REVIEW

Molecular mechanisms of antimony resistance in Leishmania

  • 1Division of Biochemistry, Central Drug Research Institute, Lucknow-226001 (UP) India
  • 2Institute of Medical Sciences, Banaras Hindu University, Varanasi, India
  • Correspondence
    Neena Goyal
    neenacdri{at}yahoo.com
  • Journal of Medical Microbiology 2007; 56(2):143–153 · https://doi.org/10.1099/jmm.0.46841-0

    View at publisher PubMed

    Abstract

    Leishmaniasis causes significant morbidity and mortality worldwide. The disease is endemic in developing countries of tropical regions, and in recent years economic globalization and increased travel have extended its reach to people in developed countries. In the absence of effective vaccines and vector-control measures, the main line of defence against the disease is chemotherapy. Organic pentavalent antimonials [Sb(V)] have been the first-line drugs for the treatment of leishmaniasis for the last six decades, and clinical resistance to these drugs has emerged as a primary obstacle to successful treatment and control. A multiplicity of resistance mechanisms have been described in resistant Leishmania mutants developed in vitro by stepwise increases of the concentration of either antimony [Sb(III)] or the related metal arsenic [As(III)], the most prevalent mechanism being upregulated Sb(III) detoxification and sequestration. With the availability of resistant field isolates, it has now become possible to elucidate mechanisms of clinical resistance. The present review describes the mechanisms of antimony resistance in Leishmania and highlights the links between previous hypotheses and current developments in field studies. Unravelling the molecular mechanisms of clinical resistance could allow the prevention and circumvention of resistance, as well as rational drug design for the treatment of drug-resistant Leishmania.

    Overview

    Leishmania, a protozoan parasite belonging to the family Trypanosomatidae, causes leishmaniasis, a group of diseases with clinical manifestations that range from self-healing cutaneous and mucocutaneous skin ulcers to a fatal visceral form [visceral leishmaniasis (VL) or kala-azar]. The disease is a significant cause of morbidity and mortality in several countries of the world. It is endemic in 88 countries, with approximately 400 000 new cases per year (Ashford et al., 1992). Now, leishmaniasis is an emerging tropical disease in the USA (Rosypal et al., 2003), with more than 500 parasitologically confirmed cases (Centers for Disease Control, 2004). Although VL (the lethal form of disease) has been reported in more than 60 countries (Sundar, 2001), nearly all of the 500 000 new cases annually of symptomatic VL occur in the rural areas of only five countries: India, Nepal, Bangladesh, Brazil and Sudan (Desjeux, 2001), with 50 % of all VL cases from the Indian subcontinent. The association of leishmaniasis with HIV has also been reported from 33 countries, with most cases in Southern Europe, where 25–70 % of adult patients with VL are coinfected with HIV (Rosenthal et al., 1995). It has been estimated that 1.5–9 % of patients with AIDS will develop leishmaniasis (Desjeux, 2000).

    The Leishmania parasite is transmitted by an invertebrate sandfly vector, Phlebotomus. The parasite leads a digenetic life cycle (Molyneux & Killick-Kendrick, 1987). Infected sandflies introduce the metacyclic forms of the promastigote stage into the bloodstream of the vertebrate host when they bite to take a meal. The promastigotes transform into an amastigote stage in the phagolysosome of reticuloendothelial cells, where they multiply in the hostile environment of the macrophage and kill the cell.

    Since vaccines against leishmaniasis are still under development (Brandonisio & Spinelli, 2002), the control of the disease relies solely on chemotherapy. The organic salt of pentavalent antimony [Sb(V)] has been the first-line drug for all forms of leishmaniasis for more than 60 years (Herwaldt, 1999). However, in recent years, a large-scale increase in clinical resistance to pentavalent antimonials has been reported (Faraut-Gambbarelli et al., 1997; Lira et al., 1999). In India, as many as 65 % of previously untreated patients fail to respond promptly or relapse after therapy with antimony drugs, due to the development of drug resistance (Sundar et al., 2000).

    Second-line drugs include pentamidine and amphotericin B, but severe side effects and high cost limit their use (Mishra et al., 1992). Miltefosine (hexadecylphosphocholine), originally developed as a neoplastic agent, has now been approved as the first oral drug for leishmaniasis. It can be used for both antimony-responding and non-responding patients (Sundar et al., 1999). Although it shows good efficacy, it is very expensive and has a long life in the body. Preliminary data from phase IV clinical trials in India involving domiciliary treatment with miltefosine along with weekly supervision suggest a doubling of the relapse rate (Sundar & Murray, 2005). This provides a warning that resistance could develop quickly in the future, and therefore plans are required to prevent it. In vitro studies have established that a single point mutation may lead to miltefosine resistance in the parasite (Perez-Victoria et al., 2003).

