Leishmania major and all other parasitic protozoa are unable to synthesize purines de novo and are therefore reliant upon uptake of preformed purines from their hosts via nucleobase and nucleoside transporters. L. major expresses two nucleobase permeases, NT3 that is a high affinity transporter for purine nucleobases and NT4 that is a low affinity transporter for adenine. nt3(–/–) null mutant promastigotes were unable to replicate in medium containing 10 μM hypoxanthine, guanine, or xanthine and replicated slowly in 10 μM adenine due to residual low affinity uptake of that purine. The NT3 transporter mediated the uptake of the anti-leishmanial drug allopurinol, and the nt3(–/–) mutants were resistant to killing by this drug. Expression of the NT3 permease was profoundly downregulated at the protein but not the mRNA level in stationary phase compared with logarithmic phase promastigotes. The nt4(–/–) null mutant was quantitatively impaired in survival within murine bone marrow-derived macrophages. Extensive efforts to generate an nt3(–/–)/nt4(–/–) dual null mutant were not successful, suggesting that one of the two nucleobase permeases must be retained for robust growth of the parasite. The phenotypes of these null mutants underscore the importance of purine nucleobase transporters in the Leishmania life cycle and pharmacology.
Leishmaniasis is a parasitic disease endemic to tropical and subtropical regions of the world, with an estimated prevalence of 12 million cases worldwide (World Health Organization, 2007). Clinical manifestations of leishmaniasis depend upon the causative species of Leishmania and the immune status of the host and range from self-healing skin lesions to mucocutaneous infection to a lethal systemic visceral form of the disease (Murray et al., 2005). Leishmania parasites are single cell protozoa that live as motile flagellated promastigotes within the gut of the sandfly vector and as non-motile amastigotes within phagolysosomes of mammalian macrophages. Extracellular promastigotes must import nutrients from the sandfly gut, whereas intracellular amastigotes must obtain nutrients from the lumen of the macrophage phagolysosome.
Purine salvage in protozoan parasites has elicited considerable interest, because none of those organisms is able to synthesize purines de novo (Carter et al., 2003) and thus must acquire these essential nutrients as either nucleosides, e.g. adenosine, or nucleobases, e.g. adenine, from their hosts (Landfear et al., 2004). Purine permeases mediate the first step in the salvage pathway, uptake of these nutrients. Furthermore these transporters are of pharmacological importance, because purine analogues such as allopurinol that are selectively toxic for the parasites enter these organisms through purine transporters (Landfear et al., 2004). All protozoan purine transporters characterized to date are members of the Equilibrative Nucleoside Transporter (ENT) family [SLC29 family in the Human Genome Organization Database (Hediger, 2004), or 2.A.57 family in the Transporter Classification Database (TCDB, 2007)]. The Leishmania major genome (Ivens et al., 2005) encodes four ENT permeases designated NT1, NT2, NT3 and NT4 (Landfear et al., 2004). Functional expression of cloned transporter genes has revealed substrate specificities for three of these permeases, NT1 (adenosine/pyrimidine nucleosides) (Vasudevan et al., 1998), NT2 (inosine/guanosine/xanthosine) (Carter et al., 2000) and NT3 (the purine nucleobases adenine, guanine, hypoxanthine and xanthine) (Sanchez et al., 2004). In addition, NT1 mediates the uptake of the toxic adenosine analogue tubercidin and NT2 the toxic inosine analogue formycin B (Iovannisci et al., 1984), and in whole cells a nucleobase transport activity, designated LmNBT1 (Al-Salabi et al., 2003), is responsible for permeation of the pyrazolopyrimidine analogue (Marr, 1991) of hypoxanthine, allopurinol, that is employed as an anti-leishmanial drug (Martinez and Marr, 1992). Nonetheless, the biological roles within the intact parasite of individual nucleobase transporters in parasite biology and drug sensitivity remain to be determined.
In this study we have examined the functional and pharmacological roles of the NT3 purine nucleobase permease in viability of both promastigotes and amastigotes by generating an nt3(–/–) null mutant (see Experimental procedures for explanation of genetic nomenclature) of L. major. We have also investigated the previously uncharacterized NT4 transporter and the nt4(–/–) null mutant and conclude that this permease exhibits low affinity adenine transport function. In addition we have demonstrated that expression of the NT3 permease is strongly regulated during the parasite growth cycle at the post-mRNA level. Furthermore, while it was possible to generate the nt3(–/–) and nt4(–/–) null mutants and the nt3(–/–)/nt4(+/–) or nt3(+/–)/nt4(–/–) null/heterozygote deletion mutants, we have not been able to obtain the nt3(–/–)/nt4(–/–) dual null mutant despite extensive attempts of targeted gene replacement performed under different experimental conditions. Although neither nucleobase permease is essential for the parasites, deletion of both the NT3 and NT4 genes together, even in the presence of intact purine nucleoside transporter genes NT1 and NT2, probably impairs parasite viability and might be a lethal event. These observations underscore the role of Leishmania nucleobase transporters in scavenging essential purines from the insect and vertebrate hosts.
