimulated with anti-CD3 antibody (A) or freeze-thawed T. cruzi in the presence of antigen presenting cells (B).After 24 h, supernatants were assayed for IFN-c production by ELISA.Data are mean+s.d. of triplicate determination and a representative result of at least three independent experiments.In each experiment, two or three mice in each group were used.*: not detected.(C) Splenocytes were isolated from T. cruzi-infected or none-infected wild-type, Myd88 2/2 and Myd88 2/2 Trif 2/2 mice, and stimulated with 1 mg/ml ionomycin plus 50 ng/ml PMA.After surface staining with APC-conjugated anti-CD4 Ab, the cells were permeabilized and then stained with PE-conjugated anti-IFN-c Ab, and analyzed by flow cytometry.The percentage of IFN-c-producing CD4 + cells of individual mice is shown.(D)T.cruzi-infected or none infected bone marrow DCs of wild-type and Myd88 2/2 Trif 2/2 were stained for expression of CD40, CD86, and I-A b , and then analyzed by FACS.doi:10.1371/journal.ppat.1000514.g001(Figure 5C).In NFATc1 expressing RAW264.7 cells, T. cruziinduced expression of Ifng, Stat1, and Irgm was enhanced, and the extent of fold-induction correlated with the NFATc1 expression level (Figure 5D).T. cruzi-induced Ifng expression was severely reduced in the presence of FK506 (Figure 5E).These findings indicate the possible involvement of NFATc1 in mediating IFN-c production in T. cruzi-infected innate immune cells.
The NFAT family of transcription factors has been shown to interact with different transcription factors to effectively induce gene activation [22].In the case of activation of the human IFNG promoter, T-bet (encoded by Tbx21) and NFAT have been shown to act synergistically on the promoter [39].T-bet is a transcription factor that is essential for IFN-c production in T cells and DCs [40].Tbx21 has also been shown to be induced by IFN-c in monocytes and DCs [41].In accordance with those studies, Tbx21 was normally induced in T. cruzi-infected Myd88 2/2 Trif 2/2 Mw (Figure 6A).Expression of NFATc1 alone weakly activated the Ifng promoter in RAW264.7 Mw, but introduction of both NFATc1 and T-bet synergistically activated the Ifng promoter in RAW264.7 Mw (Figure 6B).These findings indicate that NFATc1 mediates activation of the Ifng promoter together with T-bet in innate immune cells, like the case in T cells.

