Axenic Leishmania amazonensis Promastigotes Sense both the External and Internal Arginine Pool Distinctly Regulating the Two Transporter-Coding Genes

Leishmania (L.) amazonensis uses arginine to synthesize polyamines to support its growth and survival. Here we describe the presence of two gene copies, arranged in tandem, that code for the arginine transporter. Both copies show similar Open Reading Frames (ORFs), which are 93% similar to the L. (L.) donovani AAP3 gene, but their 5′ and 3′ UTR's have distinct regions. According to quantitative RT-PCR, the 5.1 AAP3 mRNA amount was increased more than 3 times that of the 4.7 AAP3 mRNA along the promastigote growth curve. Nutrient deprivation for 4 hours and then supplemented or not with arginine (400 µM) resulted in similar 4.7 AAP3 mRNA copy-numbers compared to the starved and control parasites. Conversely, the 5.1 AAP3 mRNA copy-numbers increased in the starved parasites but not in ones supplemented with arginine (p<0.05). These results correlate with increases in amino acid uptake. Both Meta1 and arginase mRNAs remained constant with or without supplementation. The same starvation experiment was performed using a L. (L.) amazonensis null knockout for arginase (arg -) and two other mutants containing the arginase ORF with (arg -/ARG) or without the glycosomal addressing signal (arg -/argΔSKL). The arg - and the arg -/argΔSKL mutants did not show the same behavior as the wild-type (WT) parasite or the arg -/ARG mutant. This can be an indicative that the internal pool of arginine is also important for controlling transporter expression and function. By inhibiting mRNA transcription or/and mRNA maturation, we showed that the 5.1 AAP3 mRNA did not decay after 180 min, but the 4.7 AAP3 mRNA presented a half-life decay of 32.6 +/− 5.0 min. In conclusion, parasites can regulate amino acid uptake by increasing the amount of transporter-coding mRNA, possibly by regulating the mRNA half-life in an environment where the amino acid is not present or is in low amounts.


Introduction
Leishmaniasis is a complex parasitic disease that currently affects about 12 million people and an estimated 2 million new cases per year [1]. It is caused by protozoa in the Leishmania genus, which has two distinct phases in its life cycle: the promastigote, an extracellular flagellate present at the gut of sand flies, and the amastigote that lives inside mononuclear phagocytes, mainly macrophages, in a vertebrate host.
Arginine is a key amino acid for macrophages because, being the substrate for inducible nitric oxide synthase (iNOS) to produce nitric oxide (NO), it is involved in the macrophage-defense response against pathogen infections. [2][3][4][5][6][7][8]. This amino acid is also a substrate for arginase, which catalyzes the production of urea and ornithine, a product important for polyamine pathway. This pathway is used by Leishmania to replicate and is essential for the parasite to establish infection [9][10][11][12]. It has largely been reported that macrophage or Leishmania modulation of arginine is responsible for parasite survival or its killing in the mammal host [5,[13][14][15][16][17][18][19].
Membrane transporters, present in both Leishmania and macrophages control arginine uptake [20][21][22][23][24], to sustain NO production, macrophages increase their expression of the main arginine transporter (CAT2B), which is indicative that the internal pool of arginine is not sufficient to supply arginine to iNOS [25][26][27]. On the other hand, a high-affinity arginine transporter has been described in L. (L.) donovani. This transporter is LdAAP3, and it has 480 amino acids and 11 predicted trans-membrane domains [22]. With this transporter, Leishmania seems to have mechanisms of sensing arginine decreases and responding with increased arginine uptake [28]. Therefore, the arginine-uptake control appears to be an important limiting factor to parasite survival inside macrophages [17,29].
Leishmania has a polycistronic transcription, and the control of gene expression is mainly performed through protein levels and mRNA stability [30]. In this study, we evaluated the importance of arginine transporter mRNA levels on the physiology of arginine uptake in L. (L.) amazonensis. Our data indicated that these organisms control the arginine transporter expression by regulating the transporter-coding mRNA stability. We also showed that the level of arginine transporter mRNA varies in promastigote development, and, using arginase-deficient mutants, we showed that possible changes in the internal arginine pool could be responsible for altering the transporter-coding mRNA levels.

