Subversion Mechanisms by Which Leishmania Parasites Can Escape the Host Immune Response: a Signaling Point of View
Martin Olivier,1,2,3* David J. Gregory,1,2 and Geneviève Forget4
Centre for the Study of Host Resistance at the Research Institute of the McGill University Health Centre,1 Department of Microbiology and Immunology,2 Department of Experimental Medicine, McGill University, Montréal,3 Centre de Recherche du CHUQ, Université Laval, Sainte-Foy, Québec, Canada4
The obligate intracellular parasite Leishmania must survive the antimicrobial activities of its host cell, the macrophage, and prevent activation of an effective immune response. In order to do this, it has developed numerous highly successful strategies for manipulating activities, including antigen presentation, nitric oxide and oxygen radical generation, and cytokine production. This is generally the result of interactions between Leishmania cell surface molecules, particularly gp63 and LPG, and less well identified macrophage surface receptors, causing the distortion of specific intracellular signaling cascades. We describe some of the signaling pathways and intermediates that are repressed in infected cells, including JAK/STAT, Ca2+-dependent protein kinase C (PKC) isoforms, and mitogen-activated protein kinases (especially ERK1/2), and proteasome-mediated transcription factor degradation. We also discuss protein tyrosine phosphatases (particularly SHP-1), intracellular Ca2+, Ca2+-independent PKC, ceramide, and the suppressors of cytokine signaling family of repressors, which are all reported to be activated following infection, and the role of parasite-secreted cysteine proteases.
Source: Clinical Microbiology Reviews, April 2005, p. 293-305, Vol. 18, No. 2
Leishmaniasis, caused by protozoan parasites of the Leishmania genus, is an infection encountered in tropical and subtropical regions of the world. Leishmania parasites are propagated by different species of the sandfly vector (genus Phlebotomus or Lutzomya), depending on the region. Old World species of Leishmania, such as L. donovani and L. major, cause pathology from southern Europe to Africa, the Middle East, and throughout southern Asia, whereas New World species (e.g., L. mexicana, L. amazonensis, and L. chagasi) are found throughout South and Central America and as far north as the southern states of the United States. Clinical manifestations differ widely, depending on the Leishmania species. The majority of mortality results from the visceral form of leishmaniasis, caused by L. donovani or L. chagasi; 90% of annual cases are reported in Bangladesh, Brazil, India, Nepal, and Sudan. The most common manifestations are cutaneous lesions; 90% of new cases of cutaneous leishmaniasis occur in Afghanistan, Brazil, Iran, Peru, Saudi Arabia, and Syria, and they are caused principally by L. major and the L. mexicana subgenus (64, 164). The Viannia subgenus is encountered solely in the Americas and is responsible for the clinical form named mucocutaneous leishmaniasis, which is characterized by facial disfigurement. In 2000, it was estimated that more than 12 million individuals were infected by the various Leishmania species in 88 countries, with an estimated 1.5 to 2 million clinical cases (164).
Two distinct developmental stages of Leishmania are recognized. Promastigotes are found within the sandfly and have an elongated shape and long flagellum. Promastigotes can be further classified as procyclic promastigotes, which multiply in the gut of the sandfly, or as the infective metacyclic promastigotes, which are found in the mouth parts and anterior gut and do not divide. These differentiate into round or oval amastigotes, which lack flagella, once in the host.
In its mammalian host, the lifestyle of Leishmania is that of an obligate intracellular pathogen infecting the hematopoietic cells of the monocyte/macrophage lineage, which it enters by phagocytosis. Since this cell type is specialized for the destruction of invading pathogens and priming of the host immune response, Leishmania has had to evolve a range of sophisticated mechanisms to subvert normal macrophage function. These include preventing the activation of deadly antimicrobial agents such as nitric oxide (NO) and also inhibition of many of the cytokine-inducible macrophage functions necessary for the development of an effective immune response. This enables the parasite to evade the innate immune response and to divide within the phagolysosome of the infected macrophage, from where it can spread and propagate the disease within the host. In this review, we focus on the molecular mechanisms whereby Leishmania can subvert host surveillance by altering the macrophage signal transduction machinery, thereby modulating the macrophage environment in its favor. It should be noted that studies generally use only one, or sometimes two, species of Leishmania at a time and a single developmental stage, and this gives the impression that the mechanisms revealed apply to all species. However, given the diversity of pathologies caused by Leishmania spp., it is inevitable that significant differences exist in the mechanisms of host cell manipulation. These differences may account for some of the conflicting results described in this review; our understanding is now at a stage where more studies directly comparing the effects of different species on macrophages would be extremely useful.
