NO-MEDIATED APOPTOSIS-LIKE CELL DEATH IN LEISHMANIA AMASTIGOTES

Apoptosis or type I PCD is marked by morphological characteristics occurring in the dying cell including cell shrinkage, oligonucleosomal DNA fragmentation, chromatin condensation leading to the appearance of pyknotic nuclei, and controlled disintegration of the cell into so-called apoptotic bodies (Kerr, Wyllie and Currie, 1972; Clarke, 1990). NO-mediated cell death of Leishmania axenically-grown amastigotes was first assessed by Lemesre et al. in 1997. NO-mediated DNA fragmentation exhibiting features of apoptosis was further demonstrated in axenic amastigotes incubated with several NO donors (acidified sodium nitrite, nitrosylated albumin, SNAP, DETA/NONOate) by monitoring the genomic DNA status of treated versus untreated parasites. Nuclear DNA fragmentation into oligonucleosomal-sized fragments (720, 360 and 180 bp), a typical phenotypical characteristic of apoptotic cells, was readily visible in agarose gel in the case of NO-treated amastigotes and was confirmed by the use of the in situ TUNEL technique (Fig. 2A and B; Holzmuller et al. 2002). This characteristic of the DNA status during the course of cell death by apoptosis was also found in Leishmania undergoing programmed cell death induced by other stimuli (Moreira et al. 1996; Das et al. 2001; Sereno et al. 2001; Arnoult et al. 2002; Lee et al. 2002; Zangger et al. 2002; Sudhandiran and Shaha, 2003; Lindoso et al. 2004; Paris et al. 2004; Sen et al. 2004a,b; Verma and Dey, 2004; Gallego et al. 2005; Jayanarayan and Dey, 2005; Marquis et al. 2005; Singh et al. 2005; Tavares et al. 2005). The use of the molecular DNA intercalatant YOPRO-1, which penetrates specifically apoptotic cells exhibiting the phospholipid phosphatidylserine on their surface, which is normally hidden within the plasma membrane, was a new criterion defining the cell death induced by NO as apoptosis-like (Fig. 2C; Holzmuller et al. 2002). The Annexin V apoptotic detection method was used by other investigators to demonstrate the plasma membrane phospholipids reorganisation in Leishmania undergoing cell death (Mittra et al. 2000; Arnoult et al. 2002; Sudhandiran and Shaha, 2003; Paris et al. 2004; Singh et al. 2005; Tavares et al. 2005). Nevertheless, we must also take into account that living amastigotes can exhibit phosphatidylserine (PS) on their surface. Exposed PS participates in amastigote internalization and induction of the macrophage alternative activation pathway increasing intracellular Leishmania amastigotes growth (de Freitas Balanco et al. 2001).

Fig. 2. NO-mediated programmed cell death exhibiting morphological features of apoptosis in Leishmania amastigotes. Control versus NO donor-treated axenic amastigotes DNA banding pattern in ethidium bromide-stained agarose gel (A), recognition of apoptotic DNA strand breaks by the in situ Terminal deoxynucleotidyl Transferase-mediated dUTP Nick-End Labeling (TUNEL assay kit, Alexis Biochemicals) in control versus NO donor-treated axenic amastigotes (B) and in intracellular amastigotes within control versus NO producing mouse macrophages (D), apoptotic fluorescent YOPRO-1 probe staining in flow cytofluorometry analysis of control versus NO donor-treated (6 hours with 5 mM acidified sodium nitrite) axenic amastigotes (C) and in situ in intracellullar amastigotes within control versus NO producing dog macrophages (E). Evolution of footpad lesion size and nitrate/nitrite levels in sera of Balb/C and C57/Black6 mice infected with Leishmania amazonensis (F) and in situ TUNEL in histological thin cryosections counterstained with Giemsa of footpad lesions of Balb/C (G) and C57/Black6 (H) mice.