    Because there are very few drugs in the pipeline, resistance to first-line drug(s) has a very big impact on the treatment of leishmaniasis. Therefore, an understanding of the molecular and biochemical mechanisms of clinical resistance is essential, not only for the development of molecular probes or PCR-based diagnostics to monitor the development and spread of drug resistance in the field, but also for rational drug design for the treatment of drug-resistant Leishmania. Prevention and circumvention of resistance towards antimonials has become a World Health Organization priority ().

    The mechanism of resistance to antimony in the field is unknown, and most of our understanding stems from work on laboratory mutants, mostly of Leishmania tarentolae, in which resistance has been introduced in vitro by the selective pressure of heavy metals, principally arsenite (Dey et al., 1994; Légaré et al., 1997), and which are found to be cross resistant to Sb(III) (Haimeur et al., 2000; Brochu et al., 2003). While evaluating resistance mechanisms in the field, it should be kept in mind that L. tarentolae is quite different in sensitivity to antimony to species that infect mammals (Table 1). Further, the promastigote cell lines selected for Sb(V) resistance may have been selected for resistance to an m-chlorocresol preservative instead of Sb(V) (Ephros et al., 1997), as this stage is not sensitive to pentavalent antimonials. Alternatively, Sb(V) preparations could be partially reduced to Sb(III) due to prolonged storage at acidic pH or in culture media containing thiols (Ferreira Cdos et al., 2003; Frezard et al., 2001). In recent years, our knowledge regarding the mechanism of action of pentavalent antimonials has been generated very rapidly; hence, the present review focuses on the possible mechanisms which may lead to antimony resistance in Leishmania and highlights the links between previous hypotheses and the current developments in field studies.

    Table 1.

    Sensitivities to tri- and pentavalent antimonials of promastigote and amastigote stages of Leishmania species

    nd, Not done. Sb(V) in the form of SAG containing pentavalent antimony; Sb(III) in the form of potassium antimony tartrate containing trivalent antimony.

    The primary mechanism of resistance is the decrease in the active drug concentration within the parasite cell. The parasite may lower the drug level by a variety of mechanisms, including decreased uptake, increased efflux, inhibition of drug activation, inactivation of active drug by metabolism or sequestration, and finally by developing a bypass mechanism. The mechanism of antimony resistance has been found to be multifactorial.

    Uptake of antimony

    To be active against Leishmania, antimony has to enter the host cell, cross the phagolysomal membrane and act against the intracellular parasite.

    The route of entry of antimonials [Sb(V)] into Leishmania (or into macrophages) is not well understood, although pentavalent arsenate [As(V)], a metal related to Sb(V), is known to enter via a phosphate transporter (Rosen, 2002). The transport of antimony was first studied by using [125Sb]Pentostam [Sb(V)] in both stages of the Leishmania mexicana and Leishmania donovani parasites (Berman et al., 1987b; Croft et al., 1981), but more recently MS approaches have been used to demonstrate the accumulation of two forms of antimony, i.e. Sb(V) and Sb(III), in both stages of the parasite. In a number of species, the accumulation of Sb(V) is higher in axenic amastigotes than in promastigotes (Brochu et al., 2003; Croft et al., 1981). It has been speculated that Sb(V) enters via a protein that recognizes a sugar moiety-like structure shared with gluconate, as gluconate has been shown to inhibit competitively the uptake of Sb(V) in axenic amastigotes (Brochu et al., 2003). Further, neither As(V) nor phosphate can compete with the uptake of Pentostam, ruling out the possibility of the use of an As(V) transporter by Sb(V). Although Sb(V) is accumulated by both stages of the parasite, at pharmacological concentrations, it has no antileishmanial activity (Roberts & Rainey, 1993; Roberts et al., 1995; Sereno & Lemesre, 1997; Sereno et al., 1998). However, in some studies, axenic amastigotes have been found to be as sensitive to Sb(V)as intracellular parasites (Callahan et al., 1997; Ephros et al., 1999; Shaked-Mishan et al., 2001). This discrepancy needs to be resolved. Indeed, several factors, such as species, axenization status, pH and thiol concentration, are likely to affect the drug assay and/or the rate of Sb(V) reduction (Carrio et al., 2000; Ferreira Cdos et al., 2003). The high antileishmanial activity of Sb(III) against both stages of Leishmania and the selective activity of Sb(V) against the intracellular parasite further support the hypothesis that the reduction of Sb(V) to Sb(III) is necessary for activity.

