Unlike all other organisms, parasitic protozoa of the family Trypanosomatidae maintain a large cellular pool of proline that, together with the alanine pool, serve as alternative carbon sources as well as reservoirs of organic osmolytes. These reflect adaptation to their insect vectors whose haemolymphs are exceptionally rich in the two amino acids. In the present study we identify and characterize a new neutral amino acid transporter, LdAAP24, that translocates proline and alanine across the Leishmania donovani plasma membrane. This transporter fulfils multiple functions: it is the sole supplier for the intracellular pool of proline and contributes to the alanine pool; it is essential for cell volume regulation after osmotic stress; and it regulates the transport and homoeostasis of glutamate and arginine, none of which are its substrates. Notably, we provide evidence that proline and alanine exhibit different roles in the parasitic response to hypotonic shock; alanine affects swelling, whereas proline influences the rate of volume recovery. On the basis of our data we suggest that LdAAP24 plays a key role in parasite adaptation to its varying environments in host and vector, a phenomenon essential for successful parasitism.
Leishmania are the causative agents of various cutaneous, mucocutaneous and visceral human diseases. These organisms cycle between the mid-gut of female sand flies, where they proliferate as extracellular flagellated promastigotes, and phagolysosomes of mammalian macrophages, where they develop as aflagellated intracellular amastigotes. During this life cycle, the environment of the organisms changes from relatively alkaline, sugar- and amino acid-rich, to acidic, fatty acid- and amino acid-rich [1–3]. The abundance of amino acids is a feature of both environments and, accordingly, the parasites display adaptive mechanisms that favour utilization of such compounds [3–6]. In particular, Leishmania exhibit large amino acid pools, of which alanine, proline and glutamate are the main components [5,7,8]. The parasites use these amino acids as alternative sources of carbon and as osmolytes.
In addition to their obvious function in protein synthesis, amino acids play fundamental roles in numerous processes, such as energy and nitrogen metabolism, cell growth and nucleobase synthesis. Neutral amino acids can also function as osmolytes, serving to counteract volume perturbations following a shift in extracellular osmolarity [6,9]. Furthermore, amino acids influence the differentiation of parasitic protozoa residing inside their hosts. The amino acids involved in all these processes are drawn from cellular pools, a feature of both prokaryotic and eukaryotic cells. Amino acid transporters mediate the pool supply by transporting amino acids across membranes. In the present study we provide evidence for multiple roles of a new Leishmania proline/alanine transporter in amino acid homoeostasis, metabolism and osmoregulation.
Volume regulation is a homoeostatic process that maintains thermodynamic equilibrium across the plasma membrane. Typically, cells swell in hypotonic environments and then release KCl, nonessential organic osmolytes and water to reduce cell volume, a process termed RVD (regulatory volume decrease). Conversely, in hypertonic environments, cells shrink due to osmotic water outflow, elevating the concentration of all cellular constituents, and to restore cell volume the cells accumulate organic osmolytes, a process termed RVI (regulatory volume increase). In general, neutral amino acids serving as organic osmolytes during RVD and RVI move through transporters and channels. Accordingly, in response to low water potentials, plants accumulate compatible solutes and up-regulate expression of proline/betaine transporters (ProTs) . Moreover, bacterial transporters, such as ProP, also a proline/betaine transporter, not only transport compatible solutes but also act as osmosensors . Mammalian cells employ the neutral amino acid transport system SNAT2 (sodium-coupled neutral amino acid transporter 2) to rapidly accumulate glutamine, proline, alanine and glycine and enable RVI upon exposure to hypertonic environments [12–14]. Indeed, expression of a betaine transporter of the SLC6 (solute carrier family 6) neurotransmitter family, BGT, is regulated by osmolarity and considered to participate in osmoregulation in the kidney . Similar to other eukaryotes, Leishmania and Trypanosoma spp activate massive amino acid efflux to counteract rapid swelling upon hypotonic stress [16–18]. In Leishmania, both the extent of swelling and rate of RVD are determined by the cellular pool of amino acids, cations and water flux. Although amino acids regulate the level of swelling, water flow through aquaporin 1 affects both the extent of swelling and rate of RVD [18,19]. Furthermore, it has been argued that only 50% of RVD is regulated by amino acids in Trypanosoma cruzi, whereas the rest is by water flow through the vacuolar aquaporin 1 channel and potassium [20,21]. This notwithstanding, in both Leishmania and Trypanosoma spp., a signal transduction pathway involving cAMP and phosphatidylinositol 3-kinase participates in regulation of RVD. For example, inhibition of the T. cruzi phosphatidylinositol 3-kinase homologe TcVps34 results in reduced RVD after hypotonic stress [22–24].