Results
Characterization of the two L. (L.) amazonensis arginine transporter-coding DNA sequences A DNA probe based on the AAP3 ORF sequence of L. donovani [22] was used to screen a L. (L.) amazonensis genomic-cosmid DNA library [31]. The partial DNA sequence of the selected cosmid revealed the presence of two copies in tandem from a putative homologous gene. The ORF regions of the two copies showed 93% similarity to the AAP3 ORF in L. donovani (not shown). A northern blot analysis showed the presence of two distinct mRNAs for the gene (5.1 kb and 4.7 kb) (not shown). We named these transcripts 4.7 AAP3 mRNA and 5.1 AAP3 mRNA. The mRNA identities were confirmed by sequencing RT-PCR products that were obtained using oligo-dT reverse transcription and primers based on the cosmid sequence. These sequences were deposited in GenBank with accession numbers of HQ912026 (5.1 AAP3) and HQ912027 (4.7 AAP3).
The ORFs of each copy, that are 98% identic were clone into pYES2 plasmid and the recombinant plasmids were used to transform a Saccharomyces cerevisiae mutant lacking amino-acid transporter coding region: GAP1/ YHR039W (Material and Methods S1). The transformed yeast recovered the growth capacity in medium containing L-arginine confirming the transporter character of the protein encoded by those sequences ( Figure S1).
The two transcribed copies of the gene presented different 59 untranslated regions (59UTRs). This allowed for the design of specific primers to differentially quantify each copy by quantitative reverse transcription PCR (qRT-PCR). The alignment of the 59UTRs from the two copies, their differences and primers positions are shown in Figure S2.
Arginine uptake by L. (L.) amazonensis promastigotes correlates with the arginine transporter transcript abundance It is known that L. donovani promastigotes are sensitive to arginine starvation, and they respond with an increase in both arginine transporter expression and transport rate [28]. However, it is not completely clear at which level of protein-expression regulation this control occurs. We performed an arginine starvation on mid-log and stationary L. (L.) amazonensis promastigotes for 4 h at 25uC and then evaluated arginine uptake. Initially, we could note that at time 0 that means the physiological condition in each phase the arginine uptake in midlog phase parasites is lower than the one detected in the stationary phase parasites. When starved, the mid-log phase parasites showed an increase in the arginine uptake compared to the control parasites at time 0 (p,0.05). As this behavior was not detected in starved stationary phase parasites, we can conclude that stationary phase parasites do not respond to starvation ( Figure 1A). However, the increase in arginine uptake was not observed when the mid-log phase parasites were incubated for the same time in the presence of arginine (400 mM) ( Figure 1A), but the arginine uptake of the stationary phase parasites decrease, in relation to time 0, when incubated in the presence of arginine ( Figure 1A). We also evaluated mRNA level at midlog parasites, and the increase of arginine uptake correlated with an increase in the relative copy-number of the 5.1 AAP3 mRNA ( Figure 1B, p,0.05), suggesting the existence of at least one pretranslational mechanism for controlling protein expression. Moreover, only the 5.1 AAP3 mRNA was sensitive to the amino acid starvation. No differences were observed in the 4.7 AAP3 mRNA copy-number or the mRNAs coding for arginase and Meta1, all normalized by the GAPDH mRNA copy-number ( Figure 1B).
To evaluate differences in the amino acid-starvation response between mid-log and stationary parasites, we determined the amount of AAP3 mRNA from the 1 st to the 10 th day of a culture growth curve as described in methods. We compared the mRNA expression amounts of the 5.1 AAP3 mRNA and the 4.7 AAP3 mRNA normalized to GAPDH mRNA. The arginine transporter mRNA was increased more than 10 times in the stationary parasites compared to the log-phase parasites ( Figure 2). Interestingly, both copies increased the mRNA level in the stationary phase, although the 5.1 AAP3 mRNA was at least 30 fold more abundant than the 4.7 AAP3 mRNA at the log-phase, reaching 100 fold in the stationary phase.
Arginine-transporter-coding transcripts stability are affected in axenic promastigotes experiencing arginine deprivation Treatment with actinomycin and sinefungin causes an inhibition of transcription and trans-splicing mRNA-maturation processes in the parasites [32,33]. RNA obtained from a timecourse treatment of mid-log phase promastigotes with actinomycin D and sinefungin, preceded or not by 4 hours arginine starvation was used in qRT-PCR experiments, as described in Material and Methods. The 5.1 AAP3 mRNA showed no decay after 180 min of treatment in cells submitted to arginine starvation but the presence of the amino acid induces a degradation ( Figure Figure 3B and 3C). The qRT-PCR data were normalized by SSUrRNA copy-number, a RNA that is not sensitive to the inhibitor drugs ( Figure 3D). The arginine transporter transcript abundance also assesses the sensing of the arginine internal pool: an output relying on L. amazonensis genetic manipulation targerting the arginase-coding gene.
To evaluate the influence of an internal pool of arginine on its uptake, we used a L. (L.) amazonensis arginase-null mutant (arg -), which does not use arginine to produce ornithine and present higher amounts of arginine in their cytoplasm, requiring polyamines supplementation (Laranjeira da Silva, submitted). The mutant and WT parasites (5610 7 /mL), at initial stationary phase, were submitted to amino acid starvation and then arginine uptake was evaluated. Although both parasites responded to amino acid starvation, the WT parasites presented greater arginine uptake than the arg - (Figure 4). We performed the same assay using a knockout mutant that is genetically complemented with the arginase ORF (arg -/ARG) showing a partial recovery in arginase activity (Laranjeira da Silva, submitted). Interestingly, this mutant also presented a partial recovery in arginine uptake, compared to the WT. On the other hand, the complemented mutant, that contained the arginase ORF without the correct glycosomal compartmentalization signal (arg -/argDSKL) and did not present any arginase activity (Laranjeira da Silva, submitted), showed a arginine uptake similar to the argmutant (Figure 4).