Discussion
In the present study, we analyzed TLR-independent innate immune responses against the intracellular protozoan parasite T. cruzi.T. cruzi-infected Myd88 2/2 Trif 2/2 mice displayed normal Th1 responses and normal DC maturation.A comprehensive analysis of gene expression profiles of T. cruzi-infected DCs identified IFN-c as a TLR-independent gene which mediated DC maturation and Th1 responses even in the absence of TLR signaling.T. cruzi infection induced an increase in intracellular Ca 2+ level in DCs and macrophages, which led to NFATc1 activation and IFN-c induction in a TLR-independent manner.In Nfatc1 2/2 DCs, T. cruzi-induced IFN-c production and DC maturation was impaired.These findings demonstrate that NFATc1 is responsible for TLR-independent innate immune responses during T. cruzi infection.
The family of TLRs has been established to be critical for the innate recognition of T. cruzi [14].TLR signaling pathways consist of two major components mediated by MyD88 and TRIF [43].Myd88 2/2 mice show a high susceptibility to T. cruzi infection [15,16], while mice deficient in both MyD88 and TRIF are even more susceptible to T. cruzi infection [17].These findings indicate that TLR-dependent recognition of T. cruzi is crucial to the host defense against the parasite.In this regard, TLR-dependent induction of IFN-b might be responsible for high susceptibility to T. cruzi infection in Myd88 2/2 Trif 2/2 mice in spite of the normal Th1 responses [17].
IFN-c was identified as a gene induced in T. cruzi-infected Myd88 2/2 Trif 2/2 DCs.IFN-c production by DCs was first demonstrated in IL-12-stimulated CD8a + lymphoid DCs [44].Subsequently, CD11c low B220 + NK1.1 + cells were shown to produce high amounts of IFN-c in response to IL-12 or a TLR9    ligand [45,46].These CD11c low B220 + NK1.1 + cells have been shown to be a subset of NK cells [29][30][31].Thus, IFN-c production from DCs as well as Mw is controversial [47,48].In order to exclude the possibility that our DC or Mw preparations were contaminated NK cell subsets, we isolated CD11c high B220 2 NK1.1 2 cells and analyzed IFN-c production.These cells showed a severely reduced level of IL-12/IL-18induced IFN-c production compared with NK cells.In addition, IL-12/IL-18-induced IFN-c expression was less than T. cruziinduced expression in these cells (our unpublished data).In Nfatc1 2/2 FLDCs, IL-12/IL-18-induced IFN-c expression was not impaired (our unpublished data).These results indicate that T. cruzi-induced IFN-c production in DCs is mediated by a pathway distinct from the IL-12 (or the TLR9 ligand)-induced one in NK subsets.
Protozoan parasites including T. cruzi require Ca 2+ for their survival within the host cells [19].In addition, T. cruzi evokes elevation of intracellular Ca 2+ concentration in the host cells to establish the invasion [49,50].Our findings indicate that T. cruziinduced activation of host Ca 2+ signaling mediates IFN-c production.In the host cells, the family of NFAT transcription factors, which is activated by calmodulin/calcineurin, is known to bridge Ca 2+ to promote gene expression [51].The role of NFAT proteins has been well characterized in T lymphocytes, and can induce activation of the Ifng gene [22,52].However, the role of NFAT proteins in innate immune cells remains unclear.Several reports indicate NFAT proteins are activated in macrophages [24,53].In addition, cyclosporin A, which blocks calcineurindependent NFAT activation, has been shown to inhibit DC functions [54,55].In accordance with these reports, in the present study NFATc1 was activated in Myd88 2/2 Trif 2/2 Mw in response to T. cruzi infection.Furthermore, analysis using FLDCs derived from Nfatc1 2/2 embryos demonstrated that NFATc1 mediates T. cruzi-induced IFN-c production and DC maturation.These findings establish a new signaling pathway mediating an innate immune response during T. cruzi infection.It is well established that the TLR-dependent pathway initiates innate immune responses against pathogens.In addition, in the case of invasion of protozoan parasites triggering activation of intracellular Ca 2+ signaling, NFATc1 mediates the TLR-independent innate immune responses through induction of IFN-c.Bradykinin B 2 receptor has been shown to mediate T. cruzidependent generation of inositol 1,4,5-trisphosphate, leading to elevated level of intracellular Ca 2+ [56].However, several other mechanisms that induce intracellular Ca 2+ influx have been proposed in T. cruzi infection [18].Identification of critical molecules that lead to NFATc1 activation during T. cruzi infection would be a future issue to be addressed.It would be also interesting in the future to analyze whether the NFAT pathway is involved in innate immune responses against other protozoan parasites such as Toxoplasma and Leishmania species.
In this study, we focused on DCs and Mw, which initiate Th1 responses.However, in an in vivo condition, T. cruzi are expected to invade into several other types of cells than DCs and Mw.Therefore, it is possible that the invasion of T. cruzi into cells of non-innate immune cell populations indirectly influences Th1 polarization in vivo.In the future, we should analyze whether the NFAT family of transcription factors is involved in these processes.
In summary, in the present study we revealed a new TLRindependent mechanism for the interaction between protozoan parasites and host innate immunity.Ca 2+ is critical for both living organisms, and therefore the parasite utilizes host Ca 2+ for its benefit.On the host side, Ca 2+ signaling leads to activation of NFATc1 to eliminate the parasite.It would be interesting in the future to analyze the precise role of NFATc1 in protozoan parasite infection using innate immune cell-specific NFATc1-deficient mice.