Previously, we showed biochemically that Leishmania donovani takes up proline via three systems: a promastigote-specific system A that has low specificity and affinity for proline and is cation-dependent; a system B that is more selective, promastigote-specific and cation independent; and system C, an amastigote-specific and cation-independent transporter . Similarly, proline transport in T. cruzi is considered to be mediated by more than a single transporter, each possessing distinct biochemical properties . In the present study we identify that the low affinity high capacity proline transporter in L. donovani (system A) is encoded by LdAAP24 and also translocates alanine. We show that LdAAP24 plays a critical role in amino acid homoeostasis and in response to hypotonic stress. LdAAP24 knockout resulted in a complete loss of the cellular pool of proline, and alanine concentrations were reduced. Notably, this null mutant also exhibited an impaired response to hypotonic shock. More detailed analysis of wild-type hypotonic shock responses revealed that, whereas alanine plays a dominant role in cell swelling, proline is required for effective volume recovery during RVD in promastigotes. In addition we show that LdAAP24 is a multipurpose transporter, also regulating arginine and glutamate transport as well as the cellular levels of arginine. In summary, this study provides the first genetic evidence that alanine and proline transport regulate osmotic stress responses and amino acid homoeostasis.
3H-labelled amino acids were from Amersham. NEM (N-ethylmaleimide) , G418, hygromycin B, phleomycin, M-199 (medium-199) and non-labelled L-amino acids were from Sigma. FBS (fetal bovine serum) was from Biological Industries. All other reagents were of analytical grade.
Leishmania cell culture
A clonal line of L. donovani 1SR was used in all experiments . To ensure the clonal nature of the cell line, fresh cultures were inoculated using single colonies of promastigotes taken from M-199 agar plates. Promastigotes were grown at 26°C in M-199 supplemented with 10% FBS. Cells were starved of proline and alanine by centrifugation and resuspension in M-199 lacking glucose and amino acids; other amino acids were supplemented individually. Amino acids were removed from the FBS by overnight dialysis in PBS at 4°C.
Yeast strains and growth conditions
The proline transport-deficient Saccharomyces cerevisiae strain 22574d  was used in complementation and transport assays. Cells were grown in SD media (0.17% yeast nitrogen base without amino acids, 0.5% ammonium sulphate and 2% glucose). In complementation assays, cells were plated on minimal buffered (pH 6.1) medium  supplemented with proline (1 g/l) and incubated at 30°C.
Generation of antibodies, Western blot analysis and immunofluorescence
A 264 bp DNA fragment encoding 88 amino acids from the N-terminus of LdAAP24 was amplified and then cloned into the expression vector pGEX 4T-1 (GE Healthcare). The resultant GST (glutathione transferase)–LdAAP24(1–88) chimaera was expressed in Escherichia coli strain E-151. A total of 1 litre of log-phase E. coli cells were exposed to 1 mM IPTG (isopropyl β-D-thiogalactopyranoside) for 3 h to induce protein expression. Cells were centrifuged and resuspended in 1:50 (v/v) bacterial protein extraction reagent (B-per-78248 Thermo Scientific) supplemented with 0.03 μg ml−1 DNase and protease inhibitor cocktail (P-8340, Sigma). The resulting cell extract was cleared by centrifugation before incubation with 50 mg glutathione–agarose beads (Sigma) for 30 min at 4°C. The beads were washed three times with ice-cold PBS. Protein was eluted using 10 mM GSH in 50 mM Tris, pH 8. This purified protein was injected into rabbits. Rabbit sera were processed by Sigma (Israel) to generate anti-LdAAP24 antibodies.
For Western blot analysis, 108 log phase promastigotes were centrifuged, washed twice in ice-cold PBS and re-centrifuged. The resulting pellet was resuspended in Laemmli buffer and then sonicated for 2 s to ensure denaturation. Total protein was subjected to SDS/PAGE (mini-PROTEAN; Bio-Rad Laboratories) before being transferred to a nitrocellulose membrane. Membranes were blocked with 10% (w/v) non-fat dried skimmed milk powder in PBST (PBS containing 0.1% Tween 20), incubated with anti-LdAAP24 (1:2000 dilution) or anti-HSP90 antibodies for 1 h at room temperature (22°C), washed and then incubated with secondary goat anti-rabbit HRP (horseradish peroxidase) antibodies (1:10000 dilution).
For immunofluorescence, mid-log promastigotes were washed twice in PBS and then fixed in 1% formaldehyde/PBS on a slide for 10 min before permeabilization by exposure to 0.2% Triton X-100/PBS for 10 min. Cells were incubated with blocking solution [10% (v/v) non-fat dried skimmed milk powder/PBST] for 30 min at room temperature, incubated with anti-LdAAP24 antibodies (1:200 dilution) for 1 h and then incubated with secondary polyclonal goat anti-rabbit IgG fluorescent antibodies (1:500 dilution; Dy-light 549 Red; Jackson) for 1 h in the dark at room temperature. Finally, cells were washed in PBST and supplemented with 5 μl DAPI (4′,6-diamidino-2-phenylindole; 0.5 μg/ml; Fluka). Fluorescence analyses were carried out using a fluorescent microscope (Axiovert 200M; Zeiss).
DNA and RNA work
For S. cerevisiae complementation and transport assays, PCR-amplified LdAAP24 (LinJ10_V3.0760/70), TcAAP24 (Tc00.10470535040) and TbAAP24 (Tb927.8.5450) ORFs (open reading frames) were cloned into the yeast expression vector pDR197  between the BamHI and EcoRI sites. S. cerevisiae was transformed according to Dohmen et al. .