    Sb(III) also accumulates in Leishmania amastigotes and promastigotes. The differential accumulation of Sb(III) and Sb(V) in both stages of the parasite suggests that the two forms of antimony do not enter by the same route. Interestingly, the accumulation of Sb(III) is competitively inhibited by the related metal As(III), whereas the accumulation of Sb(V) is not (Brochu et al., 2003). This strongly suggests that Sb(III) and As(III) enter the cell by the same route as that in yeasts and mammals (reviewed by Rosen, 2002).

    In both prokaryotes and eukaryotes (yeast and mammalian), aquaglyceroporins (AQPs) are known to transport trivalent metalloids (Tsukaguchi et al., 1998, 1999; Sanders et al., 1997; Liu et al., 2002; Wysocki et al., 2001). Recently, in Leishmania, an AQP (AQP1) has been identified and demonstrated to mediate the uptake of trivalent antimony in the parasite (Gourbal et al., 2004). Transfection of AQP1 is also able to sensitize a Sb(V)-resistant field isolate of L. donovani to sodium stibogluconate (SAG), due to increased accumulation of As(III) and Sb(III), thereby indicating its role in natural antimony resistance. Overexpression of AQP1 also renders parasites hypersensitive to Sb(III). Interestingly, this observation has been confirmed by a recent differential gene expression study, in which the expression of AQP1 was down-regulated at both the promastigote and the intracellular amastigote stage in antimony-resistant clinical isolates from Nepal (Decuypere et al., 2005). The mRNA abundance of AQP1 has also been shown to be decreased in antimony-resistant mutants of several Leishmania species (Marquis et al., 2005).

    The interspecies differences in sensitivity towards Sb(V) and Sb(III) have not been shown to be correlated with differences in the accumulation of antimony (Brochu et al., 2003), which is another very interesting area of research. One possibility for the higher sensitivity of amastigotes towards Sb(V) is a higher accumulation of Sb(V), which emphasizes the need to study the critical step, i.e. the reduction of Sb(V) within the parasite.

    Mechanism of action, and reduction of the metal

    Although antimonials have been in clinical use as the first-line treatment for about 60 years, there are still some crucial aspects of Sb(V) metabolism in Leishmania that remain uncharacterized. Early work suggested that antimonials probably act by inhibiting glycolysis and fatty acid β-oxidation (Berman et al., 1985, 1987a). However, specific targets in these pathways have not been identified. Recent evidence has shown that antimony kills the parasite by a process of apoptosis involving DNA fragmentation and externalization of phosphatidylserine on the outer surface of membrane (Sereno et al., 2001; Sudhandiran & Shaha, 2003; Lee et al., 2002). However, these effects do not operate via a classical caspase-mediated pathway. Further, Sb(III) has also been shown to inhibit trypanothione reductase (TR) (Cunningham et al., 1994) and glutathione synthetase (Wyllie & Fairlamb, 2006) in vitro. Exposure to Sb(III) also results in rapid efflux of trypanothione (TSH) and GSH from isolated amastigotes and promastigotes in vitro (Wyllie et al., 2004). Therefore, antimonials may compromise the redox potential of the cell by inducing loss of intracellular thiols coupled with accumulation of disulphides, and by inhibiting TR (Cunningham & Fairlamb, 1995). In animal infections, the mode of action of Sb(V) depends on host factors that also include T-cell subsets and cytokines (reviewed by Murray, 2001). Stibogluconate has also been found to be a potent inhibitor of protein tyrosine phosphatase, which leads to an increase in cytochrome response (Pathak & Yi, 2001).

    As Sb(III) is highly active against both stages of the parasite, and Sb(V) is active mostly against intracellular amastigotes (and in some cases against axenic amastigotes also) (Table 1), it is generally agreed that Sb(V) needs to be activated to Sb(III). Further, the site (macrophage or amastigote) and the exact mechanism of reduction (enzymic or non-enzymic) are still unclear. One study has shown that axenically grown amastigotes, but not promastigotes, reduce Sb(V) to Sb(III) (Shaked-Mishan et al., 2001), while other studies have suggested that reduction most likely takes place in the macrophages (Roberts & Rainey, 1993; Sereno et al., 1998). Interestingly, Sb(V) at concentrations up to 1 mg ml−1 has no effect on the growth and viability of macrophages, but Sb(III) even at ∼25 μg ml−1 kills 50 % of THP-1 macrophages (Wyllie & Fairlamb, 2006). Since IC50 values for axenic amastigotes vary from 3.2 to 100 μg Sb(III) ml−1 depending upon the Leishmania species (Table 1), it is likely that macrophages do not reduce Sb(V) to Sb(III) at a significant level, and that the conversion occurs at both sites, i.e. macrophage and parasite, generating higher lethal concentrations of Sb(III) within the parasite. Certainly a proportion of Sb(V) may be converted to Sb(III) in human (Burguera et al., 1993; Goyard et al., 2003) and animal models (Lugo de Yarbuh et al., 1994).

