Arginase Is Essential for Survival of Leishmania donovani Promastigotes but Not Intracellular Amastigotes

ABSTRACT Studies of Leishmania donovani have shown that both ornithine decarboxylase and spermidine synthase, two enzymes of the polyamine biosynthetic pathway, are critical for promastigote proliferation and required for maximum infection in mice. However, the importance of arginase (ARG), the first enzyme of the polyamine pathway in Leishmania, has not been analyzed in L. donovani. To test ARG function in intact parasites, we generated Δarg null mutants in L. donovani and evaluated their ability to proliferate in vitro and trigger infections in mice. The Δarg knockout was incapable of growth in the absence of polyamine supplementation, but the auxotrophic phenotype could be bypassed by addition of either millimolar concentrations of ornithine or micromolar concentrations of putrescine or by complementation with either glycosomal or cytosolic versions of ARG. Spermidine supplementation of the medium did not circumvent the polyamine auxotrophy of the Δarg line. Although ARG was found to be essential for ornithine and polyamine synthesis, ornithine decarboxylase appeared to be the rate-limiting enzyme for polyamine production. Mouse infectivity studies revealed that the Δarg lesion reduced parasite burdens in livers by an order of magnitude but had little impact on the numbers of parasites recovered from spleens. Thus, ARG is essential for proliferation of promastigotes but not intracellular amastigotes. Coupled with previous studies, these data support a model in which L. donovani amastigotes readily salvage ornithine and have some access to host spermidine pools, while host putrescine appears to be unavailable for salvage by the parasite.

validate new therapeutic targets and to better understand host-parasite interactions that impact these putative targets is acute.
The polyamine pathway in Leishmania consists of four enzymes: arginase (ARG), ODC, spermidine synthase (SPDSYN), and S-adenosylmethionine decarboxylase (ADOMETDC). ARG, the first and committed step in polyamine biosynthesis, converts arginine to ornithine, which is subsequently metabolized to the diamine putrescine by the catalytic action of ODC. The enzyme SPDSYN then generates spermidine via the addition of an aminopropyl group that is donated from decarboxylated S-adenosylmethionine, the product of ADOMETDC. Spermine, a prevalent polyamine of higher eukaryotes, is not made or metabolized further by Leishmania (33), and there is no spermine synthase (SPMSYN) homolog in the leishmanial genome (34). ODC, SPDSYN, and ADOMETDC have all been validated as essential for survival and growth of the promastigote form of L. donovani, as gene knockouts of each enzyme confer polyamine auxotrophy to the parasite and mutants cannot grow in the absence of appropriate polyamine supplementation (33,35,36).
Although ARG has not yet been validated as crucial for L. donovani promastigote survival, null mutants at the ARG locus have been created in L. mexicana, L. major, and L. amazonensis (37)(38)(39), species that are etiologic agents of cutaneous leishmaniasis. These Δarg strains are all auxotrophic for polyamines and exhibit significantly reduced infectivity levels in mice compared to wild-type parasites, although they are all still able to establish infections (38)(39)(40)(41). These findings imply that cutaneous amastigotes are able to access ornithine and/or polyamines to some extent from the phagolysosome but demonstrate nevertheless that the parasite ARG is required for maximum infection by these Leishmania species. In contrast, Δodc and Δspdsyn knockouts in L. donovani both elicit striking diminutions in parasite loads in infected mouse organs, although the impact of the Δodc lesion on parasite burdens in both liver and spleen is much greater than that of the Δspdsyn genetic alteration (42,43). The basis for the observed variations in infection levels among the visceralizing L. donovani Δodc and Δspdysn mutants and the cutaneous Leishmania Δarg species is unknown.
