Parasitic protozoa belonging to the order Kinetoplastida contain trypanothione as their major thiol. Trypanothione reductase (TR), the enzyme responsible for maintaining trypanothione in its reduced form, is thought to be central to the redox defence systems of trypanosomatids. To investigate further the physiological role of TR in Leishmania, we attempted to create TR‐knockout mutants by gene disruption in L.donovani and L.major strains using the selectable markers neomycin and hygromycin phosphotransferases. TR is likely to be an important gene for parasite survival since all our attempts to obtain a TR null mutant in L.donovani failed. Instead, we obtained mutants with a partial trisomy for the TR locus where, despite the successful disruption of two TR alleles by gene targeting, a third TR copy was generated as a result of genomic rearrangements involving the translocation of a TR‐containing region to a larger chromosome. Mutants of L.donovani and L.major possessing only one wild‐type TR allele express less TR mRNA and have lower TR activity compared with wild‐type cells carrying two copies of the TR gene. Significantly, these mutants show attenuated infectivity with a markedly decreased capacity to survive intracellularly within macrophages, provided that the latter are producing reactive oxygen intermediates.
Parasitic protozoa of the order Kinetoplastida are the causative agents of several medically important tropical diseases including sleeping sickness (Trypanosoma brucei rhodesiense and T.b.gambiense), Chagas disease (T.cruzi) and visceral (kala‐azar) and cutaneous (oriental sore) leishmaniasis (e.g. Leishmania donovani and L.major, respectively). Among many other metabolic peculiarities, trypanosomatids maintain their intracellular redox balance by a mechanism that is completely different from that of their insect vectors and mammalian hosts. They lack glutathione reductase (GR) which in nearly all other organisms is responsible for the maintenance of an intracellular reducing environment important for the reduction of disulfides, the detoxification of peroxides and the synthesis of DNA precursors (Fairlamb et al., 1985; Schirmer and Schulz, 1987). Instead, they possess a unique system using N1, N8‐bis(glutathionyl)spermidine (trypanothione) and its metabolic precursor N1‐glutathionylspermidine as their main thiols. These glutathione conjugates are kept in the reduced state by trypanothione reductase (TR) (Fairlamb et al., 1985; Shames et al., 1986; Fairlamb and Cerami, 1992). TRs are members of the NADPH‐dependent flavoprotein oxidoreductase family and are structurally and mechanistically related to GRs (Shames et al., 1986; Krauth‐Siegel et al., 1987). Human GR and all parasite TRs have mutually exclusive substrate specificities (Shames et al., 1986; Henderson et al., 1987; Krauth‐Siegel et al., 1987; Cunningham and Fairlamb, 1995; Krauth‐Siegel and Schöneck, 1995), providing a route for the design of selective inhibitors against the parasite enzymes.
During its infective cycle in the vertebrate host, Leishmania must survive the rigorous oxidizing environment of the macrophage. TR and its subordinate thiols are proposed to play a vital role in maintaining an intracellular reducing environment and in protecting these parasites against oxidative damage, arising both internally as a result of their aerobic metabolism and externally by the immune response of the mammalian host (Fairlamb et al., 1985; Shames et al., 1986). Selective inhibition of TR is therefore potentially an attractive strategy to incapacitate these parasites.
In view of the pivotal role and uniqueness of TR in the management of oxidative stress, we have attempted to generate attenuated Leishmania strains with decreased or null TR activity that could eventually be useful for the development of live vaccines. In order to generate attenuated strains and to investigate further the physiological role of TR, we have genetically engineered TR mutants by gene disruption in pathogenic Leishmania species and analysed the phenotype of the resulting mutants with respect to intracellular survival of the parasites within macrophages.