Otherwise, by using our in vitro culture system for axenically-grown amastigotes (Lemesre, 1994; Sereno and Lemesre, 1997), we showed that trivalent antimony induced Leishmania amastigotes cell death (Sereno et al. 2001). Changes upstream of DNA fragmentation included generation of oxidative molecules among which was nitric oxide that was primarily concentrated in the parasitophorous vacuole (Sudhandiran and Shaha, 2003). As an indirect proof of potential coordination of natural and chemotherapeutic anti-leishmanial molecules, we demonstrated recently that antimonial-resistant amastigotes were less susceptible to NO-mediated PCD (Holzmuller et al. 2005b). These data suggest that trivalent antimony could act both as an anti-leishmanial molecule and as a macrophage activating compound (Carter et al. 1989). Activation of mouse macrophages by the classical pathway leads to L. amazonensis intracellular amastigotes apoptosis-like death mediated by L-arginine-derived nitrogen oxides (Fig. 2D; Holzmuller et al. 2002). Very recently, Sousa-Franco et al. (2006) found that Balb/C peritoneal macrophages, which are unable to eliminate L. amazonensis without previous activation with cytokines or lipopolysaccharide (LPS), can kill L. guyanensis amastigotes through an apoptotic process that is independent of NO and is mediated by reactive oxygen species. As a whole, the oxidative machinery of macrophages leads to induction of Leishmania amastigotes' programmed cell death. Nevertheless, depending on the Leishmania species considered, the innate microbicidal mechanisms may be sufficient or supplemented by mechanisms triggered through the classical activation pathway of the macrophage by either acquired immunity or chemotherapeutic molecules. Finally, as previously suggested in cured Leishmania-infected dogs (Vouldoukis et al. 1996), integration of the above data strengthens the idea that usual chemotherapy and protective immunity are linked as they both lead to generation of reactive oxygen and nitrogen species (ROS an RNS) within the host cell. Under these conditions Leishmania parasites undergo a programmed cell death exhibiting morphological features of metazoan apoptosis and that can be considered as their preferred mode of death inside their mammalian hosts.

This raises the question of the occurrence of Leishmania apoptotic-like cell death in vivo. Fig. 2F shows the outcome of the infection by L. amazonensis promastigotes in both a susceptible and a resistant mouse model. In Balb/C mice (i.e. Leishmania-susceptible mice) the footpad lesion grows continuously during the time-course of the experiment without development of a Th1-type cellular immune response, as indicated by the absence of increasing levels of NO-end products in the serum (Fig. 2F). By contrast, regression of the footpad lesion size in C57/Black6 mice (i.e. Leishmania-resistant mice) correlates with increased detection of nitrites in the serum indicative of NO production, due to an efficient Th1 cellular immune response (Fig. 2F). The in situ TUNEL technique reveals amastigotes dying by apoptosis in histological thin sections of the regressive lesion (Fig. 2H). Interestingly, apoptotic amastigotes were also observed in the spleen and the liver of hamsters infected with L. chagasi, but the potential stimuli involved are only suggested and discussed (Lindoso et al. 2004). Nevertheless, according to the available data, we can reasonably postulate that phenotypically-defined apoptosis is the natural type of cell death induced in vivo by nitric oxide, or more generally by both ROS and RNS, as a consequence of a Th1-type polarised cellular immune response involved in Leishmania resistance.