LdAAP24 was knocked out by gene replacement to generate ∆ldaap24 cells as follows. The 952 bp comprising part of LdAAP24 (LinJ10_V3.0760) 5′-flanking region was cloned upstream to a hygromycin resistance cassette between the 5′-SalI and 3′-HindIII sites in plasmid pKOH . The 821 bp comprising LdAAP24 (LinJ10_V3.0770) 3′-flanking region was cloned downstream of the same hygromycin-resistance cassette between the 5′-BamHI and 3′-XbaI sites. The resulting construct that contained LdAAP24 5′- and 3′-flanking regions surrounding the hygromycin resistance cassette was amplified by PCR and electroporated into L. donovani promastigotes using standard conditions . Colonies expressing the construct were selected on M-199 plates containing 50 μg/ml hygromycin B. Similar cloning was performed to generate a second construct containing the same 5′- and 3′-flanking regions surrounding the G-418 resistance cassette in the pKON plasmid . Hygromycin B-resistant colonies from step one were electroporated with the second construct and L. donovani colonies expressing both constructs were selected on M-199 plates containing 50 μg/ml of both G-418 and hygromycin B. Primers used in this procedure are detailed in Supplementary Table S1 (at http://www.biochemj.org/bj/449/bj4490555add.htm). Three colonies of doubly resistant cells were screened by PCR using forward primers from within the LdAAP24 or hygromycin/neomycin-resistance genes and a reverse primer located downstream of the 3′-UTR (untranslated region) to confirm that the constructs were incorporated into the correct loci in the genome. A rescue vector was constructed by cloning the LdAAP24 ORF into pNUS HnB . Successful replacement and rescue was validated by Northern and Western analyses. LdAAP24 knockout procedure was repeated independently twice.
For Northern blot analysis, total RNA was extracted using the Tri reagent method (MRC). Total RNA (5 μg) was subjected to electrophoresis in a 1.2% formaldehyde agarose gel in 1× Mops buffer. RNA was transferred to a Hybond N membrane (Amersham Biosciences) overnight in 10× SSC and cross-linked by 3 min of UV irradiation. Membranes were probed overnight at 50°C with LdAAP24 ORF, the latter labelled with Dig-dUTP using the PCR DIG probe synthesis kit (Roche Diagnostics). Probe hybridization and mRNA detection was carried out using the DIG Northern starter kit according to the manufacturer's instructions (Roche Diagnostics).
For standard S. cerevisiae uptake studies, 22574d cells were grown to logarithmic phase. Cells at D600=0.8 were harvested, washed twice in water and resuspended in 0.6 M sorbitol to a final D600 of 8. The pH was adjusted by diluting the cells using 0.33 vol. of 200 mM phosphate buffer (pH 4.5–7.5)/0.6 M sorbitol to a final D600 of 6. Prior to transport measurements, the cells were supplemented with 100 mM glucose and incubated for 5 min at 30°C. To start the transport reaction, 130 μl of this cell suspension was added to 130 μl of matching buffer containing 37–92.5 kBq labelled L-[3H]proline and specified amounts of unlabelled amino acids. The transport activity of S. cerevisiae mutants transformed with the empty vector pDR197 served as background and this value was subtracted from all observed rates. Samples were removed after 30, 60, 120, 180 and 300 s, transferred to 4 ml of ice-cold potassium phosphate buffer (50 mM, at appropriate pH), filtered on glass fibre filters and washed with 9 ml of the same buffer. The uptake of 3H was determined by liquid scintillation spectrometry. Transport reactions were repeated independently and the reported values represent the means of at least three experiments.
For standard uptake studies in L. donovani promastigotes, cells were grown to logarithmic phase, washed twice in ice-cold Earl's buffer and concentrated to 108 cells/ml. The cell suspension was mixed with reaction mixture (Earl's buffer, 5 mM glucose, 10 mM Tris and 10 mM succinate) at the relevant pH to a final volume of 600 μl. Cells were then pre-incubated at 30°C for 10 min. Transport was started by mixing the cell suspension with a specified concentration of non-labelled amino acids and 3H (or 14C for glutamate) L-amino acid substrates in Earl's buffer at the relevant pH. At 30, 60, 120, 180 and 300 s, 100 μl of the mixture was removed and put directly on to 24 mm GF/C glass microfibre filters (Whatman 1822 024). Filters were washed twice in ice-cold Earl's buffer at the relevant pH and soaked in scintillation liquid. Uptake of 3H or 14C was determined by liquid scintillation counting. In all time-course analyses, i.e. Figures 1(B), 3(E), 5(A), 5(B), 6(A), 8 and 10, representative values of individual experiments are shown, but correlations between the independent repeats was r>0.9.