Inactivation of the L.donovani TR gene by gene disruption
The TR gene is single copy in the Leishmania genome (Taylor et al., 1994) and, since Leishmania is generally considered to be diploid and no sexual crosses have been achieved, two successive rounds of gene targeting should be sufficient to create null mutants for TR. In an attempt to inactivate both TR alleles in L.donovani, we have created two disruption constructs. These consist of the insertion of the neomycin phosphotransferase (neo) or the hygromycin phosphotransferase (hyg) expression cassettes (Papadopoulou et al., 1994a) into the unique BalI site of the 1.25 kb PCR fragment of the TR coding region subcloned into the pSP72 vector (Promega) in order to form TR::neo and TR::hyg, respectively (see Figure 1A).
In a first round of targeting, the TR::hyg disruption construct was transfected into L.donovani. Cells growing in the presence of hygromycin B were cloned on semi‐solid agar and the DNA of clones was digested with SacII and analysed by Southern blots hybridized to specific TR and hyg probes. Two SacII fragments of 2.6 and 2.9 kb were recognized by the TR probe in wild‐type L.donovani (Figure 1B, lane 1). Integration of the hyg cassette into the TR gene should replace one of the 2.9 kb fragments with two new restriction fragments of 1.3 and 2.83 kb (see Figure 1A) while leaving the 2.6 kb fragment intact. If only one allele is disrupted, then the 2.6 and 2.9 kb bands corresponding to the second non‐targeted TR allele should be visible as well. Hybridization to TR and hyg probes (Figure 1B, lanes 2) confirmed that the expected single disruption with the hyg gene had occurred. Likewise, transfection of wild‐type cells with the TR::neo construct resulted in successful disruption of one TR allele, generating the predicted 3.8 kb SacII fragment (Figure 1B, lanes 3).
A clone of the TR/TR::hyg transfectant was selected for the second round of targeting using the TR::neo disruption construct. Transfectants resistant to G418 and hygromycin B were obtained and clones were analysed by hybridization with probes specific for TR, neo and hyg. The neo gene was indeed integrated into the TR locus because the predicted 3.8 kb SacII fragment was obtained following hybridization to neo and TR probes (Figure 1B, lanes 4). Although TR gene disruption by neo and hyg took place, as clearly shown from the hybridization studies, one TR allele remained intact since the 2.9 kb genomic SacII fragment was still present in the double targeted mutant (Figure 1B, lane 4). Moreover, a novel 3.1 kb SacII band hybridizing to the TR probe was detected in this transfectant (Figure 1B, lane 4), most likely resulting from a genomic rearrangement. Digestions with other restriction enzymes also revealed new fragments hybridizing to the TR probe that did not co‐migrate with bands present in the wild‐type strain (not shown), indicating that a genomic rearrangement must have occurred upstream of the TR gene.
To determine how these rearranged fragments might have arisen, chromosomes of the TR/TR::hyg/TR::neo transfectant were resolved by CHEF electrophoresis and the blot was hybridized to the appropriate probes. In wild‐type L.donovani, TR is located on a 520 kb chromosome (Figure 1C), which is in contrast to the 1.1 Mb chromosome reported for another isolate of L.donovani (Taylor et al., 1994). In addition to a 520 kb chromosome on which the TR gene is normally located, a new chromosome of ∼1200 kb hybridized to a TR‐specific probe (Figure 1C, lane 3). The newly generated TR allele was the one targeted by the neo gene (see Figure 1C, lane 3), thus leaving an intact TR allele at the original chromosomal location (520 kb). In additional independent transfection experiments, we were again unable to generate a null mutant, as the TR locus became trisomic (not shown). Thus, our attempts to generate a TR null mutant in L.donovani have, therefore, failed. These results suggest that the TR gene is possibly essential for survival of L.donovani promastigotes.
In order to explain the events leading to genomic translocation and aneuploidy for TR and to determine the size of the translocated region, we have hybridized a CHEF blot of wild‐type L.donovani and TR disruption mutants to several probes covering a region of ∼35 kb containing the TR locus. As a first probe, we used a 4.1 kb ClaI–EcoRI fragment (probe 1, Figure 2A) containing a large part of the TR gene and the putative sequences used for the rearrangement at the 5′ end as indicated from the SacII digestion (see Figures 1B and 2A). Even at relatively high stringency, a positive hybridization signal was obtained at the level of a 1200 kb chromosome in the L.donovani wild‐type and TR/TR::hyg strains (Figure 2C, lanes 1 and 2), indicating that sequences homologous to those present at the TR locus on the 520 kb chromosome were also found on a 1200 kb chromosome.