Another important surface molecule is the glycoprotein gp63 (promastigote surface protease). This is a zinc-dependent metalloprotease with a wide range of substrates, including casein, gelatin, albumin, hemoglobin, and fibrinogen (104). While around 10-fold less abundant than LPG, gp63 is still found throughout the promastigote surface (100, 123). However, its shorter length means that it is essentially buried under a sea of LPG. Like LPG, gp63 is down-regulated in the amastigote form (141). This reduced expression may be counteracted by the absence of LPG on the amastigote surface, meaning that gp63 is no longer masked and may therefore play an important role in amastigote survival and modulation of the host response (102, 123).
The most abundant promastigote surface molecule is glycosylinositol phospholipid (GIPL), a class of GPI-linked glycolipids. These molecules are 10 times more abundant than LPG, but their small size keeps them close to the parasite membrane, so it is unclear what role they play in interaction with the host (45, 101). Unlike LPG, which is continually shed, GIPL has a long half-life and so is believed to play a protective role at the promastigote surface (127).
The completion of the L. major database has revealed putative 65 cysteine peptidases; some of these clearly have roles in promoting parasite survival and disease progression, particularly during L. mexicana infection (for a recent review, see reference 109). At least 50% of the cysteine peptidase activity is localized within the lysosomes of the parasite, so is unlikely to pay a direct role in modulating macrophage signaling (69). However, it is also clear that a significant number of cysteine peptidases are released following amastigote death and as a result of their unusual intracellular trafficking pathway (17, 69). As described below, these released proteases can modulate macrophage activity by acting directly on the host cell surface or following entry into the macrophage endoplasmic reticulum from the phagosome.
This list is far from exhaustive, and it is likely that other surface molecules play important roles in modulating specific aspects of the host immune response. In particular, amastigotes express some poorly characterized, ß-mercaptoethanol-activated metalloproteases (141), and they are even able to incorporate membrane lipids from the host cell (100). Furthermore, other secreted molecules, such as proteophosphoglycans and acid phosphatases, have been directly linked to parasite survival and pathogenicity (89). This review will discuss the role of some of these molecules in subversion of the host defense response.
Parasite surface molecules also play an important role during attachment to the macrophage. In vivo opsonization of Leishmania metacyclic promastigotes by C3b and C3bi permits the interaction with the macrophage complement receptor 1 (CR1) and CR3, respectively. However, since C3b is rapidly converted to C3bi by gp63, it appears that CR3 is the more important receptor, and interaction with CR1 is only transient (73). Attachment via CR3 rather than CR1 is advantageous to the parasite, since it will not trigger the oxidative burst during phagocytosis (108). Promastigotes can also attach to the macrophage via the mannose-fucose receptor, which binds to mannan residues of LPG (10). LPG can also interact with C-reactive protein (CRP), an early inflammatory product, and thus triggers phagocytosis via the CRP receptor (31) without leading to the macrophage activation that is usually seen following CRP receptor-mediated phagocytosis (12). Furthermore, gp63 and LPG interact with fibronectin receptor and CR4, respectively (15, 155), although LPG appears to play only a minor role during attachment and internalization (reviewed in reference 38). More recently, a number of Leishmania surface molecules that play a role during the initial Leishmania-macrophage interaction have been identified, although the macrophage receptors for these molecules are not yet clear. For instance, Chiang and Sefton (27) have identified an ICAM-related molecule, ICAM-L, that may be necessary for the interaction between the parasite and the murine macrophage cell line J774. A recent study has reported that the greater infectiousness of metacyclic promastigotes is due in part to elevated surface phosphatidylserine (157), while blocking antibodies against another parasite surface molecule, GIPL, have been shown to inhibit attachment of L. major (154). Amastigotes can also be internalized in an Fc receptor-dependent fashion following opsonization with specific antibodies (57). The large number of receptors implicated suggests a degree of redundancy among parasite-macrophage interactions, although it appears that several interactions are necessary for internalization (15).
Following their attachment to the macrophage, Leishmania promastigotes are internalized to the relatively benign environment of the endosome, where they begin to differentiate into amastigotes. Unlike amastigotes, promastigotes are vulnerable to degradation by the acidic and hydrolytic environment of the phagolysosome. They must therefore retard endosome maturation and phagosome-endosome fusion, a process dependent on LPG (39). This retardation has been observed by the absence or delayed arrival of late endosomal markers such as rab7 and LAMP-1 (142) and may be related to the LPG-dependent accumulation of F-actin (67). The mechanism is not completely understood, but it has been shown to depend upon calcium presence (156) and protein kinase C (PKC) inhibition (66). Furthermore, LPG appears to change the shape of membranes, leading to steric repulsion between the phagosome and the endosome (107). The delay in phagolysosome maturation provides a window during which promastigotes can differentiate into the more resistant amastigotes. This is consistent with our earlier observation that parasite survival and infectiousness in different strains of mice correlate with macrophage phagocytic activity (120). Interestingly, while LPG is important for the survival of promastigotes in the early stages of uptake, it is not expressed by amastigotes, indicating that its role is transient and confined to only the early stages of infection.
