Leishmania braziliensis infections are often associated with exaggerated immune responses that can sometimes lead to severe disease associated with high levels of IFN-γ and TNF-α. To explore the role played by dendritic cells (DCs) in these responses, we characterized DCs that were exposed to L. braziliensis. We found that DCs cultured with L. braziliensis parasites up-regulated DC activation markers and produced IL-12 and TNF-α. However, not all DCs in the culture became infected, and an analysis of infected and uninfected DCs demonstrated that the up-regulation of activation markers and IL-12 production was primarily confined to the uninfected (bystander) DCs. Further studies with Transwell chambers and parasite fractions indicated that the activation of bystander DCs was mediated by a soluble parasite product, in a type 1 IFN- and MyD88-independent, but TNF-α-dependent fashion, and that the activated DCs were more efficient at presenting Ag than control DCs. In contrast, L. braziliensis-infected DCs failed to up-regulate activation markers, but exhibited a dramatic enhancement in their ability to produce TNF-α in response to LPS as compared with uninfected DCs. These findings uncover a dual role for DCs in L. braziliensis infection: T cell activation by bystander DCs due to enhanced Ag-presenting capacity following exposure to soluble parasite products, and increased production of TNF-α by infected cells that may contribute to the local control of the parasites, but concomitantly induce immunopathology.
Leishmaniasis is caused by several different species of protozoan parasites, and the clinical manifestations of the infection vary widely depending upon the Leishmania species involved and the immune response of the host. In all cases, control of the parasites is associated with the expansion of CD4+ Th1 cells that produce IFN-γ, which promotes destruction of the parasites within infected cells (1, 2, 3, 4). In the case of Leishmania braziliensis infections, however, an exaggerated Th1 response can lead to a severe immunopathologic form of leishmaniasis, known as mucosal (ML)3 disease, characterized predominantly by severe tissue damage in the nasopharyngeal region and disfiguring facial lesions. The pathogenesis of this infection remains poorly understood, although TNF-α is thought to play a contributing role in the tissue damage seen in these patients (5, 6, 7, 8, 9, 10). Thus, following stimulation, PBMC from ML patients secrete higher levels of TNF-α than those from individuals with simple cutaneous leishmaniasis (6, 7). In addition, combining drug therapy with pentoxifylline, a drug that down-modulates TNF-α production, enhances resolution of ML lesions (8, 9).
Dendritic cells (DCs) play a pivotal role in promoting resistance to leishmaniasis, both by activating CD4+ T cells and promoting their differentiation into Th1 cells by producing IL-12 (11, 12). Thus, a reasonable place to start investigating the pathogenesis of L. braziliensis infections would be to study the interactions of these parasites with DCs. The response of DCs to infection with several other Leishmania parasites has been examined, although the results of these studies have often yielded conflicting results. For example, Leishmania major amastigotes, but not promastigotes, were reported to activate murine DCs, whereas activation of human DCs required an additional stimulation through CD40 (11, 13, 14, 15, 16). In the case of New World Leishmania species, some studies have indicated that the parasites activate DCs, whereas others fail to see DC activation by the parasites (17, 18, 19, 20, 21, 22). Differences in the parasite stage, parasite species or strain, and source of DCs may explain some of these divergent results. Moreover, most previous studies did not differentiate between infected DCs within the cultures and those that were exposed to the parasites, but not infected.
In the present study, we investigated whether interactions between DCs and L. braziliensis might contribute to the pathogenesis of ML. We found that, whereas DCs infected with L. braziliensis were not activated, uninfected DCs in the same culture (bystander DCs) were stimulated by a secreted product from L. braziliensis. This stimulation led to the up-regulation of costimulatory molecules, production of IL-12 and TNF-α, and the enhanced ability of DCs to activate T cells. We found that bystander DC activation was TNF-α dependent, but MyD88 and type 1 IFN independent. In contrast, L. braziliensis-infected DCs failed to up-regulate DC activation markers, although they produced TNF-α. Further studies found that L. major and Leishmania mexicana also activated bystander DCs, but in contrast to L. braziliensis these bystander DCs produced much lower levels of TNF-α. Taken together, these results demonstrate that Leishmania parasites can activate bystander DCs, and indicate that one factor contributing to the development of ML following L. braziliensis infection may be the ability of these parasites to induce bystander DCs to produce high levels of TNF-α.