Mice
All animal experiments were conducted in accordance with guidelines of the Animal Care and Use Committee of Osaka University and Kyushu University.Myd88 2/2 , Trif 2/2 , Ifngr1 2/2 , and Nfatc1 2/2 mice were generated as previously described [17,42].Each mouse strain was backcrossed to C57BL/6 for at least five generations, and then used to generate double or triplemutant mice.

Preparation of macrophages (Mw), dendritic cells (DCs), and antigen presenting cells (APC)
To isolate peritoneal Mw, mice were i.p. injected with 2 ml of 4% thioglycolate medium (Sigma), and peritoneal exudate cells were isolated from the peritoneal cavity at three days post injection.The cells were incubated for 2 h, then washed three times with HBSS.The remaining adherent cells were used as peritoneal Mw in experiments.To prepare bone marrow-derived DCs or Mw, bone marrow cells were prepared from the femur and tibia, and cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS), 100 mM 2-mercaptoethanol (2ME), and 10 ng/ml GM-CSF (Pepro Tech) or 30% L cell culture supernatant, respectively.After six days, the cells were used as bone marrow DCs or bone marrow Mw in experiments.To prepare fetal liver-derived DCs (FLDCs), fetal liver (FL) were obtained from 12.5 days post-coitum murine embryos, and FL cells were dissociated by pipetting, passed through a nylon mesh, and then cultured in RPMI 1640 medium supplemented with 10% FBS, 100 mM 2ME, 20 ng/ml GM-CSF, 10 ng/ml Flt3 ligand (Pepro Tech), and 10 ng/ml SCF (Pepro Tech) as described [57].After eight days, the floating cells were used as FLDCs in experiments.RAW 264.7 cells were transfected with the NFATc1 (pcDNA3) expression plasmid.The cells expressing NFATc1 were selected in the presence of 0.4 mg/ml G418 and cloned.Splenocytes from wild-type mice were irradiated (30 Gy) and used as APC.

Parasite and experimental infection
The trypomastigote stage of T. cruzi Tulahuen strain was maintained in vivo in 2/2 mice by passages every other week or in vitro in LLC-MK 2 cells by passages every four days.For in vitro experiments, 5610 5 Mw or DCs were infected with 1.5610 6 trypomastigotes.For in vivo experiments, mice were i.p. injected with 6610 1 trypomastigotes or PBS.Epimastigotes of Tulahuen strain were grown at 26uC in liver infusion tryptose liquid medium, supplemented with 2.5% hemoglobin and 10% fetal calf serum.

Intracellular cytokine staining
Splenic cells were isolated from T. cruzi-infected mice at the indicated time point and stimulated with PMA and ionomycin for 4 h in the presence of 10 mg/ml brefeldin A. In experiments to detect IFN-c production from CD11c + cells, splenic cells were infected with T. cruzi (1:1) for 12 h, and further cultured for 6 h in the presence of 10 mg/ml brefeldin A. After staining of surface CD11c, CD4, CD8 or NK1.1, the cells were fixed with CytopermCytofix (BD Biosciences) for 20 min and incubated with PE-conjugated anti-IFN-c Ab.Flow cytometric analysis was performed on FACSCantoII (BD Biosciences).

Measurement of cytokine production
For in vivo experiments, mice were i.p injected with T. cruzi, and CD4 + T cells were isolated from the spleen at the indicated days after the infection.2.5610 5 CD4 + T cells were stimulated with anti-CD3 Ab or freeze-thawed T. cruzi in the presence of 2.5610 5  APC for 24 h.The culture supernatants were collected and diluted at 1:5.ELISA was performed with anti-mouse IFN-c Ab, avidin-HRP, and TMB solution purchased from eBioscience.Optical densities were determined at 450 nm wavelengths with reference at 570 nm.Levels of IFN-c were calculated from the standard curve by using purified mouse IFN-c purchased from eBioscience.

Flow cytometry
Bone marrow DCs or FLDCs were infected with T. cruzi for 6 h, washed and then cultured for 24 or 48 h.The T. cruzi-infected cells were stained with the combination of PE-conjugated anti-CD11c and the indicated antibodies at 4uC for 20 min, and washed.Flow cytometric analysis was performed on FACSCalibur or FACSCant II flow cytometer (BD Biosciences) and using FlowJo software (Tree Star).CD11c high cells and NK cells were sorted using FACS Aria (BD Biosciences).The instrumental compensation was set in each experiment using single color, 2color or 4-color stained samples.

