Arginine Homeostasis and Transport in the Human Pathogen Leishmania donovani

Abstract

Arginine is an essential amino acid for the human pathogen Leishmania but not to its host. Thus, the mechanism by which this protozoan parasite regulates cellular homeostasis of arginine is critical for its survival and virulence. In a previous study, we cloned and functionally characterized a high affinity arginine-specific transporter, LdAAP3, from Leishmania donovani. In this investigation, we have characterized the relationship between arginine transport via LdAAP3 and amino acid availability. Starving promastigotes for amino acids decreased the cellular level of most amino acids including arginine but also increased the abundance of both LdAAP3 mRNA and protein and up-regulated arginine transport activity. Genetic obliteration of the polyamine biosynthesis pathway for which arginine is the sole precursor caused a significant decrease in the rate of arginine transport. Cumulatively, we established that LdAAP3 expression and activity changed whenever the cellular level of arginine changed. Our findings have led to the hypothesis that L. donovani promastigotes have a signaling pathway that senses cellular concentrations of arginine and subsequently activates a mechanism that regulates LdAAP3 expression and activity. Interestingly, this response of LdAAP3 to amino acid availability in L. donovani is identical to that of the mammalian cation amino acid transporter 1. Thus, we conjecture that Leishmania mimics the host response to amino acid availability to improve virulence.

l-Arginine is a metabolically versatile cationic amino acid that provides a precursor function for the synthesis of a variety of bioactive molecules in all organisms. These molecules include nitric oxide, polyamines, phospho- and methylarginine, and proteins. Most of these molecules are essential for cell growth, and a few, like nitric oxide, are also used to protect hosts from pathogen invasion (1, 2). To enable simultaneous feeding of multiple and disparate biosynthetic pathways, cells must maintain ample cellular levels of arginine at all times. A variety of mechanisms for arginine homeostasis have been identified in mammals and fungi (3, 4). However, only very limited information on arginine homeostasis in protozoan parasites and its role in pathogenesis is available. Our laboratory aims to address these questions in the human pathogen Leishmania. Protozoan parasites of the genus Leishmania are the causative agents of a wide spectrum of human and veterinary diseases. Infections with Leishmania vary in their clinical manifestations depending upon the species ranging from self-healing skin lesions to visceral pathogenesis, which is invariably fatal if untreated (5). Leishmania species exhibit a digenetic life cycle that includes both extracellular promastigote and intracellular amastigote forms. The extracellular promastigotes develop in the alimentary tract of sand flies, whereas amastigotes reside within macrophage phagolysosomes (6, 7).

Arginine is an essential amino acid for Leishmania (8), and therefore its metabolism and homeostasis depends on supply from external pools. This amino acid has only two functions in the parasite serving as a monomeric precursor for polypeptide synthesis and to funnel carbon atoms into the polyamine biosynthetic pathway (9). Recently, we cloned and functionally characterized a high affinity arginine transporter, LdAAP3, whose protein is localized to the plasma membrane of Leishmania donovani (10). LdAAP3 belongs to the trypanosomatid-specific group of amino acid permeases, which forms a distinct clad within the amino acid auxin permease superfamily (10, 11). The L. donovani genome contains two adjacent LdAAP3 genes in the complementary strand of chromosome 31 (LinJ31_V3.0900 and LinJ31_V3.0910). These genes contain identical 5′-untranslated regions (UTRs)³ and open reading frames (ORFs) but completely different 3′-UTRs. Northern analysis indicated that gene 1 produces a 10-kb transcript, whereas gene 2 produces a 5- and a 7-kb transcript. LdAAP3 is specific for arginine, whereas arginine and lysine are cotransported via the same permease in mammalian cells (3).

Mammalian cells synthesize arginine from proline or glutamate via ornithine and citrulline. However, this endogenous synthesis of arginine is not sufficient to feed pathways that use it as a precursor molecule; therefore, mammalian cells must import extracellular arginine. Arginine, as well as lysine, is taken up via four cationic amino acid transporters (Cat1, -2a, -2b, and -3) that differ in their affinities for arginine and lysine (3). Cat1 exhibits the highest affinity for arginine and is thought to play the major role in arginine homeostasis. Starvation of mammalian cells for amino acids augments Cat1 activity by increasing Cat1 mRNA stability and translation (3, 12), thereby increasing Cat1 abundance.