Discussion
The data presented in this study show that L. (L.) amazonensis can control arginine uptake when promastigotes parasites are amino acid starved. These observations are similar to those made in L. donovani [28]. Adding to that data, we showed that the internal pool of the amino acid is also important to regulate the uptake. Furthermore, we showed that the higher concentration of one of the AAP3 transcripts is due to a stabilization process in the mature mRNA and not to an increase in the transcriptional rate or mRNA trans-splicing maturation. Most eukaryotes generally control their gene expression at the transcriptional level; however, Leishmania lacks this control mechanism because its transcription is polycistronic [34]. Besides, there are no known RNA polymerase II  promoter regions that have binding sites for transcriptional regulatory factors. However, these organisms can control gene expression at the mRNA maturation level (poly-adenylation/transsplicing coupled processes) or by changing mRNA half-lives in different conditions [30]. This digenetic organism experiences different environmental conditions, such as pH, temperature and nutrient availability, when it cycles between invertebrate and mammalian hosts [35]. Our findings suggest a possible mechanism for the parasite to overcome the different requirements due to environmental changes that they will find in insect gut along their development, going from a nutrient rich media after insect blood meal through an deprived ambient until the next feed.
A possible physiological explanation for the presence of two copies of the AAP3 gene is that each copy could be differentially regulated according to the environmental conditions of the parasite's differentiated stage. It is interesting that one of the arginine transporter mRNAs (4.7 AAP3 mRNA) presented the same typical decay behavior observed for GAPDH mRNA when promastigotes were treated with actinomycin and sinefungin. The other transcript (5.1 AAP3 mRNA) remains stable even 180 min after blocking of transcription and trans-splicing. This could explain the reason why this copy presents at least 30-fold more copies than the 4.7 AAP3 mRNA. Previous reports show that the mRNA stability in these organisms is altered by amino acid availability, but the molecular mechanisms by which this occurs are still unclear [36][37][38][39]. As arginine internal pool is also important in regulating arginine uptake, we may speculate if, in some way, the amino acid concentrations in extra-or intracellular medium could regulate AAP3 mRNA decay.
It is well known that arginine plays a critical role as substrate for macrophages to kill Leishmania via iNOS activity [19], but macrophage's arginine internal pool is not enough to provide substrate for either Th1 or Th2 responses [40], highlighting the importance of the arginine uptake control. As our results suggests, stationary phase promastigotes can uptake higher amounts of arginine than parasites in log phase. If so, the arginine influx into macrophages may be buffered if recently phagocytized Leishmania promastigotes, with increased arginine transport, sequester arginine from the macrophages. This is especially true when considering that macrophage CAT2 has no sensitivity to arginine concentrations [41].
L. (L.) major arginase-null mutants show a delay in starting lesions in in vivo infections [42]. The L. (L.) amazonensis arginasenull mutants also presented a delay in in vivo infectivity (Laranjeira da Silva, submitted). Under arginine starvation, the arginine uptake in the argmutants was less pronounced than in the WT parasites. This infective-capacity impairment may be attributed to both arginine uptake and/or arginase activity. The uptake pattern of argtended to be reverted in the mutant that was genetically complemented with the WT arginase ORF, but the pattern did not revert when the genetic complementation was done with an arginase ORF without the correct glycosomal addressing signal. This supports the idea that arginase exerts some control on arginine uptake and arginase activity, and arginine uptake is crucial to parasite survival inside macrophages. In addition, the lower arginine uptake that occurred in response to starvation in the null mutants indicates a mechanism to increase arginine transporter mRNA. This may occur through detecting possible changes in the internal arginine pool because the disruption of one arginine pathway decreases responses to amino acid starvation, which is also described in L. donovani ornithine decarboxylase or spermidine synthase-null mutants [28].