    Both reduced GSH and reduced TSH have been found to be responsible for non-enzymic reduction of Sb(V) to Sb(III) (Ferreira Cdos et al., 2003; Frezard et al., 2001; Yan et al., 2003a, b). The conditions which are found within the phagolysosome in which Leishmania resides, i.e. acidic pH (pH 5) and elevated temperature (37 °C), favour this conversion. Further, all antimony is reduced when the molar ratio of GSH : Sb(V) is equal to or more than 5 : 1, indicating a stoichiometry of 5 : 1 in the reaction between GSH and Sb(V). As the rate of reduction is very low, the physiological relevance of this conversion is still open to question. Interestingly, promastigotes contain higher intracellular concentrations of TSH and GSH than amastigotes (Ariyanayagam & Fairlamb, 2001; Wyllie et al., 2004), and both stages maintain an intracellular pH value close to neutral (Glaser et al., 1988). The non-enzymic reduction of Sb(V) to Sb(III) fails to account for the insensitivity of promastigotes to Sb(V).

    In bacteria and yeast, metal reduction can be mediated by enzymes (Rosen, 2002), and this may be the case in Leishmania too. Shaked-Mishan et al. (2001) have reported that the reduction of Sb(V) to Sb(III) occurs in both stages of the parasite, but that the activity is much higher in the amastigote stage, which explains the severalfold higher sensitivity of the amastigote to Sb(V) and the role of reduction in sensitization of the parasite. Further, the ability to reduce Sb(V) to Sb(III) is lost in Pentostam-resistant mutants, supporting the role of reducing activity in antimony resistance.

    Recently, a parasite-specific enzyme, namely thiol-dependent reductase (TDR)1 has been shown to catalyse the enzymic reduction of pentavalent antimonials to trivalent (Denton et al., 2004). The enzyme is a tetramer protein, containing domains of the omega class of the glutathione S-transferases (GSTs), and uses GSH as the reductant. Although TDR1 has been found to be highly abundant in the amastigote stage of the parasite, a direct relationship between the enzyme activity and antimony sensitivity in Leishmania amastigotes cannot be established.

    Arsenate reductases are ubiquitous in prokaryotes and eukaryotes, and are essential for conferring resistance to arsenate (Mukhopadhyay & Rosen, 2002). ScAcr2p is an arsenate reductase homologue in Saccharomyces cerevisiae that reduces arsenate to arsenite, both in vivo and in vitro (Mukhopadhyay & Rosen 1998, 2001; Mukhopadhyay et al., 2000). Recently, the arsenate reductase homologue LmACR2 from Leishmania major has been identified and characterized. The enzyme has been shown to catalyse the reduction of Sb(V), thus increasing the sensitivity of Leishmania cells to Sb(V) (Zhou et al., 2004). Like ScAcr2p, it also requires GSH and glutaredoxin as co-factors for enzyme activity and is inhibited by As(III), Sb(III) and phenylarsine oxide. Purified LmACR2 also has the capacity to reduce both As(V) and Sb(V). However, unlike ScAcr2p, structurally LmACR2 is a monomer. LmACR2 functionally complements the arsenate-sensitive phenotype of an arsC deletion strain of Escherichia coli or a ScACR2 deletion strain of S. cerevisiae, confirming its role. Most importantly, transfection of LmACR2 in Leishmania infantum promastigotes augments Pentostam sensitivity in intracellular amastigotes, confirming its physiological significance. It is also possible that more than one mechanism is responsible for drug activation.

    As macrophages do not efficiently reduce Sb(V) to Sb(III) (Wyllie & Fairlamb, 2006), parasites are not exposed to lethal amounts of Sb(III) within macrophages; therefore, reduction of pentavalent antimony becomes the critical event, and the loss of reductive activity in the parasite could lead to resistance. However, this has yet to be shown in clinical resistance.