Interestingly, the polyamine pathway of Leishmania is partitioned between the cytosol and the glycosome (37), a peroxisome-like microbody unique to trypanosomatids (44,45). ARG is located in the glycosome, while ODC, SPDSYN, and ADOMETDC are cytosolic enzymes (37). This discrete cellular segregation of ARG and the downstream polyamine pathway enzymes may afford better control of polyamine biosynthesis and/or facilitate spatial partitioning of the amino acid arginine for polyamine versus protein biosynthesis. Mislocalization of ARG to the cytosol in both L. mexicana and L. amazonensis dramatically reduced parasite burdens, implying that the glycosomal milieu, although not critical for ARG function in promastigotes (37), is indispensable for proper ARG function in amastigotes (39,46). To address the differential impacts of Δarg lesions in cutaneous strains on infectivity with those in genes encoding downstream polyamine enzymes in L. donovani and to assess the importance of glycosomal targeting of ARG in both activity and infection, we created and characterized L. donovani Δarg strains. We found that ARG is essential for polyamine biosynthesis de novo and for promastigote growth in unsupplemented media. The Δarg deletion triggered a one order of magnitude reduction in parasite burdens in livers of mice after a 4-week standard infection but did not impact parasite loads in spleens. A comparison of murine infectivity in Δarg, Δodc, and Δspdsyn lines supports a model in which L. donovani amastigotes in the phagolysosome scavenge ornithine efficiently and salvage spermidine to some extent, while putrescine acquisition is basically insufficient to support amastigote maintenance.

RESULTS
L. donovani ARG. The coding and flanking sequences for the ARG locus of L. donovani (LdBPK_351490.1) were obtained from www.genedb.org. A multisequence alignment showed that the L. donovani ARG open reading frame (ORF) is 97%, 96%, and 95% identical to the predicted ARG proteins of L. major, L. mexicana, and L. amazonensis, respectively (see Fig. S1 in the supplemental material). Like other previously reported Leishmania ARG sequences (37,39), the L. donovani gene encodes a C-terminal tripeptide that mediates the translocation of the protein to the glycosome. It should be noted that the ARG genes from L. donovani and L. major, both Old World Leishmania species, encompass an AKL C-terminal triad, while the ARG gene from L. mexicana and L. amazonensis, both New World Leishmania species, encode an SKL C-terminal tripeptide. Both AKL and SKL are archetypical topogenic signals for targeting proteins to the glycosome (47,48). The L. donovani ARG protein is 43% and 39% identical to the two human ARG enzymes, ARGI and ARGII, respectively.
Creation and genotypic characterization of ⌬arg mutants and add-back cell lines. To investigate the functional role of ARG in parasite growth and infectivity, gene deletion mutants were created by double targeted gene replacement from wild-type L. donovani. ARG/arg heterozygotes were first generated by replacing one copy of ARG with either a HYG or PHLEO drug resistance marker, and the expected chromosomal rearrangements were confirmed by PCR. Both ARG/arg heterozygotes were then subjected to a second round of transfection to generate Δarg null mutants, and two clonal Δarg null mutant cell lines, derived from a HYG heterozygote, were selected for further analysis. In addition, add-back lines were made via insertion of the ARG coding region or a mutated arg gene lacking the C-terminal AKL glycosomal topogenic signal into the ribosomal locus of a Δarg cell line. Both PCR and Western blot analyses were employed to confirm the predicted genotypes of the mutant parasites (Fig. 1). The PCR analysis demonstrated that the ARG coding region could not be detected in the ⌬arg background but was present in wild-type, heterozygous, and add-back parasites (Fig. 1A). Western blot analysis with lysates harvested from either promastigotes or axenic amastigotes confirmed the absence of ARG protein from ⌬arg parasites and the presence of ARG protein in wild-type, heterozygous, Δarg[ARG], and Δarg[argΔAKL] cells (Fig. 1D). Somewhat surprisingly, the heterozygous parasites expressed very little ARG protein compared to the much more robust expression of ARG protein in wild-type parasites (Fig. 1D), an observation that could be ascribed to allele-specific expression differences. Immunofluorescence studies confirmed that the wild-type and Δarg [ARG] parasites expressed ARG in the glycosomes and that the Δarg[argΔAKL] cell line produced Arg protein that mislocalized to the cytosol (Fig. 2).
