Introduction

Leishmaniasis is a vector-borne parasitic disease caused by protozoan of the genus Leishmania. Leishmania parasites exist as extracellular flagellated promastigotes within sandfly vectors and as intracellular amastigotes in infected mammals. Disease manifestation depends on the infecting species and ranges from subclinical infections, to self-healing cutaneous lesions, to life-threatening infections of visceral organs. Visceral leishmaniasis (VL), which is caused by L. donovani and/or L. chagasi/infantum, is characterized by chronic parasitization of the spleen, liver and bone marrow.

Most of the studies on the immunology of VL have been carried out in murine models.

Like in humans, L. donovani infections in mice also result in the parasitization of the spleen, liver and bone marrow. However, the parasite only establishes chronic infection in the spleen and bone marrow [1]. Indeed, infection in the liver is self-resolving within 6–8 weeks and relies on the development of a Th1-dominated granulomatous response [2], characterized by high IFN-γ production. The Th1 response is induced by IL-12 secreted by dendritic cells (DCs) [3] and is crucial for parasite control and resolution of the disease [3, 4].

Although disease resolution occurs in the liver, the spleen and the bone marrow remain chronically infected with Leishmania [5]. As yet we know very little about the immune response in the spleen and the requirements for controlling parasite growth in this organ. IL-12, Th1 responses and CD8+ T cells are definitely involved in partially limiting parasite growth during chronic infection [4, 6]. In contrast, TNF-α, which is a key cytokine essential for parasite clearance in the liver [7], is responsible for the disruption of the microarchitecture of the splenic marginal zone that occurs during chronic infection [5, 8]. This disruption severely hampers the migration of DCs [8] and naïve T-cells to the T-cell area [9], contributing to the immune suppression observed at later stages of infection. Furthermore, the spleen mounts a regulatory response dominated by CD4+ T-cells co-producing IL-10 and IFN-γ [10], cells that have also been observed in human VL patients [11]. To date, the strategies adopted by Leishmania to evade protective immune responses in the spleen and the reasons why IFN-γ-producing Th1 responses are unable to control parasite multiplication in the spleen remain unknown. Of numerous mechanisms of immune evasion that have been proposed in the past [1215], the induction of immune suppressive IL-10 responses is the most prominent one.

IL-10 in experimental VL

IL-10 has been implicated as a key regulator during infection with a variety of parasitic, bacterial, viral and fungal pathogens (reviewed in [1618]). Although IL-10 responses are typically generated to balance excessive Th1 and CD8+ T-cell responses and prevent immunopathology, overproduction of IL-10 has been shown to inhibit proinflammatory responses leading to susceptibility to infectious pathogens such as Malaria [19, 20], Leishmania [21, 22], lymphocytic choriomeningitis virus (LCMV) [23, 24] and Mycobacteria spp. [25].

Various studies have now shown that one of the major contributing factors to disease progression in leishmaniasis is indeed IL-10. IL-10 receptor blockade in mice infected with L. donovani nearly abrogates infection [26]; and Il10 −/− mice are highly resistant to VL [27]. Several cell populations have been shown to express IL-10 during Leishmania infections including, natural regulatory T cells (Tregs) [12], Th1 cells [10, 13], NK cells [28], macrophages [29], B cells [28] and DCs [30]. Nevertheless, the cellular sources of IL-10 responsible for mediating disease progression and parasite persistence remain unknown.

Chronic infections in humans and in mice have been recently associated with the generation of CD4+ T-cells co-expressing IL-10 and IFN-γ [10, 11]. These cells also appear to be involved in parasite persistence in a model of cutaneous leishmaniasis [13] and to display strong host-protective functions in Toxoplasma infections [31]. Leishmania-specific IFN-γ+ IL-10+ Th1 responses are first detected 3 weeks after infection in L. donovani-infected mice [10]. Because of the lack of a specific marker characterizing these cells, formal evidence that they are in fact the major players in the induction of disease progression is still lacking.

Regulatory T cells (Tregs) have also been shown to be involved in parasite persistence during chronic Leishmania infections [12]. However, IL-10 production by these cells or an increase in Tregs frequency during chronic disease was never detected in L. donovani infections [10]. We do not exclude, though, that these cells could play an IL-10-independent role in the establishment of persistent infection in VL.