We tested the possibility of whether homologous genomic repeated sequences could also be used for the downstream rearrangement site. No homology between the upstream and the putative downstream point of rearrangement has been detected. Indeed, the only rearranged fragment hybridizing to probe 1 (even at longer exposure) was the 3.3 kb XhoI–XhoI′ representing the upstream rearrangement in the double TR‐targeted mutant (Figure 2B and D, lane 3). To try to pinpoint the putative downstream rearrangement site, we used a large probe of 23 kb located 9 kb downstream of the TR gene (probe 2, Figure 2A). No rearranged fragments were observed using probe 2 in the double targeted TR mutant compared with wild‐type (data not shown), suggesting that the second rearrangement occurred further downstream. The intense hybridization signal (Figure 2E, lane 3) observed at the level of the 1200 kb chromosome for the TR/TR::hyg/TR::neo mutant indicated that the region covered by probe 2 was translocated along with TR.
Decreased intracellular survival of the L.donovani TR/TR::hyg and TR/TR::hyg/TR::neo mutants
Not only the single but also the double targeted L.donovani TR mutants that we have generated by gene disruption contain one wild‐type TR allele. To look at the effect of the loss of one TR copy in our mutants, we first examined TR mRNA levels in both promastigote and amastigote stages of the parasite. L.donovani stationary phase promastigotes were differentiated in vitro into amastigotes as described in Materials and methods. Differentiation into amastigotes was monitored by hybridizing RNAs (not shown) to an amastigote A2‐specific probe (Charest and Matlashewski, 1994). Northern blot analysis of total RNAs isolated from L.donovani wild‐type and TR mutants hybridized to a TR‐specific probe showed that parasites either with the TR/TR::hyg or the TR/TR::hyg/TR::neo background produce lower amounts of TR mRNA than wild‐type cells (Figure 3A). Densitometric scanning of the RNA bands and normalization with the tubulin RNA levels used as control indicated that there is 3‐ to 5‐fold less RNA in the TR mutants than in wild‐type Leishmania. Furthermore, no stage‐specific expression of the TR gene was observed in wild‐type Leishmania (see Figure 3A).
To ascertain whether decreased levels of TR mRNA in the disruption mutants correlated with decreased TR activity, we have measured enzymatic activities in control and TR mutant strains. A 56% decrease of TR enzymatic activity was detected in the single and the double targeted L.donovani TR mutants compared with the control cells (see Table I). As TR is the enzyme that maintains trypanothione in its reduced form (Fairlamb et al., 1985), we examined whether a decrease in TR activity correlates with an alteration in thiol levels. No statistically significant differences in levels of trypanothione, glutathionylspermidine or glutathione were observed between controls and TR‐targeted mutants (Table I).
Given the putative importance of TR in the detoxification of lethal oxygen metabolites that are produced by the macrophage following infection, we investigated whether our TR disruption mutants with less TR activity than wild‐type cells have the same ability to survive intracellularly. To test this, we used an in vitro system to infect either murine or human macrophages with Leishmania promastigotes. Both the L.donovani TR/TR::hyg and TR/TR::hyg/TR::neo mutants were tested. In the case of human monocytes differentiated into macrophages, a significant decrease in the percentage of parasitized cells was observed with the single and double targeted TR mutants as early as 24 h following infection (Figure 4A). Moreover, although the infection rate was maintained throughout the experiment for the control L.donovani‐neo transfectant, this was not the case for the TR/TR::hyg and TR/TR::hyg/TR::neo mutants where infectivity levels decreased steadily to reach only 12% after 72 h (Figure 4A). Similarly, the number of amastigotes per macrophage decreased drastically from nine at 6 h infection to ∼0.5 amastigote/cell after 72 h for the TR/TR::hyg and TR/TR::hyg/TR::neo mutants (see Figure 4B). Indeed, the loss of one TR allele has dramatic consequences on the intracellular viability of the parasite. To prove that the observed phenotype was due solely to the disruption of TR, we have transfected into the L.donovani TR/TR::hyg a neo expression vector carrying a 2.2 kb insert containing the TR gene of L.donovani infantum and 650 bp of upstream sequence (see Materials and methods). This recombinant strain regained its ability to survive inside macrophages to approximately wild‐type levels (Figure 4), strongly suggesting that the loss of one TR copy was responsible for the decreased intracellular survival observed in the mutants.