    Efflux of the drug

    As a resistance mechanism, the efflux of a drug or its active derivative is very common in bacteria, yeasts and fungi, and various pathogenic protozoa, e.g. Plasmodium falciparum, Entamoeba histolytica, Giardia lamblia, Trypanosoma cruzi, and Trichomonas vaginalis. This may be the case in Leishmania too. There are two types of ABC transporter known to be responsible for multi-drug resistance (MDR) in cancer cells: P-glycoprotein (P-gp) and multi-drug resistance-related protein (MRP). P-gp is encoded by the mdr1 gene, which confers resistance to many hydrophobic drugs (MDR), and is characterized by reversion with verapamil and cyclosporine A. In Leishmania, MRP also confers MDR, although this cannot be reversed by conventional MDR modulators; the protein responsible is known as MRP1.

    In Leishmania, both classes of ABC transporters have also been reported to be amplified in various species in response to different drugs under laboratory conditions (reviewed by Ouellette et al., 1998, 2001; Leandro & Campino, 2003). The Leishmania MDR1 confers resistance to drugs such as daunorubicin and vinblastine, but there is no evidence of a relationship between the MDR1 protein and antimony resistance (reviewed by Ouellette et al., 2001; Klokouzas et al., 2003).

    Analysis of the complete Leishmania genome () has revealed eight putative protein homologues belonging to the MRP1 family, known to be involved in thiol-associated efflux and metal resistance in mammalian cells (reviewed by Brost & Elferink, 2002; Haimeur et al., 2004). Two of them appear to be involved in antimony resistance in the parasite. The first one is PGPA (renamed as MRPA). However, Leishmania MRPA is functionally distinct from mammalian MRP, as resistance is not conferred to pentavalent antimonials, zinc and cadmium, or the typical multi-drug-resistant P-gp substrates vinblastine and puromycin (Ellenberger & Beverley, 1989). The gene has been found to be amplified in a number of laboratory mutants of Leishmania species selected for resistance to Sb(III), Sb(V) and As(III) (Callahan & Beverley, 1991; Ferreira-Pinto et al., 1996; Haimeur et al., 2000; Légaré et al., 2001; Ouellette et al., 2001). Its role in antimony resistance has been confirmed by transfection studies (Callahan & Beverley, 1991; Papadopoulou et al., 1994; Légaré et al., 1997). However, this transporter is not responsible for the drug efflux across the plasma membrane. Rather, it confers resistance by sequestration of metal–thiol conjugates, a mode of metal detoxification in yeast cells (Légaré et al., 2001; Rosen, 2002; Tamas & Wysocki, 2001). MRPA is localized in membrane vesicles that are close to the flagellar pocket, the site of endo- and exocytosis in the parasite (Légaré et al., 2001). Overexpression of MRPA has been reported to decrease influx of antimony rather than increase efflux (Callahan et al., 1994), and this may be due to a dominant negative effect through interaction with other membrane proteins. Thus, MRPA is an intracellular transporter rather than an efflux transporter, thereby suggesting that MRPA may play a major role in antimony resistance (Weise et al., 2000; Webster & Russel, 1993).

    Recently, it has been shown by DNA microarray that MRPA is overexpressed in the axenic amastigote stage of Sb(III)-resistant L. infantum (El Fadili et al., 2005). Transfection of MRPA confers Sb(III) resistance upon promastigotes and Sb(V) resistance upon the intracellular stage of L. infantum. However, MRPA has not been found to be upregulated in a comparative transcriptomic study of antimony-resistant L. donovani field isolates (Decuypere et al., 2005). Further, no reports are available regarding the amplification of ABC transporter gene(s) in field isolates. Thus, it is still of great interest to determine whether or not drug-resistant field isolates adopt the same strategies to resist antimony as those of laboratory mutants.

    Recently, a second ABC transporter protein, PRP1, involved in antimony resistance has been isolated by functional cloning selecting for pentamidine resistance (Coelho et al., 2003). This gene has been shown to confer cross-resistance to antimony. The localization of this protein and the mechanism by which it confers resistance remain to be determined.

    Another transporter that confers antimony resistance by an active extrusion system independent of MRPA is also present in L. tarentolae laboratory mutants (Dey et al., 1994). Using everted vesicles enriched for the plasma membrane, it has been shown that a metal efflux pump is present in the Leishmania plasma membrane. Like MRPA, this efflux pump also recognizes the metal conjugated to thiols such as GSH and TSH (Mukhopadhyay et al., 1996) and requires ATP. The identity of this efflux pump is still unknown, even 10 years after its discovery. Further, it also appears that this efflux system does not play a significant role in antimony resistance, as the transport kinetics of the vesicles prepared from sensitive and resistant isolates are similar (Dey et al., 1996).