Leishmania-specific antibodies have also been shown to induce IL-10 expression by macrophages through ligation of FcγRs with IgG-opsonized parasites [29]. Since humoral responses are first detected at day 21 pi [32], which coincides with the loss of control on parasite growth in the spleen, it is possible that antibody-induced IL-10 expression by macrophages also contributes to the establishment of a persistent infection in the spleen of L. donovani-infected mice. Indeed, antibody-induced macrophage-derived IL-10 has been shown to play a pivotal role in disease progression in a low dose model of L. major [33]. Moreover, FcγrIII −/− mice are also resistant to infection with L. mexicana [34]. This literature is in agreement with studies in L. donovani-infected B-cell deficient mice (μMT mice). Interestingly, μMT mice are significantly less susceptible to L. donovani and are able to cure infection within 35 days [35]. However, this increased resistance may not only rely on the absence of antibody production and therefore lack of induction of IL-10 by macrophages. B cells have also been reported to express IL-10 during VL [28] and could theoretically contribute to parasite persistence. Although IL-10+ Th1 responses are thought to be the major source of IL-10 during chronic VL, the contribution of B-cell and macrophage-derived IL-10 to parasite persistence requires further clarification.

Since Il10 −/− mice show a certain degree of resistance to L. donovani during acute infection [27], it is possible that the downregulation of leishmanicidal responses at different stages of infection may be caused by different sources of IL-10. Several cells were recently shown to express IL-10 mRNA during acute infection; these include NK cells, CD4 T cells, CD8 T cells and B cells. Nevertheless, the relative amount of IL-10 produced by each cell population and its contribution toward establishing Leishmania susceptibility are yet to be determined.

IL-10 in human VL

In humans, plasma levels of IL-10 are strongly linked to susceptibility to a variety of infectious diseases, such as malaria [36], leprosy [37] and HIV [38]. Persistent L. donovani infections are also associated with increased levels of IL-10 in the serum of VL patients [39], however, the cellular sources of IL-10 have been elusive. Recently, a study demonstrated that individuals with VL display an accumulation of IL-10 mRNA in CD25Foxp3 T cells within the spleen [11]. Besides the technical limitations involved in human studies, inter-individual variations in IL-10 production due to single nucleotide polymorphisms (SNPs) in the IL-10 promoter [4042] also restrict our understanding of the mechanism involved. This inter-individual difference in IL-10 production may be one of the possible factors responsible for the variation in susceptibility to infection with visceralizing Leishmania species observed in humans [43]. Indeed, the IL10-819C/C genotype was lately found to be associated with higher levels of IL-10 [44] and with increased susceptibility to L. braziliensis infection. To overcome these obstacles, we recently employed a novel experimental model to study human IL-10 expression in VL using transgenic Il10 −/−/hIL10BAC mice [45]. In these mice, human IL-10 is under genomic regulatory control, although we cannot rule out the possibility that there are control elements missing in the human IL10 BAC. This topic is discussed in more detail by Hedrich et al. in this issue. Sequence analysis of the human IL-10 BAC revealed a SNP haplotype block (IL-10-819T), which has previously been associated with low IL-10 production [46]. Il10 −/−/hIL10BAC mice rapidly cleared L. donovani infection and displayed the same resistance phenotype as Il10 −/− mice. Interestingly, Il10 −/−/hIL10BAC mice were able to recapitulate IL-10 production in myeloid cells and were therefore resistant to endotoxin-induced shock; yet they failed to generate antigen-specific, IL-10-secreting Th1 responses following L. donovani infection. This observation suggests that IL-10 expression may be regulated in a cell-specific manner and that T-cell-derived IL-10 seems to be the major factor involved in the establishment of persistent infections. Nevertheless, we still need to identify all cellular sources of human IL-10 in our model in order to determine the cell-type-specific contribution to disease pathogenesis during L. donovani infection.

L. donovani and CD8+ T-cell responses

Another mechanism of evasion of host-protective immunity is the capacity of L. donovani to escape CD8+ T-cell responses [14]. Unlike the experimental model of cutaneous leishmaniasis, in which the role of CD8+ T cells has been controversial in the past [4750], CD8+ T cells have clearly been shown to be essential for the control of primary infections [2, 6, 5153] and to be the main mediators of resistance to rechallenge in the mouse model of VL [2]. These cells have also been associated with cure in VL patients [54]. Several vaccination studies involving various Leishmania species have further highlighted the critical role played by CD8+ T cells in the host’s-protective immunity [32, 5559]; however, the mechanisms involved remain elusive.

We have recently characterized the antigen-specific CD8+ T-cell responses in L. donovani-infected mice using transgenic parasites expressing the model antigen ovalbumin [14]. In this study, we showed that L. donovani evades CD8+ T-cell responses by limiting clonal expansion (200- to 10,000-fold less expansion than in other infection models [6064]) and by inducing functional exhaustion during chronic disease. Furthermore, effector CD8+ T cells appear to undergo only 5–7 rounds of division, and only 20% of the cells that survive clonal contraction display an effector phenotype (80% of the cells express markers that are typical for central memory CD8+ T cells). Since strong clonal expansion is needed to control infection and determines the magnitude of the memory response [61], it would be of interest to identify the mechanisms by which L. donovani parasites interfere with the expansion of parasite-specific CD8+ T cells. Reasons for limited expansion may be multifactorial, either resulting from a reduced capacity of dendritic cells to process and present antigen and/or due to a suppressive environment.