The intracellular survival of L.donovani TR/TR::hyg and TR/TR::hyg/TR::neo mutants was also studied in the murine macrophage‐like J774 cell line. However, in this cell line, the mutant's intracellular survival was not affected and parasites grew similarly to the control (not shown). To try to explain this intriguing difference, we looked at macrophage function related to intracellular parasite killing that could influence the phenotype of the TR mutants, such as the production of superoxide anion. The oxidation of cytochrome c was measured in the supernatants of cells stimulated with Leishmania, or phorbol myristate acetate as a positive control, as described in Materials and methods. Our results indicate that, following stimulation with Leishmania cells, the murine macrophage‐like J744 cell line shows no detectable levels of superoxide anion, in contrast to the human macrophages that produce high amounts of this reactive oxygen metabolite (see Table II). Thus, the intracellular survival of Leishmania with diminished TR activity seems to decrease only in macrophages capable of producing superoxide anions.
Phenotype of L.major TR/TR::hyg mutant
To test whether the results observed in L.donovani would be similar in other pathogenic species, we have inactivated one allele of the L.major TR gene by targeted gene disruption. Since isogenic DNA appears to be necessary for efficient gene targeting in Leishmania (Papadopoulou and Dumas, in preparation), we amplified the TR gene of L.major by PCR and introduced the hyg expression cassette into the BalI unique site of the gene. The resulting construct was then transfected into L.major. As shown by Southern blot hybridization of total XhoI DNA digests of L.major wild‐type and TR/TR::hyg transfectant, the integration of the hyg gene into one TR allele leads to an increase in size of one genomic fragment from 8 to 9.2 kb (Figure 5B, lane 2). Less TR activity was found in this mutant as observed with L.donovani disruption mutants (Table I). The L.major TR/TR::hyg disruption mutant was tested further for its capacity to infect human macrophages. As observed for L.donovani, the L.major transfectant missing one TR allele showed an important decrease in parasite infectivity. Indeed, 1.1 amastigotes/cell were detected at 72 h post‐infection in L.major TR/TR::hyg mutant compared with five amastigotes/cell at the beginning of infection (Figure 5C).
Trypanothione and its related enzymes are thought to be involved in many cellular functions similar to those of glutathione in mammalian cells, such as maintenance of an intracellular reducing environment and defence against damage by oxidants, certain heavy metals (Mukhopadhyay et al., 1996) and possibly xenobiotics (for a review, see Fairlamb and Cerami, 1992). The role of TR in L.donovani and T.cruzi has been investigated previously by overexpressing TR using transfections (Kelly et al., 1993). Such transfectants were equally susceptible to inhibition by hydrogen peroxide (H2O2) as controls, and metabolized H2O2 at comparable levels, suggesting that the ability to regenerate trypanothione from trypanothione disulfide is not the rate‐limiting step in the metabolism of H2O2 (Kelly et al., 1993). To investigate further the physiological role of TR, we attempted to generate TR‐knockout mutants by gene disruption. Recently, the function of other Leishmania genes has been established using gene targeting procedures (Bello et al., 1994; Papadopoulou et al., 1994b, 1996; Mottram et al., 1996).