LdAAP24 and its orthologues mediate proline transport in S. cerevisiae
Assaying proline incorporation into proteins
Promastigotes at 106 log phase were centrifuged, resuspended in fresh M-199 and pre-incubated for 1 h at 26°C. L-[3H]proline (5 μCi/ml) was added to the medium. Aliquots (1 ml) were harvested at indicated time points, washed in ice-cold PBS and added to 5 ml ice-cold 10% TCA (trichloroacetic acid). The suspension was vacuum filtered on to GF/C glass fibre filters (Whatman 1822 024). Filters were washed initially in 10% ice-cold TCA and then in ice-cold 96% ethanol. Incorporation of 3H-labelled L-proline was determined by liquid scintillation spectrometry.
Assaying cellular amino acid pools
Promastigotes [5×108 or 5×109 (for analysis in HPLC or amino acid analyser respectively)] log phase were centrifuged, washed twice and resuspended in ice-cold PBS. The cells were mixed with 1 ml of 1 M ice-cold PCA (perchloric acid), vortexed and incubated for 10 min at room temperature. Samples were sent to Aminolab (Weizmann Science Park, Rehovot, Israel) for analysis by an amino acid analyser. For analysis using HPLC, the cell suspension was centrifuged again (20000 g, 4°C, 10 min). The supernatant was collected and supplemented with 232 μl of 5 M KOH to neutralize the pH. Additional centrifugation was performed under the same conditions, and the supernatant was collected and analysed according to the method of Fekkes et al. . HPLC analysis was carried out at the Medical Biochemistry Laboratory (Rambam Medical Center, Haifa, Israel). The intracellular amino acid concentration was calculated assuming the cell volume of 108 promastigotes to be 4.2 μl, as determined previously .
Volume regulation assays
Promastigotes (5×107) log phase at D550=0.5 were centrifuged, washed twice in isotonic chloride buffer (300 mOsm ), resuspended in isotonic buffer and kept on ice. Cells were pre-incubated for 5 min at 26°C and then transferred to a spectrophotometer vessel. To start the reaction, sterile double distilled water and isotonic buffer at 26°C were added in the relevant ratios such that the cells were exposed to the desired osmolarity (150, 75 and 60 mOsm). D values (550 nm) were measured at the indicated time points at 26°C.
LdAAP24, TbAAP24 and TcAAP24 mediate proline transport in S. cerevisiae
L. donovani LdAAP24 (accession number EU391425), and its closest homologues in Trypanosoma brucei, TbAAP24 (Tb927.8.5450), and T. cruzi, TcAAP24 (Tc00.1047053504069.120), are syntenic genes that encode proteins comprising 488, 488 and 476 amino acids respectively. Each contains 11 predicted transmembrane domains and belongs to the AAAP (auxin/amino acid permease) family ([39,40] and http://tritrypdb.org/tritrypdb/). To investigate their function, these proteins were expressed in S. cerevisiae mutants that were impaired in the uptake of various amino acids. As shown in Figure 1(A), all three permeases mediated growth of the proline-transport-deficient strain 22574d on selective concentrations of proline . Consistent with growth on proline, expression in strain 22574d was associated with linear L-[3H]proline uptake over time (Figure 1B). As expected, strain 22574d transformed with empty vector did not transport proline. Notably, LdAAP24- and TcAAP24-expressing 22574d cells took up proline more effectively than yeast expressing TbAAP24. Next, we performed kinetic analysis of proline uptake by LdAAP24- and TcAAP24-expressing 22574d cells and observed low affinity for proline with apparent Km values of 0.63±0.12 and 1.6±0.32 mM respectively (Figures 2A and 2C). Proline uptake by S. cerevisiae expressing TcAAP24 was optimal at an acidic pH (Figure 2D), whereas uptake by yeast expressing LdAAP24 exhibited less pronounced pH dependence, with the optimum shifted to neutral pH (Figure 2B), highlighting functional differences between the two transporter homologues. Competition assays corroborated preferential recognition of proline and revealed weak competition by alanine when competitors were in 10-fold excess (Supplementary Table S2 at http://www.biochemj.org/bj/449/bj4490555add.htm). Taken together, these data demonstrate that LdAAP24, TbAAP24 and TcAAP24 each encode a permease that mediates proline uptake when expressed in S. cerevisiae.