One of the main questions that arise is whether infected DCs can process and present Leishmania antigens, provide adequate co-stimulation and secrete the cytokines that are necessary for the priming of naïve CD8+ T cells. Since Leishmania preferentially resides in macrophages, most of the studies in the past have mainly investigated the interaction between the parasite and macrophages. In these cells, Leishmania appears to interfere with the cells’ antigen-presenting capacity by disrupting lipid rafts [65], inhibiting phagosome maturation [66, 67] and cleaving epitopes by secreting endopeptidases [68]. Recent studies, however, have indicated that Leishmania may also interfere with the antigen cross-presenting machinery of dendritic cells [69] and actively downregulate various signaling events in DCs (reviewed in [70]). These observations suggest that infected DCs may not be able to optimally deliver the 3 signals necessary for effective priming of CD8+ T cells and that limited expansion of antigen-specific CD8+ T cells could result from antigen paucity and from poor activation of antigen-presenting DCs. A recent study, though, has shown that DCs infected in vitro with L. braziliensis, despite producing high amounts of TNF-α, fail to upregulate activation markers upon infection [71]. Yet, bystander, uninfected DCs upregulate activation markers, produce IL-12 and are more efficient in presenting antigen than control DCs. The authors suggest that this bystander DC activation together with increased TNF-α production by infected DCs could be responsible for the exaggerated T-cell responses observed in some patients infected with L. braziliensis. It is possible that a similar mechanism also operates during VL and bystander DCs overstimulate effector CD8+ T cells rendering them more susceptible to apoptosis. This would explain the low percentage of effector CD8+ T cells that survive clonal contraction and also the high percentage of multifunctional cells observed already during expansion, cells that are typically seen at later stages of infection (reviewed in [72]). Further investigations are needed to dissect the strategies utilized by the parasite to subvert DC activation and antigen processing.

The splenic environment may also contribute in controlling antigen-specific CD8+ T-cell expansion. The cytokine environment in the spleen has been extensively studied in the mouse model of VL. L. donovani induces a strong proinflammatory response at the onset of infection that is characterized by the production of key proinflammatory cytokines such as TNF-α, IFN-γ, and IL-12 [3, 4]. IL-12 can be detected as early as 5 h after infection and is mainly produced by splenic CD8+ DCs [73]. IL-12 production, though, has been shown to be transient, with the levels of IL-12 being similar to those observed in naïve mice by day 3 post-infection, probably because of DCs exhaustion [3]. Cytokines such as IL-12 and type I IFN have been shown to be crucial for stimulating strong clonal expansions of antigen-specific CD8+ T cells and for the development of their effector functions [74, 75]. A reduced exposure to inflammatory cytokines results in either impaired effector functions or limited clonal expansion, depending on the timing of the deprivation of cytokines [76, 77]. Thus, lack of IL-12 production by splenic DCs could also contribute to limited expansion of Leishmania-specific CD8+ T cells, since maximal antigen-presenting capacity by splenic DC is observed at day 6 after infection. The effect of IL-12 deprivation during CD8+ T cell priming could be acting synergistically with the presence of anti-inflammatory cytokines. In fact, IL-10 mRNA is readily detectable at 24 h after infection with L. donovani [28]. In experimental VL, IL-10 appears to be expressed mainly by natural killer cells (NK) during the first 2 weeks of infection [28]. However, other cell types, such as T cells [10, 12, 22, 28], macrophages [29], DCs [7880], and B cells [28], have also been reported to express this cytokine following Leishmania infection. IL-10 is a potent anti-inflammatory molecule that exerts immunosuppressive effects on antigen-presenting cells and is able to directly restrict the magnitude of CD8+ T-cell expansion [81]. However, the role of IL-10 during the priming of CD8+ T cells is still controversial. This cytokine has also been shown to enhance priming of CD8+ T cells [82, 83]. Recently, another immunosuppressive cytokine, TGF-β, has been shown to control effector CD8+ T-cell number during clonal expansion, by lowering Bcl-2 expression and selectively promoting the apoptosis of short lived effector cells [84]. Although TGF-β appears to be involved in the establishment of chronic infection in several Leishmania models [85, 86], it does not seem to play an important role during L. donovani infections [87], suggesting that it is probably not involved in controlling the expansion of CD8+ T-cell responses.