Nutritional assessment of the L. donovani ⌬arg mutant. Growth assays confirmed that the Δarg lesion conferred a conditionally lethal growth phenotype to L. donovani. Only the Δarg cell line was unable to grow in the absence of ornithine or polyamine supplementation, while wild-type, Δarg[ARG], and Δarg[argΔAKL] parasites grew at comparable rates and to similar densities in polyamine-free medium (Fig. 3A). The phenotypic consequences of the genetic lesion in Δarg parasites could be avoided by supplementation of the medium with the downstream metabolite putrescine ( Fig.  3A and B), while spermidine, the penultimate polyamine in Leishmania, did not rescue Δarg proliferation at concentrations of up to 1,000 M (Fig. 3B). The concentrations of ornithine and putrescine that restored Δarg cell growth were determined by incubation of the knockouts in medium containing different ornithine or putrescine levels ( Fig. 3B  and 4). An initial screen indicated that optimal growth of Δarg parasites could be achieved at 100 M putrescine, whereas supplementation of the medium with 100 M ornithine had essentially no growth-stimulatory effect (Fig. 3A). Only modest growth could be achieved at 500 M to 1 mM ornithine (Fig. 3B). Incubation of Δarg promastigotes in medium containing serial dilutions of either ornithine or putrescine revealed effective concentrations of additive that achieved 50% maximum cell densities (EC 50 s) of 750 Ϯ 229 M for ornithine and 7.5 Ϯ 5 M for putrescine (Table 1 and Fig. 4A). Because ornithine and putrescine incorporation by the intracellular amastigote form is ultimately key for the clinical validation of ARG as a potential therapeutic target, the requirement for ornithine and putrescine was also assessed in axenic amastigotes. The EC 50 s calculated for ornithine and putrescine were 383 Ϯ 147 M and 6.25 Ϯ 3.86 M, respectively (Fig. 4B).
Previous studies with L. mexicana promastigotes revealed that transport capacities for putrescine and ornithine are similar and thus cannot explain the difference in ornithine and putrescine concentrations that are required to efficiently bypass a Δarg lesion (37). However, ornithine levels are much higher than those for putrescine in L. mexicana (37), suggesting that ODC is the rate-limiting step in polyamine production. To test this conjecture, we investigated whether overproduction of ODC in the Δarg parasites could reduce ornithine requirements for parasite proliferation by constructing a Δarg derivative line, the Δarg[ODC] strain, that overexpressed ODC. Overproduction of the ODC protein in the Δarg[ODC] line was confirmed by Western blotting (Fig. 5A). Growth analysis in serial dilutions of ornithine revealed that the Δarg[ODC] parasites indeed required smaller amounts of ornithine than the parental Δarg parasites (Fig. 5B).
Effect of ARG inhibitors on promastigote proliferation. Because ARG is essential for promastigote proliferation, the arginine analogues N -hydroxy-L-arginine (NOHA),  Goat anti-rabbit Oregon greenconjugated secondary antibody was used to detect the ARG primary antibodies. Parasites were also stained with DAPI (B, E, H, and K) and photographed using differential interference contrast (DIC) (C, F, I, and L).
were less effective growth inhibitors. An EC 50 of ϳ2.5 mM was obtained for ABH, while BEC displayed no growth-inhibitory effect on promastigote growth up to 5 mM (data not shown). Interestingly, NOHA exhibited efficacy against L. major (38) but not against L. mexicana promastigotes (46). The relatively poor efficacy of the ARG enzyme inhibitors, at least for L. mexicana and L. donovani, can be ascribed to potentially poor uptake of the charged compounds (46,49).