Targeted gene disruption of NT3
To determine the biological function of the L. major NT3 transporter, a null mutant was generated by two sequential rounds of targeted gene replacement using HYG and BLEO antibiotic resistance markers (Fig. 1A). Southern blot analyses of genomic DNA from the nt3(–/–) cell line showed the predicted Asp718 fragments of 4.3 and 11 kb following hybridization with a probe from the 3′ flanking region of NT3 (Fig. 1B). Furthermore, hybridization of nt3(–/–) genomic DNA with the NT3 open reading frame (ORF) on Southern blots demonstrated the absence of this gene, and polymerase chain reaction (PCR) analysis employing oligonucleotide primers outside the integration matched with primers within each integration construct generated PCR fragments of the predicted sizes, thus further confirming that the nt3(–/–) null mutant had been generated (data not shown).
Functional characterization of nt3(–/–) null mutant
Previously we have demonstrated that NT3 functions as a high affinity purine nucleobase permease when expressed in Xenopus oocytes (Sanchez et al., 2004). Furthermore, we have obtained similar results when NT3 was expressed in a nucleobase transport-deficient mutant of Saccharomyces cerevisiae (data not shown). Nevertheless, to assess the role of NT3 in nucleobase transport within the intact parasite, we analysed the uptake of adenine (Ade), guanine (Gua), hypoxanthine (Hyp) and xanthine (Xan) in the nt3(–/–) null mutant (Fig. 2). Whereas wild-type cells (filled circles) were able to take up all purine nucleobases, the nt3(–/–) null mutant promastigotes (filled triangles) were impaired in the uptake of the four purine nucleobases. Nonetheless, a residual uptake for adenine and possibly for xanthine was observed in nt3(–/–) null mutant cells, although for unknown reasons the residual uptake rates varied in multiple experiments and for xanthine were sometimes not significantly above the background level of binding observed in formaldehyde-treated cells (e.g. see Fig. 7 below). Complementation of the nt3(–/–) cells with an episome expressing NT3 (nt3(–/–)[pNT3], solid squares) restored wild-type levels of uptake for all four purine nucleobases. Therefore, NT3 is the principal transporter for purine nucleobases in L. major, but there is at least one other permease that mediates the uptake of adenine.
Allopurinol is an isomer of hypoxanthine and an anti-leishmanial drug (Das et al., 2001; Momeni et al., 2002) that is thought to kill the parasite by metabolism to the nucleoside triphosphate followed by incorporation into RNA (Nelson et al., 1979). To determine whether NT3 is the allopurinol transporter, we monitored uptake of [3H]-allopurinol by wild-type and nt3(–/–) null mutant parasites (Fig. 3A). Whereas wild-type parasites (filled circles) took up 1 μM allopurinol, uptake of the drug above the background level detected in formaldehyde-treated cells was almost abolished in nt3(–/–) parasites (Fig. 3A, filled triangles). Allopurinol transport activity was restored to nt3(–/–) mutants that were transfected with an episomal expression vector encompassing the NT3 gene (Fig. 3A, filled squares). To confirm these results, we also expressed NT3 in Xenopus oocytes. Indeed, oocytes injected with NT3 cRNA mediated the uptake of allopurinol significantly above that observed in oocytes injected with water (Fig. 3B).
The preceding results suggested that the nt3(–/–) null mutant might be relatively resistant to killing by allopurinol. When incubated with graded concentrations of allopurinol, nt3(–/–) null mutants were 10 to 20-fold more resistant (EC50 = 70.1 μM) than wild type (EC50 = 7.3 μM) or than nt3(–/–)[pNT3] parasites (EC50 = 3.8 μM) (Fig. 3C). Hence, NT3 is the allopurinol transporter and a determinant of drug sensitivity.
Ability of nt3(–/–) null mutants to grow on purine nucleobases
Because NT3 is the major nucleobase permease, we monitored the ability of nt3(–/–) null mutants and nt3(–/–)[pNT3] parasites to grow in RPMI medium supplemented with 10 μM adenine, guanine, hypoxanthine, and xanthine (Fig. 4). The growth of nt3(–/–) mutants was dramatically impaired in all four nucleobases and was similar to wild-type parasites cultured in the absence of any purine source. However, the nt3(–/–) null mutants did grow at a slow rate on adenine (see growth between days 10 and 14, Fig. 4, Ade), consistent with the residual adenine uptake in this mutant (Fig. 2, Ade). In contrast, complementation of the null mutant with an NT3–expressing episome (nt3(–/–)[pNT3], Fig. 4, filled circles) restored growth on all four purines to levels observed with wild-type parasites (data for wild type not shown to avoid overcrowding the figure). Hence NT3 is required for growth when nucleobases serve as the purine source at low micromolar concentrations similar to the levels present in human serum (Hammond and Gutteridge, 1984). However, nt3(–/–) null mutants grew at higher non-biological concentrations of purine nucleobases (e.g. 100 μM xanthine) or when purine nucleosides that permeate through the NT1 or NT2 transporters were present (data not shown).