Stationary phase WT parasites have an increased expression of the arginine transporter, in relation to mid-log phase parasites, but did not respond to amino acid starvation. On the other hand, stationary phase argmutant responds to arginine starvation like mid-log phase WT parasites ( Figure 1A and Figure 4). A possible explanation is that the mutant uses less arginine present in the culture media, keeping a higher concentration compared to the WT growing media. The data suggests that the maximum expression of the transporter may have been already reached at WT stationary phase. Achieving maximum transporter expression may be a response to decreases in nutrient concentrations in the culture medium over time, and it may represent an adaptation to the low-nutrient availability found inside the fly mid-gut after blood digestion was completed. Another mechanism could be that nutrient depletion drives the modifications in promastigote parasites that induce differentiation in the infective stage [43]. Thus, arginine depletion may represent a signal to metacyclogenesis, although we did not observe changes in Meta1 mRNA due to amino acid starvation. This is possibly because we only observed them for 4 hours.
The results presented in this study lead us to conclude that arginine uptake is controlled by transporter-coding mRNA levels in L. (L.) amazonensis. They also suggest a mechanism that senses internal arginine concentrations and controls arginine uptake by increasing arginine transporter expression. This may represent a part of metacyclogenesis for achieving the infective stage.
The mRNA expression and arginine uptake were evaluated along the growth curve by maintaining the parasites at log phase by sub-culturing them every 24 h with the same initial cell ratio (5610 5 parasites/mL), as previously described [44]. AAP3 gene cloning Based on sequences described by Shaked-Mishan et al. [22], we amplified the L. (L.) donovani AAP3 ORF from genomic DNA, purified as described previously [45]. We used this amplicon as a template to construct a radioactive probe (0.1 mCi). This probe was constructed using a 32 P-dCTP (10 mCi/mL; 3,000 Ci/mmol; GE Healthcare, UK) and Amersham Megaprime DNA Labeling Systems (GE Healthcare, UK) following manufacturer's standard protocol. This probe was used to screen a L. (L.) amazonensis cosmid library [31] (kindly provided by S.R. Uliana, ICB-USP), and cosmid DNA was printed from bacterial-containing plates onto a nylon membrane [46]. Hybridization was performed at 42uC overnight followed by two separate, 20-min washes at ambient temperature and 50uC with SSC 2X SDS (0.1%). The nylon membrane was exposed to X-ray film (Kodak) and developed according the manufacturer's protocol. Selected clones were recovered in SOB growth media, and cosmid DNA was isolated by alkaline lysis [45]. Sequencing of the cosmid DNA was performed by the Sanger dideoxy protocol as described previously [47].
RNA purification, cDNA synthesis and qRT-PCR RNA was extracted with Trizol Reagent (Invitrogen) using the manufacturer's protocol. Reverse transcription was performed with a random primer protocol (Fermentas, M-MuLV RT) using total RNA (2 mg). The obtained cDNA was diluted in water and used in quantitative Real-Time PCR (qRT-PCR) with primers (Table 1) designed to differentially amplify the 59UTR region of the two copies of the AAP3 gene. The expected products were cloned and sequenced to validate the PCR. Known amounts of the cloned-DNA products were calculated (number of molecules), and they were used in the qRT-PCR to produce the standard curve. The following protocol was used in the qRT-PCR: 50 total cycles encompassing an association/fragment extension step at 61uC for 50 s and a denaturation step for 20 s at 94uC. A 7300 System (Applied Biosystems, USA) was used to run the qRT-PCR. The primers used to amplify GAPDH (internal control), arginase and SSUrRNA are described elsewhere [44].