Materials and Methods
Female wild-type C57BL/6 and C3H mice, and C3H type 1 IFN receptor-deficient and OTII transgenic mice were purchased from The Jackson Laboratory. MyD88-deficient mice were provided by L. Turka (University of Pennsylvania, Philadelphia, PA). Animals were maintained in a specific pathogen-free environment and tested negative for pathogens in routine screening. All experiments were conducted following the guidelines of the University of Pennsylvania institutional animal care and use committee.
Parasites and DC culture
L. braziliensis parasites were isolated from a ML leishmaniasis patient from the endemic area of Corte de Pedra, Brazil, and typed by Random amplification of polymorphic DNA PCR (23). L. braziliensis, L mexicana (MNYC/BZ/62/M379), and L. major parasites (MHOM/IL/80/Friedlin) were grown until stationary phase in Grace’s insect culture medium (Life Technologies) supplemented with 20% heat-inactivated FBS (HyClone) and 2 mM l-glutamine. Soluble Leishmania Ag (SLA) was prepared, as previously described (24), tested for endotoxin using the Limulus amebocyte lysate test, and used at a concentration of 10 μg/ml. Endotoxin levels were below the level of detection. Anti-TNF-α (R&D Systems) mAb was used in some cultures in a concentration of 10 μg/ml. LPS was used at a concentration of 10 ng/ml. Bone marrow-derived DC were cultured, as previously described (25). Briefly, bone marrow DC precursors were differentiated for 8–10 days in the presence of 20 ng/ml GM-CSF in RPMI 1640 containing 10% FBS, 100 U/ml penicillin/streptavidin, 0.05 mM 2-ME, and 2 mM l-glutamine. On days 8–10 of culture, DCs were harvested and infected with stationary-phase promastigotes at different parasite:DC ratios. After 2 h, cells were washed twice to eliminate extracellular parasites, and incubated for 18 h at 37°C. For the Transwell experiments, DCs (1 million) were placed on one side of the chamber, and Leishmania (1, 2, or 5 million) were placed on the other side. After an incubation period of 18 h at 37°C, DCs were harvested and analyzed by flow cytometry, as described below. For these experiments, we used 0.4-μm pore-size filters (Corning Glass) to impair the ability of parasites to infect DCs, but allow the passage of soluble factors through the pores.
CFSE labeling of parasites and flow cytometric analysis of DCs
Parasite labeling with CFSE (Invitrogen) was performed, as previously described (26). Briefly, parasites were washed twice in PBS and resuspended at 5 × 107/ml in 1 ml of PBS, with 5 mM CFSE, and incubated at 37°C for 10 min in the dark. Parasites were then washed in 10 ml of PBS and 10% FBS, and resuspended in RPMI 1640. For flow cytometry, DCs were harvested, stained with fluorochrome-conjugated Abs for surface markers (CD11c, CD80, CD86, and class II (e-Bioscience)), and fixed by using 2% formaldehyde. For intracellular staining, the fixed cells were permeabilized with a solution of saponin and stained for 30 min at 4°C using fluorochrome-conjugated Abs against IL-12 p40, TNF-α, or isotype control Abs.
Samples were acquired on a FACSCalibur flow cytometer (BD Pharmingen), and analysis was performed using FlowJo software (Tree Star). Analysis gates were based on live cells, and in some cases live CD11c+ cells, as indicated in the figure legends.
DC and T cell coculture and proliferation assays
Naive OTII CD4+ T cells were isolated by negative depletion using MACS columns, labeled with CFSE, as previously described (27), and then cultured with DCs that had been pulsed overnight with OVA (100 μg/ml) or OVA plus SLA (10 μg/ml), at a ratio of 10 CD4+ cells:DC. Cells were cultured at 37°C for 4 days and then collected and stained for CD4. CFSE dilution was assessed by flow cytometry, as described above. For the Transwell experiments, immune cells were obtained from the spleens of infected C57BL/6 mice (1 million parasites/footpad) for 20 wk. Immune CD4+ T cells were isolated by negative depletion and cultured with DCs that had been exposed overnight to parasite-secreted products in 0.4-μm pore-size Transwells (Corning, NY) at a ratio of 10 CD4+ cells:DC. After 4 days, supernatants were harvested, and IL-2 and IFN-γ concentrations were measured by sandwich ELISAs, as previously described (28).