Fungi respond to amino acids starvation by activating a general amino acid control mechanism that activates both amino acid metabolic and uptake pathways (13). These microbial organisms take up arginine via three biochemically and genetically distinct permeases; Can1, which transports arginine, Gap1, which transports all amino acids, and Alp1, which mediates arginine transport when overexpressed (14). Lyp1 has also been reported to transport arginine in addition to lysine; however, when overexpressed, only lysine transport increased (14). Gap1 and Can1 seem to play a major role in cellular arginine homeostasis in Saccharomyces cerevisiae.

The fact that fungi and mammals are sensitive to amino acid deprivation by employing mechanisms that regulate arginine transport emphasizes the significant role arginine permeases have in maintaining cellular arginine homeostasis in eukaryotes. In the present work, we have examined the relationships among LdAAP3-mediated arginine transport, amino acid starvation, and activity of metabolic pathways downstream of arginine. We discovered that starvation of L. donovani promastigotes for amino acids stimulated arginine transport by increasing LdAAP3 abundance. Mutations that prevent the use of arginine as a precursor for polyamine synthesis down-regulate arginine transport to reduce the cellular pool of this amino acid. The results of this work indicate that arginine transport in L. donovani is regulated by a mechanism that senses changes in the level of the intracellular pool of arginine.

EXPERIMENTAL PROCEDURES

Leishmania Cell Culture

Two lines of L. donovani, MHOM/S.D./00/1S; 1SR (15) and DI700 (16), were used. Promastigotes of both lines were grown and maintained as described previously (15, 17). Two null mutant derivatives of DI700, one that lacks ornithine decarboxylase (ODC (16)) and one deficient in spermidine synthase (SPDSYN (18)) were used in experiments in which the role of the polyamine pathway in arginine transport and homeostasis was evaluated. The Δodc and Δspdsyn knockouts were routinely propagated in DME-l-CS medium (DME-l with 10% chicken instead of bovine serum) supplemented with either 200 μm putrescine or 200 μm spermidine, respectively (16, 18).

Starving Promastigotes to Amino Acids

Mid-exponential phase promastigotes (∼1 × 10⁷ cells/ml) were washed twice in Earl's salt solution and then resuspended to a final density of 1 × 10⁸ cells/ml in the same solution supplemented with 5 mm glucose and 1% dialyzed heat-inactivated fetal calf serum. Starvation was carried out at 26 °C for either 2 or 4 h and terminated by moving the cells to ice. The starved cells were used for transport assays and analyses of cellular amino acid content without any additional washes.

Transport Assays

Uptake of 20 μm [³H]arginine (600 mCi/mmol), 20 μm [³H]lysine (1.0 Ci/mmol), or 1.0 mm [³H]proline (8.0 mCi/mmol) into mid log phase parasites was determined by the rapid filtration technique of Mazareb et al. as reported (19). To determine initial rates of transport, transport measurements were performed on 1 × 10⁸ promastigotes exposed to radiolabel for up to 2 min. The amount of radiolabel associated with the cells was linear with time over the 2-min time course of the transport assay.

Antibodies

A 16-mer peptide, 5′-DKHPSGEQGNHLHKNG-3′, synthesized to the hydrophilic NH₂ terminus of LdAAP3 protein was conjugated at its COOH terminus to keyhole limpet hemocyanin and injected to rabbits with 5 mg/ml protein. Rabbits were boosted four times with 5 mg/ml of the conjugated peptide. Injections and bleedings were carried out as a contracted service by Sigma-Aldrich. Antiserum was used at 1:1000. Rabbit antiserum against L. donovani HSP83 (15) was used as protein loading marker.

Western Blot Analyses

Parasite lysates and enriched membranes were prepared as previously reported (20), and Western blot analyses with anti-peptide antibody against LdAAP3 and HSP83 were carried out as described (21) using luminol reagent for peroxidase-conjugated antibody detection (Santa Cruz Biotechnology, Santa Cruz, CA).