Survival of nt3(–/–) null mutant in sandflies
To interrogate the role of the NT3 transporter in promastigotes within their natural biological environment, Phlebotomus duboscqi sandflies were fed on blood meals containing wild-type, nt3(–/–) and nt3(–/–)[pNT3] parasites, and the numbers of parasites present within the insect guts were determined between 2 and 14 days post feeding (supplemental Fig. S1). While there was considerable variation in the number of parasites present in different flies fed upon the same parasite lines, the average parasite burden in flies infected with nt3(–/–) null mutant was significantly lower (P < 0.05, asterisks and brackets in Fig. S1) than that of flies infected with wild-type parasites on days 10 and 14. Complementation of the nt3(–/–) null mutant with episomally expressed NT3 (nt3(–/–)[pNT3]) restored the parasite burden to levels similar to wild-type parasites on day 10 but not 14. Hence deletion of the NT3 gene impairs but does not abrogate parasite viability in the sandfly. Similar results were obtained in another independent experiment.
Survival of nt3(–/–) null mutant as amastigotes inside macrophages
To determine the role of NT3 in the intracellular amastigotes stage, bone marrow-derived macrophages from BALB/c mice were infected with the same three parasite lines employed in the sandfly infections, and intracellular parasites were counted at 4 h, 3 days and 5 days post infection. The nt3(–/–) null mutants were found to be as efficient at infecting and surviving within macrophages as wild-type parasites at all time points examined (Fig. 5A and B). The stimulation of macrophages with lipopolysaccharide and interferon-γ as reported previously (Tovar et al., 1998a), or application of different multiplicities of infection (Spath et al., 2000), did not induce any significant difference of invasiveness or survival of nt3(–/–) null mutants (data not shown). Hence the NT3 permease does not appear to play a crucial role in transformation into or survival of amastigotes, suggesting that other purine permeases such as the NT1 and NT2 nucleoside transporters may be able to support purine salvage in this life cycle stage by importing purine nucleosides present within the macrophage phagolysosomes.
Effect of cell density on nucleobase transport
In initial experiments in wild-type parasites, we noted that hypoxanthine uptake was greatly reduced in stationary phase compared with logarithmic growth phase promastigotes. To investigate this phenomenon systematically, transport of [3H]-hypoxanthine was performed over the growth curve from 2.6 × 106 to 6.8 × 107 cells ml−1 (Fig. 6A). Above a density of ∼2 × 107 cells ml−1, uptake rates for hypoxanthine decreased dramatically as a function of cell density reaching very low levels in stationary phase parasites. In other experiments, the increase in uptake between days 1 and 2 was not always observed and there was a monotonic decrease of uptake with increasing cell density. In contrast, although the level of NT3 mRNA fluctuated somewhat in different samples, it did not decrease systematically as a function of cell density (Fig. 6B) in a way that reflected the pronounced decrease of hypoxanthine uptake. To investigate whether downregulation of hypoxanthine uptake was due to reduction in the level of NT3 permease as cells obtain high density, we expressed a fusion protein in wild-type parasites in which green fluorescent protein (GFP) was attached to the amino terminus of NT3 (GFP-NT3). Immunoblots (Fig. 6C) revealed that while GFP-NT3 was robustly expressed at low cell density (1.8 × 106 cells ml−1, day 1), increasing densities resulted in decreasing levels of fusion protein representing an approximately ninefold reduction between days 1 and 5. Soluble GFP (28 kDa), probably generated in vivo or in vitro by cleavage of the GFP-NT3 fusion protein, also decreased somewhat with increasing cell density. Equal loading of the protein fraction was verified by probing the blot with an antibody against α-tubulin. Furthermore, we were able to corroborate by direct GFP fluorescence that GFP-NT3 was downregulated in the membrane of promastigotes as the cell density increased, while α-tubulin remained present at unaltered levels (Fig. 6D). Similar results showing downregulation of NT3 were observed for GFP-NT3 expressed in the nt3(–/–) null mutant background (data not shown).
Cumulatively, these results indicate that NT3 protein, but not NT3 mRNA, is strongly downregulated as promastigotes reach the stationary phase of their growth cycle. A similar downregulation of transport activity has been observed for transport of adenosine by NT1 (R. Valdés, D. Rodriguez-Contreras and S. Landfear, unpublished) and transport of inosine and guanosine by NT2 (N. Carter and B. Ullman, unpublished). However, a thorough examination of the regulation of the NT1 and NT2 permeases by cell density has not been performed to date.
Functional characterization of the NT4 transporter
A new member of the ENT family was identified in the L. major genome database using the NT3 sequence as query (Landfear et al., 2004). The entry denominated NT4 (LmjF11.0550), whose amino acid sequence revealed 33% identity to NT3, encodes a 59.8 kDa protein with 11 predicted transmembrane domains. Initial expression of NT4 in Xenopus oocytes suggested that NT4 might be a nucleobase transporter (data not shown). The functional characterization of NT4 was further performed in the nt3(–/–) null mutant, because this mutant represents a homologous system with greatly reduced purine nucleobase transport. As shown in Fig. 7, the expression of NT4 from an episome in the nt3(–/–) null mutant (nt3(–/–)[pNT4], filled squares) restored the uptake of adenine but not guanine, hypoxanthine or xanthine. While there is residual adenine uptake in the nt3(–/–) null mutants, the expression of NT4 was able to increase the levels of adenine uptake to that of wild-type parasites (Fig. 7, Ade). These results were reproducible in three independent experiments. These experiments suggest that NT4 possesses adenine transport activity that may be responsible, at least in part, for the residual adenine uptake observed in the nt3(–/–) null mutant.