Intracellular calcium measurements
This assay was performed as described [58].In brief, bone marrow Mw plated on glass-bottom dishes were incubated in serum-free RPMI 1640 supplemented with 2 mM Fluo-4 AM, the increase in fluorescent intensity of which indicates increased Ca 2+ level, at 37uC for 30 min.The cells were then washed to remove the free extracellular dye, and were maintained in culture medium during the whole experiment.The analysis of changes of basal intracellular calcium concentrations in response to T. cruzi infection was performed using an IX71 fluorescence microscope (Olympus).

Luciferase assay
RAW 264.7 cells were transfected with the indicated expression plasmids together with the reporter plasmid IFN-c-Luc and the internal control plasmid phRG-TK by Nucleofection (Nucleofector Kit V; Amaxa).After 18 h, the cells were infected with T. cruzi for 18 h, and the luciferase activities of whole cell lysates were measured using the Dual-luciferase reporter assay system (Promega) and Lumat LD 9507 (Berthold).

Statistical analysis
Differences between control and experimental groups were evaluated by the Student's t-test.then infected with T. cruzi for 30 min.T. cruzi-infected cells were stained with anti-NFATc1 antibody (red) and DAPI (blue).The right panels show the percentage of nuclear translocated NFATc1.Average of number of cells with nuclear NFATc1 (% in total cells counted) in twelve fields from three independent experiments (four fields in each experiment) in 6400 magnification is shown.Found at: doi:10.1371/journal.ppat.1000514.s012(0.10 MB PDF)

Supporting Information
Figure S13 Generation of CD11c + cells from Nfatc1 2/2 fetal liver cells.Fetal liver cells from 12.5 d.p.c.wild-type and Nfatc1 2/2 embryos were cultured with 20 ng/ml GM-CSF, 10 ng/ml Flt3 ligand, and 10 ng/ml SCF for 8 days.Expression of CD11c was analyzed and shown to be comparable between both genotypes.CD11c + cells were enriched by MACS (Miltenyi Biotec) and used for experiments as FLDCs.Found at: doi:10.1371/journal.ppat.1000514.s013(0.03 MB PDF) Figure S14 Response of Nfatc1 2/2 FLDCs to LPS and T. cruzi.(A) Fetal liver DCs from 12.5 d.p.c.wild-type and Nfatc1 2/2 embryos were stimulated with 100 ng/ml LPS for 24 h.LPSstimulated FLDCs were stained with the combination of PEconjugated anti-CD11c and the indicated antibodies at 4uC for 20 min, and washed.Flow cytometric analysis was performed on FACSCalibur.(B) Fetal liver DCs from 12.5 p.d.c.wild-type and Nfatc1 2/2 embryos were infected with T. cruzi for the indicated periods.Total RNA was extracted, and used for real-time RT-PCR analysis using primers specific for Il12b.All data were normalized to the corresponding gene Eef1a1 encoding elongation factor-1a (EF1a) expression, and the fold difference relative to the Eef1a1 is shown.Found at: doi:10.1371/journal.ppat.1000514.s014(0.02 MB PDF)

Figure 1 .
Figure1.Th1 response and DC maturation in T. cruzi-infected Myd88 2/2 Trif 2/2 mice.(A, B) Splenic CD4 + T cells were isolated from wildtype, Myd88 2/2 and Myd88 2/2 Trif 2/2 mice after 6 days (A) or the indicated days (B) of T. cruzi infection or PBS injection, and stimulated with anti-CD3 antibody (A) or freeze-thawed T. cruzi in the presence of antigen presenting cells (B).After 24 h, supernatants were assayed for IFN-c production by ELISA.Data are mean+s.d. of triplicate determination and a representative result of at least three independent experiments.In each experiment, two or three mice in each group were used.*: not detected.(C) Splenocytes were isolated from T. cruzi-infected or none-infected wild-type, Myd88 2/2 and Myd88 2/2 Trif 2/2 mice, and stimulated with 1 mg/ml ionomycin plus 50 ng/ml PMA.After surface staining with APC-conjugated anti-CD4 Ab, the cells were permeabilized and then stained with PE-conjugated anti-IFN-c Ab, and analyzed by flow cytometry.The percentage of IFN-c-producing CD4 + cells of individual mice is shown.(D)T.cruzi-infected or none infected bone marrow DCs of wild-type and Myd88 2/2 Trif 2/2 were stained for expression of CD40, CD86, and I-A b , and then analyzed by FACS.doi:10.1371/journal.ppat.1000514.g001