Another survival strategy used by Leishmania parasites is the inhibition of the hydrolytic enzymes and other destructive molecules that are secreted into the phagolysosome. Two newly discovered Leishmania molecules, named peroxidoxins LcPxn1 and -2 (4), and a superoxide dismutase (54) are believed to deplete nitrite derivatives and reactive oxygen intermediates (ROI), which are the most important microbicidal small molecules. Furthermore, there is evidence that LPG itself can enhance promastigote survival by neutralizing vacuolar ROI (25). LPG may also protect against lysosomal enzymes, perhaps by its strong negative charge and galactose-mannose repeating units (41). The proteolytic activity of the surface molecule gp63 is optimal at the acidic pH found in phagolysosomes, supporting the suggestion that it targets lysosomal enzymes (143). However, its role is questionable, as parasites with mutations in the six gp63 genes are still capable of survival, differentiation, and replication within macrophages (70).
Sequestering itself insides the cells of the host allows Leishmania to escape many of the immune responses that would otherwise be directed against it. However, it is also necessary to inhibit numerous macrophage functions, particularly those involved in immune surveillance and macrophage activation, at either the protein or gene expression level. One study of 245 macrophage genes showed that 37% were repressed at least twofold following in vitro infection with amastigotes (18), although larger-scale microarray studies have suggested that promastigotes induce and repress to similar extents (26, 136). However, these studies, and our unpublished observations following infection in vivo, all show reduced expression of numerous genes with important roles in various immune, cell physiological, and signaling functions. We describe some of these functions, and how they are affected by Leishmania, in detail in the following paragraphs.
) and/or lipopolysaccharide (LPS) (126, 127). However, it seems that IFN-
and LPG can synergize to generate NO when administered simultaneously to naive macrophages (126, 127). This suggests that contact between the parasite and the macrophage prevents the macrophage from responding to subsequent exposure to IFN-
produced by lymphoid cells. Inhibition of NO production may result from the production of interleukin-10 (IL-10) and/or transforming growth factor ß (TGF-ß), inactivation of the JAK/STAT pathway, activation of phosphotyrosine phosphatases, and/or ceramide production, as discussed below.
In contrast to mice deficient for NO production, mice deficient for the generation of ROI can ultimately control the infection, after an initial period of increased susceptibility (111), indicating that ROI play a less important role in parasite clearance. However, ROI generation is also inhibited by L. donovani infection (19, 117, 118). Inhibition appears to be dependent on the surface molecules LPG and gp63 (38, 148) and has been shown to involve abnormal PKC activity (118).
Some studies have shown that L. donovani inhibits antigen presentation by repressing major histocompatibility complex (MHC) class II gene expression, both basal and particularly following stimulation with IFN-
(35, 133, 134). In contrast, macrophages infected with L. amazonensis have been shown to express normal levels of MHC class II (84, 124). Antigen presentation may be inhibited in this case by interfering with the loading of antigens onto MHC class II molecules (50, 124) or by sequestration of the MHC class II molecule and/or antigens within the phagolysosome (79, 85). De Souza Leao and colleagues demonstrated a third level of inhibition, at least for the case of L. amazonensis, when they observed direct endocytosis of MHC class II molecules by amastigotes themselves, followed by cysteine peptidase-dependent degradation (40). Consistent with the phagosomal location of the Leishmania, MHC class II appears to be more important than class I for resistance, although class I does have a role to play, at least in some situations (158). Mice defective in MHC class I presentation are resistant to infection with L. major, but MHC class II–/– mice are susceptible (68, 88).
Antigen presentation depends upon cellular communication through costimulatory molecules such as B7/CD28 and CD40/CD40L. It has been demonstrated that B7-1 of L. donovani-infected macrophages could not be further expressed in response to LPS stimulation (75) and that this inactivation process was prostaglandin dependent (137). It seems that inhibition of CD40/CD40L ligation is responsible for the absence of iNOS and macrophage microbicidal activities (72, 147) and that cure of L. major infection depends upon an active CD40/CD40L ligation (23, 62). Recent findings suggest that p38-dependent signaling triggered by CD40 interaction is altered in infected macrophages, and this may lead to diminished iNOS expression (2). In contrast, repression of MHC class II gene expression appears to involve a cyclic AMP-independent mechanism (83).
]) or in T-lymphocyte activation (IL-12). LPS-induced IL-1ß secretion has been reported to be inhibited in L. donovani-infected (134, 135) and LPG-exposed (47) macrophages. LPG seems to repress IL-1ß transcription by acting through a promoter repression sequence (60). In contrast, IL-1
transcription is induced by L. major, but this is not reflected in increased secretion, indicating that a downstream repression mechanism counteracts the induction (61). Interestingly, the induction of IL-1
appears to be Myd88 dependent, suggesting a role for Toll-like receptors. TNF-
production is also repressed in infected macrophages treated with LPS (36). More recently, this has been shown to involve IL-10 and PKC inhibition (8).