Kinetic parameters of proline transport by S. cerevisiae 22574d expressing LdAAP24 or TcAAP24
LdAAP24 is a bona fide L. donovani proline/alanine transporter
There are two identical adjacent copies of LdAAP24 in the Leishmania infantum genome on the reverse strand of chromosome 10 that generate three mRNA bands sized 5.7, 5 and 4 kb in promastigotes (Figure 3C, +/+ lane). The relatively long transcripts resemble transcripts of other AAP genes, such as lysine (LdAAP7) and arginine (LdAAP3) transporters [41,42] in Leishmania and the mammalian cation amino acid transporter 1 . To facilitate our investigation of LdAAP24 function, we raised polyclonal antibodies against a recombinant peptide containing the first 88 amino acids of the LdAAP24 N-terminus (see the Experimental section). This antiserum recognized with high specificity two proteins of 42 and 39 kDa present in the promastigote extract (Figure 3D, +/+ lane). To characterize functionally LdAAP24 in L. donovani promastigotes, we deleted the two LdAAP24 copies of both alleles (all four LdAAP24 ORFs) using the gene replacement technique originally developed by Cruz et al. . Both copies of LdAAP24 on the same chromosome were replaced with the gene coding for hygromycin resistance to obtain heterozygous mutants and subsequently both copies on the second allele were replaced with the gene coding for neomycin resistance to generate null mutants (Figure 3A). PCR analysis demonstrated that the two antibiotic resistance genes were inserted in the correct orientation in the genome and amplification of the LdAAP24 ORF showed that all copies were absent from the genome, indicating that the replacements were successful, resulting in the knockout mutant termed ∆ldaap24 (Figure 3B). Northern and Western blot analyses confirmed that heterozygous ldaap24/hygro expressed half of the endogenous mRNA, whereas ∆ldaap24 did not express LdAAP24 transcript and protein (Figures 3C and 3D). Importantly, transport analysis validated that promastigotes of heterozygous ldaap24/hygro exhibited half and Δldaap24 promastigotes exhibited almost no proline transport activity compared with wild-type (Figure 3E, compare filled squares with filled circles). Ectopic expression of wild-type LdAAP24 in Δldaap24 restored proline uptake to a level identical with that determined in heterozygous ldaap24/hygro parasites (Figure 3E).
Generating the L. donovani LdAAP24 knock out line ∆ldaap24
Immunofluorescence microscopy studies performed using the anti-LdAAP24 antiserum revealed that LdAAP24 is located at the surface membrane of wild-type L. donovani promastigotes (Figure 4A). No antibody binding on the plasma membrane was observed in Δldaap24 mutants (Figure 4B). Fluorescence observed near the flagella pocket in the knockout cells indicates non-specific antiserum binding, because these cells do not express LdAAP24 at all (see Figures 3C and 3D). Similar to wild-type, ectopically expressed LdAAP24 protein is present on the plasma membrane of Δldaap24 promastigotes (Figure 4C). Cumulatively, these data establish that LdAAP24 encodes a bona fide proline transporter in the plasma membrane of L. donovani cells.
LdAAP24 is localized at the plasma membrane of L. donovani promastigotes
The transport analysis of the ∆ldaap24 knockout mutant described above (Figure 3E) was performed using 1 mM proline and so did not rule out the existence of additional proline transporters with higher or lower affinities. To determine whether promastigotes express other proline transporters, we measured proline incorporation into proteins in wild-type and ∆ldaap24 cells (Figure 5A). Proline incorporation was lower in the ∆ldaap24 mutant, but significant enough to postulate the existence of one or more additional proline transporter(s), probably high-affinity transporter(s). To corroborate this premise, we measured the transport of 1 μM proline in the absence and presence of the sulfhydryl inhibitor NEM. ∆ldaap24 cells were observed to transport proline at a rate of 11 pmol/min per 108 cells, and this transport was inhibited completely by NEM (Figure 5B), supporting the fact that the L. donovani genome encodes at least one other permease that transports proline. The existence of at least two distinct proline transport systems agrees with our previous biochemical analysis  that indicated in promastigotes a proton-driven, low specificity, cation-dependent transport system, which we termed system A, as well as a selective, cation-independent system termed system B. To delineate which transport system is encoded by LdAAP24, we examined the cation dependence of proline uptake in wild-type and ∆ldaap24 cells. Proline transport in wild-type cells was observed to be dependent on the presence of cations (K+ and Na+), whereas in ∆ldaap24 cells transport was cation-independent (Figures 5C and 5D). These data not only confirm the existence of two biochemically distinct proline transport systems as postulated in Mazareb et al. , but also identify LdAAP24 to be the cation-dependent system A.
LdAAP24 is the system A proline transporter
The competition uptake analyses performed using S. cerevisiae expressing LdAAP24 suggested that proline is not the sole substrate of LdAAP24 (Supplementary Table S2). Kinetic analysis of the initial alanine transport rate in L. donovani promastigotes indicated an apparent Km of 1 mM for alanine (results not shown). In agreement with LdAAP24 also functioning as an alanine transporter, transport of 1 mM alanine was significantly decreased in ∆ldaap24 cells relative to wild-type cells, 1.15±0.22 compared with 3.37±1.14 nmol/min per 108 cells in wild-type cells. Moreover, LdAAP24 expression in ∆ldaap24 cells restored this alanine transport by ~50% (Figure 6A). Furthermore, proline competed strongly with alanine transport in wild-type promastigotes, but not in ∆ldaap24 cells (Figure 6B), supporting that alanine is indeed a substrate of LdAAP24. Of note, these results indicate that alanine uptake in promastigotes is mediated by more than one transporter, the other being proline insensitive. In summary, our findings establish that LdAAP24 encodes a low specificity, cation-dependent proline transporter (system A) that also transports alanine.
