An acidocalcisomal exopolyphosphatase from Leishmania major with high affinity for short chain polyphosphate.

We report the cloning, overexpression, purification, and characterization of the Leishmania major exopolyphosphatase (LmPPX). The product of this gene (LmPPX), the first related to polyphosphate (polyP) metabolism isolated from an eukaryotic organism different from yeast, has 388 amino acids and a molecular mass of 48 kDa. LmPPX differs from other exopolyphosphatases previously investigated. Heterologous expression of LmPPX in Escherichia coli produced a functional enzyme that was similar to the yeast exopolyphosphatase with respect to its Mg(2+) requirement, optimum pH, and sensitivity to cations, amino acids, and heparin but that, in contrast to the yeast enzyme and other exopolyphosphatases investigated before, acts on polyP of short chain lengths with higher rates and affinity. LmPPX is a processive enzyme, and it does not hydrolyze pyrophosphate, ATP, or p-nitrophenylphosphate. Confocal immunofluorescence microscopy using affinity-purified antibodies against the recombinant enzyme indicated an acidocalcisomal and cytosolic localization. High levels of short chain (21.4 +/- 3.0 mm) and long chain polyP (55.9 +/- 5.6 mm) were detected in L. major promastigotes. The unique characteristics of LmPPX and L. major polyP metabolism may facilitate the development of novel antileishmanial agents.

sequestration and storage, in cell membrane formation and function, in gene activity control, in regulation of enzyme activities, in the stress response and stationary-phase adaptation, and in the formation of channels and pumps (1,2).
In many organisms the mobilization of polyP is performed primarily by the action of enzymes that catalyze the synthesis and degradation of this polymer, the polyphosphate kinase, and the endo and exopolyphosphatases, respectively (1,2). These enzymes have been described in many prokaryotic and eukaryotic organisms. In eukaryotes these enzymes are localized in a variety of cellular compartments such as cytosol, mitochondria, vacuoles, and nucleus (1). Genes encoding for an exopolyphosphatase (3) and an endopolyphosphatase (4) from Saccharomyces cerevisiae are the only ones that have been cloned from an eukaryotic organism, whereas the identification of genes encoding for polyP kinases in eukaryotes has been elusive (1).
In a number of early branching eukaryotes, such as trypanosomatid and Apicomplexan parasites (5), in the green algae Chlamydomonas reinhardtii (6), and in the slime mold Dictyostelium discoideum (7), polyP is mainly localized in acidic organelles known as acidocalcisomes. These organelles are characterized by their acidic nature, their high electron density, and their high concentration of calcium, magnesium, and other elements in addition to pyrophosphate (PP i ) and polyP (5). In Trypanosoma cruzi polyP is mobilized during cell growth and differentiation and when the parasite is submitted to environmental stresses (8). An exopolyphosphatase activity was detected in their acidocalcisomes (8).
A predicted open reading frame with sequence similarity to the yeast exopolyphosphatase was found to be present in chromosome 1 of Leishmania major (9), but the gene product has not been characterized. Leishmania, a member of the trypanosomatid protozoa, is responsible for the spectrum of diseases in humans collectively called leishmaniasis. Therapy against leishmaniasis is unsatisfactory, and the need to develop novel chemotherapeutic agents through the identification of biochemical pathways that allow survival of the parasite and are absent in the host is clear. Here we report the cloning, overexpression, purification and characterization of the L. major exopolyphosphatase (LmPPX). It is demonstrated that the product of this gene, the first related to polyP metabolism isolated from an eukaryotic organism different from yeast, differs from other exopolyphosphatases previously investigated. LmPPX has an acidocalcisomal and cytosolic localization and preferentially participates in the degradation of short chain polyPs, which are very abundant in these parasites.
Clones-Escherichia coli strains CA38 pTrxPPX1 and NR100 P9E30 ppk ϩ were kindly provided by Prof. Arthur Kornberg, Stanford University School of Medicine (Stanford, CA). A full-length cDNA clone (543D) in ZAPII vector corresponding to the LmPPX open reading frame was generously provided by the World Health Organization Leishmania Genome Initiative. The integrated pBS plasmid containing the cDNA pBS-543D was excised from the ZAPII vector for sequencing and subsequent cloning steps according to the manufacturer's protocol (Stratagene).
Genomic DNA Isolation-Promastigotes (1 ϫ 10 7 -1 ϫ 10 8 cells/ml) were harvested by centrifugation, and cell pellets were washed 2ϫ in Dulbecco's PBS. The cell pellet was resuspended in 100 l of breaking buffer (10 mM Tris-HCl, pH 8.0, 15 mM NaCl, 10 mM EDTA, pH 8.0, 0.4% SDS, and 100 g/ml proteinase K) and gently vortexed before incubating at 37°C overnight. Lysates were extracted with TE buffer (10 mM Tris-HCl, pH 7.4, 1 mM EDTA)-saturated phenol. The mixture was centrifuged at 20,800 ϫ g at 4°C for 15 min, and the upper phase (200 l) was collected. Genomic DNA was precipitated by the addition of 5 l of 3 M sodium acetate, pH 5.2, and 400 l of ethanol for 1 h at Ϫ20°C. After the pellet was formed, the DNA was washed with 70% ethanol and resuspended in deionized water before quantification by spectrophotometry.