In this work, we have generated double and single TR disruption mutants for L.donovani and L.major, but were unable to obtain TR null mutants. Instead, the TR locus became trisomic (Figure 1). Recombinant parasites possessing only one TR allele produce less mRNA (Figure 3A) which correlates with lower TR activity when compared with wild‐type levels (Table I). Despite a marked decrease in TR activity in the mutants, we found no differences in trypanothione levels between wild‐type and mutants (Table I), and this is in accordance with previous results reported by Kelly et al. (1993). Reduced TR activity does not affect the growth of Leishmania in culture media, nor does it influence the transformation to amastigotes in vitro (not shown). The loss of one TR allele by gene disruption is not compensated by either an increase in TR copy number or RNA production (Figure 3A). These observations are in agreement with two alternative approaches to modulating TR activity that involve either the extrachromosomal expression of dominant‐negative TR mutants of T.cruzi (Borges et al., 1995) in Leishmania cells (J.Tovar, M.L.Cunningham and A.H.Fairlamb, unpublished) or the homologous overexpression of antisense RNA for TR in T.cruzi (Tovar and Fairlamb, 1996). In both of these approaches, no increase in TR copy number or levels of TR mRNA from the endogenous gene was observed.
Interestingly, Leishmania possessing only one wild‐type TR allele showed a dramatic decrease in their capacity to survive within human macrophages. Indeed, 72 h following infection either with the L.donovani or L.major TR‐targeted mutants, the number of amastigotes inside human macrophages decreased by 17‐ and 5‐fold, respectively (Figures 4 and 5). This decreased survival inside macrophages may be explained as follows. In standard culture medium, one TR allele is sufficient to maintain wild‐type levels of trypanothione. However, in the case where the demand for trypanothione is high, such as during infection of the macrophage where reactive oxygen metabolites are produced, the capacity to regenerate reduced trypanothione via TR needs to be of wild‐type levels since many of the steps of the antioxidant defence system involve trypanothione (Henderson et al., 1987; Fairlamb, 1989). If TR activity levels drop, as is the case for the TR mutants, decreased intracellular survival will occur as a result of a less efficient defence against the oxidative stress. This model is consistent with our observations that TR mutants grow well inside a murine macrophage cell line incapable of producing reactive oxygen derivatives, whereas they are eliminated rapidly inside human macrophages that produce high amounts of superoxide anion (Table II). This, we believe, is a formal proof that one of TR's physiological roles in the parasite is to combat oxidative stress inside macrophages. The ability of the TR gene alone to restore the mutant infectivity phenotype (Figure 4) strongly supports that TR is indeed involved in defence against oxidative damage. Moreover, as we never observed any recombination/repair event of the hyg disruption in L.donovani TR/TR::hyg transfected with a TR‐expressing plasmid (not shown), we can conclude that we are clearly complementing the TR disruption rather than any putative neighbouring gene.
In view of the claimed pivotal role of TR in the defence against oxidative stress in Leishmania, our inability to generate TR null mutants was not overwhelmingly surprising. Indeed, following successful integration of both hyg‐ and neo‐selectable markers into the two existing TR genomic alleles by gene disruption, a third allelic locus was created not only in species pathogenic to humans but also in the lizard parasite L.tarentolae (unpublished data). The genomic rearrangements leading to formation of a third locus for TR have been studied by hybridization experiments. In L.donovani TR/TR::hyg/TR::neo mutant, the third TR copy arose via translocation of a large segment of DNA to a new chromosome. The exact region within the TR locus translocated from the 520 to the 1200 kb chromosome is unknown. Our results demonstrated that homologous sequences, most likely of small size (based on the hybridization signal intensity), are present both in the vicinity of the 5′‐flanking sequences of the TR locus and on a 1200 kb chromosome (Figure 2C). Indeed, the rearrangement point located upstream of the TR gene has been mapped to a region lying within a 3.1 kb SacII fragment (Figure 1B) and it is likely that generation of the third TR allele may have arisen by homologous recombination through these sequences. Several of the genomic rearrangements observed in Leishmania are indeed mediated by homologous recombination at the level of repeated sequences (Ouellette et al., 1991; Grondin et al., 1993). Since the downstream rearrangement point has not yet been identified, at this stage we cannot exclude the possibility that the whole region from the TR gene to the telomeric end has been translocated into a larger chromosome. However, this seems unlikely because we have mapped the TR gene at 380 kb from the telomere (S.Tamar and B.Papadopoulou, unpublished), yet we observed no change in the size of the 1200 kb chromosome by CHEF between the wild‐type and the mutant in which the rearrangement took place (Figure 2C).