Furthermore, we recently described the capacity of naturally secreted antigens easily purified from culture supernatant of L. infantum promastigotes (LiESAp), successfully cultivated in completely defined medium (Lemesre, 1994; Merlen et al. 1999), to protect dogs, a natural host for visceral leishmaniasis, against experimental L. infantum infections (Lemesre et al. 2005). We show that vaccine-induced protection correlates with an early establishment of a strong and long-lasting predominantly Th1-type cellular immune response specifically directed against LiESAp as demonstrated by anti-LiESAp IgG2 reactivity, LiESAp-specific lymphocyte proliferation assays and enhanced NO-mediated anti-leishmanial activity of canine monocyte-derived macrophages (CM-DM) in response to higher IFNγ production by T cells. The use of both in situ TUNEL and in situ YOPRO-1 techniques (Fig. 2E) indicates Leishmania amastigote death by apoptosis inside T lymphocytes-activated CM-DM derived from dogs immunised with LiESAp (Holzmuller et al. 2005a). These studies confirm that the NO-mediated apoptosis of intracellular Leishmania amastigotes that we previously demonstrated in a murine laboratory experimental model, also occurs in a canine model, a natural reservoir for L. infantum/L. chagasi, the etiologic agents of visceral leishmaniasis, in response to cell-mediated protective immunity.

BIOCHEMICAL PATHWAYS INVOLVED IN NO-MEDIATED PROGRAMMED CELL DEATH IN LEISHMANIA

Apoptosis defined in higher eukaryote cells (i.e. mammalian cells) is regulated by two well-characterized executive pathways (reviewed in Green, 2000). The first involves the proteolytic activation of caspases (reviewed in Hengartner, 2000), and could be considered as an evolutionary step acquired by metazoa. The second one is more complex and involves the mitochondrion, with outer membrane permeabilisation leading to the release into the cytosol of mitochondrial intermembrane space proteins that either induce caspase activation, e.g. cytochrome c, or promote the induction of caspase-independent pathways, e.g. apoptosis inducing factor (AIF) (reviewed in Zamzami and Kroemer, 2001). This second pathway can be considered as inherited during the nucleated cell/bacteria-derived mitochondrion symbiosis and is characterized by apoptosis regulators belonging to the Bcl-2 protein family (Reed, 1994; Henkart and Grinstein, 1996; Gross, McDonnell and Korsmeyer, 1999; Martinou and Green, 2001). Caspase activities seem to be essential for the induction of the typical nuclear features of apoptosis, such as chromatin condensation and oligonucleosomal DNA fragmentation, whereas they are not required, in several circumstances, for the induction and execution of PCD (Sperandio, de Belle and Bredesen, 2000). Moreover, evidence is now accumulating that non-caspase proteases including cathepsins, calpains, granzymes and the proteasome complex, also have roles in mediating and promoting cell death (Orlowski, 1999; Johnson, 2000).

Caspase-like activities have been described in Leishmania through the cleavage of specific substrates or the use of specific inhibitors. In 2001, Das et al. made the first report of a caspase-like-regulated cell death in L. donovani promastigotes exposed to ROS. They demonstrated a significant increase in the ability of parasite lysates to cleave a substrate for the CED-3/CPP32 group of proteases. Pre-treatment of cells with a specific inhibitor of this group of proteases reduces the number of cells showing apoptosis-like features and inhibits the cleavage of a poly(ADP)ribose-polymerase (PARP)-like protein (Das et al. 2001). Camptothecin, an inhibitor of topoisomerase I, induces formation of ROS in L. donovani, which increases cytosolic calcium levels and decreases both intracellular pH and potassium levels inducing the apoptotic process through activation of caspase 3-like proteases CED-3/CPP32 (Sen et al. 2004a). Endogenous ROS formation causes subsequent elevation in the level of lipid peroxidation that are potentially involved in the loss of mitochondrial membrane potential and cytochrome c release (Sen et al. 2004b). Activation of CED-3/CPP32 and ICE group of proteases occurs downstream of mitochondrial injuries (Sen et al. 2004b). Interestingly, novobiocin, an inhibitor of topoisomerase II, also induces a CED-3/CPP32-regulated apoptosis but without inducing ROS (Singh et al. 2005). This suggests that different biochemical pathways could be exploited by Leishmania parasites to die by apoptosis. This hypothesis is strengthened by the partial effect of cell permeable caspase inhibitors on amphotericin B-induced apoptosis in Leishmania promastigotes and axenic amastigotes (Lee et al. 2002). Moreover, drug resistance may influence the biochemical pathway involved in PCD since PARP cleavage was evident in the wild type strain but not in the arsenite resistant strain of L. donovani undergoing apoptosis (Jayanarayan and Dey, 2004).