    Thiol metabolism

    Thiol metabolism has a central role in the maintenance of an intracellular reducing environment so that the cell can defend itself against the damage caused by oxidative stress inside the macrophage, oxidants, certain heavy metals and, possibly, xenobiotics (Meister & Anderson, 1983). As antimony causes oxidative stress (Lecureur et al., 2002), a reducing environment within the cell and the presence of thiols become important for antimony resistance.

    Arsenite- or antimony-resistant laboratory mutants of all Leishmania species exhibit significantly increased levels of intracellular thiols, namely cysteine, GSH, spermidine and TSH, suggesting a role for thiols in resistance (Mukhopadhyay et al., 1996; Legare et al., 1997; Haimeur et al., 2000). TSH, the major thiol, is found only in trypanosomatids, and is a conjugate of GSH and spermidine (Fairlamb & Cerami, 1992). The synthesis of these two precursors determines the level of TSH. The γ-GCS gene, encoding γ-glutamylcysteine synthetase, which catalyses the rate-limiting step in GSH biosynthesis, has been found to be amplified in arsenite-resistant L. tarentolae (Grondin et al., 1997), while the gene ODC, which encodes ornithine decarboxylase, an enzyme involved in the regulation of spermidine biosynthesis, is also overexpressed (Haimeur et al., 1999; Guimond et al., 2003). This suggests that a lowering of intracellular thiol concentrations may result in the attenuation of the resistant phenotype. This proposed hypothesis is confirmed by inhibition studies. The inhibition of the γ-GCS and ODC genes by their specific inhibitors, l-buthionine-(sr)-sulphoximine (bso) and dl-α-difluoromethylornithine (DFMO), respectively, results in the reversal of arsenite or antimony resistance in laboratory mutants (Arana et al., 1998; Grondin et al., 1997; Haimeur et al., 1999; Légaré et al., 2001). Although the combination of BSO and DFMO sensitizes the resistant cells, the residual level of resistance is still higher than that in wild-type cells, suggesting that GSH and TSH alone are not sufficient to confer metal resistance. Overexpression of either ODC or γ-GCS in L. tarentolae wild-type cells results in increased thiol levels, almost equivalent to those of resistant mutants, but the transfectants do not exhibit arsenite resistance (Grondin et al., 1997). While co-transfection of ODC or γ-GCS with MRPA in wild-type cells results in arsenite resistance (Haimeur et al., 1999, 2000), this acquired resistance in transfectants is also reversed by the thiol depletor BSO (El Fadili et al., 2005). This therefore establishes that MRPA and increased TSH concentrations act synergistically, and that TSH availability is the limiting factor in both the transport of drug conjugates and resistance to arsenite and/or antimony (Ouellette et al., 1991). In Leishmania tropica and Leishmania mexicana cell lines, an increase in TSH is not associated with either the amplification of γ-GCS or overexpression of ODC (Légaré et al., 1997).

    Interestingly, resistance to Sb(V) in L. donovani clinical isolates (India) is also reversed in animal models by treatment with BSO (Carter et al., 2003, 2005). It is also noteworthy that in these resistant isolates, the expression of γ-GCS is also increased significantly. However, in another study on L. donovani isolates from Nepal, expression of γ-GCS and ODC was significantly decreased in resistant isolates (Decuypere et al., 2005). Therefore, there is a need to study the level of thiols in clinical isolates and determine their role in natural antimony resistance.

    Recent studies have shown that antimony-resistant isolates down-regulate the expression of γ-GCS of macrophages (Carter et al., 2006), probably by down-regulating host NF-κB, which is known to regulate γ-GCS expression (Jang & Surh, 2004). This would result in the reduction of intramacrophage GSH levels and promote an intracellular oxidative environment (Wyllie & Fairlamb, 2006), thereby minimizing the intramacrophage reduction of Sb(V) to its toxic form Sb(III). These results clearly indicate that SAG resistance in L. donovani is associated with manipulation of both host and parasite thiol levels.