Northern Blot Analyses

Total RNA from L. donovani promastigotes was prepared and subjected to Northern blotting as described before (21). Blots were hybridized to the LdAAP3 ORF and the 3′-UTR of LdAAP3 gene copy 2 using probes that had been PCR-amplified from L. donovani genomic DNA as described previously (10).

Determination of Cellular Amino Acid Content

1 × 10⁹ promastigote cells in growth medium or after starvation were washed once with the same volume of ice-cold phosphate-buffered saline (1,200 × g for 10 min) and then twice with 2 ml of ice-cold phosphate-buffered saline for 2 min each. The washed cell pellet was then resuspended in 1 ml of ice-cold phosphate-buffered saline plus 1 ml of 5 n perchloric acid, vortexed, and incubated on ice for an additional 10 min. Perchloric acid lysates were centrifuged in a microcentrifuge at 14,000 rpm for 10 min at 4 °C, and 232 μl of 5 n KOH was added to the supernatant to titrate the sample pH to 7.0. Additional centrifugation was performed under the aforementioned conditions, and 200-μl aliquots were analyzed for amino acid content by the method of Fekkes et al. (22). The analyses were carried out at the Medical Biochemistry Laboratory at Rambam Medical Center in Haifa. Intracellular amino acid concentrations were calculated based on a previously determined promastigote cell volume of 4.2 μl per 10⁸ cells (23).

Transfection Assays

The pSNBR-ODC plasmid harboring the ODC ORF (16) was transfected into Δodc promastigotes using standard electroporation conditions for Leishmania promastigotes (16, 24). Transfectants were selected in 100 μg/ml G418 in the absence of putrescine.

RESULTS

Amino Acid Starvation Up-regulates LdAAP3 Expression and Activity

When mid log phase L. donovani promastigotes were transferred from Earl's-based medium 199 to Earl's salt solution supplemented with only 5 mm glucose, the initial rate of arginine transport increased as a function of time of starvation (Fig. 1A). Arginine transport rates increased 2-fold after 2 h and 5-fold after 4 h of amino acid deprivation. Addition of all amino acids to the starvation medium at their medium 199 concentrations almost completely inhibited this up-regulation of arginine transport rates, indicating that the augmentation in arginine transport rates could be ascribed to decreased amino acid availability (Fig. 1B). Moreover, incubating promastigotes under amino acid starvation conditions in the presence of cycloheximide inhibited starvation-induced transport up-regulation, indicating that up-regulation require de novo protein synthesis (Fig. 1C). Starvation-induced arginine transport up-regulation is an energy-dependent process, because addition of glucose to the starvation solution was essential for observed up-regulation of arginine transport (Fig. 1D). The same effect on starvation-induced arginine transport was achieved when glucose was replaced by proline as the sole source of metabolic energy (data not shown).

FIGURE 1.

Starvation to amino acids up-regulates l-arginine transport in time and protein synthesis-dependent manner. Mid-logarithmic phase L. donovani promastigotes were washed twice and suspended in Earl's salt solution supplemented with 5 mm glucose and 1% dialyzed fetal calf serum. Starvation was initiated by incubating these cells at 26 °C. At the indicated time points, cells were subjected to 2-min arginine uptake assays and the initial rate of transport was calculated as described under “Experimental Procedures.” A, time course of starvation. Results expressed as mean ± S.D. of at least four experiments; B, 4-h starvation with (plaid pattern) or without (■) externally added amino acids at their medium 199 concentrations; C, starvation done in the presence (gray) or absence (■) of 10 μg/ml cycloheximide; D, logarithmic phase promastigotes were starved for 2 h with (□) or without (▴) 5 mm glucose. ■, control unstarved promastigotes. The results in B and C represent one of two independent experiments with essentially identical results.