Statistical analysis was performed using two-tailed Student’s t test. Differences were considered significant at p < 0.05.
L. braziliensis parasites infect and activate DCs
We first examined whether L. braziliensis parasites could infect and survive within murine DCs. Bone marrow-derived DCs were cultured with L. braziliensis promastigotes, and after 2 h we observed (by optical microscopy) that ∼70% of the DCs were infected. The percentage of infected DCs remained the same over 96 h, whereas the number of parasites per DC increased (Fig. 1,A). These results indicate that L. braziliensis parasites are able to infect and multiply within murine DCs. We next examined the activation status of the DCs that were cultured with L. braziliensis, by analyzing the expression of DC activation markers (class II, CD80, and CD86) and the production of proinflammatory cytokines (IL-12 and TNF-α). We observed a slight up-regulation of class II and CD86 on DCs taken from cultures infected with L. braziliensis (Fig. 1,B). Moreover, there was increase in the frequency of IL-12 and in TNF-α-producing DCs in cultures infected with L. braziliensis (Fig. 1 C).
Bystander DCs, but not infected DCs, up-regulate class II and costimulatory molecules
Because not all of the DCs were infected with L. braziliensis, we next wanted to focus specifically on those DCs that were infected. Therefore, we infected DCs with parasites that had been labeled with CFSE, which allowed us to analyze DC activation in both infected (CFSE bright) and bystander (CFSE dim) DC populations within the same culture (Fig. 2, A and B). Surprisingly, slightly lower expression of class II and CD86 was observed in infected DCs when compared with DCs from control cultures that were not exposed to parasites (Fig. 2, A and B). In contrast, uninfected bystander DCs expressed higher levels of class II, CD80, and CD86. Furthermore, when the ability of infected and bystander DCs to produce IL-12 was tested, it was the bystander DC population that contained cells positive for IL-12, whereas TNF-α-producing cells were observed in both the infected and bystander populations (Fig. 2,C). Part of the difference seen in activation markers between infected cells and control DCs may be due to preferential infection by the parasites of slightly more immature DCs, which would be consistent with decreased phagocytosis by activated DCs (29, 30). To see whether the activation of bystander cells was related to infected cell frequency within a culture, a dose-response curve was performed in which DCs were exposed to different numbers of parasites. The ability of bystander DCs to up-regulate class II, CD86 and produce TNF-α was found to be related to the frequency of infected cells within the cultures (Fig. 2, D and E).
L. braziliensis secretes a molecule that activates DCs and enhances their Ag-presenting capability
The activation of the bystander DCs that we observed could be due to the secretion of a factor from the infected DCs, or due to a parasite molecule present in the culture. To differentiate between these possibilities, we used Transwell chambers, in which uninfected DCs could be separated from parasites by a membrane that allowed passage of soluble molecules, but not the parasites. When DCs were cultured in a Transwell chamber without parasites, we saw no DC activation. However, the presence of L. braziliensis parasites in the chamber separated from the DCs by a membrane led to increased expression of activation markers and the production of IL-12 and TNF-α (Fig. 3, A and B). To determine whether the bystander DCs were activated by other species of Leishmania, we performed Transwell assays with L. major and L. mexicana. We found that similar to L. braziliensis, DCs exposed to L. major and L. mexicana exhibited an increase in activation markers (Fig. 3,C). However, when we examined TNF-α production, we found that neither L. major nor L. mexicana stimulated bystander DCs to produce detectable levels of TNF-α (Fig. 3,D). The ability of secreted products from the parasites to activate DCs was decreased as the number of parasites added to the Transwell was decreased (Fig. 3,E). These results suggest that a soluble parasite factor can activate DCs in the absence of any infection. To confirm this result, we made a soluble fraction of L. braziliensis (SLA) and tested its ability to activate DCs. As seen in Fig. 3 F, L. braziliensis SLA induced up-regulation of class II and CD86.