It is not yet clear whether antigen-specific CD8+ T-cell responses originated in this type of environment have intrinsic defects as a result of defective priming, and thus become dysfunctional over the course of infection. Alternatively, but not mutually exclusive, the induction of exhausted T cells could results from other factors, such as the suppressive splenic environment during chronic infection, decreased levels of IL-21 [88, 89] and/or the constant presence of high levels of antigen [90]. Functionally impaired CD8+ T-cell responses are generated in many chronic infections in humans and in mice, for example hepatitis B [91], hepatitis C [92], HIV [93], LCMV [94]. The exact mechanisms responsible for the generation of exhausted responses are being extensively investigated. Several pathways have been suggested to result in T-cell exhaustion. The inhibitory receptor programed death 1 (PD-1) and its ligand B7-H1 have been shown to control the initiation and reversion of anergy, to inhibit T-cell functions, and to be the key pathway in the induction of exhaustion [9597]. During experimental VL, conventional CD11chi splenic DC increasingly express the inhibitory molecule B7-H1 and fail to upregulate the co-stimulatory molecule CD80 [14]. Upregulation of B7-H1 on DC has been observed during several chronic infections and in a wide range of tumors [97100]. B7-H1 is definitely involved in the induction of T-cell exhaustion in VL, as in vivo blockade of B7-H1 during chronic L. donovani infection increased the survival of antigen-specific CD8+ T cells [14]. Since B7-H1 is thought to inhibit T-cell proliferation by ligation with the PD-1 receptor [101] and to induce programed cell death of effector T cells through ligation with a yet unknown receptor [102], increased survival of CD8+ T cells after B7-H1 blockade may result from restoration of the proliferative capacity or inhibition of induced cell death of effector CD8+ T cells. Interestingly, in contrast to what has been recently reported in the literature [96, 97], in vivo blockade of B7-H1 during chronic VL only transiently restored the functional capacity of exhausted antigen-specific CD8+ T-cells. This suggests that additional pathways may be involved in the suppression of cytokine production by CD8+ T-cells during chronic L. donovani infection. Exhausted CD8+ T cells have been shown to co-express multiple inhibitory receptors under the regulation of the transcriptional repressor Blimp-1 [103, 104]. Thus, various inhibitory pathways may control T-cell exhaustion. Indeed, the simultaneous blockade of PD-1 and LAG3 [104] and/or PD-1 and CTLA-4 [105] showed synergistic effects in lymphocytic choriomeningitis virus (LCMV) and hepatitis C virus (HCV) infection, respectively.

The suppressive splenic environment during chronic VL could also act synergistically with the various inhibitory pathways. A study has demonstrated a synergistic effect between TGF-β and the B7-H1/PD-1 axis in suppressing CD8+ T-cell responses [106]. Furthermore, TGF-β has also been shown to mediate virus-specific CD8+ T-cell deletion during chronic LCMV infection [107]. Although TGF-β does not seem to play an important role during chronic L. donovani infections [87], it is expressed during other Leishmania infections [85, 108] and may therefore play a role in these models. However, IL-10 is highly expressed in both mouse and human VL [11, 2628, 39, 44, 109]. This cytokine has been shown to inhibit expansion and suppress cytokine production in several chronic infections [24, 81, 110]. Thus, IL-10 could potentially synergistically act with the B7-H1/PD-1 axis in the inhibition of CD8+ T-cell responses and be responsible for the suppression of cytokine production observed during chronic VL.

B7-H1 blockade results in a significant decrease in the parasite burden in L. donovani-infected mice. The mechanism of protection is not clear and might not merely rely on IFNγ production, as blockade failed to fully restore cytokine production. In a previous study, we have shown that therapeutic intervention with antigen-specific CD8+ T-cells in mice chronically infected with L. donovani dramatically reduced the parasite burden [111]. Our recent findings consolidated these results by showing that OT-I CD8+ T-cells rescued from cell death by blocking in vivo B7-H1 or by superinfecting mice with recombinant vaccinia virus expressing SIINFEKL resulted in significant host protection in mice infected with ovalbumin-transgenic L. donovani [14]. Many chronic infections generate dysfunctional CD8+ T-cell responses, and the rescue of these responses results in host protection. Thus, it is important to understand the pathways responsible for the induction of exhaustion and deletion of antigen-specific T cells.

Conclusions

Pathogens have developed sophisticated mechanisms to evade immune responses and to establish chronic infections. Leishmania parasites have evolved several strategies that allow them to survive and persist inside the host; key among these are the induction of immunosuppressive IL-10 responses and the generation of functionally impaired antigen-specific CD8+ T cells. The exact pathways involved in immune escape and in the establishment of persistent infections are mostly unknown. A better understanding of the immune evasive mechanisms adopted by the various pathogens could help in the development of novel therapeutic interventions, which are much needed especially for neglected diseases like leishmaniasis, where resistances to drug therapy are constantly emerging and no effective vaccine is available.