Furthermore, in our experiments with nitric oxide, there was no evidence of CED-3/CPP32 proteases family activation in the L. amazonensis amastigote NO-induced apoptosis as indicated by the lack of effect of specific caspase inhibitors Z-VAD-fluoromethylketone and Z-DEVD-fluoromethylketone, or a general cysteine-protease inhibitor [(2S,3S)-trans-epoxysuccinyl-L-leucylamido-3-methylbutane ethyl ester E-64d (Holzmuller et al. 2002). No inhibitory effect was observed with the caspase inhibitors in L. mexicana axenic amastigotes undergoing apoptosis whereas they were active in amastigotes isolated from mouse macrophages (Zangger et al. 2002). The authors hypothesised either uptake of macrophage caspases by the parasite or the involvement of Leishmania cathepsins in the DEVDase activity measurement used to provide evidence of caspases 3 and 7 (Zangger et al. 2002). This shows the limits of the specificity of the substrates used to characterize caspase-like activities in PCD in Leishmania. Moreover, to date no caspase homologues have been demonstrated in any unicellular eukaryotes (Aravind, Dixit and Koonin, 2001). However, genes encoding for metacaspases, which belong to an ancestral metacaspase, paracaspase and caspase super-family have been identified in protozoa (Uren et al. 2000; Szallies, Kubata and Duszenko, 2002; Mottram et al. 2003). The functions of these metacaspases, in particular in promoting cell death, and sensitivity to classical caspase inhibitors have yet to be elucidated. In fact, to our present knowledge, cathepsin or calpain-like cysteine protease activities strengthen the evolutionary hypothesis of a less complex ancestral biochemical pathway mediating PCD in Leishmania. In contrast to the lack of effect of caspase inhibitors, we observed a significant inhibitory effect on NO-induced apoptosis in L. amazonensis amastigotes with specific proteasome inhibitors, in particular lactacystin and calpain inhibitor I (Holzmuller et al. 2002). The delay in NO-mediated amastigote apoptosis-like death in the presence of reversible proteasome inhibitors supports the view that protease activities of the proteasome complex are involved in promoting apoptosis-like changes in NO-exposed amastigotes (Holzmuller et al. 2002). Recent studies have pointed out in two Leishmania species the existence of an active proteasome, one similar to the proteasomes of other eukaryotes (Robertson, 1999; Christensen et al. 2000; Paugam et al. 2003). Protease activities of this proteasome, in particular calpain, could participate in the cleavage of PARP-like proteins inhibiting DNA repair and favouring nuclear events of apoptosis, as demonstrated in human neuroblastoma cells (McGinnis et al. 1999). More recently, calpain inhibitor I was shown to interfere with apoptotic DNA fragmentation in L. donovani promastigote death induced by miltefosine (Paris et al. 2004). Moreover, Arnoult et al. (2002) have identified a calpain-like sequence in the kinetoplastid database and suggested that L. major cysteine proteinases, inhibitable by both broad caspase and cysteine protease inhibitors, are calpain-like proteases.

Finally, as caspase activities have only been indicated by the use of inhibitors or substrates that are also effective on other cysteine proteases, oxidative stress-mediated PCD in Leishmania could be executed by cysteine proteases belonging to the cathepsin or the calpain families, or to a new cysteine protease family generated during the evolution of protozoans. In fact, Leishmania parasites contain different cysteine proteases, among which cathepsin B- and L-like proteases play an important role in the proliferation and differentiation processes (Frame, Mottram and Coombs, 2000) and could participate in the regulation of parasite death or survival in the host (Selzer et al. 1999). After the activation of the death programme, cysteine proteases related to the calpain family could promote the induction of nuclear apoptosis-like features in Leishmania parasites. Sequential involvement of both cathepsin and calpain families could represent a prototype of the caspase cascade occurring in metazoan apoptosis. Further investigations that must consider both the Leishmania species studied and the apoptotic stimuli used, are needed to determine either the precise role of cysteine proteases in executing PCD or the existence of other biochemical pathways controlling PCD.