Sequence and Phylogenetic Analysis-DNA sequence data were generated at the High Throughput Sequencing and Genotyping unit of the Keck Center for Comparative and Functional Genomics at the University of Illinois at Urbana-Champaign. Sequence analysis was performed using the Biology Workbench 3.0 utility (biology.nsca.uiuc.edu) and the Wisconsin Package (version 10.0-UNIX, Genetics Computer Group, Madison, WI.). The PSORT program was from Kenta Nakai, University of Tokyo, Tokyo, Japan. Phylogenetic analysis was done using programs within the PHYLIP software package (version 3.5c, J. Felsenstein, 1993, distributed by the author) and other resources provided by the Canadian Bioinformatics Resource (www.cbr.nrc.ca).
Expression of LmPPX in E. coli-To overexpress LmPPX, E. coli strain DH5␣ was transformed with pET-LmPPX. Transformed cells were grown in Luria-Bertani medium at 37°C, and gene expression was induced by the addition of 1 mM isopropyl-1-␤-d-galactopyranoside when the cell density reached an A 600 of ϳ0.6. The cells were harvested after 1-3 h of incubation, and pellets were kept frozen at Ϫ80°C until further use.
Purification of Recombinant Exopolyphosphatase-Bacterial cell pellets were resuspended (2.5 ml/g of wet weight of pellet) in lysis/binding buffer (300 mM NaCl, 250 mM sucrose, 50 mM Tris, and 5 mM imidazole, pH 7.2), incubated with 10 mg/ml lysozyme for 15 min on ice, and then sonicated 3 times at 10% amplitude for 10 s at 4°C with 20-s intervals. The lysate was incubated with benzonase diluted 1:1,000 for 15 min on ice to reduce its viscosity before being centrifuged at 16,000 ϫ g for 15 min to separate the pellet and supernatant fractions. The pellet was used for isolation of inclusion bodies and further purification of recombinant protein under denaturing conditions in the presence of 8 M urea for antibody production purposes. For biochemical studies, the recombinant protein in its native form was purified from the supernatant. Purification was performed in a batch mode under native or denatured conditions in which 8 M urea was included in all the buffers used. Briefly, supernatants containing the native or denatured protein solubilized from the inclusion bodies were re-centrifuged under the same conditions and filtered using a 0.45-m Millipore filter before being mixed with the nickel nitrilotriacetic acid resin (4:1) under gentle rotation for 1 h at 4°C. The mixture was loaded into an empty column, and the flow-through was collected. The column was subsequently washed twice with 5 ml of binding buffer and then once with 10 ml of the same buffer with the concentration of imidazole changed to 10 mM. Recombinant protein was eluted twice with 500 l (1 ml total) of 500 mM imidazole. Immediately after elution, 2 mM dithiothreitol was added to the purified protein, which was kept frozen at Ϫ80°C for further use. To remove the polyhistidine tag, thrombin cleavage was done as described by the kit manufacturer (Novagen).
Preparation of Antibodies-Purified exopolyphosphatase (100 g) emulsified in Freund's complete adjuvant, was injected subcutaneously in a rabbit followed by another injection 2 weeks later with the protein (100 g) emulsified in Freund's incomplete adjuvant. At 6 weeks after the initial injection, the rabbit was boosted with another 100 g of the purified protein, this time in PBS containing a 10 mg/ml suspension of Al(OH) 3 . Serum was collected before the initial injection (pre-immune serum) and 10 days after each boost. The antiserum was separated into aliquots and stored at Ϫ80°C. Affinity purification of the antibody using a HiTrap protein G Hp column was performed as described elsewhere (10), and according to the instructions of the manufacturer.
SDS-PAGE and Western Blotting-Electrophoresis was performed as described by Laemmli (11) under reducing conditions. Electrophoresed proteins were transferred to nitrocellulose membranes, which were first incubated with polyclonal antibody generated against the recombinant exopolyphosphatase, anti-LmPPX (1:20,000), and then with horseradish peroxidase-conjugated anti-rabbit IgG antibody (1:30,000). Pre-immune serum was used as control at the same concentration. For detection of the histidine tag, the anti-His antibody and the horseradish peroxidase-conjugated anti-mouse IgG antibody used were diluted 1:1,000 and 1:10,000, respectively. Immunoblots were visualized on blue-sensitive x-ray film (Midwest Scientific, St. Louis, MO) using the ECL chemiluminescence detection kit according to the instructions of the manufacturer.