To determine whether the increase in arginine transport was due to a change in LdAAP3 abundance, polyclonal antibodies against a 16-residue peptide from the hydrophilic NH₂-terminal region of LdAAP3 were raised (see “Experimental Procedures”). The polyclonal antisera specifically recognized a 46-kDa protein in a membrane-enriched fraction of promastigotes (Fig. 2A). This 46-kDa band is slightly smaller than the expected molecular mass of LdAAP3 (52 kDa), possibly due to the hydrophobic nature of LdAAP3. Immunoblot analysis revealed that LdAAP3 abundance increased as a function of the length of time of amino acid starvation (Fig. 2B) in proportion to the observed increase in LdAAP3-mediated transport activity (see Fig. 1A). Cycloheximide inhibited this starvation-induced increase in LdAAP3 protein (Fig. 2C), indicating that starvation-induced arginine transport up-regulation required de novo synthesis of LdAAP3. Using polyclonal antiserum against L. donovani HSP83 (15) we observed that this protein copurified with the enriched parasites membrane fraction, and its amount was proportional to the protein content of this fraction. Thus, HSP83 could be used as an additional loading control in the immunoblotting experiments with LdAAP3.

FIGURE 2.

Amino acid starvation up-regulates LdAAP3 protein abundance. Proteins from promastigotes lysates or enriched membranes were separated on 10% SDS-PAGE, transferred onto nitrocellulose paper, and reacted with anti-LdAAP3 rabbit antiserum that was raised against a 16-residue peptide from the N terminus of LdAAP3 (see “Experimental Procedures”). A, antiserum specificity was tested by preincubation with (+) or without (−) 20 μg of peptide; B, promastigotes were subjected to 2- or 4-h starvation; C, logarithmic phase promastigotes were subjected to a 4-h starvation with or without 10 μm cycloheximide. In both B and C enriched membranes were used. The results in B and C represent one of three and two independent repeats, respectively.

The effect of amino acid starvation on LdAAP3 transcript abundance was also determined. A significant increase in LdAAP3 transcript abundance was observed in parasites incubated under amino acid starvation conditions as assessed by probing Northern blots with either the LdAAP3 ORF or 3′-UTR of LdAAP3 copy 2 (Fig. 3). Significant increases in the abundance of both the 5- and 7-kb transcripts were detected. Conversely, only a minor change in the abundance of the 10-kb transcript was observed, demonstrating that prolonged amino acid starvation did not greatly affect the expression of LdAAP3 gene copy 1. Thus, starvation appears to differentially activate only one of the two LdAAP3 gene copies.

FIGURE 3.

Amino acid starvation up-regulates the expression of one of the two LdAAP3 gene copies. Total RNA from promastigotes before and after 4-h starvation was probed with DNA of the LdAAP3 ORF (left two lanes) or 3′-UTR of gene copy 2 (right two lanes). The results in this figure represent one of three independent repeats.

Starvation Affects Both the Cellular Amino Acid Pool and Arginine Transport

One possible explanation for the starvation-activated LdAAP3-mediated arginine transport augmentation is that starvation triggers a decrease in the cellular content of arginine, thereby activating a signaling pathway that stimulates LdAAP3 expression and activity. To evaluate this hypothesis, we determined the amino acid pool of L. donovani promastigotes before and after starvation (Table 1). Alanine is the most abundant amino acid (39% of the pool), followed by glutamate (18%), arginine (11%), glycine (9%), lysine (7%), and serine (5%) in exponentially growing L. donovani promastigotes that were not subjected to amino acid starvation (C₀). This amino acid distribution resembles that of L. major promastigotes as reported by Vieira and Cabantchik (25). Subjecting L. donovani promastigotes to 4-h starvation in the presence of glucose caused ≥40% reduction in the cellular concentration of most amino acids (Table 1, C₄). Interestingly, the intracellular pools of both lysine and valine, two amino acids that are essential for Leishmania, remained largely unchanged during starvation (<20%), while the concentration of two nonessential amino acids, aspartate and glutamine (8), increased during the starvation period by 110 and 39%, respectively.

TABLE 1.

Effect of starvation on the cellular amino acid pool in L.donovani promastigotes

The cellular amino acid concentrations were determined before (C₀) and after 4 h (C₄) of starvation. ΔC% = (C₄ − C₀)/C₀ × 100%. Results are expressed as mean ± S.D. (n = 3).