The capacity of Leishmania to infect macrophages without inducing proinflammatory cytokines, and then to inhibit their induction in response to various agonists, probably represents a survival mechanism whereby the parasites can inhibit a harmful inflammatory reaction. Nevertheless, these studies have been mainly performed in an in vitro context. Recent studies performed in vivo have clearly demonstrated that proinflammatory cytokines (IL-1, IL-6, and TNF-
) as well as various chemokines, a family of cytokines responsible for recruitment of inflammatory cells to the site of infection, were induced in the early stages of L. donovani and L. major infection (98). Of interest is that L. major promastigotes were shown to be better activators of proinflammatory events than L. donovani, as shown by a greater, transient recruitment of inflammatory cells. This may reflect the different pathologies caused by the two strains. In addition, both species recruit a heterogeneous population of host inflammatory cells, including neutrophils and monocytes/macrophages (98). This is of particular interest from a host defense point of view, since neutrophils have been recently shown to be important for controlling L. major infection (46, 87).
The cytokine IL-12 plays a critical role in the regulation of cellular immune responses. It is essential for T-lymphocyte activation and subsequent IFN-
secretion leading to macrophage activation and production of microbicidal molecules. It is therefore not surprising that Leishmania has developed the ability to inhibit IL-12 production. This has been shown for promastigotes of L. donovani and L. major (24), L. mexicana amastigotes (163), and the phosphoglycan portion of LPG (122) in vitro. IL-12 inhibition has been also reported to occur in L. major-infected mice (5). The intracellular mechanism is still unclear. Macrophage complement receptors and Fc
receptor, which are known to interact with Leishmania during phagocytosis, have been shown to repress IL-12 (93, 152). Furthermore, Piedrafita et al. showed that L. major LPG-mediated IL-12 repression was independent of the NF-
B transcription factor family, despite IL-12 being NF-
B responsive (122). Instead, repression may result from increased ERK1/2 phosphorylation (44); however, that report is problematic, because it contrasts with the well-documented dephosphorylation of ERK during infection (discussed below). A very recent report has shown that the abilities of L. mexicana amastigotes to degrade NF-
B and to repress IL-12 are both dependent on cysteine peptidase B activity (22). While this is purely correlative, there is no reason to assume that promastigotes of Old World species and New World amastigotes employ the same mechanisms, especially given that both species and developmental stages differ markedly in both cysteine protease and LPG type and expression.
TGF-ß production is induced by several Leishmania species in vitro and in vivo (reviewed in reference 13). Augmentation of TGF-ß secretion correlated with retarded iNOS expression and reduced NK cell activity in lymph nodes (140, 151). This is consistent with the idea that TGF-ß inhibits macrophage microbicidal action and the production of IFN-
by NK cells, although the exact role of NK cells during leishmaniasis is somewhat controversial (74, 80, 90, 138, 139, 161). A recent study demonstrated that L. chagasi induces TGF-ß production in the immediate environment of the infected human macrophage, and this may permit the local inhibition of immune responses (51). Interestingly, at least for the case of L. chagasi, the increased production appears to be a result not of increased gene expression but of cleavage of pro-TGF-ß by amastigote cysteine proteases to produce active TGF-ß (51, 145). Interaction between the macrophage and phosphatidylserine motifs on the amastigote surface has also been proposed to trigger this induction (49).
IL-10 is another anti-inflammatory cytokine produced by Leishmania-infected macrophages in vitro, apparently via interaction with the Fc
receptor (153). Its production may be responsible for the suppression of macrophage microbicidal activity involving NO, production of several cytokines (IL-1, IL-12, and TNF), and expression of costimulatory molecules such as B7-1/2 (reviewed in reference 32). Its importance in vivo is illustrated by the observation that transgenic mice constitutively expressing IL-10 are unable to control Leishmania infection (73). As for TGF-ß, IL-10 is apparently induced following recognition of amastigote surface phosphatidylserine residues by the macrophage (49).
Prostaglandin E2 (PGE2) seems to be generated by Leishmania-infected macrophages and to favor parasite survival and progression (43, 98, 131, 132). This arachidonic acid metabolite has been reported to cause inhibition of macrophage proliferation and to suppress production of TNF-
, IL-1, and reactive oxygen intermediates (6). A recent study reports that PGE2 induction in L. donovani-infected macrophages depends upon PKC activation and cyclooxygenase-2 expression (98). Interestingly, one study has correlated increased visceralization of L. donovani in malnourished mice with increased PGE2 production in the lymph nodes (1).
It is therefore clear that Leishmania parasites are capable of modulating numerous macrophage functions in order to promote survival within the host. While we have seen that the parasite surface coat is responsible for triggering many of these effects, we have not directly addressed the intracellular mechanisms by which the signals are communicated. Some of the intracellular signaling pathways that are modulated by Leishmania are discussed in the next section.