We next asked whether the bystander DCs had taken up leishmanial Ags and could stimulate T cells. To test this, we assayed the ability of DCs exposed to L. braziliensis through the Transwell membrane to activate T cells from Leishmania-immune mice. DCs were cultured in a Transwell chamber with or without L. braziliensis parasites in the other compartment. After 18 h, CD4 T cells from immune (infected for 20 wk) or naive mice were added and 4 days later the production of IL-2 and IFN-γ was assessed. DCs not exposed to parasites failed to stimulate the immune cells to produce IL-2 or IFN-γ (Fig. 4,A), and naive T cells failed to respond (data not shown). In contrast, there was a significant increase in the production of IL-2 and IFN-γ by immune CD4 T cells added to DCs that were cultured in a Transwell with L. braziliensis (Fig. 4 A).
To determine whether the bystander DCs activated by L. braziliensis exhibited an increased capacity to present Ag compared with DCs that were not exposed to the parasites, we tested the ability of these two populations to present Ag to T cells. CFSE-labeled TCR transgenic OTII T cells, which recognize OVA, were cultured with OVA-pulsed DCs or OVA-pulsed DCs that had been exposed to a soluble fraction of L. braziliensis (SLA). After 4 days in culture, the proliferation of the OTII cells was assessed by flow cytometry. As expected, we found that DCs exposed to OVA stimulated a significant increase in the proliferation of OTII cells. However, DCs that were exposed to OVA and SLA stimulated significantly better T cell proliferation (Fig. 4 B).
Infected DCs make high levels of TNF-α
Leishmania infection of macrophages has been associated with a loss of the ability of the infected cells to respond to other activating agents, such as TLR ligands. Therefore, we wanted to know whether L. braziliensis-infected DCs were able to respond to LPS. To test this, we infected DCs with CFSE-labeled parasites for 2 h, and then pulsed some of the cultures with LPS for 6 h. We found that infected DCs were not able to up-regulate class II, CD80, or CD86 in response to LPS (Fig. 5,A). Bystander DCs, which already expressed higher levels of class II, CD80, and CD86, did not further up-regulate these molecules in response to LPS. There was an LPS-induced increase in IL-12 production by infected DCs, although this increase was substantially less than that observed by uninfected DCs (Fig. 5,B). In contrast, there was a dramatic increase in the ability of infected DCs to produce TNF-α in response to LPS. Although 30% of the bystander DCs produced TNF-α, over 90% of the infected DCs did so after LPS stimulation (Fig. 5 B). These results demonstrate that whereas infected DCs may fail to increase the expression of activation markers, they not only can make TNF-α, but make more TNF-α than bystander DCs following exposure to a TLR ligand.
To study the mechanism by which Leishmania-secreted products activate bystander DCs, we investigated the role of MyD88 and type 1 IFN by using MyD88−/− and type 1 IFN receptor-deficient DCs in Transwell experiments, as described above. We found that activation proceeded independently of MyD88 and type 1 IFN (Fig. 6, A and B).
Because the data above show that both bystander and infected DCs produce TNF-α, and TNF-α is known to induce up-regulation of class II, CD80, and CD86 (31, 32), we next evaluated whether TNF-α plays a role in bystander DC activation. By blocking TNF-α activity with anti-TNF-α mAb, we found that the up-regulation of class II, CD80, and CD86 on bystander DCs was prevented (Fig. 6 C). Thus, whereas neither MyD88 nor type 1 IFN is required for bystander DC activation, both infected and bystander DCs produce TNF-α that is essential for activation of the bystander DCs.