    Spontaneous formation of Sb(III) complexed with GSH or TSH, or both, has already been demonstrated by proton nuclear magnetic resonance spectroscopy (Sun et al., 2000; Yan et al., 2003) and by MS (Mukhopadhyay et al., 1996). Since GST is elevated in mammalian cells selected for resistance to arsenite (Lo et al., 1992), it has been proposed that GST mediates the formation of metalloid thiol pump substrates in Leishmania species also. However, in Leishmania, GST is not detectable; rather, a related trypanothione S-transferase activity is observed (Vicker & Fairlamb, 2004).

    Thus, the thiols have a dual role in antimony resistance, i.e. sensitization of the parasite by the reduction of pentavalent to trivalent antimony, and promoting resistance by forming conjugates with trivalent antimony for efflux and/or sequestration.

    Changes in the cytoskeleton

    Microtubules are dynamic cytoskeleton polymers consisting of repeating α-/β-tubulin heterodimers along with γ-tubulin, and are vital for cell shape, growth and differentiation of Leishmania (Chan et al., 1991). Since arsenite has high affinity to bind tubulins and can inhibit tubulin polymerization (Ramirez et al., 1997; Jayanarayan & Dey, 2004), acquisition of resistance to arsenite or antimony may cause changes in cytoskeleton proteins.

    Recent studies have shown that the expression of α-tubulin is similar in both wild-type promastigotes and arsenite-resistant mutants. Differentiation into axenic amastigotes causes down-regulation in resistant mutants compared with the wild-type (Prasad et al., 2000). A twofold increased sensitivity of a mutant resistant to Paclitaxel (known to promote tubulin assembly) is found to decrease the expression of α-tubulin in arsenite-resistant mutant promastigotes (Prasad et al., 2000). On the other hand, the expression level of β-tubulin is higher in both stages of an arsenite-resistant mutant than in the wild-type (Jayanarayan & Dey, 2002), while γ-tubulin expression is upregulated in the amastigote stage only, and is unaltered in the promastigote stage. Although Paclitaxel treatment significantly increases the expression of β-tubulin in resistant promastigotes, it has no effect on γ-tubulin expression in either strain, either before or after differentiation (Jayanarayan & Dey, 2002). Further, arsenite treatment has been shown to decrease the expression of α-and β-tubulin in wild-type promastigotes, while expression remains unaltered in an arsenite-resistant mutant (Jayanarayan & Dey, 2004). Since tubulin synthesis is regulated by the unpolymerized tubulins, and arsenite has been shown to inhibit microtubule polymerization in the parasite, arsenite may decrease the synthesis of tubulins by inhibiting polymerization. It is noteworthy that phosphorylation of α- and β-tubulin is highly increased in the arsenite-resistant mutant (Prasad & Dey, 2000). Phosphorylation of tubulins could directly affect the dynamics of tubulin assembly and could regulate and affect several signal-transduction pathways (Gundersen & Cook, 1999); therefore, tubulins have an important role in metal resistance.

    Other mechanisms

    The levels of heat-shock proteins of varying molecular masses are known to increase in the presence of arsenite (Lawrence & Robert-Gero, 1985). Overexpression of Hsp70 is capable of bestowing a heat-inducible increase in resistance to peroxide (Miller et al., 2000). Recently, Hsp70 and Hsc70 have been identified by functional cloning (Brochu et al., 2004). However, transfection of an HSP gene does not confer resistance directly, but rather increases the tolerance of the cell to metals. Further, it is still not known how Hsp70 provides resistance and whether this mechanism operates in field isolates also.

    It is known that antimony kills Leishmania by inhibiting enzymes of the glycolytic pathways, and by effects on thiol redox potential, DNA fragmentation and programmed cell death (Berman et al., 1985, 1987a; Sereno et al., 2001; Wyllie et al., 2004). It seems that the modes of action of antimony and NO are closely correlated, because both act by oxidative stress, depletion of enzymes of energy pathways, and eventual apoptosis (Holzmuller et al., 2002). Like antimony, NO also induces the expression of stress proteins (Hsp70, Hsp81 and Hsp65), and this is thought to be a protective mechanism against its toxic effect (Adhuna et al., 2000). Interestingly, Sb(III)-resistant L. infantum amastigotes have been found to be cross resistant to NO when NO is supplied extracellularly by an NO donor or intracellularly via macrophage activation (Holzmuller et al., 2005).

    In another study, L. tarentolae promastigotes selected for resistance to SAG, and cross resistant to other antimony-containing drugs and arsenite, showed amplification of a locus containing an ORF of 770 aa. However, the gene for this ORF does not have significant homology with any database sequence (Haimeur & Ouellette, 1998); therefore, it is still uncharacterized.