Regardless of the exact mechanism leading to partial trisomy at the TR locus, it is different from the sole example described previously in a virulent L.major where the chromosome number was increased to keep one dhfr–ts allele intact when attempting to generate a dihydrofolate reductase–thymidylate synthase null mutant (Cruz et al., 1993). These authors have proposed that such alterations in the genome could be a diagnostic criterion for essentiality of gene function. If this supposition is correct, then the TR gene is likely to be essential for parasite survival under conditions not involving exogenous oxidant stress.
We used several approaches to try to determine whether our inability to obtain a TR null mutant was related to the fact that the TR gene is essential. All attempts were inconclusive, but one will be discussed. Vectors carrying inserts of different sizes expressing the TR gene were transfected into L.donovani TR/TR::hyg clones. These transfectants and the TR/TR::hyg clone were subjected to a stepwise increase in hygromycin B concentration. This approach, first used in mammalian cells (Mortensen et al., 1992) and more recently in Leishmania (Gueiros‐Filho and Beverley, 1996), can lead to loss of heterozygocity. In three independent clones containing the TR plasmids and selected for 8‐fold more resistance to hygromycin B, we were indeed capable of disrupting a second (rearranged) TR allele with the hyg gene, but not of engineering a TR null mutant because a wild‐type copy was detected in these cells despite the presence of the plasmid. This would suggest that a gene (or sequence) near to TR is essential and that double disruption of the TR gene by hyg and neo would be detrimental to its expression in such a way that the region becomes trisomic. However, when the TR/TR::hyg mutant was selected for 10‐fold more resistance to hygromycin B in the absence of the TR plasmid, we did not observe any loss of heterozygocity or genomic rearrangements. This suggests that TR itself is also refractory to change and is likely to be an important gene. Our present working hypothesis is that both TR and a nearby gene or sequence are important for parasite survival. More experiments are required, however, to draw firm conclusions on this question.
Genes, such as TR, that are required for parasite intracellular survival should constitute interesting targets to inactivate in order to generate a live vaccine. There are no reliable vaccines against trypanosomatid infections and, in many cases, treatment with drugs currently available is unsatisfactory due to toxicity or limited efficacy (Ouellette and Papadopoulou, 1993). Approaches to developing live vaccines for leishmaniasis are now theoretically feasible and, recently, a L.major dhfr–ts null mutant was used successfully in vaccination of mice (Titus et al., 1995). The TR mutants we have generated in this study indeed showed an attenuated capacity to survive the oxidative stress in macrophages. Despite their interesting phenotype, these mutants probably could not be used for vaccination, given the possibility of reversion in vivo due to the presence of one intact TR copy. However, it could be possible to use this mutation in combination with others in order to obtain stable attenuated strains useful for vaccine development.