Alanine is a substrate of LdAAP24
LdAAP24 influences amino acid pool composition and osmotic stress responses
To investigate whether LdAAP24 expression influences cellular amino acid pool composition, we measured the concentrations of cellular amino acids in log phase wild-type, ∆ldaap24 and rescue L. donovani promastigotes. Proline levels were 7.9±3.7 mM in wild-type (Figure 7 and Supplementary Table S3 at http://www.biochemj.org/bj/449/bj4490555add.htm), barely detectable in null cells and restored ~50% in the rescue cells, indicating that LdAAP24 serves as the sole contributor to the cellular proline pool. In contrast, we found cellular alanine levels to be 26.5±1.2 mM in wild-type and reduced by only 35% in ∆ldaap24 cells (17±5 mM; P=0.01), but restored fully in rescue cells (Figure 7 and Supplementary Table S3), indicating that alanine pools are supported by another transporter and/or alanine biosynthesis. Remarkably, ∆ldaap24 cells also exhibited changes in the pool level of arginine, whereas all other amino acids remained unchanged (Supplementary Table S3). The cellular concentration of arginine increased from 4.1±0.4 in wild-type cells to 7.8±0.6 mM in ∆ldaap24 cells and was restored back to wild-type levels in the rescue cells. In general, the cellular concentrations of amino acids are somewhat variable; however, the trend that LdAAP24 expression in ∆ldaap24 cells restored levels towards that observed in wild-type supports the biological significance of the aforementioned pool changes. In summary, LdAAP24 expression influences not only the cellular pool levels of its substrates (proline and alanine) but also the arginine pool.
LdAAP24 influences the amino acid pool composition in promastigotes
Having established that LdAAP24 expression affects cellular amino acid pool composition, and since L. donovani pool levels have been shown to influence responses to osmolarity changes , we checked whether this transporter participates in cell volume regulation upon osmotic shock. Wild-type, ∆ldaap24 and rescue L. donovani promastigotes were exposed to hypotonic stress and cell volume changes were monitored. A 2-fold hypotonic shift (from 300 to 150 mOsM) caused wild-type cells to swell rapidly to 1.35-fold their original cell volume within 1.5 min (Figure 8A). Subsequently, within 10 min the volume of wild-type cells reached the new set point at 150 mOsM, 1.19-fold of the isotonic volume (Figure 8A). More severe hypotonic stresses, shifts to 75 mOsM (Figure 8B) and 60 mOsM (Figure 8C; note difference in time scale), resulted in wild-type cells swelling to 2- and 2.2-fold their isotonic size respectively, and slower RVD times of more than 20 and 60 min respectively. ∆ldaap24 cells responded to the 2-fold hypotonic shift similarly to wild-type cells (Figure 8A). However, more pronounced swelling of ∆ldaap24 cells to 2.3- and 2.8-fold the isotonic volume was observed after exposure to 75 (Figure 8B) and 60 mOsM (Figure 8C) respectively. Moreover, when exposed to the more severe hypotonic shifts, RVD rates were significantly slower in ∆ldaap24 than wild-type cells, with the null cells exposed to 60 mOsM exhibiting a RVD rate close to zero (Figure 8C). Notably, ectopic expression of LdAAP24 in ∆ldaap24 resulted in partially recovered RVD responses, but had little impact on the extent of swelling. These results corroborate that the extent of swelling and rate of RVD in L. donovani is proportional to osmotic stress and reveal a role for LdAAP24 in responses to osmotic shock, most likely due to the involvement of this transporter in amino acid pool regulation.
Promastigote response to hypotonic stress
To substantiate that the role of LdAAP24 in responding to osmotic stress involves transport of proline and/or alanine, wild-type L. donovani promastigotes were grown in either alanine- or proline-depleted conditions and responses to osmotic stress were evaluated. Initially, we characterized the effect of the depleted conditions on growth, cellular amino acid pools and LdAAP24 protein expression. Specifically, promastigotes were transferred from M-199 to modified medium lacking either proline or alanine, supplemented with 10% dialysed serum, and growth was monitored for up to 72 h at 26°C before the cells were submitted to biochemical analysis. The parasites were intact and fully viable under the depleted conditions, even though growth rates were slightly slower than in complete medium. Biochemical analysis indicated that promastigotes grown in proline-free medium exhibited a depleted proline pool, normal alanine levels and a 2.5-fold higher arginine level (Figure 9A), whereas cells grown in alanine-free medium displayed a depleted proline pool, halved alanine levels (from 26±1.9 mM to 12±2.3 mM in complete compared with alanine-free medium) and no changes in the cellular arginine pool. Notably, cellular glutamate levels were reduced by at least half in both proline- and alanine-starved cells, a change that was not observed with ∆ldaap24 (compare with Figure 7). The finding that alanine-starved cells exhibit zero cellular proline levels despite the availability of proline in the growth medium suggests that parasites cannot survive without a pool of alanine and under these conditions synthesize alanine from proline [3,45]. Notably, 72 h of proline or alanine starvation resulted in ~50% reduced LdAAP24 abundance (not shown), which was associated with reduced uptake of both amino acids in transport assays (Figure 9B).