Infectivity analyses. To determine the impact of an ARG deficiency on infectivity, parasite burdens in both liver and spleen tissues were assessed in BALB/c mice inoculated with either wild-type, Δarg, Δarg[ARG], or Δarg[argΔAKL] parasites. Mice were sacrificed 4 weeks postinfection, since a time course experiment had previously determined that hepatic parasite burdens in our mouse model peaked between two and 4 weeks while splenic parasite loads were much lower but remained at steady levels in the spleen for up to 10 weeks (Fig. S2). Previous infectivity studies with Δodc and Δspdsyn L. donovani strains were also performed 4 weeks postinoculation (42,43). The average liver parasite loads of the five mice infected with wild-type parasites varied between 10 4 and 10 5 parasites/g, and the average parasite numbers of mice infected with the two independent Δarg knockout strains were an order of magnitude lower (Fig. 6A). Statistical analysis (paired t test) revealed a significant difference in liver parasite numbers harvested from mice injected with either wild-type or knockout Parasites were incubated at a density of 5 ϫ 10 5 parasites/ml, and cell viability was evaluated after 5 days by resazurin-to-resorufin conversion. The capacity of wild-type parasites to metabolize resazurin in unsupplemented medium was equated to 100% proliferation. The experiments were set up in duplicate and repeated three times with similar results.

FIG 4
Requirements for ornithine and putrescine in Δarg promastigotes and axenic amastigotes. Δarg promastigotes (A) and Δarg axenic amastigotes (B) were incubated in serial dilutions of putrescine (black circles) or ornithine (gray squares). Supplement concentrations ranged from 1 to 2,000 M. Promastigotes were seeded at 5 ϫ 10 5 /ml in 96-well plates and incubated in DME-L at 27°C, while axenic amastigotes were seeded at 5 ϫ 10 5 /ml in 96-well plates and incubated in a modified acidic medium at 37°C. After 5 days, the ability of parasites to metabolize resazurin was assessed by fluorescence, and readings obtained with the highest supplement concentrations equated to 100% proliferation. The experiments were set up in duplicate and repeated three times with similar results. parasites (P values of 0.0261 and 0.0279 for a comparison between the wild type and the Δarg mutants 1 and 2). Parasite burdens in livers obtained from mice infected with the Δarg[ARG] add-back parasites were equivalent to those of mice inoculated with wild-type parasites, intimating that the infectivity deficit incurred by the Δarg lesion can be ascribed to ARG loss and not to some ancillary genetic alteration that occurred through the extended creation of the null line. Parasite loads in livers of mice infected with Δarg[argΔAKL] parasites expressing the misolocalized cytosolic arg were intermediate between those from mice injected with wild-type and Δarg parasites. Only inconsequential differences in parasite numbers were observed in splenic preparations among mice infected with either wild-type, Δarg knockout, or add-back parasites (Fig. 6B).

DISCUSSION
To evaluate the functional role of ARG, the first and committed step in polyamine biosynthesis in L. donovani, a Δarg knockout was created by double targeted gene replacement and characterized with respect to its growth and infection capabilities. The Δarg lesion conferred polyamine auxotrophy to the parasite that could be rescued by supplementation of the growth medium with either ornithine or putrescine. The finding that putrescine could fully rescue the deleterious consequences of a Δarg null mutation in both promastigotes and axenic amastigotes authenticates that the sole function of ARG in L. donovani is to support polyamine biosynthesis, a result similar to previous findings obtained with promastigotes of Δarg null mutants of several cutaneous species of Leishmania (37)(38)(39). In contrast, addition of spermidine to the culture medium up to a concentration of 1 mM did not enable growth of Δarg promastigotes  Fig. 4, by incubating parasites in serial dilutions of ornithine or putrescine for 5 days. Proliferation was evaluated by assessing resazurin conversion as a measure of metabolic activity. Each value represents the means and standard deviations from three independent experiments set up in duplicate. The EC 50 s for L. mexicana Δarg and L. major ⌬arg