Materials and methods
Part of the coding region for L.donovani TR (nucleotides 153–1406; Taylor et al., 1994) was amplified by PCR using degenerate oligonucleotide primers. The full primer sequences used were 5′‐TGC GTT AAC GTC(G) GGT(C) TGC GTC(G) CC‐3′ and 5′‐GCA GAG CTC CTC G(C)GC C(G)GA G(C)GT C(G)GG‐3′ containing the cloning sites HpaI and SacI (underlined). The same primers were used to amplify part of the TR gene from L.major total DNA. The HpaI–SacI PCR products (1250 bp for L.donovani and ∼1350–1400 bp for L.major) were subcloned into the SmaI–SacI sites of pSP72 vector (Promega). To make the TR gene knockout constructs, neo and hyg cassettes (Papadopoulou et al., 1994a) were introduced as SmaI–EcoRV fragments into the unique BalI site of the TR gene from both species. Constructs were linearized by digestion with BamHI–EcoRV and purified by electrophoresis on LMP agarose (Gibco‐BRL) before transfection. A cosmid clone (Ldi E7‐2) containing the TR locus was isolated by screening an L.donovani infantum genomic cosmid library made in the vector CL‐hyg (Ryan et al., 1993) and gridded on high density filters with an L.donovani TR‐specific probe. A 2.2 kb SphI–SphI fragment containing the TR gene coding sequence and 650 bp of upstream sequences was subcloned into pGEM7‐neo expression vector (Papadopoulou et al., 1992) and transfected into L.donovani TR/TR::hyg single mutant strain to test for the specificity of the TR mutant phenotype.
Cell culture and transfections
Leishmania major MHOM/IL/67/JERICHO II (50122) and L.donovani donovani MHOM/IN/80/DD8 (50212) strains were obtained from ATCC. Both strains were grown in SDM‐79 medium (Brun and Schönenberger, 1979) supplemented with 10% fetal bovine serum (FBS) (Multicell, Wisent Inc.) and 5 μg/ml of haemin. Two to 3 μg of linearized DNA from the targeting constructs were transfected into L.donovani promastigotes by electroporation as described previously (Papadopoulou et al., 1992). Conditions for L.major transfections were as described by Descoteaux et al. (1994). Transfectants were selected with 40 μg/ml of G418 (Geneticin, Gibco‐BRL) or 160 μg/ml of hygromycin B (CalBiochem) on SDM‐agar plates (1%). The mouse monocyte–macrophage‐like cell line J774A.1 obtained from ATCC (TIB‐67) was cultured in Dulbecco‘s modified Eagle's medium (DMEM, Gibco‐BRL) supplemented with 10% FBS. Human peripheral blood monocytes were isolated from heparinized venous blood supplied by the Canadian Red Cross. Cells were centrifuged over a Ficoll‐Paque gradient (Pharmacia) as previously described (Olivier et al., 1992). After several washes, cells were resuspended in RPMI 1640 medium (Gibco‐BRL) containing 10% human serum (Gibco‐BRL). In order to differentiate monocytes into macrophages, 3 × 106 peripheral blood leucocytes, counted using trypan blue, were adhered and cultured for 5 days at 37°C in a humidified atmosphere of 5% CO2/95% air. To transform L.donovani promastigotes into amastigotes in vitro, the parasites were grown up to stationary phase at room temperature in SDM medium in which HEPES was replaced by MES to adjust the pH to 5.0, and then switched overnight at 37°C in 5% CO2 atmosphere (Charest and Matlashewski, 1994).
In vitro macrophage infection
The capacity of the Leishmania TR mutants to infect murine and human macrophages in vitro was tested in comparison with a control Leishmania‐neo transfectant, as follows. Macrophages from murine and human origins were seeded (200 μl/well, 5 × 104 cells/ml) into 8‐well chamber slides, and were infected with recombinant parasites, at a parasite to cell ratio of 20:1 for a period of 6 h. Following this incubation, the non‐engulfed parasites were removed by 3–5 washes with warm medium and chambers were replenished with 500 μl of fresh culture medium. The level of infection was determined at 6, 24, 48 and 72 h by optical microscopy examination following Diff Quick staining of cell preparations (Muyombwe et al., 1997). Values are expressed either as the percentage of macrophages that were infected by Leishmania or as the total number of amastigotes within 100 macrophage cells.