A cell transmutes a stimulus on the outside of the plasma membrane into changes in the cell's physiological program by means of intracellular signaling pathways. These are usually triggered by the ligation of an external ligand, such as a cytokine or a Leishmania surface molecule, to a receptor on the cell surface. This ligation causes activation of the receptor, commonly by phosphorylation and/or conformational changes, resulting in activation of second messengers within the cytosol. These second messengers are often protein kinases, which then phosphorylate other kinases to continue a cascade that ultimately results in the activation of effector molecules, such as transcription factors or actin filaments, and causing a change in the cell's behavior. It should be emphasized that the activity of an intracellular pathway is normally determined by a balance of both positive and negative regulation. Activation of a given kinase cascade will often result in the activation of its opposing phosphatases, in a classic example of negative feedback. Many prokaryotic and eukaryotic pathogens, including Leishmania, have evolved various strategies to exploit host cell signaling regulatory mechanisms by distorting this balance between positive and negative influences. In the following paragraphs, we describe some of the most important observations made to date concerning modulation of host cell signaling by Leishmania infection (summarized in Fig. 1).
Ca2+- and PKC-Dependent Pathways
One of the first second messengers that was reported to be modulated by the Leishmania parasite is calcium (Ca2+). Leishmania infection was shown to augment the intracellular Ca2+ concentration of phagocytes (41, 117). The surface molecule LPG seems to chelate Ca2+ (41) and to contribute to the rapid elevation of intracellular Ca2+ concentration (91); however, a causal link between these two properties has not been established. The increased intracellular calcium concentration appears to be a result of augmented Ca2+ uptake by infected cells in response to depletion of intracellular Ca2+ stores (91). While cells infected with L. donovani or L. mexicana have shown altered Ca2+-dependent responses, such as chemotaxis and production of ROI (14, 117), it is hard to establish a clear sequence of events linking the two. For example, inhibition of ROI production may also be due to a significant reduction of the inositol-1,4,5-triphosphate (IP3) (117). This may be due to the action of a Leishmania acid phosphatase responsible for dephosphorylation of IP3 (33), or the elevated Ca2+ concentration itself could activate a host IP3 phosphatase as part of a calcium-dependent cascade (116).A number of downstream regulatory enzymes require Ca2+ in order to be fully active (for example, the serine/threonine phosphatase calcineurin [also called PP-2B or PPP3]). Elevated intracellular Ca2+ concentrations are therefore consistent with increased calcineurin activity (91). Some isoforms of PKC are also Ca2+ dependent, so it is perhaps surprising that there is reduced PKC activity in L donovani-infected macrophages, and this contributes to parasite survival (37, 106, 118). Promastigote LPG has been shown to cause this inhibition (37); however, amastigotes, which lack LPG, can also inhibit PKC in infected human monocytes, indicating the availability of alternative mechanisms (118). LPG-mediated repression may be explained by the observations that it interferes with binding of regulators, including Ca2+ and diacyl glycerol, and can also block PKC membrane insertion (reviewed in reference 38). Interestingly, reduced sensitivity of PKC to diacyl glycerol was first demonstrated biochemically in cells that had been infected with L. donovani amastigotes, and it correlated with reduced generation of oxygen radicals (118). Various species of Leishmania and purified gp63 have been reported to inhibit MARCKs-related protein, a substrate of PKC associated with the cytoskeleton and involved in vacuole dispersal (28, 29). GIPL, another Leishmania surface glycolipid, may also inactivate PKC, but this process needs further investigation (105).
Interestingly, inhibition of Ca2+-dependent isoforms of PKC, in particular PKCß, is accompanied by induction of calcium-independent PKC
(9). Both inhibition and activation are apparently a result of increased ceramide synthesis following infection (52). This ceramide has also been proposed to contribute to activation of phosphotyrosine phosphatase (PTP) and inhibition of mitogen-activated protein (MAP) kinases (discussed below).
One study has also demonstrated that addition of recombinant IL-10 favors macrophage PKC inhibition during L. donovani infection and that addition of anti-IL-10 antibodies prior to infection prevented inhibition of PKC (8). This is surprising, since PKC modulation was previously shown to be a direct result of interaction between the macrophage and the parasite (9, 49, 98). However, it is very possible that IL-10 secreted from infected macrophages inhibits the PKC activity of neighboring, naive macrophages, thus weakening their ability to take part in an immune activation and perhaps making them more prone to subsequent infection.
-Mediated Signaling
inducible. It is therefore not surprising that Leishmania can inhibit JAK2/STAT1 signaling, which is the principal pathway downstream of the IFN-
receptor. Infected macrophages are defective in their ability to phosphorylate JAK1, JAK2, and STAT1 upon IFN-
stimulation (11, 114). In the case of L. donovani promastigotes, this inactivation process has been shown to depend upon PTP activation, especially activation of the PTP SHP-1 (11) (Fig. 1). One study has also shown that negative regulation of the IFN-
receptor contributes to JAK2/STAT1 pathway inactivation in L. donovani-infected cells (130). However, this does not seem to be the case following L. mexicana amastigote infection, where IFN-
-dependent regulation of MHC class I is not affected by infection, indicating that the primary signaling lesion lies downstream of the IFN-
receptor (78).