mutants have been reported elsewhere (*, reference 37; **, reference 38). ( Fig. 3B). Because spermidine is efficiently transported into Leishmania (50-52), the observation that spermidine could not rescue the polyamine auxotrophy of the L. donovani Δarg mutant implies that putrescine is an essential metabolite for proliferation of L. donovani promastigotes. Coupled with previous findings that the L. donovani Δspdsyn mutant accumulates putrescine as a result of the genetic block but also display polyamine auxotrophy (35), our data establish for the first time that both of the polyamines found in L. donovani, putrescine and spermidine, are each of their own accord critical for the survival and growth of the parasite. That spermidine by itself is indispensable for parasite proliferation is unsurprising, since the polyamine is a component of two vital downstream reactions. Spermidine is a constituent of trypanothione, the thiol reductant in Leishmania and related trypanosomatids that replaces the role of glutathione in mammalian systems (17)(18)(19)(20), and hypusine, an unusual amino acid that is the result of a posttranslational modification in eukaryotic translation initiation factor 5A and is essential for its proper function (14)(15)(16). In contrast to spermidine, a specific function for putrescine is unknown other than that the diamine serves as a general aliphatic polycation. Finally, the fact that supplementation of the growth medium with spermidine cannot rescue a Δarg deficiency in L. donovani (Fig.  3B), together with our previous observation that spermine supplementation cannot rescue a Δodc deficiency (33), proves the absence of a back-conversion pathway from spermine to spermidine and putrescine in Leishmania that is operative in mammalian cells (53,54). The basis for why higher concentrations of ornithine than putrescine are required to rescue the phenotypic consequences of a Δarg deficiency in both promastigotes and axenic amastigotes is not known. Ornithine and putrescine uptake rates into L. mexicana promastigotes (37) are roughly equivalent, although an analysis of ornithine and putrescine uptake into L. donovani promastigotes has not been undertaken. Because L. mexicana promastigotes contain vastly more ornithine (ϳ130 nmol/10 7 parasites) than putrescine (ϳ2.5 nmol arginine/10 7 parasites) (37), we speculate that ODC is the rate-limiting enzyme in polyamine biosynthesis, converting only small amounts of stockpiled ornithine into putrescine. Our finding that overexpression of ODC in the Δarg background reduced the levels of ornithine required to enable growth (Fig. 5) Statistical analyses were performed using the paired t test and revealed significant differences in liver infectivity between wild-type and Δarg mutant 1 parasites, as well as between the wild type and the Δarg mutant 2 strain (denoted with an asterisk). supports this conjecture. Little is known about the regulation of polyamine synthesis in trypanosomatids, although it has long been established that mammalian cells utilize an intricate regulatory system to ensure low and stable polyamine levels (55)(56)(57)(58). Our supposition that ODC is a rate-limiting enzyme in Leishmania together with the recent discovery of the controlled expression of prozyme, necessary for the activity of ADOMETDC, in African trypanosomes (24,(59)(60)(61)(62) suggests that polyamine biosynthesis is also regulated in trypanosomatids.
The cellular localization of ARG does not appear to be crucial for its function in L. donovani promastigotes or axenic amastigotes, as add-back parasites expressing glycosomal or cytosolic mislocalized ARG both exhibit polyamine prototrophy (Fig. 3). Similar results have also been observed in L. mexicana and L. amazonensis promastigotes (37,39). However, mislocalization of ARG to the cytosol in both L. mexicana and L. amazonensis, unlike L. donovani (Fig. 6), dramatically reduced parasite burdens, implying that the glycosomal milieu, although not critical for ARG function in promastigotes, is indispensable for proper ARG function in amastigotes of the two species (39,46). Why the cellular environment appears to be important for ARG function in intracellular L. mexicana and L. amazonensis amastigotes but not extracellular parasites is unclear but could be due to limiting availability of arginine in the host cell (63).