Proline and alanine starvation influences amino acid pool composition and LdAAP24 transport activity
Having characterized cells starved of proline or alanine, we subjected each population to hypotonic stress and monitored ensuing volume changes. Cells starved of proline responded to a 2-fold hypotonic shift in a similar manner to cells grown in full medium (Figure 8D). When the hypotonic shift was increased to 75 mOsM (Figure 8E) or 60 mOsM (Figure 8F), proline-depleted and non-starved cells swelled equally, but the RVD rates of the former were significantly slower. Similarly, cells starved of alanine responded to a 2-fold hypotonic shift comparably to cells grown in full medium. However, when the hypotonic shift was increased to 75 mOsm (Figure 8E) or 60 mOsm (Figure 8F), the volume of the alanine-depleted cells was significantly smaller than that of promastigotes grown in full medium. Indeed, under these conditions, cells starved of alanine swelled only 1.90-fold, whereas non-starved cells and proline-starved cells swelled to 2.6-fold the isotonic size. As with proline-starved cells, the RVD rates of alanine-starved cells after shifting to 75 and 60 mOsm were slower than those of non-starved promastigotes. Since alanine-starved cells did not swell up so much, these cells reached the new set volumes within 20 and 60 min in 60 and 75 mOsm respectively (Figures 8E and 8F), comparable with the times taken by non-starved cells to attain the same set points. Taken together, the data support that proline is indeed important for RVD in response to hypotonic stress, whereas alanine influences the extent of swelling.
LdAAP24 affects other transport systems
The observation that the LdAAP24 knockout increased arginine levels and proline and alanine starvation altered the cellular pools of arginine and glutamate (Figures 7 and 9) prompted us to assess a role for proline and LdAAP24 in transport of these amino acids. We measured initial transport rates of glutamate and arginine in wild-type, ∆ldaap24 and rescue cells in the absence and presence of proline. In the absence of any competitor, uptake rates of glutamate and arginine were comparable in the three cell types. However, when proline was present in 10-fold excess, the transport of each amino acid was inhibited by greater than 50% in wild-type cells but was unaffected in ∆ldaap24 cells (Figures 10A and 10B, left-hand and centre bars). The proline sensitivity of glutamate and arginine transport was partially restored in the rescue cells (Figures 10A and 10B, right-hand bars). In contrast, addition of 10-fold excess alanine to the assay mixtures of arginine (Figure 10C) or glutamate (results not shown) had no inhibitory effect on transport. Expressing LdAAP24 in S. cerevisiae 30.537a and 21983c, which lack glutamate and arginine transporters respectively, did not support growth on either glutamate or arginine nor did LdAAP24 mediate transport of these radiolabelled amino acids, validating that they are not substrates of LdAAP24 (results not shown). Hence LdAAP24 must directly or indirectly influence glutamate and arginine transport and homoeostasis in a proline-dependent manner.
LdAAP24 affects other transport systems
In the present study we identify and characterize a new neutral amino acid transporter, LdAAP24, which translocates proline and alanine across the plasma membrane of L. donovani promastigotes. We have provided genetic evidence demonstrating that this transporter is a key player in determining the size and content of the cellular amino acid pool. Specifically, we found that alanine, proline and glutamate are the major pool constituents in wild-type promastigotes, whereas in mutant promastigotes where the two LdAAP24 genes have been deleted, the amino acid pool lacks proline, alanine levels are lower and the concentration of arginine is elevated. To date, the only amino acid transporter shown to influence cellular amino acid homoeostasis in eukaryotic non-plant cells is the mammalian neutral amino acid transporter SNAT2, which translocates glutamine, proline, alanine and glycine in mammalian cells, the host of Leishmania. In fibroblasts, glutamine, glutamate, glycine and threonine are the major constituents of the cellular amino acid pool. Silencing SNAT2 expression in these cells using RNAi (RNA interference) was observed to reduce the cellular level of glutamine, the most abundant amino acid in the pool . Notably, despite the different compositions of Leishmania and fibroblast amino acid pools, in both cells the response to osmotic shock was disturbed when the amino acid transporter involved in homoeostasis was knocked out; in promastigotes, LdAAP24 deletion affected hypotonic stress responses and in fibroblasts, SNAT2 silencing compromised responses to hypertonic shifts.
The present study corroborates that amino acids are involved in both swelling and volume recovery in L. donovani; alanine availability affects the extent of cell swelling, whereas proline influences the RVD rate. Given that proline is an abundant constituent of the amino acid pool in promastigotes, it was surprising to discover that proline has a role only in RVD and not in swelling. Taking the results of the present study together with earlier reports [16,19,20], we suggest that, in Trypanosomatids, RVD is mediated either by water and proline fluxes, or by proline-associated aquaporin 1 activation. Previously the response of Leishmania to a 2-fold shift in osmolarity (300–150 mOsM) has been studied [7,18], a hypotonic shift that induces modest swelling of 1.3-fold. In the present study we establish that promastigotes can tolerate significantly larger osmotic gradients that incur more extreme volume changes. Cells were exposed to 4- and 5-fold hypotonic shifts that resulted in volume increases of 2- and 2.2-fold respectively, but remained viable. Even in these more extreme hypotonic conditions, the RVD rates were fast and cell volume was reduced to the same set point exhibited by cells exposed to only a 2-fold hypotonic shift (Figure 8). Moreover, we observed that Δldaap24 cells were capable of tolerating a volume increase of approximately 2.5-fold (Figure 8C). To our knowledge, this is the largest hypotonic swelling of viable cells reported to date. It is likely that the rigid promastigote cytoskeleton comprising subpelicular microtubules is responsible for this large tolerance to extreme volume changes. Enhanced sensitivity to osmotic stress was exhibited by Δldaap24 cells only upon hypotonic shifts of 4-fold or greater, with the mutants unable to activate RVD upon a 5-fold shift (Figure 8C). These observations support that the effect of LdAAP24 on hypotonic tolerance is probably due to its role in allowing sufficient accumulation of intracellular proline.