Amino acid	C0	C4	ΔC
	mm		%
Asp	0.35 ± 0.22	0.75 ± 0.44	114
Gln	0.69 ± 0.41	0.96 ± 0.46	39
Val	1.48 ± 0.46	1.79 ± 0.15	21
Lys	3.79 ± 2.05	3.4 ± 2.63	−10
Asn	0.40 ± 0.07	0.25 ± 0.11	−38
Glu	9.75 ± 2.63	5.51 ± 2.02	−43
Arg	5.98 ± 4.47	3.07 ± 3.45	−49
Ser	2.60 ± 0.63	1.19 ± 0.22	−54
Ala	21.34 ± 4.33	9.54 ± 1.30	−55
Gly	5.11 ± 2.37	1.95 ± 0.35	−62
Ile	0.48 ± 0.19	0.17 ± 0.03	−65
Met	0.36 ± 0.15	0.1 ± 0.01	−72
Leu	1.42 ± 0.81	0.38 ± 0.11	−73
Thr	0.53 ± 0.21	0.12 ± 0.09	−77
Phe	0.13 ± 0.06	0.03 ± 0.02	−80
Tyr	0.12 ± 0.06	0.02 ± 0.01	−83
His	0.22 ± 0.13	0.03 ± 0.01	−86

The cellular concentration of arginine decreased after 4 h starvation by almost 50%, from 6 mm to 3.1 mm. Because arginine is essential for Leishmania and is the sole precursor for polyamine synthesis, it was not surprising that, to compensate for the loss of cellular arginine, arginine transport increased significantly (Figs. 1A and 4A). Conversely, transport of lysine whose cellular concentration remained unchanged increased only slightly during starvation (38%, Fig. 4B). Amino acid starvation also produced a minor effect on proline transport (increased by 29%, Fig. 4C). Proline pools were not evaluated in these analyses, because our analytical method cannot detect proline and therefore proline content is missing in Table 1. However, glutamate is the precursor of both proline and glutamine, and, because the cellular level of glutamine increased after starvation (Table 1), it is likely that proline pools were also enhanced and, therefore, there was no requirement for promastigotes to increase their proline transport capacity.

FIGURE 4.

Transport of l-lysine or l-proline is not affected by amino acid starvation. Logarithmic phase promastigotes before (■) and after 4-h starvation (□) were subjected to transport assays of arginine (A), lysine (B), and proline (C).

To investigate whether the up-regulation of LdAAP3 protein in response to general amino acid starvation was amino acid-specific, L. donovani promastigotes were incubated in the starvation buffer to which specific amino acids were added. Addition of arginine to the starvation buffer inhibited LdAAP3 up-regulation (Fig. 5). Conversely, addition of other amino acids such as glycine, glutamine, or aspartate had no effect on LdAAP3 abundance up-regulation. These data suggest that the effect of starvation on LdAAP3 expression and activity is specific for arginine availability.

FIGURE 5.

Only arginine inhibits the starvation-induced LdAAP3 protein abundance up-regulation. Logarithmic phase promastigotes were starved for 4 h in the presence of the indicated amino acids at their medium 199 concentrations. Enriched membranes of these cells were separated on 10% SDS-PAGE, transferred to nitrocellulose paper, and probed with anti LdAAP3 antiserum as in Fig. 2. The relative density of each band is illustrated below the blot. Relative intensity = the density of each LdAAP3 band divided by the density of its corresponding hsp83 band. The results in this figure represent one of two independent repeats.

LdAAP3 Senses the Polyamine Synthesis Pathway

Because the sole metabolic function of arginine, aside from serving as a building block for protein synthesis (9), is in the biosynthesis of polyamines, we conjectured that genetic ablation of the polyamine biosynthetic pathway would suppress arginine uptake to avoid overflow of intracellular arginine. To evaluate this supposition, arginine uptake rates were determined in Δodc and Δspdsyn L. donovani mutants (16, 18). Addition of either putrescine or spermidine to the growth medium of wild-type L. donovani promastigotes diminished arginine transport rates by 7 and 30%, respectively (Fig. 6A). Arginine transport in the Δodc and Δspdsyn strains was ∼50% of that of wild-type parasites (Fig. 6B). This decrease in arginine uptake capacity was independent of the external supply of polyamines, because incubating the two null lines for 4 days without polyamine supplementation did not affect their arginine transport capacity (Fig. 6B).

FIGURE 6.