NO MOLECULAR TARGETS ASSOCIATED WITH LEISHMANIA PROGRAMMED CELL DEATH

S-nitrosylation reactions with thiol groups on cysteines represent the prototypic molecular NO redox-based mechanism of interaction with proteins (reviewed in Stamler, Lamas and Fang, 2001). In this regard, there is unquestionable evidence that the active site of caspases can be S-nitrosylated, which results in loss of enzyme function (Dimmeler et al. 1997; Liu and Stamler, 1999). Although caspase homologues have, to date, not been formally demonstrated in any unicellular eukaryotes (Aravind et al. 2001), we could hypothesise that if caspase-like or rather metacaspase activities (Uren et al. 2000; Szallies et al. 2002; Mottram et al. 2003) exist in Leishmania, they represent high affinity NO molecular targets and that NO-mediated PCD is under control of other molecules. Moreover, recent studies demonstrate inactivation of Leishmania cysteine protease by NO (Fig. 3, A and B; Salvati et al. 2001; Ascenzi et al. 2004). Although the exact role of cysteine proteases in Leishmania killing is unclear, it has been demonstrated that Leishmania can not grow within macrophages in the presence of cysteine protease inhibitors (Mottram et al. 1996). Cysteine proteases are involved in several biological functions of the parasite (Mottram, Brooks and Coombs 1998). For example, cathepsin-B and -L-like proteases are essential for Leishmania virulence (Denise et al. 2003). Furthermore, calpain-like cysteine proteases are present in Leishmania (Arnoult et al. 2002; Mottram, Coombs and Alexander, 2004) and they are described as playing a crucial role in NO-mediated cell injuries (Volbracht et al. 2005). In fact, considering the definition of apoptosis characterized in metazoans, almost all potential effector molecules belong to the cysteine protease super-family and represent targets inhibitable by NO. Nevertheless, kinetics of inactivation of protozoan cysteine proteases by NO exhibit second-order to pseudo-first-order reaction kinetics depending on NO concentration (Bocedi et al. 2004). Based on the observation of differential inhibition of L. amazonensis cysteine proteases of amastigotes incubated with NO donors (Holzmuller et al. unpublished data), we could also suggest differential sensitivity to NO depending on the catalytic site conformation. In this view, further research is needed to analyse the potential role of cysteine proteases in promoting NO-mediated Leishmania PCD.

Fig. 3. NO molecular targets in Leishmania. Proteinase activities in gelatin-copolymerised SDS-PAGE stained by Coomassie blue of L. amazonensis amastigotes protein lysates either incubated in the presence or absence of 5 mM acidified sodium nitrite during gel revelation steps (A) or prepared from axenic parasites cultured for 4 hours in acidified PBS in the absence or presence of 5 mM sodium nitrite (B). Cis-aconitase activity of N2- versus NO-treated promastigotes (P) and axenic amastigotes (A) of L. amazonensis (L. amaz.) and L. mexicana (L. mex.) by multilocus enzyme electrophoresis (MLEE) (C). Apoptosis of L. amazonensis axenic amastigotes cultured in the presence of NO donor in a medium supplemented by either cis-aconitate (product of cis-aconitase) or 1,3-bisphosphoglycerate (product of glyceraldehyde-3-phosphate dehydrogenase) (D). MLEE profiles revealed with the cis-aconitase hydratase (E.C.4.2.1.3), glyceraldehyde-3-phosphate dehydrogenase (E.C.1.2.1.12), 6-phosphogluconate dehydrogenase (EC1.1.1.44) of wild-type amastigotes (LiWT) and amastigotes resistant to 50 mM (LiNOR50) and to 100 mM (LiNOR100) DETA/NONOate (E).