There exists an extraordinary positional variation in the impacts that specific lesions in polyamine biosynthesis exert on L. donovani infection capacity. Among the three null L. donovani mutants deficient in polyamine biosynthesis that have been tested in mice, the Δodc deletion effectively obliterates the capacity of the parasite to trigger a mouse infection (42), the Δspdsyn lesion shows reductions of three and two orders of magnitude in parasite burdens in liver and spleen, respectively (43), and the Δarg mutation confers a statistically significant 10-fold reduction in liver parasite loads but does not compromise splenic parasite numbers (Fig. 6). All determinations of parasite loads in these experiments were performed 4 weeks postinfection, the interval in which both hepatic and splenic infections were maximal in our system (see Fig. S2 in the supplemental material). It should be noted, however, that although it has long been known that hepatic infections of visceralizing Leishmania strains are self-limiting in mice, splenic parasite numbers in murine infections tend to increase throughout month-long infections (64)(65)(66)(67)(68). The basis for the discrepancies in the time courses of splenic infections between our laboratory and others (64-68) is unclear and was not pursued further, because the objective was to compare the impacts of specific genetic defects in the polyamine pathway upon infection.
Although it is problematic to compare infectivity phenotypes between visceral and cutaneous strains of Leishmania, a similar statistically meaningful reduction in parasite virulence, as determined by footpad lesion size, was also observed with Δarg L. mexicana, L. major, and L. amazonensis mutants, although all three cutaneous Δarg strains retained significant infection capacity (38)(39)(40)(41). The growth deficits of the Δodc and Δspdsyn L. donovani mutants cannot be due to a lack of polyamine transport capacity, because axenic amastigotes are able to efficiently import both putrescine and spermidine, the products of ODC and SPDSYN, respectively (50,51). Proficient putrescine uptake is also functionally demonstrated by the observation that Δarg axenic amastigotes grew well in medium supplemented with micromolar concentrations of putrescine (Fig. 4). Most likely, the divergence in the infection capacities among the Δarg, Δodc, and Δspdsyn L. donovani mutants reflects the relative ornithine and polyamine pools within the phagolysosome of infected macrophages. It is logical to propose that the parasite has access to ample ornithine to bypass a Δarg lesion, essentially no putrescine to evade the consequences of a Δodc deficiency, and intermediate spermidine pools that enable partial bypass of a Δspdsyn mutation. It is intriguing that Δarg axenic amastigotes require much higher concentrations of ornithine than putrescine for optimal growth (Fig. 4B), thus the physiological concentrations of ornithine would have to be high in the phagolysosome to circumvent the arg deletion. The model of putrescine and relative spermidine deficiencies in the host cell is supported by the observation that macrophages rapidly convert these smaller polyamines to spermine (69). Previous Leishmania infection models have proposed that an increased host ARGI activity in infected macrophages provides polyamines for salvage and parasite proliferation (70,71); however, our findings imply a refined model where ornithine, rather than polyamines, is available for salvage by the parasite (Fig. 7). This hypothesis can be tested in the future by measuring nutrient concentrations in macrophages and, ideally, phagolysosomes.

MATERIALS AND METHODS
Materials, chemicals, and reagents. Resazurin, G418, hygromycin, putrescine, and ornithine were purchased from VWR International (Radnor, PA). Phleomycin was procured from Thermo Fisher Scientific Cell lines, cell culture, and assessment of growth phenotypes. All genetically manipulated parasites were derived from the wild-type LdBob strain of L. donovani (72) that was originally obtained from Stephen M. Beverley (Washington University, St. Louis, MO). Wild-type LdBob and genetically manipulated derivative strains were routinely cycled between the promastigote and axenic amastigote stages at 26°C, pH 7.4, and 37°C, pH 5.5, respectively, using previously reported cell culture conditions, in order to maintain infectivity (73,74). Promastigotes were incubated in completely defined DME-L (75), while amastigotes were grown in an M199-based medium (73). Wild-type, Δarg[ARG], and Δarg[argΔAKL] parasites were routinely cultured in medium with no ornithine or putrescine supplementation, while the Δarg cells were maintained in medium supplemented with 50 g/ml hygromycin, 50 g/ml phleomycin, and 100 M putrescine.