-dependent NO generation (G. Forget et al., submitted for publication). Ghosh et al. reported that infection of macrophages with L. donovani results in increased synthesis of ceramide (52). This appears to result in reduced phosphorylation of ERK1/2, apparently via activation of an endogenous phosphatase, with resultant repression of NF-
B and AP1, reduced NO generation, and enhanced parasite survival (52, 53).
It seems that other MAP kinase family members are also inactivated following Leishmania infection. For example, the p38 MAP kinase has distorted activity in L. major-infected cells following stimulation with a specific anti-CD40 antibody, which mimics the phagocyte-T-lymphocyte interaction (2). The authors of that study suggested that p38 inactivation contributes to the inhibition of iNOS induction and NO generation.
Ceramide-mediated inactivation of ERK1/2 MAP kinases and resulting inhibition of the transcription factors AP-1 and NF-
B have been also proposed to explain the absence of NO generation by L. donovani-infected cells (53). These studies are supported by the observation that all three MAP kinases subfamilies (ERK1/2, p38, and JNK) are not induced when L. donovani is phagocytosed by naive macrophages (125). This is due in part to LPG, at least in the case of ERK1/2, as LPG-deficient promastigotes do trigger ERK1/2 phosphorylation (125). However, this does not necessarily mean that LPG is directly responsible for the inactivation of these kinases but might mean merely that, in the absence of LPG, the interaction of Leishmania with its host cell is modified. Using LPG-deficient L. donovani promastigotes, we have observed that the PTP SHP-1 was still induced by these parasites, resulting in inactivation of signaling and transcription factors related to JAK2 and MAP kinase signaling pathways (M. Olivier et al., unpublished data). The observation that ERK1/2 is activated in the absence of LPG (125) may be a secondary effect of other, LPG-independent processes, such as chemokine and PGE2 secretion. Finally, it seems that p38 activation is important for the control of Leishmania infection, as the parasite survival rate is diminished in macrophages subjected to treatment with anisomycin, which activates p38 (71). This is not surprising, since the use of any compounds that could augment signaling pathways leading to macrophage activation should demonstrate protective action against Leishmania infection. Using the PTP inhibitor peroxovanadium, we demonstrated, in vitro and in vivo, that modulation of host cell signaling by targeting negative regulatory molecules can confer almost full protection against the development of cutaneous and visceral leishmaniasis (99, 119).
A very recent study has proposed another potential mechanism, in which L. mexicana amastigotes cause the rapid, cysteine peptidase-dependent degradation of ERK and JNK but not p38 (22). At first sight this new report contrasts remarkably with the earlier studies, in which ERK1/2 phosphorylation, but not abundance, was affected by infection. However, those studies all used promastigotes of L. major or L. donovani or amastigotes of L. donovani, and the only other report of ERK phosphorylation following infection with New World Leishmania (L. amazonensis) amastigotes did not address total ERK abundance (94). It is therefore possible that a marked difference in the mechanism of ERK inactivation between Leishmania species has been revealed. The choice of mechanism may depend on divergent evolution between New and Old World strains or on another factor. Direct comparative studies involving a range of species, differing in geographical origin and clinical outcome, are necessary to resolve this question.
SHP-1 is responsible for the negative regulation of many signaling pathways in all hematopoietic cell types, by acting in a variety of fashions. For instance, SHP-1 can bind to receptors and dephosphorylate them directly; it can also associate with a receptor and dephosphorylate other members of the receptor binding complex. The PTP can also interact with other cytosolic proteins and tyrosine dephosphorylate them or their associated proteins (48). The majority of documented SHP-1 effects are the result of the inhibition by dephosphorylation of various kinases and their signaling pathways. SHP-1 plays a vital role in limiting the activation of the JAK/STAT pathways following cytokine receptor stimulation. Dephosphorylation of JAK1/2, TYK2, and STAT1
, -2, -3, -5
/ß, and -6 has already been documented (34, 58, 81, 129, 166). In viable motheaten mice, whose SHP-1 phosphatase activity is deficient, increased nuclear translocation of the transcription factor NF-
B has been reported (46, 76, 95), which seems to provoke an exacerbated inflammatory response. SHP-1 has been also reported to activate (82, 165) or repress (113; Forget et al., submitted) MAP kinases, as well as to induce Src kinase activity (144).