To further confirm adenine uptake by NT4, we measured uptake of increasing concentrations of [3H]-adenine (0–1000 μm) by both the nt3(–/–) null mutant and the nt3(–/–)[pNT4] complemented strain. While the rate of uptake of adenine increased with increasing substrate concentrations, it was not possible to attain maximal velocity at any adenine concentration. Non-saturable adenine uptake was also observed when NT4 was expressed in Xenopus oocytes (data not shown). These results suggest that NT4 is a low affinity adenine permease.
Targeted disruption of NT4 gene and phenotypic characterization
To determine the biological role of the NT4 permease in L. major, a nt4(–/–) null mutant was generated by sequential targeted gene replacement using SAT and BSD antibiotic resistance markers flanked by sequences 5′ and 3′ of the NT4 ORF (Fig. 1C). Southern blot analysis of genomic DNA from the doubly transfected line showed that the ∼5 kb PvuII band that hybridized to the NT4 ORF in wild-type and nt4(+/–) heterozygous deletion parasites was absent after integration of the two selectable marker constructs (Fig. 1D), confirming that this line was a nt4(–/–) null mutant. Because this null mutant retained intact NT3 genes that encode the high affinity nucleobase permease, the mutant was not impaired in the ability to take up purine nucleobases. To ascertain whether these mutants were impaired in growth or infectivity as amastigotes, BALB/c bone marrow-derived macrophages were infected with wild type, nt4(–/–) null mutants and nt4(–/–)[pNT4] complemented null mutants and intracellular parasites were counted at 4 h, 3 days and 5 days post infection. All lines produced similar numbers of intracellular parasites at 4 h (∼75 parasites per 100 macrophages, Fig. 5C), indicating they were equally invasive, but the nt4(–/–) null mutant exhibited an ∼50% reduction in intracellular amastigotes at days 3 (not shown) and 5 (Fig. 5D) compared with wild-type parasites or complemented mutants. These results were reproducible in three independent experiments. Hence the nt4(–/–) null mutants were viable as both promastigotes and amastigotes but did exhibit a quantitative reduction in survival inside macrophages.
Examination of the nt4(–/–) null mutant by phase contrast microscopy suggested that these parasites were somewhat larger than either wild-type or the complemented nt4(–/–)[pNT4] promastigotes. Indeed protein determination revealed that the nt4(–/–) promastigotes contained 1.71 ± 0.03 times as much protein per cell as wild-type promastigotes (n = 3) but that this relative difference decreased to 1.26 ± 0.06 for nt4(–/–)[pNT4] promastigotes. The student t-test confirmed that these values represented a statistically significant difference (P < 0.05) between the nt4(–/–) null mutant and either wild-type parasites or complemented mutants but that the difference between wild-type and complemented mutants was not significant.
Expression of NT4 permease as function of cell density
To determine whether NT4 was regulated during growth of promastigotes to high density in a manner similar to NT3, uptake of 250 μM adenine was examined in the nt3(–/–) null mutant (Fig. 8A). In this cell line the high affinity adenine uptake due to NT3 will be absent, allowing selective monitoring of low affinity transport. In contrast to the dramatic reduction in uptake of hypoxanthine observed in wild-type parasites (Fig. 6A), low affinity adenine uptake increased approximately threefold in the nt3(–/–) at stationary phase.
To examine expression of NT4 mRNA, total RNA was isolated from wild-type parasites at increasing cell densities. Northern blots probed with the NT4 ORF revealed that NT4 mRNA was downregulated as a function of cell density (Fig. 8B), decreasing ∼2.5-fold from day 1 to day 5 of growth. Expression of GFP-NT4 protein in a nt4(–/–) background was monitored using immunoblots probed with anti-GFP antibody (Fig. 8C) and by fluorescence microscopy (Fig. 8D). GFP-NT4 protein experienced an approximately threefold decrease in level on day 5 compared with day 1 as determined by densitometry of the Western blot, and the fluorescence signal observed in stationary phase cells was also decreased.
Attempts to generate a nt3(–/–)/nt4(–/–) dual null mutant
As promastigotes and amastigotes of both the nt3(–/–) and the nt4(–/–) null mutants were viable, an important question is whether a null mutant deficient in both nucleobase permeases would be impaired in either life cycle stage. To address this question, we attempted to generated the dual null mutant beginning with both the nt3(–/–) and nt4(–/–) mutants and performing targeted replacement of the remaining nucleobase transporter gene. Using the constructs described above, we were able to obtain both the nt3(–/–)/nt4(+/–) and nt3(+/–)/nt4(–/–) null/heterozygous deletion mutants, as determined by PCR analysis and Southern blots of the transfectants. However, multiple attempts to target the remaining NT3 or NT4 allele resulted in transfected lines, 133 of which were analysed, that had correctly integrated the final targeting construct but had nonetheless retained a copy of the wild-type gene and were thus of a nt3(–/–)/nt4(+/–/–) or nt3(+/–/–)/nt4(–/–) genotype. Figure 9 shows results from one such transformation in which the two alleles of the NT4 gene have been correctly replaced in an nt3(–/–) background, first by a BSD (lane 3) and then by a SAT (lane 4) targeting construct. PCRs of genomic DNA, employing primers within the resistance marker and from the flanking regions of the NT4 ORF but outside the targeting construct, produced the predicted amplification products confirming that targeted replacement of both NT4 alleles had been accomplished (Fig. 9A). However, a copy of the NT4 gene was retained in the final transformant (lane 4) as determined by genomic Southern blot (Fig. 9B). These gene deletion experiments were performed under a variety of experimental conditions to provide purines that should support parasite growth even in the absence of functional nucleobase transporters. Thus, selective medium in various experiments contained 100 μM xanthine that supports growth in the absence of the either the NT3 or NT4 genes, possibly by passive diffusion across the plasma membrane, or 100 μm adenosine or inosine that permeates through the intact NT1 and NT2 transporters respectively. In addition, in some experiments we provided daily supplementation with 100 μM inosine or xanthosine in liquid medium followed by thin layer chromatography of the medium to confirm the continued presence of purine nucleosides during the selection process, and clonal transformants were obtained by dilution cloning. None of these supplementation protocols generated an nt3(–/–)/nt4(–/–) dual null mutant.