Effect of externally added polyamines on arginine transport. A, L. donovani clone DI700 promastigotes were grown for at least a week without (■) or with 0.2 mm putrescine (gray) or 0.2 mm spermidine (□) in the growth medium. Initial rates of arginine transport were determined as in Fig. 1; (B) Promastigotes of DI700 wild-type (gray), Δodc mutant grown with 0.2 mm putrescine (■) and Δspdsyn mutant grown with 0.2 mm spermidine (□) in the growth medium (t = 0). These cells were washed and resuspended in M199 without the polyamines and further incubated at 26 °C for 4 days (t = 4). Promastigotes at from both time points were subjected to arginine transport assays as described in Fig. 1; (C) The relative change in the cellular pool of arginine (■) and ornithine (gray) were determined in DI700 wild-type grown with 0.2 mm of either putrescine or spermidine, Δodc mutant grown with 0.2 mm putrescine and Δspdsyn mutant grown with 0.2 mm spermidine (indicated in the figure). 100% corresponds to 5.2 mm arginine and 16.8 mm ornithine in wild-type cells grown without polyamines. Results are expressed as mean ± S.D. (n = 3).

Kinetic analysis of arginine transport in promastigotes of wild-type indicated K_m and V_max values of 10.4 μm and 310 pmole/min/10⁸ cells, respectively (Table 2). Ablation of the spdsyn gene reduced the V_max value of arginine transport by 42% but had no effect on the K_m value in cells supplemented with spermidine in their growth medium. In Δodc mutants that were grown in the presence of putrescine, V_max value decreased even more, by 68% and K_m value by 40% (Table 2). These data indicate that ablation of either spdsyn or odc genes affected primarily the V_max value of arginine transport by LdAAP3.

TABLE 2.

Kinetics of arginine transport in Δodc and Δspdsyn L. donovani promastigotes

Results are expressed as mean ± S.D. (n = 2).

Strain	Vmax	Km
	nmol[(min)(108 cells)]−1	μm
Wild type	0.31 ± 0.02	10.44 ± 2.79
ΔSpdSyn	0.18 ± 0.05	11.33 ± 2.11
Δodc	0.10 ± 0.02	6.24 ± 4.04

To examine the relationship between arginine transport capability and intracellular pools, arginine and ornithine levels were determined in wild type and appropriately supplemented mutant parasites (Δodc and Δspdsyn cells were supplemented with putrescine and spermidine, respectively). Growth of wild-type parasites in the presence of 0.2 mm putrescine increased the cellular concentration of arginine by ∼20% but did not affect the cellular ornithine content (Fig. 6C). Conversely, addition of 0.2 mm spermidine to the culture medium reduced both ornithine and arginine pool by 50% (Fig. 6C). This correlated with the decrease in arginine transport capacity in parasites growing with spermidine (Fig. 6A). The cellular arginine pool of Δodc mutants that grew in 200 μm putrescine was ∼50% that of wild-type parasites maintained in the same growth medium, but ornithine pools were equivalent (Fig. 6C), presumably because Δodc cells continue to synthesize ornithine from arginine. The low level of cellular arginine in the Δodc mutants correlated with the lower level of arginine transport in these cells (Fig. 6B). The Δspdsyn null mutant displayed profoundly reduced pools of both arginine and ornithine. Arginine and ornithine pools in the Δspdsyn parasites were 18 and 34%, respectively, of those of wild-type parasites grown in the same medium. This finding was unexpected, because the rate of arginine transport in these cells was similar to that of the Δodc parasites grown with putrescine (Fig. 6B).

Even though genetic ablation of the polyamine pathway from arginine reduced arginine transport rates, it had no effect on the abundance of LdAAP3 protein (Fig. 7). In contrast, ectopic expression of the wild-type odc gene from an episome within the Δodc background increased both the cellular levels of LdAAP3 protein in enriched membranes and the rate of LdAAP3-mediated arginine transport by 2-fold (Fig. 7).

FIGURE 7.

Ectopic expression of wild-type odc inside Δodc mutants increase arginine transport activity and LdAAP3 protein abundance. Δodc null mutants or mutants transfected with pSNBR-odc were subjected to arginine transport assays as in Fig. 1 or Western blot analysis as in Fig. 2. The results in this figure represent one of two independent experiments.