Some studies have demonstrated that the inhibition of IFN-
-dependent phosphorylation cascades following infection is due to activation of host cell tyrosine phosphatases (11, 119). Injection of PTP inhibitors, the bis-peroxovanadium compounds, into mice infected with L. major or L. donovani results in control of the infection. Both cutaneous lesions and parasitic load were significantly reduced in L. major-infected mice, whereas L. donovani-infected mice were completely protected (119). That in vitro and in vivo study was the first demonstration that modulation of host PTP by the parasite plays a pivotal role in the progression of the infection. The bis-peroxovanadium-mediated protection was shown thereafter to depend upon iNOS induction and NO generation, as well as to involve augmented innate inflammatory response that must have played an important role in the containment of the infection (99).
In a parallel study, we have demonstrated that macrophage PTP activity is activated very rapidly, within 5 min of exposure to L. donovani promastigotes, and this correlates with a rapid, general tyrosine dephosphorylation of high-molecular-weight proteins (11; Forget et al., submitted). SHP-1 is the most important component of this PTP activation (Fig. 2) (Forget et al. submitted). Furthermore, SHP-1 directly associates with JAK2 following L. donovani infection, and this interaction could explain in part the inability of JAK2 signaling to be triggered in response to IFN-
stimulation (11). More recently, using immortalized macrophage cell lines deficient for SHP-1, we have firmly established that this PTP is responsible for inactivation of JAK2 and ERK1/2 MAP kinases. Protection of these kinases in SHP-1-deficient macrophages allows these cells to respond to IFN-
stimulation by generating NO, even when they are infected (Forget et al., submitted). Corroborating our findings, another group has reported that SHP-1 interacts strongly with MAP kinases upon L. donovani infection (113). They showed that infected macrophages did not respond to phorbol myristate acetate, an artificial PKC activator, as reflected in inhibition of ERK1/2 phosphorylation, Elk-1 inactivation, and c-fos mRNA expression. A striking difference between these two studies is that the inhibition of ERK was observed 17 h postinfection, whereas interaction between JAK2 and SHP-1 and consequent down-regulation of JAK2 were observable within 1 h. Differences between cell lines and their status of differentiation may partially explain this discrepancy. However, the capacity of the parasite to rapidly inhibit the host cell functions is essential for its survival. In addition to our studies using promastigotes, amastigotes of L. amazonensis have also been shown to rapidly inactivate ERK1/2 (94). Comparing the capacity of L. donovani infantum promastigotes and amastigotes to trigger host PTP activity, we found that both can rapidly and similarly induce this negative regulatory activity in macrophages (I. Abu-Dayyeh and M. Olivier, unpublished data). Recently, it has been proposed that the elongation factor EF-1
of L. donovani could be responsible for the later induction of SHP-1, observed 16 h postinfection (115). However, SHP-1 is activated so rapidly that a receptor-mediated process must also be necessary in the early stages (11, 94). The particular receptors involved may be revealed by a strategy using small interfering RNAs against different negative regulatory receptors of the immunoglobulin family.
B) known to be involved in that expression were not altered or diminished in SHP-1-deficient mice. In fact, infection resulted in increased NO generation, which must have contributed to killing of the parasite. This was in fact supported by the finding that aminoguanidine, an inhibitor of iNOS, completely reversed the protection from L. major infection observed in SHP-1-deficient animals (46). In addition, we recently obtained similar in vivo results concerning the role of SHP-1 in the control of visceral leishmaniasis conferred by L. donovani infection. Of interest, using a pathogenomic approach, we found that a majority of the host genes involved in signaling and the immune response that were repressed following infection in vivo were negatively regulated by PTP SHP-1 (Olivier et al., unpublished data).
SOCS genes are not expressed in naive macrophages. Their transcription is induced in response to various cytokines, hormones, and growth factors, such as IFN-
, that activate the JAK/STAT pathways (reviewed in reference 56). Once transcribed, the SOCS proteins will inhibit the JAK/STAT pathways in a negative feedback loop. Their modes of action vary, but as an example of how this type of regulator works, it has been reported that CIS seems to binds to phosphorylated JAK via its SH2 motif. This may then block binding and thus phosphorylation of STAT (97).
One study has shown that SOCS3 is induced in L. donovani-infected human phagocytic cells (7). SOCS3 mRNA expression was induced in a transient fashion by live or heat-killed parasites, but it was not induced by LPG alone and was not dependent on phagocytosis or production of proinflammatory cytokines (TNF-
or IL-1), stimuli that are known to trigger SOCS expression (7). While the data are convincing, our earlier studies were unable to show induction of any SOCS family members in L. donovani-infected macrophages (Olivier et al., unpublished data). Furthermore, the functional consequences of SOCS3 induction have not been addressed; further studies showing the role played by this inhibitor family in subversion of macrophage functions by Leishmania will be of great interest.