The difficulty encountered generating the dual null mutant suggested either that the parasites require one copy of either NT3 or NT4 for viability or that they at least suffer a significant disadvantage in viability if both genes are deleted. To further address this issue, we transfected the nt3(–/–)/nt4(+/–) promastigotes with an episomal copy of the NT3 or NT4 gene and subsequently attempted to target for deletion the remaining chromosomal allele. Potentially, such experiments would produce a dual mutant maintaining an episomal NT3 or NT4 allele, and subsequent efforts to cure the episome would help to determine whether or not a dual null mutant is viable in the absence of complementation. However, all transfectants derived from such complemented mutants also retained a genomic copy of the final targeted gene, and we were thus unable to generate a dual null mutant even in the presence of a complementing episome, despite the fact that the complementing episome restored transport function.
Purine transporters in Leishmania parasites
One remarkable distinction between Leishmania parasites and the vertebrates that serve as their hosts is the inability of these protozoa to synthesize the purine ring de novo and their consequent reliance upon purine salvage from their hosts. There is considerable interest in exploiting this biochemical disparity by developing novel therapies based upon either inhibition of purine salvage by the parasite or development of subversive substrates for purine salvage, often purine analogues such as allopurinol, that exhibit selective toxicity for the parasite (Martinez and Marr, 1992). Purine nucleoside and nucleobase transporters serve a crucial role in purine salvage by initiating purine uptake. Previously, the NT1 (adenosine/pyrimidine) (Vasudevan et al., 1998) and NT2 (inosine/guanosine/xanthosine) (Carter et al., 2000) nucleoside transporters have been identified, and nt1(–/–), nt2(–/–) and nt1(–/–)/nt2(–/–) null mutants have been generated in Leishmania donovani (Liu et al., 2006) and demonstrated to be viable despite ablation of the cognate nucleoside transport activities. More recently, we have identified the NT3 purine nucleobase transporter (Sanchez et al., 2004) and the uncharacterized but related NT4 permease (Landfear et al., 2004) of L. major, thus completing the annotation of all ENT family members present within the Leishmania genome (Ivens et al., 2005). However, the role of these latter two transporters in the intact parasite has not been determined by direct genetic approaches. The study reported here has thus focused upon characterization of the NT3 and NT4 genes and their roles in the biology of L. major promastigotes and amastigotes.
The NT3 purine nucleobase transporter
Genetic ablation of both NT3 alleles was achieved by targeted gene replacement. Infection of P. duboscqi sandflies revealed that the nt3(–/–) null mutant was viable but significantly impaired in ability to sustain an infection within the insect gut compared with either wild-type or episomally NT3-complemented mutants. Despite the inherent variability from one fly to the next, these results suggest that nucleobases serve as significant purine sources within the insect vector but that adenine or purine nucleosides must be present in sufficient quantity to maintain a residual infection in the absence of the NT3 permease. In contrast, the ability of nt3(–/–) null mutants to invade and survive within mammalian macrophages does not appear to be impaired, suggesting that purine nucleosides and/or adenine may be present at sufficient levels to support survival of this mutant within the macrophage phagolysosome. NT3 mRNA is expressed in both promastigotes and amastigotes, as determined by quantitative real time PCR (data not shown), and uptake of nucleobases has been demonstrated in intact amastigotes of Leishmania mexicana (Al-Salabi and de Koning, 2005). The ability of nt3(–/–) mutants to survive within macrophages suggests that mutations impairing the NT3 permease, which also mediates the uptake of allopurinol, could lead to allopurinol resistance in a clinical setting.
The NT4 transporter
We have also examined the biochemical and biological roles of the previously uncharacterized NT4 permease. Preliminary studies had demonstrated low affinity transport for purine nucleobases when this transporter was expressed in Xenopus oocytes, and the residual low affinity adenine transport activity present in nt3(–/–) null mutants suggested that this uptake might be attributable to the NT4 protein. Indeed, expression of an episomal copy of the NT4 gene in the nt3(–/–) null mutant was able to restore adenine transport levels to those observed in wild-type parasites (Fig. 7), although this activity was of low affinity similar to the residual activity in the uncomplemented null mutant and not like the high affinity activity present in wild-type parasites that is attributable to the NT3 permease. As most transporters characterized are saturable for their natural substrates, it is possible, perhaps likely, that NT4 has other higher affinity substrates that have not been identified yet. Indeed another member of the ENT family, hENT4 from humans, has a broad substrate specificity encompassing a range of organic cations but not nucleosides or nucleobases (Engel and Wang, 2005).