DISCUSSION

Leishmania promastigotes maintain a large cellular pool of arginine, ∼6 mm, that funnels carbon skeletons into polyamines and polypeptides, the sole metabolic pathways in the parasite to which the amino acid is a precursor. Because arginine is an essential amino acid for Leishmania, we hypothesized that the homeostatic mechanisms that regulate arginine pools and flux depend largely upon arginine uptake. In these investigations, we determined the relationships between arginine transport via LdAAP3 and amino acid availability in L. donovani. Major observations were: (a) starving promastigotes for amino acids decreased the cellular level of most amino acids including arginine; (b) amino acid starvation up-regulated arginine uptake via LdAAP3; (c) amino acid starvation increased abundance of both LdAAP3 mRNA and protein; and (d) disconnecting the polyamine synthesis pathway from the pool of arginine significantly decreased arginine transport and its cellular pool, but not LdAAP3 protein abundance. In all experiments with wild-type L. donovani, LdAAP3 expression and activity increased whenever the cellular level of arginine decreased. However, when polyamine biosynthesis was genetically disrupted, as in the Δodc and Δspdsyn knockouts, the rate of transport decreased to a new steady state that reduces the cellular pool of arginine. Thus, we propose that promastigotes have a signaling pathway that senses cellular concentrations of arginine and subsequently activates a mechanism that regulates LdAAP3 expression.

After amino acid starvation, two of the three LdAAP3 mRNAs, the 5- and 7-kb transcripts accumulated, indicating that only one of the two LdAAP3 copies was overexpressed. Because gene expression in Leishmania is regulated post-transcriptionally (27), it is likely that increased mRNA stability determined the level of abundance of the two LdAAP3 transcripts. It is interesting that overexpression of only the second LdAAP3 gene copy is observed upon amino acid starvation. This second gene copy expresses both the 5- and 7-kb transcripts, both of which are up-regulated by amino acid starvation, whereas the 10-kb transcript that is transcribed from LdAAP3 gene 1 does not appear to be significantly up-regulated. What regulates this selective up-regulation of LdAAP3 gene copy 2 awaits further analysis, but these data establish that this second gene copy encodes a functional LdAAP3 permease. Correlating with the increase in LdAAP3 mRNA, the abundance of LdAAP3 protein and arginine transport activity increased in parallel, intimating that the observed increase in transport activity was due to a change in the number of available permease molecules in the parasite membrane rather than an increase in the turnover rate of LdAAP3. Hence, the starvation-induced arginine transport up-regulation is controlled by mRNA abundance.

Strikingly, this response of L. donovani LdAAP3 to amino acid availability is identical to that of the mammalian arginine/lysine cation amino acid transporter 1 (CAT1 (3)). The CAT1 permease is the major supplier of extracellular arginine in mammalian cells and is expressed almost ubiquitously with adult liver being a prominent exception (3). Amino acid starvation increases the expression of CAT1 with the concomitant accumulation of two CAT1 transcripts of 7.4 and 3.4 kb. The two CAT1 mRNAs accumulate with different kinetics and stability with the 7.4-kb mRNA exhibiting considerably greater stability in amino acid-depleted than in amino acid-replete cells (28). Nuclear run-on and other experiments indicated that mRNA stability is the mechanism responsible for CAT1 transcript up-regulation (28). Following mRNA accumulation, starvation also induced an increase in CAT1 protein abundance, which resulted in a significant increase in arginine transport activity. These studies indicate that CAT1 mRNA stability is the mechanism responsible for CAT1 response to change in amino acid availability. The observation that LdAAP3 and CAT1 respond similarly to amino acid starvation further supports our notion that mRNA stability regulates starvation-induced LdAAP3 expression and activity.

Mechanisms that sense changes in amino acid availability have been described in fungi, including Neurospora crassa, Aspergillus nidulans, and S. cerevisiae (13). In S. cerevisiae, amino acid starvation or imbalance activates a mechanism termed “general amino acid control” that regulates not only amino acid biosynthetic genes of distinct metabolic pathways, but also aminoacyl tRNA synthetases, pathway-specific activators, and amino acid transporters (13). The response is mediated by derepressed translation of the protein kinase GCN4 that finally induces expression of target genes. Furthermore, several plasma membrane-localized sensors that monitor extracellular nutrient availability have been identified in yeast (29). Extracellular amino acids are sensed by the amino acid sensor Ssy1, a protein highly homologous to functional amino acid transporters (29–32). When extracellular amino acids bind Ssy1, a signal transduction pathway that activates the transcription factors Stp1 and Stp2 that subsequently activate expression of several amino acid transporter genes is induced (33, 34).