plays an important role in the induction of many IFN-
-inducible genes, which are repressed by Leishmania. Consistent with this role, STAT1
phosphorylation (114) and DNA binding activity (130; Forget et al., submitted) are markedly reduced in L. donovani-infected cells. Various studies in different contexts have shown that the use of proteasome inhibitors could stabilize STAT phosphorylation levels. However, none have clearly addressed the hypothesis that proteasome action was directed against STATs. Some have reported that proteasomes were affecting receptor-JAK stability or JAK activity (21, 59, 159, 160, 169). In a recent study, we have observed that the inability of IFN-
to activate STAT1
nuclear translocation in Leishmania-infected cells is the consequence of rapid and sustained STAT1
protein diminution that begins very early after the interaction between the pathogen and the host cell (Forget et al., submitted). Importantly, the same results were obtained following infection with a variety of different Leishmania species (L. major, L. donovani, L. mexicana, or L. braziliensis). This phenomenon seems to be specific to STAT1
, since STAT3 levels were unchanged. Using a PTP inhibitor and SHP-1-deficient macrophages, we were able to show that PTPs were not involved in L. donovani-mediated STAT1
inactivation. Proteasome inhibitors were shown to rescue STAT1
protein degradation. Our study further revealed that protein kinase C
(PKC
)-dependent signaling could be implicated in this proteasome-mediated STAT1
inactivation process. Together, these results argue for a direct role of the proteasome pathway in the specific proteolysis of STAT1
in macrophages infected with Leishmania, representing a new mechanism whereby pathogens could subvert microbicidal actions.
Intracellular parasites living within the harsh environment of phagocytes have developed strategies permitting their rapid physiological adaptation, escape from first-line defense systems, and the capacity to inhibit several functions of their host cells. This review has addressed some of the mechanisms by which Leishmania achieves this by manipulating signaling pathways of the host macrophage. Such strategies to manipulate the host immune response are by no means unique to Leishmania. There is increasing evidence that a great number of unicellular and pluricellular pathogens have also evolved means to inactivate or exacerbate immune cell functions, contributing to their survival and the development of pathogen-specific pathologies. It is also likely that mechanistic details will differ, sometimes markedly, between promastigotes and amastigotes and also between different species of Leishmania. These differences are likely to be of great importance in explaining the widely different clinical manifestations of leishmaniasis. However, it is clear that the common trait of these modulations is the manipulation of the host cell signaling system. As signaling pathways can be pharmacologically manipulated, a better knowledge of their role and the mechanisms whereby they regulate host immune cell functions and pathogen growth should permit the development of new therapies to control infectious agents.
M.O. is supported by grants from the Canadian Institutes in Health Research (CIHR) and is member of a CIHR Group in Host-Pathogen Interaction. M.O. is the recipient of a Fonds de la Recherche en Santé du Québec (FRSQ) Senior Scholarship and is a Burroughs Wellcome Fund Awardee in Molecular Parasitology. G.F. was the recipient of a CIHR Ph.D. studentship. D.J.G. is the recipient of a CIHR Training Grant postdoctoral award from the Centre for the Study of Host Resistance at McGill University Health Centre.
* Corresponding author. Mailing address: McGill University, Department of Microbiology and Immunology, Duff Medical Building, Room 610, 3775 University St., Montréal, Québec H3A 2B4, Canada. Phone: (514) 398-5592/1302. Fax: (514) 398-7052. E-mail: [email protected] .
..................................................
FIG. 1. Signaling events leading to the induction or inhibition of macrophage functions during Leishmania infection. Leishmania internalization within the macrophage is a receptor-mediated event, and this initial host-pathogen interaction is responsible for the rapid activation and deactivation of several signaling pathways leading to macrophage functions (e.g., phagocytosis, chemokine secretion, and prostaglandin secretion). SHP-1 negatively affects JAK2, Erk1/Erk2 MAP kinases, NF-
B, IRF-1, and AP-1, thus inhibiting IFN-
-inducible macrophage functions (e.g., nitric oxide, IL-12 production, and immunoproteasome formation). STAT1
degradation by proteasome is PKC
dependent. Other phosphatases (e.g., IP3 phosphatase and calcineurin) and surface parasite molecules (e.g., LPG) play a pivotal role in the alteration of various second messengers (e.g., PKC, Ca2+, inositol lipids, and inositol phosphates), regulating important phagocyte functions (e.g., NO and superoxide production). M
, macrophage; DAG, diacyl glycerol; PLC, phospholipase C; fMLPR, formyl peptide receptor; PIP2, phosphatidyl inositol 4,5-biphosphate.

..................................................
FIG. 2. Activation of SHP-1 in Leishmania-infected macrophages. Naive macrophages or macrophages infected for 30 min with L. donovani were stained with propidium iodide (PI) (red) or antibody specific for SHP-1 (green). SHP-1 is spread evenly throughout the cytoplasm in control cells but adopts a more punctate distribution following infection. Foci of SHP-1 are also visible in the nuclei of infected cells. Nil, no infection; Ld, L. donovani infection.

..................................................
Source: Clinical Microbiology Reviews, April 2005, p. 293-305, Vol. 18, No. 2