Growth phenotypes were established by incubating wild-type and mutant L. donovani promastigotes or axenic amastigotes seeded in a volume of 100 l in 96-well plates at a density of 5 ϫ 10 5 /ml in the absence or presence of various concentrations of ornithine, putrescine, or spermidine as specified. Growth experiments with spermidine added to the medium necessitated replacement of fetal bovine serum with chicken serum to avoid polyamine oxidase-mediated toxicity (35,76). After 5 days, 10 l of 250 M resazurin was added to each well to assess cell density, and plates were incubated for an  (88,89), while promastigotes are able to synthesize ornithine and polyamines endogenously. Putrescine and spermidine, but not ornithine, are essential for parasite growth and survival, and gene deletion mutants depend on polyamine supplementation of the media. (B) L. donovani amastigotes reside inside the phagolysosome (shaded gray) of host macrophages. Our data support a model wherein ornithine salvage is hypothesized to be efficient, while spermidine and especially putrescine salvage pools are too limited to support robust infections. Although amastigotes have putrescine and spermidine transporters (50,51), both polyamines have been shown to be rapidly metabolized to spermine in macrophages (69) and thus are not accessible to the intracellular parasite. Spermine cannot meet the polyamine requirements of Leishmania parasites (33). additional 4 h. Conversion of resazurin to resorufin was evaluated on a BioTek Synergy plate reader by monitoring fluorescence (excitation wavelength, 579 nm; emission wavelength, 584 nm). Graphs were prepared using GraphPad Prism version 4.0 or 6.0f for Mac (GraphPad Software, La Jolla, CA).
Creation of ⌬arg parasites. The Δarg knockouts were generated in the LdBob background by double targeted gene replacement (77,78). To construct drug resistance cassettes for the replacement of ARG, the 5= ARG flanking region was amplified by PCR using the following primers containing HindIII and SalI restriction sites (underlined): forward primer, GCATAAGCTTGATCCCTCAAGCACATTGTGAAG; reverse primer, GCATGTCGACTGACCCTGTCACCACCAGAC. The resulting PCR product was first subcloned into the pCR 2.1-TOPO vector, excised with HindIII and SalI endonucleases, and inserted into the HindIII and SalI sites of pX63-HYG and pX63-PHLEO vectors to generate pX63-HYG-5=F and pX63-PHLEO-5=F, respectively. The 3= flanking region was then amplified by PCR using the following primers containing SmaI and BglII sites (underlined): forward primer, GGGCCCGCGCATGTGTTCCTTCCAAG; reverse primer, GCATAGATCTGGGAGATGTCAGAGGAGAGGAGAG. The PCR product was first ligated into the pCR 2.1-TOPO vector and then inserted into the SmaI/BglII site of pX63-HYG-5=F to generate pX63-HYG-Δarg. Because of the presence of an SmaI site within the PHLEO coding region of pX63-PHLEO, a different cloning strategy was employed to insert the 3=-flanking region of ARG into pX63-PHLEO-5=F. The 3=-flanking region was amplified by PCR using a forward primer containing the BamHI restriction site, GCATGGATCCGCGCATGTGTTCCTTCCAAG, and the above-noted reverse primer containing the BglII site. The resulting PCR product was subcloned into the pCR 2.1-TOPO vector and then introduced into the BamHI/BglII sites of pX63-PHLEO-5=F to generate pX63-PHLEO-Δarg. The correct orientations of the 5=and 3=-flanking regions within the gene-targeting plasmids were verified by limited nucleotide sequencing of the insert junctions.