It is yet not clear whether an Ssy1-like sensing mechanism also exists in Leishmania. However, our observation indicated that induction of LdAAP3 transcript and protein abundance was due to a mechanism that is specific for arginine and senses intracellular, not extracellular, arginine. Furthermore, whereas in both fungi and yeast activation of gene expression by amino acid availability is accomplished at the transcription level, mammalian CAT1 and leishmanial LdAAP3 are activated primarily by posttranscriptional processes.

Our finding that two distinct permeases, CAT1 and LdAAP3 (see Table 1 in Ref. 10), which belong to phylogenetically disparate organisms, are regulated by similar mechanisms is intriguing. We hypothesize that Leishmania are mimicking mammalian cells response to amino acid starvation as part of their mechanism of adaptation to their hosts. This further supports the hypothesis raised by Shaked-Mishan et al. (10) that LdAAP3 contributes to Leishmania virulence.

L. donovani mutants that lack the genes that encode key enzymes in the polyamine synthetic pathway are auxotrophic for polyamines and can only be maintained via the exogenous provision of polyamines synthesized downstream of the mutations (16, 18). Because wild-type promastigotes grow in polyamine-deficient medium, arginine is the sole source precursor for polyamine synthesis. Because arginine is no longer a polyamine precursor in the mutants transport diminished. Furthermore, ectopic expression of wild-type odc in the Δodc mutants restored the use of arginine as the sole source precursor for polyamine synthesis as well as arginine transport activity. Thus, it appears that LdAAP3 “senses” lack of precursor function in the mutants and accordingly sets the level of the cellular pool of arginine; high (∼6 mm) in wild-type and low (≤3 mm) in the null mutants. The fact that supplementing wild-type promastigotes with either putrescine or spermidine in the growth medium only slightly changed the rate of arginine transport indicates the prioritization that promastigotes assign to the polyamine precursor function of arginine. Interestingly, whereas addition of putrescine to the growth medium had no effect on arginine metabolism, spermidine significantly reduced their levels in both wild-type and mutants. This implies that spermidine biosynthesis is a regulation point for polyamines metabolism. How this regulates arginine metabolism and transport is an interesting question that needs to be addressed in the future.

Whenever arginine transport activity increased, either due to amino acid starvation or ectopic expression of odc, it was due to an increase in LdAAP3 protein abundance. In contrast, when arginine transport was suppressed as in the polyamine biosynthesis null mutants, there was no change in the abundance of this protein. Thus, it is likely that up-regulation and down-regulation of LdAAP3 activities are regulated by different mechanisms; one requires net synthesis of LdAAP3, and the other may involve allosteric regulation of LdAAP3 activity.

The abundance-independent suppression of arginine transport in L. donovani resembles the negative regulation of arginine transport via the arginine transporter CAN1 in S. cerevisiae (4, 35). In these cells, a small G protein, Rheb, negatively regulates arginine transport by directly interacting with Can1 via a farnesylated C terminus binding site. Interestingly, the L. infantum genome (closely related to L. donovani) contains three Rheb homologues (LinJ10_V3.0960, LinJ10_V3.1250, and LinJ27_V3.0620) that are 30–35% identical and 50% similar to S. cerevisiae Rheb. Furthermore, recent proteomic analysis done in our laboratory revealed that L. donovani promastigotes and amastigotes express the protein products of all three genes (26). However, none of them contain the farnesyl binding site that is found in the S. cerevisiae Rheb. Thus, whether L. donovani Rhebs are involved in regulating LdAAP3 activity is an open question that is currently under investigation.

In conclusion, this work reveals how Leishmania responds to arginine availability and highlights the importance of arginine transport. Because promastigotes represent the invasive stages, we propose that Leishmania mimic hosts' responses to amino acid availability to improve virulence.