Figure 2 .
Figure2.Expression of IFN-c-inducible genes in T. cruzi-infected Myd88 2/2 Trif 2/2 DCs.(A) Bone marrow DCs from wild-type, Myd88 2/2 , Trif 2/2 , and Myd88 2/2 Trif 2/2 mice were infected with T. cruzi for 3 or 6 h, and total RNA was isolated and mRNA expression of Stat1, Irgm and Ifng was quantified by real-time RT-PCR and normalized to the level of elongation factor-1a (EF1a).*: not detected (B, C) Bone marrow DCs from wild-type mice were pretreated with 1 mg/ml cycloheximide (CHX) for 5 min, then infected with T. cruzi for the indicated periods.Next, mRNA expression of Stat1, Irgm (B) and Ifng (C) was analyzed by real-time RT-PCR.*: not detected.Data are representative of three independent experiments.(D, E) Bone marrow DCs from wild-type and Ifngr1 2/2 mice were infected with T. cruzi for 3 or 6 h.Next, mRNA expression of Stat1, Irgm (D) and Ifng (E) was analyzed.*: not detected (F) CD11c high B220 2 NK1.1 2 cells were isolated from the spleen of wild-type and Myd88 2/2 Trif 2/2 mice, and then T. cruziinduced expression of Ifng was analyzed.Data are mean+s.d. of triplicate determination and a representative result of at least three independent experiments.(G) Splenocytes from wild-type mice were infected with T. cruzi for 18 h.After surface staining with APC-conjugated anti-CD11c Ab, the cells were permeabilized and then stained with PE-conjugated anti-IFN-c Ab, and analyzed by flow cytometry.Representative results are shown from four independent experiments.The percentage of IFN-c-producing CD11c + cells of individual mice is shown.doi:10.1371/journal.ppat.1000514.g002

Figure 3 .
Figure 3. IFN-c dependent DC maturation and Th1 response in T. cruzi infection.(A) Bone marrow DCs of wild-type, Ifngr1 2/2 , and Myd88 2/2 Trif 2/2 Ifngr1 2/2 mice were infected with T. cruzi for 6 h, then washed and cultured for 48 h.The cells were analyzed for expression of CD40, CD86, and I-A b using flow cytometry.Representative results are shown from three independent experiments.(B) Wild-type, Ifngr1 2/2 , and Myd88 2/2 Trif 2/2 Ifngr1 2/2 mice were infected with T. cruzi.At six days after infection, CD4 + T cells were isolated from the spleen, and then stimulated with freeze-thawed T. cruzi in the presence of antigen presenting cells.After 24 h, supernatants were collected and assayed for IFN-c production by ELISA.The values are the means+s.d. of four independent experiments each carried out in triplicate.#:P,0.05.doi:10.1371/journal.ppat.1000514.g003