Other molecular targets, involved in both glycolysis and citric acid cycle, have been shown to be inhibited by NO in Leishmania parasites, in particular cis-aconitase (Fig. 3C; Lemesre et al. 1997) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Mauel and Ransijn, 1997). Interestingly, NO seems to disrupt crucial steps of the energetic metabolism to induce cell death. Up-stream impairment of mitochondrial respiration (Szabo and Salzman, 1995; Lemesre et al. 1997), NO inhibition of glycolysis through ADP ribosylation of GAPDH (Zhang and Snyder, 1993; Mauel and Ransijn, 1997), concomitant with depletion of cellular NAD+ pools (Radons et al. 1994), could represent the death signal initiating PCD. In fact, apoptosis is a form of cell death that requires energy. By blocking the metabolic pathway involved in the conversion of the source of energy (i.e. glucose), we can postulate that NO action forces Leishmania parasites to use the stocks of energy to ensure a silent cell death. It makes sense if we consider the NO disruption of the citric acid cycle as the effector signal promoting PCD since inhibition of respiration may contribute to the NO pro-apoptotic effect (Boyd and Cadenas, 2002; Moncada and Erusalimsky, 2002). A biologically indirect argument supporting our hypothesis is that amastigotes appeared to be less sensitive than corresponding promastigotes to NO action (Lemesre et al. 1997). This difference was consistent with the relatively weak development of the mitochondria in amastigotes compared to promastigotes (Mukkada et al. 1985). In terms of co-evolution of Leishmania inside the host cell, decreased mitochondrial development could represent an adaptative strategy to manage NO toxicity and related cell death induction. Furthermore, NO-mediated PCD in Leishmania amastigotes is abolished by supplementation with the products of these two enzymes: either cis-aconitate for cis-aconitase or 1,3-bisphosphoglycerate for GAPDH (Fig. 3D). This suggests that these two enzymes are key targets involved in NO-mediated apoptosis. We recently demonstrated that L. infantum amastigotes selected in vitro for their NO resistance over-express both cis-aconitase and GAPDH (Fig. 3E; Holzmuller et al. 2006). In fact, parasitic over-expression of NO molecular targets may protect the amastigotes both directly and indirectly. Directly, enzymes act as NO scavengers and consequently detoxify the cell. Indirectly, increased GAPDH could prevent ATP depletion and consequently cell death, by engaging anaerobic glycolysis as observed in NO-treated glucose-fed human epithelial cells (Le Goffe et al. 2002). Moreover, over-expression of cis-aconitase, which is considered as a two-faced protein, i.e. acting firstly as an enzyme and secondly as an iron regulatory protein (Beinert and Kennedy, 1993), would increase regulation of iron homeostasis, which plays a crucial role in tumour cell protection from the pro-apoptotic effect of NO (Feger et al. 2001). In addition to cis-aconitase and GAPDH, we evidenced over-expression of 6-phosphogluconate dehydrogenase (6PGDH) (Fig. 3E; Holzmuller et al. 2006). The 6PGDH is the third enzyme of the pentose phosphate pathway (PPP), which generates NADPH and ribulose-5-phosphate (Barrett, 1997). One hypothesis is that the principal function of PPP in Leishmania amastigotes is the production of NADPH, which is known to protect the parasite Trypanosoma brucei against oxidative stress (Dardonville et al. 2003).

As a whole, data on NO molecular targets open the way for further investigations on the molecular characterization of NO-mediated PCD in Leishmania. In particular, it would be of interest to elucidate the NO targeting of mitochondrial enzymes, correlated with subsequent release of calcium from the de-energized organelle to the cytosol (Richter et al. 1994), leading to activation of calpain-related enzymes for the execution of the nuclear features of cell death. This could therefore define a PCD specific to protozoan parasites and represent the first evolutionary step to apoptosis defined in metazoa.