Uptake assays
We adapted a protocol from dos Santos et al. [48] for the uptake assays. Briefly, promastigotes in the mid-log phase (2610 7 parasites/mL) or in the initial stationary phase (4.5610 7 ) were washed twice with cold Earle's Based Salt Solution (EBSS) (LGC Biotecnologia, SP, Brazil) and resuspended at 2610 8 parasites/ mL. We combined 50 mL of this mixture (10 7 parasites) with 50 mL of 3 H-Arginine (40 mM; 1 mCi/mL; GE Healthcare, UK) at 25uC. Uptake was stopped at different times by adding 50-mM ice-cold arginine (Ajinomoto, Tokio, Japan) (200 mL). Parasites were then washed twice with EBSS (200 mL), and radioactivity was measured by liquid scintillation spectrometry in a 2100 TR Packard Tri-Carb Liquid Scintillation Counter (PerkinElmer, USA).

Starvation assay
We used a protocol described by Daslyuk et al. [28] to starve the promastigotes, with one difference: the parasites were kept for 4 h at 25uC. Controls were performed at time 0 by putting the parasites on ice or incubating the parasites in the presence of arginine (400 mM).

Actinomycin and Sinefungin treatments
RNA half-lives were determined using actinomycin D (Sigma-Aldrich, MO, USA) (10 mg/mL) and sinefungin (Sigma-Aldrich, MO, USA ) (2 mg/mL) as described by Stewart & Clayton [32]. At different times, the parasites were placed in EBSS, and the treatments were stopped by lysing the parasites with Trizol Reagent (Invitrogen, USA) for RNA extraction.

Statistical data analysis
Statistical significance was determined by Student's t test (p,0.05).

Supporting Information
Material and Methods S1 Methodological description of genetic complementation of yeast mutant. (DOC) Figure S1 Genetic complementation of yeast mutant certifies the AAP functional character of L. amazonensis amino acid transporter ORFs. pYES2 plasmid DNA carrying 5.1 AAP3 and 4.7 AAP3 ORFs complemented Saccaromyces cerevisiae mutant GAP1/YHR039W. The transformed yeast clones, ORF 4.7 and ORF 5.1, obtained as described in Material and Methods S1 were able to growth in the minimal medium supplemented with 1 mg/m of L-arg (A). The mutant transformed with the recipient plasmid pYES2 alone, only grew in a medium containing ammonium as (NH4+) as nitrogen source (B). (TIF) Figure S2 Sequence alignment of the 59UTRs from 5.1 AAP3 mRNA and 4.7 AAP3 mRNA. Cyan-colored box represents the Spliced-Leader RNA sequence. Gray box represents the ORF beginning. Blue boxes show the primers for amplifying the 5.1 AAP3 mRNA, and the green boxes show the primers for amplifying the 4.7 AAP3 mRNA. (TIF)