Unexpectedly, in addition to its role in osmotic shock responses, we discovered that LdAAP24 also influences transport and metabolism of other amino acids in promastigotes. Its expression affects transport rates of glutamate and arginine when proline, but not alanine, is present. Heterologous expression of the L. donovani arginine transporter LdAAP3 in S. cerevisiae revealed that arginine transport is not sensitive to proline . However, in accord with the premise that LdAAP24 somehow inhibits arginine transport, a significant (2-fold) increase in the cellular concentration of arginine was observed in Δldaap24 cells. In general, alterations in arginine pool levels are surprising, since arginine is an essential amino acid and current genome data do not indicate any pathway for arginine biosynthesis in Leishmania . We envision three possible models for how LdAAP24 affects the homoeostasis and transport of non-substrate amino acids. In the first model, LdAAP24 interacts physically with other transporters. More specifically, we speculate that its long hydrophilic N-terminus forms heteromeric complexes with other transporters and proline acts as an allosteric ligand mediating such interactions. A long hydrophilic N-terminus is a feature common to all Leishmania AAPs . In the second model, LdAAP24 activity induces a signalling pathway that attenuates the transport of other amino acids. The third model is based on parasite bioenergetics, since proline uptake is driven by the PMF (protonmotive force) . Rapid proline transport is coupled with inward translocation of protons that most likely reduce the size of the PMF across promastigote plasma membranes, thereby affecting transport of other PMF-driven systems. Currently, we are assessing these models. The models demand that a sensing mechanism exists that monitors the amino acid composition of the cellular or extracellular environments. Notably, we observed that the effect of LdAAP24 on arginine, glutamate and glycine transport required the presence of both LdAAP24 itself and its substrate proline. In light of this, we propose that attenuation of amino acid transport by LdAAP24 is activated by sensing extracellular proline and speculate that these mechanisms serve to maintain high cellular levels of proline and alanine.
Mazareb et al.  described, using biochemical studies, the existence of a promastigote-specific transport system A that has low specificity and affinity for proline and is cation-dependent . On the basis of competition analyses it was speculated that system A is a general amino acid permease. In the present study we establish that system A is encoded by LdAAP24 that preferentially translocates proline and alanine. Notably, the low specificity and affinity for proline was obvious only in wild-type promastigotes, where a 10-fold excess of amino acid competitors was sufficient to inhibit proline transport (compare Supplementary Table S2 with ). In S. cerevisiae expressing LdAAP24, amino acid competitors had to be in 100-fold excess in order to observe the same inhibition profile (results not shown). These data suggest that the phenotype of LdAAP24 is different when expressed heterologously compared with in its natural environment.
In summary, the present study identifies a new neutral amino acid transporter in the unicellular family of Trypanosomatid pathogens that plays multiple roles in osmoregulation, amino acid transport and amino acid homoeostasis. In light of its versatility, we hypothesize that LdAAP24 plays a vital role in host–pathogen interactions and thus represents a potential therapeutic target. In support of this premise, recent reports indicated that knocking down TbAAP24 (AAT6, Tb927.8.5450) expression using RNAi induced a >5-fold resistance to eflornithine in T. brucei [48–50]. Eflornithine is an efficient inhibitor of ornithine decarboxylase and currently represents a first-line drug against sleeping sickness. These reports taken together with the present study raises the idea that a combination therapy, where an agent that manipulates TbAAP24 expression/activity is combined with eflornithine, could treat sleeping sickness more effectively.
Ehud Inbar initiated this work and performed most of the experiments, including Figures 1(B), 2–5 and 7–9. He was also involved in writing the paper. Doreen Schlisselberg performed the experiments in Figures 6 and 10. Marianne Suter Grotemeyer performed the analyses with yeast that are illustrated in Figures 1(A) and 1(B). Doris Rentsch supervised all the yeast experiments and was involved in the discussions and writing the paper. Dan Zilberstein supervised the experiments in Leishmania (Doreen Schlisselberg and Ehud Inbar are his graduate students) and wrote the majority of the paper.
We thank Professor Isabel Roditi of Bern University for the knockout constructs and Dr Tanya Gottlieb for editing this paper prior to submission.
This work was supported by The Israel Science Foundation, founded by The Academy of Sciences and Humanities [grant number 402/08], and by the Swiss National Science Foundation [grant number CRSII3 127300].
Present address: Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892, U.S.A.