The activation state of DCs can be distinguished based on the expression of class II, costimulatory molecules, as well as cytokine production (33). Before infection, DCs survey the tissues and the lymphoid organs as immature cells. However, when exposed to pathogens, DCs up-regulate expression of surface markers associated with Ag presentation, such as class II, CD80, and CD86, and produce proinflammatory cytokines, such as IL-12 and TNF. In leishmaniasis, DCs not only play a key role in the development of a protective immune response to Leishmania, but also act as a host cell for the parasites. This dual role means that DCs will need to be activated to initiate the immune response, but at the same time the parasites will presumably want to inhibit activation to promote parasite survival. Many studies have examined the response of DCs to Leishmania parasites, but the use of different parasite species and life-cycle stages, as well as different sources of DCs, has led to a wide range of results (11, 17, 18, 19, 20, 21, 34, 35, 36, 37). In addition, most of these studies examined DC activation in cultures that contained both infected and uninfected DCs. In this study, we monitored the response of infected vs bystander DCs, and found that whereas infected DCs appear immature, bystander DCs up-regulate class II, CD80, and CD86 expression and secrete IL-12 and TNF-α. Our data suggest that infected DCs are relatively immature, and in other contexts these semimature DCs have been reported to be more likely to induce tolerance than promote an immune response (33). Thus, it is likely that the bystander DCs may be more important for the initial T cell response, which is consistent with our finding that the bystander DCs were capable of stimulating T cell responses. We also found that bystander DC activation was not unique to L. braziliensis, and was observed with L. major and L. mexicana, although bystander DCs exposed to L. braziliensis made significantly more TNF-α. Consistent with these results, noninfected DCs appear to mature in vivo following infection and present leishmanial Ags (21, 38, 39). Taken together, these data indicate infected and bystander DCs play distinct roles in leishmaniasis.
Leishmania is known to inhibit several macrophage functions, such as activation, cytokine release, and Ag presentation (17, 40, 41, 42, 43). For instance, down-regulation of class II expression and the inability to produce IL-12 have been observed in several studies (16, 43, 44, 45). Also, TLR-induced up-regulation of costimulatory molecules, as well as TNF-α and IL-12 production, was impaired in L. major-, Leishmania chagasi-, Leishmania donovani-, and L. mexicana-infected macrophages, and in the case of L. mexicana, disruption of NF-κB signaling was observed (44, 46, 47, 48). Less well understood are the interactions between Leishmania and DCs. Although some studies documented impairment in DC function, others found increased DC activation, IL-12 production, and NF-κB signaling after Leishmania infection (11, 15, 18, 19, 21, 36, 37). We found not only that infected DCs remain immature after Leishmania infection, but that they are also impaired in the up-regulation of class II, costimulatory molecules, and IL-12 after stimulation by LPS. Our results show that infected and bystander DCs respond differently to infection, indicating that understanding the response of DCs to Leishmania requires analyzing both infected and noninfected cells within a culture. Some of the conflicting results addressing the question of whether DCs are activated by Leishmania may be due to the fact that the impact of Leishmania on DC activation is often assessed on populations that contain both infected and uninfected cells. In this study, we show that infected DCs are less responsive to the TLR ligand LPS, as measured by their ability to up-regulate expression of class II, CD80, CD86, and IL-12, but remain responsive, as assessed by their ability to produce high levels of TNF-α upon LPS stimulation. In another study, it was shown that concomitant activation with LPS and IFN-γ may overcome this inhibition, although even under these conditions, fewer infected cells made IL-12 when compared with DCs from uninfected cultures (22).
Of particular interest is the increased percentage of TNF-α-producing cells within the L. braziliensis-infected DC population. TNF-α is a key cytokine mediating T cell-mediated inflammation. It is involved in leukocyte recruitment by increasing expression of adhesion molecules on vascular endothelium and increasing angiogenesis (49, 50). Although TNF-α promotes increased macrophage activation, and contributes to control of Leishmania parasites, deleterious consequences of excessive TNF-α production have been reported in many conditions, such as rheumatoid arthritis or psoriasis (51, 52, 53, 54, 55). In humans, TNF-α plays an important role in the pathogenesis of ML and cutaneous leishmaniasis due to L. braziliensis infection (6, 7). The high levels of TNF-α and IFN-γ secreted by mononuclear cells from these patients is positively correlated with lesion size (56), and the use of drugs that down-modulate production of TNF-α in combination with antimony increases the rate of healing and allows the cure of refractory cases of ML and cutaneous disease (8, 9). Although monocytes, CD4+, and CD8+ T cells are known to contribute to TNF-α production in patients with ML leishmaniasis (6), the role of DCs in this process has not been explored. Our data show that both bystander and infected DCs are another source of TNF-α following infection with L. braziliensis. Future studies with DCs from both patients and uninfected individuals will be required to extend these studies to human leishmaniasis.