The plasmids pX63-HYG-Δarg and pX63-PHLEO-Δarg were digested with HindIIII and BglII to liberate 6-kb and 5.5-kb linear fragments, designated X63-HYG-Δarg and X63-PHLEO-Δarg, respectively. X63-HYG-Δarg and X63-PHLEO-Δarg were purified from DNA agarose gels using the Wizard SV gel and PCR clean-up system according to the manufacturer's protocol and then transfected into parasites using standard electroporation conditions (79). Two heterozygous cell lines, ARG/Δarg::HYG and ARG/Δarg:: PHLEO, were generated first and selected in semisolid agar plates containing 50 g/ml hygromycin or 50 g/ml phleomycin, respectively. The genotypes of the two heterozygous cell lines were confirmed by PCR and then subjected to a second round of transfection. The ARG/Δarg::HYG heterozygotes were transfected with the PHLEO-Δarg fragment, and the ARG/Δarg::PHLEO heterozygotes were transfected with the X63-HYG-Δarg fragment. Potential homozygous Δarg::HYG/Δarg::PHLEO (Δarg) knockouts were selected in semisolid agar plates containing 50 g/ml hygromycin, 50 g/ml phleomycin, and 200 M putrescine. Several clones were picked and parasite cultures expanded. The homologous gene replacements were confirmed by PCR (Fig. 1). Two Δarg knockout clones, derived from the hygromycin-resistant heterozygotes (here designated ARG/arg), were picked for further analysis.
Complementation of ⌬arg cell lines. To generate a stably complemented add-back line, ARG was introduced into Δarg parasites by homologous recombination at the ribosomal locus (80). Two different constructs were generated, one containing the wild-type ARG sequence (Δarg[ARG]) and one containing a mutated version lacking the C-terminal AKL tripeptide (Δarg[argΔAKL]), an archetypical signal that targets proteins for glycosomal localization in Leishmania (47,48,81). The ARG coding region was amplified with the following primers, each containing an SfiI restriction site, GGCCNNNNNGGCC (underlined): forward primer GAGGCCACCTGGGCCTCATCATGGAGCACGTGCAG and reverse primer GAGGCCA GCCCGGCCCGCACACACACATCTACAGTTTGG. A mutated version of the ARG coding region was constructed by using a reverse primer where the nucleotides encoding the C-terminal AKL triad were replaced with a stop codon. The following reverse primer was used, with the SfiI site underlined and the stop codon in boldface: GAGGCCAGCCCGGCCCTAGCTCTTACGTGGGGTGTAAAGAAGTG. The SfiI fragments were subcloned into pRP-M (80), yielding pRP-M-ARG (pRP-ARG) and pRP-M-argΔAKL (pRP-argΔAKL). The plasmids were linearized by digestion with AvrII and purified with the Wizard SV gel and PCR clean-up system (Promega) according to the manufacturer's protocol. Δarg parasites were transfected with these constructs using standard electroporation conditions (79), and integrated constructs were selected in bulk culture in 20 g/ml G418. Clonal isolates from the bulk cultures were obtained by plating on semisolid agar plates. Individual colonies were picked and expanded in medium containing 20 g/ml G418. Genotypes and phenotypes of the Δarg[ARG] and Δarg[argΔAKL] parasites were verified by PCR, Western blot analysis, and localization studies.
ODC overproducers in the ⌬arg background. The construction of the ODC overexpression vector pSNBR [ODC] has been described previously (82). The Δarg[ODC] parasites were generated by transfecting Δarg parasites with the pSNBR[ODC] plasmid using standard electroporation conditions (79). Parasites harboring the episomal construct were selected with 100 g/ml G418 in bulk culture and the correct phenotype verified by Western blot analysis.
PCR and Western blotting. Genomic DNA from the wild type, ARG/arg heterozygotes, two independent Δarg clones, and the Δarg[ARG] and Δarg[argΔAKL] add-backs was prepared for Southern blot analysis using the DNeasy kit (Qiagen Inc., Valencia, CA) according to the manufacturer's protocol. The following primers were utilized to amplify a 900-bp ARG coding region fragment: ATGGAGCACGTGCA GCAG and CACGTGCGATTCTGTAGCG. Additional primers were designed to detect the replacement of the ARG gene with either the PHLEO or HYG drug resistance cassette. The forward primer (CCATCGTG CACGTCTATGAG) was designed from sequence upstream of the deletion construct, and reverse primers were selected from the PHLEO (GAAGTCGTCCTCCACGAAGT) or HYG (CCCGCAGGACATATCCAC) coding region. The presence and size of the PCR products were determined on ethidium bromide-stained agarose gels.