    Conclusions

    Drug resistance is a major impediment to successful treatment of leishmaniasis (Guerin et al., 2002; Sundar et al., 2000). Unfortunately, the analysis of resistance mechanisms in field isolates has lagged behind studies of laboratory-derived resistant mutants. Resistance is an interplay between the uptake, efflux and sequestration of active molecules (Croft et al., 2006). In laboratory-derived resistant cell lines, Leishmania cells are exposed to Sb(III), the active molecule; therefore, the resistance mechanisms involved are mainly based on decreasing the concentration of the drug within the cell. This could be achieved either by decreasing Sb(III) uptake by decreased expression of AQP1 or by increasing efflux/sequestration of the active drug in conjugation with thiols by increasing the levels of thiols and by amplification of transporters. Indeed, both mechanisms have been demonstrated in several resistant laboratory mutants (El Fadilli et al., 2005; Grondin et al., 1997; Haimeur et al., 2000; Légaré et al., 1997; Marquis et al., 2005) (Fig. 1a). In natural antimony resistance, parasites are exposed to Sb(V), and there is an extra resistance step involved, which is to minimize the conversion of the Sb(V) prodrug to Sb(III), at both the host and the parasite level (Fig. 1b). Based on a few studies conducted on field isolates, it has been demonstrated that the resistant parasite modulates the expression of γ-GCS in the host macrophage (Carter et al., 2006), thereby decreasing the host GSH concentration and also host-mediated conversion of Sb(V) to Sb(III). The parasite also down-regulates the expression of key enzymes for the synthesis of the GSH and spermidine moieties of TSH (Decuypere et al., 2005), thereby decreasing the parasite thiol-mediated activation of the drug. Further, it also restricts the entry of macrophage-generated Sb(III) by lowering the expression of AQP1 (Gourbal et al., 2004), an uptake channel (reviewed by Ouellette et al., 2004). Therefore, it may be proposed that natural resistance to antimony results from a lower accumulation of the active drug within the parasite. Finally, the Sb(III) that enters the parasite may be taken care of either by sequestration or by efflux of thiol–Sb(III) conjugate. However, these steps still need to be confirmed with more isolates. Thus, the mechanism of natural antimony resistance is multifactorial, and may differ from laboratory resistance.

    Figure image not available in archive
    Fig. 1.

    Schematic representation of proposed mechanism of antimony resistance in Leishmania. (a) Laboratory-generated Sb (III)-resistant mutants; (b) natural Sb (V)-resistant clinical isolates. The entry of Sb (V) into the parasite occurs by an unknown transporter, and entry of Sb (III) through AQP1. In Sb (III)-resistant cell lines, the level of TSH is increased due to a severalfold increase in the activity of ornithine decarboxylase (ODC) and γ-glutamylcysteine synthetase (γ-GCS), the rate-limiting enzymes for thiol biosynthesis. The detoxification pathway includes complex formation of Sb (III) with TSH and subsequent sequestration via amplified MRPA and/or efflux by unknown pumps. In natural Sb (V) resistance, a lower expression of ODC putatively leads to lower thiol biosynthesis, thus inhibiting activation of Sb (V). Decreased expression of AQP1 also restricts the entry of Sb (III) into the parasite. Intracellular Sb (III) may be dealt with either by sequestration or by efflux of thiol–Sb (III) conjugates, while that of macrophages is conjugated with GSH and effluxed by host ABC transporters. Other abbreviations: TSH, reduced trypanothine; TS2, oxidized trypanothione; Sb(TS)2, conjugate of Sb (III) with trypanothione; MRPA, ABC transporter.

    Sb(V)-resistant strains are emerging throughout the Indian subcontinent, and a new generation of molecular-based drug-susceptibility tests is urgently required to set up feasible and efficient monitoring strategies. So far, the only reliable method for monitoring the resistance of isolates is the technically demanding in vitro amastigote macrophage model (Carter et al., 2001; Lira et al., 1999), which involves counting the stained parasite. Field isolates expressing reporter genes such as firefly luciferase can greatly facilitate the analysis of resistance (Ashutosh et al., 2005). Further, no molecular markers of resistance are available for currently used antileishmanial drugs, i.e. SAG. These aspects urgently need to be developed to monitor regular drug use, drug response and the spread of resistance.

    Acknowledgements

    The work was supported by a grant from the Department of Biotechnology (DBT) (no. BT/PR2792/Med/14/383/2001). Financial support from the Council of Scientific and Industrial Research (CSIR) to A. is gratefully acknowledged. This is Central Drug Research Institute (CDRI) communication no. 7102.

    References