Although null mutants at the NT4 locus could be readily generated, the reduced survival of nt4(–/–) amastigotes within macrophages suggests that this permease is required for optimal viability in this life cycle stage. Possibly reduced survival within macrophages is related to a currently unknown transport activity of NT4.
Regulation of the NT3 and NT4 permeases as a function of cell density
Experiments reported in Fig. 6 reveal that the NT3 permease is strongly downregulated as promastigotes proceed along their growth curve to stationary phase but that this regulation occurs at the level of protein but not mRNA accumulation. A similar but less pronounced downregulation of GFP-NT4 protein is also apparent in stationary phase parasites (Fig. 8C), although NT4 mRNA is downregulated as well (Fig. 8B). Downregulation of a folate transporter in Leishmania infantum as a function of cell density, but not its cognate mRNA, has been observed previously (Richard et al., 2004) and the results reported here suggest that a variety of other transporters in Leishmania parasites, exemplified by NT3, may be similarly regulated at the post-mRNA level. It is also noteworthy that the NT1 adenosine/pyrimidine and the NT2 guanosine/inosine/xanthosine transport activities are strongly downregulated by inclusion of their cognate ligands in the culture medium (Seyfang and Landfear, 1999) and that a similar substrate-mediated downregulation of the myo-inositol transporter induces a corresponding reduction in the level of the permease protein, when quantified on immunoblots, but not its mRNA (A. Seyfang and S. Landfear, unpubl. result). It is possible that both substrate and growth curve modulation of nucleobase and nucleoside transporter expression share common molecular regulatory components.
Combined roles of NT3 and NT4 transporters
While it was possible to generate null mutants at either the NT3 or NT4 locus, and the null/heterozygous mutants nt3(–/–)/nt4(+/–) and nt3(+/–)/nt4(–/–) were viable, we were not successful in generating the nt3(–/–)/nt4(–/–) dual null mutant despite extensive efforts employing different purine sources. Over 100 transformants were analysed in these experiments, all of which had integrated the NT3 or NT4 deletion construct at the correct locus resulting in targeted replacement of the desired gene. Nonetheless, all these transformants retained a copy of the final targeted allele as revealed by either Southern blot or PCR analysis of genomic DNA. These results cannot be explained by innate aneuploidy of the wild-type parasites for the chromosome encompassing the NT3 or NT4 genes (e.g. three copies of the chromosome), because we were able to delete sequentially two alleles of each gene and generate null mutants for the individual genes (Fig. 1). The genomes of Leishmania parasites are plastic with regard to generation of multiple alleles (Beverley, 1991), and gene duplication can be detected in targeted gene replacement experiments when parasites are subjected to selective pressures that favour survival of cells that have retained a wild-type allele (Cruz et al., 1993). We have not attempted to characterize in detail the extra copies of the NT3 and NT4 genes that were present in the transformants obtained in the attempted gene disruption experiments.
The inability to generate a bona fide null mutant for a particular gene or genes accompanied by the consistent observation of a retained wild-type allele has been advanced as suggestive evidence that the targeted gene may be essential and that the null mutant would thus not be viable (Cruz et al., 1993; Mottram et al., 1996; Tovar et al., 1998b). Nonetheless, in the absence of more conclusive evidence it is not possible to rigorously demonstrate that a null mutant is not viable. An alternative strategy would be to complement a mutant, in this case one of the null/heterozygous mutants, with an episome encompassing the gene to be deleted and then to target the remaining chromosomal allele. The ability to cure the chromosomal null mutant of the episome, either by removing selective pressure for the drug resistance encoded by the episome or by applying negative selection to an episome carrying a negative selectable marker such as the herpes virus thymidine kinase (LeBowitz et al., 1992), would then constitute evidence that the gene is not essential. Conversely, inability to cure the chromosomal null mutant of the episomal expression vector would provide positive evidence for the essential nature of the targeted gene. However, in the experiments described herein, we have not been able to delete all chromosomal alleles of the NT3 and NT4 genes even in the presence of a complementing episome expressing either NT3 or NT4 thus contravening the strategy outlined above. This failure to support deletion of all chromosomal alleles occurs despite the fact that episomes encompassing the NT3 gene restore high affinity purine nucleobase transport to nt3(–/–) null mutants (Fig. 2) and thus complement the transport deficiency of the null mutant. Others have observed a similar failure of episomal expression constructs to support deletion of all chromosomal copies of other genes (Boitz and Ullman, 2006), but the reasons for such failures of complementation are not clear.
In conclusion, the difficulty encountered in deleting all alleles of the NT3 and NT4 genes suggests that parasites devoid of both genes are significantly impaired in viability and are thus strongly selected against in targeted gene deletion experiments. It is possible but not certain that the nt3(–/–)/nt4(–/–) null mutant may not be viable. It is not obvious why this dual null mutant should encounter difficulty surviving under the conditions employed for selection, as purine nucleosides can permeate through the intact NT1 and NT2 transporters and should thus be able to support the metabolic requirement for purines and parasite growth even in the absence of purine nucleobase uptake. However, it is possible that synergy exists between the requirement for purine nucleobases and another potentially unknown function of the NT4 protein and that the absence of these synergistic interactions contributes to the difficulty in generating the dual null mutant. Overall the results reported here reveal that the NT3 and NT4 nucleobase transporters, both individually and in combination, play significant roles in the biology of L. major parasites.