DCs can be activated by several pathways, including MyD88-dependent TLR signaling and via type 1 IFN- and TNF-α-stimulated pathways (31, 32, 33, 57). The adaptor molecule MyD88 is required for resistance to Leishmania, and type 1 IFN has been shown to up-regulate expression of NO synthase in vivo and ameliorate cutaneous leishmaniasis lesions (21, 58, 59). Therefore, we tested whether MyD88 and type 1 IFN would mediate DC activation by Leishmania-secreted products. We found that whereas MyD88 and type 1 IFN are not involved in bystander DC activation, products secreted from the parasites induce TNF-α production, and in an autocrine manner TNF-α induces up-regulation of class II and costimulatory molecules on bystander DCs. DCs activated by TNF-α have been reported to gain a semimature phenotype, characterized by up-regulation of class II and costimulatory molecules, but a failure to make proinflammatory cytokines (33). These DCs have been thought to be tolerogenic rather than inflammatory. However, in our studies, bystander DCs not only up-regulate class II and costimulatory molecules, but also produce TNF-α and IL-12. Moreover, they are able to process and present Leishmania Ags to CD4 T cells. We hypothesize that the differences between the semimature tolerogenic DCs induced by TNF-α alone and the DCs activated by Leishmania-secreted products might be due to additional signals induced by the parasite. Interestingly, whereas the bystander DCs respond to TNF-α, Leishmania-infected DCs were inhibited in their ability to respond. Why this is the case is unknown, although the inability of Leishmania-infected cells to respond to other signals (such as LPS) has been described before. For example, NF-κB signaling is compromised in Leishmania-infected macrophages, and because TNF-α-mediated DC activation is dependent on NF-κB, this is one possible mechanism that could be involved (48).
Finally, the parasite molecule(s) that activates bystander DCs is unknown. The most well-described Leishmania surface molecule is a lipophosphoglycan (LPG), which covers the surface of the parasite and can be released (60, 61). Many functions have been ascribed to LPG, both within the sandfly and the mammalian host (62, 63). Specifically, LPG can activate DCs by binding to TLR2, although LPG has also been reported to inhibit activation (64, 65). However, because the products secreted by Leishmania activate DCs in a MyD88-independent fashion, it may be unlikely that LPG would be involved in the bystander DC activation process. Similarly, another TLR ligand known to contribute to the activation of DCs is Leishmania DNA (58, 66). However, Leishmania DNA activates DCs by binding to TLR9, and because this process is also dependent on MyD88, we can rule out the possibility that DNA is involved in the bystander DC activation that we are observing. Thus, further studies will be required to define the parasite products that can stimulate bystander DCs.
Our findings show a dual role for DCs following infection with L. braziliensis. Although infected DCs remain immature and secrete TNF-α, bystander DCs respond to parasite factors by producing TNF-α, which in turn up-regulates class II and costimulatory molecules. These bystander DCs also produce IL-12, and thus are able to present Ag to CD4+ T cells and promote the generation of CD4+ Th1 cells. In contrast, the inability of infected DCs to up-regulate costimulatory molecules or produce IL-12 suggests that these cells are not efficient at Ag presentation. However, these DCs produced high levels of TNF-α, and at the same time provide a site for the parasites to survive, suggesting that they may promote increased immunopathology. Future studies are ongoing to characterize the mechanisms by which L. braziliensis modulates DC function, as well as identification of Leishmania products involved in DC activation. A better understanding of how Leishmania parasites inhibit DC activation will be useful in the design of new immunotherapies for leishmaniasis.
We thank Euihye Jung and Carlos Rodriguez for technical assistance, and Christopher Hunter, David Artis, and members of the Pathobiology Department for comments and criticisms.
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 National Institutes of Health Tropical Infectious Diseases Training Grant D43 TW007127 and Grants AI35914 and AI053825.
Abbreviations used in this paper: ML, mucosal; DC, dendritic cell; LPG, lipophosphoglycan; SLA, soluble Leishmania Ag.