Measurement of superoxide production in macrophage cells
Superoxide (O2−) was measured by the reduction of ferricytochrome c (Sigma type III) as described (Olivier et al., 1992). Approximately 106 adherent macrophages either from the murine cell line J774 or from human origin were incubated in RPMI medium containing 1 mg/ml of ferricytochrome c and then infected with 2 × 107 Leishmania (ratio 20 parasites/cell) for 2, 4 and 6 h at 37°C. As a control, macrophage cells were stimulated with 100 ng/ml of phorbol 12‐myristate 13‐acetate (PMA) (Sigma). Quantitation of reactive oxygen intermediates on cell supernatants was done spectrophotometrically at 550 nm. Results were expressed in nmol/h/mg cell protein.
Trypanothione reductase and thiol measurement
Preparation of cell‐free extracts to measure the TR enzymatic activities in wild‐type and TR‐disrupted transfectants was as described previously (Tovar and Fairlamb, 1996). TR activity was assayed spectrophotometrically by monitoring trypanothione‐dependent oxidation of NADPH at 340 nm (Shames et al., 1986). Alanine aminotransferase activity was measured as an internal control to correct for any variability in the extraction procedure (Tovar and Fairlamb, 1996). One unit of activity (U) is defined as the amount of enzyme required to catalyse the conversion of 1 μmol of NADPH to NADP+ per min at 27°C. Thiol content was determined by fluorescence detection following pre‐column derivatization with monobromobimane and separation by HPLC. Thiol content was determined relative to authentic standards. Exponential phase cells (typically 1 × 107 cells/ml) were harvested and washed, and 1 × 108 cells were derivatized as previously described (Shim and Fairlamb, 1988). Analyses were performed in duplicate or quadruplicate for three independent experiments.
Total genomic DNA from Leishmania was prepared as described (Bernards et al., 1981), digested with SacII, XhoI or EcoRI, resolved on 0.7% agarose gels and transferred to nylon membranes (Hybond‐N, Amersham). Total RNA from L.donovani wild‐type and TR mutants from both promastigote and amastigote stages were prepared using Trizol (Gibco BRL). Southern and Northern blots, hybridizations and washings were done following standard procedures (Sambrook et al., 1989). Intact chromosomes were prepared from Leishmania promastigotes harvested during mid‐ to late log phase, washed and lysed in situ in 1% low melting agarose plugs as previously described (Van der Ploeg et al., 1984). The blocks were loaded directly into the wells of 1% agarose gel and sealed in place. Leishmania chromosomes were separated by pulsed‐field gel electrophoresis using a Bio‐Rad CHEF‐DR III apparatus at 5 V/cm, 120° separation angle and switch times varying from 35 to 120 for 30 h and transferred to nylon membranes for Southern blot hybridization. Blots were hybridized with radiolabelled probes prepared by randomly primed synthesis with Klenow DNA polymerase and [α‐32P]dATP. The TR probes used here correspond to PCR‐generated fragments from the two species covering a large part of the gene‐coding region. The other probes used correspond to hyg‐ and neo‐coding sequences also made by PCR, to the α‐tubulin gene of T.brucei (Thomashow et al., 1983) and to the A2 amastigote‐specific gene (Charest and Matlashewski, 1994). Probe 1 corresponds to a 4.1 kb ClaI–EcoRI fragment containing a large part of the TR gene and upstream sequences, and probe 2 is a 23 kb EcoRI–EcoRI fragment located 9 kb downstream of the TR gene (see Figure 2A). These two probes were generated by digestions of the Ldi E7‐2 cosmid.
We wish to thank Drs Greg Matlashewski for probe A2 and Steve Beverley for shuttle vector CL‐hyg. This work was supported by grants from the Connaught Laboratories and the Canadian Medical Research Council Industry program (to B.P.M.Ouellette, and M.Olivier), the Canadian Medical Research Council (to B.P.) and the Wellcome Trust (to A.H.F.). B.P. is a Medical Research Council scholar and both M.Ouellette and M.Olivier are chercheur boursier junior of the Fonds de Recherche en Santé du Québec. M.Ouellette is a recipient of a Burroughs Wellcome Fund New Investigator Award in Molecular Parasitology.
- Copyright © 1997 European Molecular Biology Organization