Figure 4 .
Figure 4. Ca 2+ signaling-dependent IFN-c production in T. cruzi-infected DCs.(A) Bone marrow DCs were stimulated with live T. cruzi or T. cruzi killed by repeated freeze-thaw steps for the indicated periods.Next, total RNA was isolated and analyzed for Ifng expression by real-time RT-PCR.The fold differences of each sample relative to EF1a are shown.*: not detected.(B) Bone marrow Mw from wild-type and Myd88 2/2 Trif 2/2 mice were treated with Fluo-4AM for 30 min, and washed.Then, cells which were infected or non-infected with T. cruzi were analyzed by fluorescence microscopy at the indicated periods.Representative of three independent experiments.(C) Bone marrow DCs from wild-type and Myd88 2/2 Trif 2/2 mice were pre-treated with BAPTA-AM (100 mM) for 30 min, and washed.Then, cells were infected with T. cruzi for 3 h.Expression of Ifng was analyzed by real-time RT PCR.Bone marrow DCs from wild-type mice were stimulated with LPS (100 ng/ml) for 3 h, and analyzed for expression of Il6 and Tnf.(D) Bone marrow Mw from wild-type and Myd88 2/2 Trif 2/2 mice were stimulated with 5 mM Ca 2+ ionophore plus 50 ng/ml PMA or 5 mM Ca 2+ ionophore for the indicated periods, and analyzed for expression of Ifng.Data are mean+s.d., and representative one of three independent experiments.*: not detected.doi:10.1371/journal.ppat.1000514.g004

Figure 5 .
Figure 5. T. cruzi-induced activation of NFATc1 in Myd88 2/2 Trif 2/2 DCs and Mw.(A) Wild-type and Myd88 2/2 Trif 2/2 bone marrow Mw were infected with T. cruzi or stimulated with LPS for 6 h in the presence or absence of FK506 (0.5 mM or 2.5 mM).Next, total RNA was isolated and analyzed for Ifng, Tnf or Il6 expression by real-time RT-PCR.The fold differences of each sample relative to EF1a are shown.*: not detected.(B) Bone marrow Mw from wild-type and Myd88 2/2 Trif 2/2 mice were transfected with the NFATc1 expression plasmid.Cells were then infected with T. cruzi for 30 min., and stained with anti-NFATc1 antibody (red) and DAPI (blue).(C) RAW 264.7 cell clones (designated #3, #7, and #9) transfected with the NFATc1 expression plasmid were analyzed for expression of NFATc1 by immunoblot with antibodies specific for NFATc1 and b-actin.(D) RAW 264.7 cells expressing NFATc1 were infected with T. cruzi for 3 h.Next, total RNA was extracted and used for real-time RT-PCR analysis using primers specific for Ifng, Stat1, and Irgm.(E) RAW 264.7 cells expressing NFATc1 (clone #9) were infected with T. cruzi for 3 h in the presence or absence of FK506, and total RNA was extracted.Real-time RT-PCR analysis was performed using primers specific for Ifng.*: not detected.Data are mean+s.d., and a representative result of at least three independent experiments.doi:10.1371/journal.ppat.1000514.g005

Figure 6 .
Figure 6.NFATc1 and T-bet had a synergistic effect on activation of the Ifng promoter.(A) Peritoneal Mw from wild-type, Myd88 2/2 , Trif 2/2 , and Myd88 2/2 Trif 2/2 mice were infected with T. cruzi for 3 h, and analyzed for Tbx21 expression by real-time RT-PCR.Data are mean+s.d., and a representative result of at least three independent experiments.(B) RAW 264.7 cells were transiently transfected with either T-bet or NFATc1 expression vector, or both expression vectors together with the IFN-c reporter plasmid.After 18 h of transfection, the cells were infected with T. cruzi for 18 h and activity of the reporter analyzed by luciferase assay.Data indicate mean+s.d., and a representative result of at least three independent experiments.doi:10.1371/journal.ppat.1000514.g006

Figure 7 .
Figure 7. Defective T. cruzi-induced IFN-c response in Nfatc1 2/2 DCs.(A, B) Wild-type and Nfatc1 2/2 FLDCs were infected with T. cruzi for the indicated periods.Next, total RNA was extracted, and analyzed for expression of Ifng, Tnf and Il6.*: not detected.(C) Wild-type and Nfatc1 2/2 FLDCs were infected with T. cruzi for 6 h, then washed and cultured for 24 h in the presence or absence of rexombinant IFN-c.Surface expression of CD40, CD86, and I-A b was analyzed by flow cytometry.Data are shown mean+s.d. of triplicate samples and a representative of four independent experiments.doi:10.1371/journal.ppat.1000514.g007