Standard genetic nomenclature for Leishmania (Clayton et al., 1998) is employed for proteins (e.g. NT3) and for genes and RNAs (e.g. NT3). However, for deletion mutants, a superscript nomenclature (e.g. nt3(+/–), nt3(–/–), nt3(–/–)/nt4(+/–), nt3(–/–)/nt4(+/–/–), etc.) is employed to designate with maximum clarity the genotype of various heterozygotes.
[2,8–3H]-hypoxanthine (36.4 Ci mmol−1), [8–3H]-xanthine (18 Ci mmol−1), [2,8–3H]-adenine (50 Ci mmol−1), [8–3H]-guanine (7 Ci mmol−1), [4-Hydroxypyrazolo[3,4-d]-pyrimidine-4-one-3H]-allopurinol (1.2 Ci mmol−1) were purchased from Moravek Biochemicals (Brea, CA). All other chemicals and regents were of the highest commercial quality available.
Nucleic acid purification, blotting and hybridization
For isolation of genomic DNA and total RNA, DNAzol and TRIzol reagents (Invitrogen, Carlsbad, CA) were used according to manufacturer's instructions. Southern and Northern blots were performed by standard methods (Sambrook et al., 1989)
Construction of targeting vectors
To generate the nt3(–/–) null mutant, the two alleles were replaced sequentially with hygromycin phosphotransferase (HYG) and the phleomycin binding protein (BLEO) genes encoding resistance markers for the antibiotics hygromycin B and phleomycin, respectively, whereas generation of the nt4(–/–) null mutant employed the nourseothricin (SAT) or blasticidin (BSD) resistance marker genes. The targeting vectors for hygromycin and phleomycin selection were based on the Leishmania expression vectors pX63HYG (Cruz et al., 1991) and pX63PHLEO, a derivative of pHMPHLEO (Freedman and Beverley, 1993), whereas those for nourseothricin and blasticidin were based upon the expression vectors pCPC-SAT (Bart et al., 1997) and pXGBSD (Goyard and Beverley, 2000) respectively. Each selectable marker cassette was flanked by ∼1 kb segment of DNA immediately 5′ and 3′ of the NT3 or NT4 ORF. To generate the episomal expression constructs, the NT3 and NT4 ORFs were amplified by PCR and inserted into the EcoRI restriction site within the pX63NEORI (Valdes et al., 2004) or the BamHI site of the pX63NEO-GFP2+ (Ha et al., 1996) expression vectors respectively.
Oligonucleotide primers and PCR
For PCRs, 200 ng of genomic DNA was amplified using 10 pmol of each primer and the MasterAmp Taq DNA polymerase kit (Epicentre Biotechnologies, Madison, WI). Samples were preheated to 95°C for 3 min followed by 30 cycles of 95°C for 0.5 min, 55°C for 0.5 min, and 68°C for 2 min and then a single cycle of 68°C for 10 min. The primers employed for PCR experiments in Fig. 9 were: primer a – cagtggtgaacagcctctggtg, primer b – ggcgacgctgtagtcttcagagatg, primer c – catggccaagcctttgtctcaag, primer d – cacaccatgcccatcacatgg, primer e – ttaggcgtcatcctgtgctcccgag, primer f – cgaacaatgtacctgcctgcaatttg.
For homologous gene replacements each vector was digested with appropriate restriction enzymes to release the linear targeting construct, which was gel purified and resuspended at 1 mg ml−1 in sterile water, and 5–10 μg of fragment were used for each transfection. For complementation experiments, undigested plasmids (10 μg) were used for transfections. Electroporation was performed according to Robinson and Beverley (2003).
Expression in Xenopus oocytes
The NT3 or NT4 ORFs were subcloned into the EcoRI site of the Xenopus oocyte expression vector pL2-5 (Arriza et al., 1993) linearized, and in vitro transcribed using the mMessage mMachine T7 Ultra Kit (Ambion, Austin, TX) according to the manufacturer's instructions. Stage V-VI Xenopus oocytes were injected with 23 nL of cRNA (∼10 ng) or water and incubated for 3 days at 16°C in ND96 buffer (Sanchez et al., 2004) before performing uptake assays.
In L. major, uptake of [3H]-substrates was assayed as described in Vasudevan et al. (2001). For kinetic analysis, initial rates of uptake at each substrate concentration were determined by linear regression over the linear portion of the time-course using Prism 4.0b software (GraphPad Software, San Diego, CA), after subtracting the uptake value at time 0 from each time point. In oocytes, uptake of radiolabelled substrates was assayed as described in Sanchez et al. (2004).
To determine the effect of allopurinol (Sigma-Aldrich, St Louis, MO) on growth, 5 ml of RPMI supplemented with 10% dialysed FBS, 150 μg ml−1 haemin and 100 μM xanthosine were inoculated with 106 cells ml−1 and different concentrations of allopurinol. After 3 days the cell density was determined using a haemacytometer. The initial cell density was subtracted from the final cell density and the resulting difference expressed as a percentage of the control growth in the absence of allopurinol and plotted as a function of the concentration of allopurinol. The EC50 values were determined by fitting the data by non-linear regression to a one-site competition model using Prism 4.0b software.