WHY DO LEISHMANIA PARASITES DIE BY APOPTOSIS?

The important question still remaining is what could be the role of PCD in trypanosomatid parasites? It is fair to assume that, if unicellular organisms have retained such a very complex PCD pathway during evolution, it is because this pathway must be beneficial or essential for survival of the species or population. Two main hypotheses have been proposed to answer the fascinating question concerning the benefits of apoptosis for unicellular organisms (Debrabant et al. 2003; Debrabant and Nakhasi, 2003; Nguewa et al. 2004; Wanderley et al. 2005). First, cell death can be important for population size control in response to limited resources (Welburn, Barcinski and Williams, 1997; Al-Olayan, Williams and Hurd, 2002; Lee et al. 2002) or to avoid host death (parasitic organisms) (Heussler, Kuenzi and Rottenberg, 2001). In this case unicellular apoptotic cells show altruistic behaviour, dying for the benefit of others. The second explanation is that apoptotic cells, which will not necessary die (apoptotic mimicry), could provide signals that enhance the survival of the entire population (Lee et al. 2002; Zangger et al. 2002). For example, the evidence that phagocytosis of apoptotic cells reduces the secretion of mammalian-derived pro-inflammatory cytokines or signals as growth factors, offers the interesting speculation that the ability of intracellular Leishmania to undergo apoptosis may reduce the host immune response and favour overall parasite survival (de Freitas Balanco et al. 2001; DosReis and Barcinski, 2001).

Apoptosis in Leishmania promastigotes and axenically-grown amastigotes has been well demonstrated in vitro (Das et al. 2001; Sereno et al. 2001; Arnoult et al. 2002; Holzmuller et al. 2002; Lee et al. 2002) and then ex vivo in interacellular amastigotes (Sereno et al. 2001; Holzmuller, 2002, 2005a). For the moment, only one published study reports the occurrence of apoptosis in Leishmania amastigotes in vivo (Lindoso et al. 2004), suggesting that PCD could constitute a mechanism that regulates growth of the parasite population during Leishmania infection. Moreover, we show in this paper that apoptotic amastigotes were detected in regressive lesions of experimentally infected resistant mice, allowing us to suggest that PCD could occur in vivo as a consequence of a Th1-type polarised cellular immune response involved in Leishmania resistance. Furthermore, even if apoptotic PCD also seems to be important in vector/parasite interactions in both malaria (Hurd, Carter and Nacer, 2005) and African trypanosomiasis (Welburn and Murphy, 1998), apoptosis during Leishmania metacyclogenesis within the sandfly vector still needs to be demonstrated.

In order to demonstrate successfully the real purpose of the PCD, whether altruistic or otherwise, in Leishmania or other trypanosomatid parasites, it is essential to establish definitively that PCD really occurs in vivo in the insect vectors and/or in the mammalian hosts (inside or outside the macrophage for Leishmania). A better and more detailed understanding of the in vivo role of PCD in Leishmania, as in other unicellular parasites, is needed because it could be exploited to identify new targets for therapeutic intervention. The clarification of its potential relevance in silencing the host immune response to favour parasite survival or infection (alternative activation of macrophages, limitation of antigen presentation to the immune system, Leishmania antigen-specific suppression of the T-cell response) and in increasing the host survival time allowing the pathogen to live in pseudo-symbiosis with the host (regulation of cell population in tissues and organs, control of the growth and/or selection of parasitic population by parasite-derived signals, control of the virulence, mechanisms to promote and maintain clonality within the population) will help to define precisely the role played by PCD in the establishment and the maintenance of the Leishmania/host relationship.

Finally, the most important benefit of PCD knowledge in kinetoplastids would be the design of more active and less toxic drugs directed towards specific molecular targets of the parasites.