Growth in purine nucleobases
To evaluate the ability of the nt3(–/–) null mutant to grow in various purine nucleobases, cells were washed 2× with PBS and resuspended at 5 × 105 cells ml−1 in RPMI supplemented with 10% dialysed FBS, various purine nucleobases at a final concentration of 10 μM. Cell density was determined by monitoring the absorbance of the oxidized and reduced forms of alamarBlue™ (BioSource International, Camarillo, CA) at 570 and 600 nm (Raz et al., 1997). The per cent reduction of alamarBlue™ was calculated according to the manufacturer's instructions.
Three- to 5-day-old P. duboscqi sandflies were fed through a chick skin membrane on a mixture of heparinized mouse blood containing 2–3 × 106 promastigotes per ml obtained from 1- to 2-day-old logarithmic cultures. Blood-engorged sandflies were separated and maintained at 28°C with 30% sucrose solution. At various times after feeding, the flies were killed in a 5% soap solution and their midguts were dissected and examined microscopically for the presence and localization of promastigotes. The number of parasites per midgut was determined by placing individual midguts into a microcentrifuge tube containing 30 μl of PBS, pH 7.4, homogenizing each gut by using a Teflon-coated microtissue grinder, and counting released promastigotes in a haemacytometer.
Metacyclic promastigotes were selected from 6- to 8-day-old stationary phase cultures by agglutination of non-metacyclic parasites with peanut agglutinin (Sigma-Aldrich) as described by da Silva and Sacks (1987). BALB/c bone marrow-derived macrophages were infected at a multiplicity of 10 with metacyclic promastigotes and centrifuged at 20°C for 5 min (Courret et al., 2002). Pelleted cells were resuspended in Dulbecco's Modified Eagle medium supplemented with 4 mM l-glutamine, 0.11 g l−1 sodium pyruvate, 4.5 g l−1 glucose and 10% FBS. Cultures were seeded in 4-well Laboratory-TekII Chamber Slides (Nalgen Nunc International Naperville, IL) containing 1.0 ml of macrophage growth medium and incubated at 35°C in a humidified 5% CO2 incubator. After 4 h, adherent macrophages were washed 3 × with PBS to eliminate residual extracellular promastigotes after which fresh medium was added. The medium was changed at 3 days post infection, and on day 5 intracellular parasites were detected microscopically in methanol-fixed macrophages by nuclear staining with Diff-Quick Kit (International Medical Equipment, San Marcos, CA).
To determine the level of expression of GFP fusion proteins, 108 cells were lysed in Laemmli sample buffer and heated at 70°C for 10 min. Lysates were then resolved by electrophoresis on 4–12% gradient NuPAGE® Novex Bis-Tris gels (Invitrogen) under reducing conditions using the XCell SureLock™ Mini-Cell system according to manufacturer's protocols (Invitrogen). Subsequently, proteins were electro-transferred under denaturing conditions onto PVDF membranes (Millipore Corporation, Billerica, MA) using the XCell II™ Blot Module (Invitrogen) according to the manufacturer's instructions. Blots were blocked with 5% non-fat milk in TPBS (0.05% Tween-20, PBS pH 7.2) for 30 min at room temperature, washed 3 × for 15 min with TPBS, incubated with a 1:10 000 dilution of mouse anti-GFP (Clontech, Mountain View, CA) or anti-α-tubulin (Molecular Probes, Invitrogen) antibodies in blocking solution for 1 h at room temperature. Blots were washed 3 × for 15 min with TPBS and incubated with a 1:10 000 dilution of goat anti-mouse antibody conjugated to horseradish peroxidase (Pierce Biotechnology, Rockford, IL) for 1 h at room temperature. Membranes were washed as before and proteins were detected using the Western Lightning Chemiluminescence Reagent Plus Kit (Perkin-Elmer) employing the manufacturer's protocols and Kodak BioMax Light Film (Eastman Kodak Company).
To detect GFP fluorescence, formaldehyde fixed parasites were adhered to poly-l-lysine coated cover slips, blocked with 2% goat serum, 0.01% sodium azide, 0.01% saponin in PBS (blocking solution), and incubated with a 1:1000 dilution of the primary mouse anti-α-tubulin antibody (Molecular Probes, Eugene, OR) in blocking solution for 1 h at room temperature. Cells were rinsed 3 × with PBS and incubated with a 1:1000 dilution of goat anti-mouse IgG-Alexa Fluor 594™ (Molecular Probes, Invitrogen) in blocking solution for 1 h at room temperature in the dark. Cover slips were rinsed again and mounted onto slides using Fluoromount-G (Southern Biotechnology Associates, Birmingham, AL). Fluorescence images were obtained using a wide field deconvolution system (Applied Precision LLC., Issaquah, WA) as described previously (Nasser and Landfear, 2004).
This work was supported by grant number AI44138 from the National Institutes of Health to S.M.L and a Veterans Administration Merit Award plus grant number AI056446 from the National Institutes of Health to H.G.A.B. We wish to thank Dr David Sacks for co-ordinating and supporting the sandfly experiments reported here, Suzanne Brandt for harvesting murine bone marrow-derived macrophages, and Nicola Carter for comments on the manuscript.