Immunofluorescence Microscopy-Parasites were washed with PHEM buffer (60 mM Pipes, 25 mM Hepes, 10 mM EGTA, 2 mM MgCl 2 , pH 7.2) and fixed with 4% freshly prepared paraformaldehyde and 0.1% glutaraldehyde in PHEM buffer for 10 min at room temperature and 50 min at 4°C. Cells were allowed to adhere to poly-L-lysine-coated coverslips, washed with PBS, and permeabilized with 0.3% Triton X-100 in PBS for 5 min. Coverslips were blocked with PBS containing 3% bovine serum albumin (BSA), 1% cold fish gelatin, 2% normal goat serum, and 50 mM NH 4 Cl for 60 min at room temperature. Samples were first incubated with an affinity-purified rabbit polyclonal antibody against LmPPX (1:50) and with a mouse polyclonal antibody against T. cruzi pyrophosphatase (1:20) (12) diluted in PBS, 3% BSA. Samples were washed with PBS, 3% BSA and incubated with a fluorescein-conjugated goat anti-rabbit antibody and a rhodamine-conjugated goat anti-mouse antibody, both diluted 1:100 in PBS, 3% BSA for 60 min at room temperature. Coverslips were washed with PBS, 3% BSA and then with PBS before being mounted in glass slides using Vectashield mounting media. Micrographs were obtained with an Olympus Fluoview FV300 laser-scanning confocal microscope using optical sections 0.1 m.
Synthesis of [ 32 P]Polyphosphate-For extraction of polyP from E. coli, a strain that overexpresses the polyphosphate kinase (NR100 P9E30 ppk ϩ ) was used. Bacterial cells were grown overnight in Luria-Bertani medium in the presence of 100 g/ml ampicillin and 25 g/ml kanamycin and then inoculated (5% inoculum) in fresh medium in the presence of antibiotics and grown for ϳ60 min or to an A 600 of 0.5 at 37°C. At this time, 50 M isopropyl ␤-D-thiogalactoside was added to the culture to induce the overexpression of the polyphosphate kinase. After 60 min of growth, the cells were washed 2ϫ in MOPS minimum medium (13) containing 10 mM K 2 HPO 4 . The culture was incubated in the same medium in the presence of the antibiotics and [ 32 P]phosphate at a final concentration of 10 Ci/ml for 2 h at 25°C. After labeling, the cells were washed 2ϫ in the same medium without the labeled P i , and the final pellet was resuspended in a lysis buffer containing 50 mM Tris-HCl, pH 7.5, and 4 M guanidine isothiocyanate. Long chain polyP was isolated according to Ault-Riché et al. (14).
Exopolyphosphatase Activity and PolyP Analysis-Exopolyphosphatase activity was assayed by measuring the rate of inorganic phos-phate release using the EnzCheck phosphate assay kit as described before (15) with the microtiter plate modification (16) and different polyphosphates as substrates. The sensitivity of this method was calibrated for the different buffers used. In the cases where o-vanadate and ammonium molybdate were tested, phosphate release was detected by the method of Taussky and Shorr (17). For the assays with [ 32 P]polyP, 600 M polyP 300 (8 Ci/mol) was added to a reaction mixture containing 250 mM sucrose, 50 mM Tris-HCl, pH 7.5, 2 mM MgCl 2 , and 1 g/ml purified LmPPX. Aliquots of 50 l were taken at different intervals of time up to 120 min and added to a mixture of phenol:chloroform to stop the reaction. PolyP was isolated from the aqueous phase and analyzed by electrophoresis in 20% polyacrylamide, 7% urea gel as described before (18).
Cell Fractionation-For analysis of polyP levels in different fractions of L. major, the cells were harvested and washed 2ϫ in a buffer containing 116 mM NaCl, 5.4 mM KCl, 0.8 mM MgSO 4 , 5.5 mM D-glucose, and 50 mM Hepes, pH 7.4, and resuspended in the same buffer. The cell suspension was lysed by sonication, and the lysate was centrifuged at 15,000 ϫ g for 10 min at 4°C. The pellet was separated, and the supernatant was centrifuged again at 100,000 ϫ g to separate a cytosolic fraction. PolyP levels were determined from the amount of P i released upon treatment with an excess of rPPX1 as previously described (8). The intracellular concentration of polyP in L. major was calculated taking into account the promastigote cell volume reported before (19). This value corresponds to 4.74 l/mg protein.

RESULTS
Isolation of the LmPPX Gene-A gene encoding a protein (accession number AAC2464) with sequence similarity to the yeast exopolyphosphatase was found to be present in chromosome 1 of L. major (9). The deduced amino acid sequence of this gene is ϳ40% identical to the sequences of the exopolyphosphatase of S. cerevisiae (S. cerevisiae exopolyphosphatase 1, Ref. 3) and the putative exopolyphosphatases of Schizosaccharomyces pombe (CAB16267) and Methanococcus jannaschii (U67509). Greater sequence identity (ϳ60%) was found with a putative exopolyphosphatase from Trypanosoma brucei (AC013485). Translation of the ORF of LmPPX yielded a polypeptide of 388 amino acids with a predicted mass of 48 kDa. No apparent signal sequence was detected in the LmPPX polypeptide using PSORT. Comparison of the amino acid sequence of LmPPX to the NCBI Conserved Domain Data base indicated that the polypeptide consisted of an N-terminal DHH domain followed by a C-terminal DHHA2 (DHH-associated domain type 2) domain. This domain structure is shared by a number of other eukaryotic, archaeal, and eubacterial proteins, several of which have been identified as exopolyphosphatase enzymes by sequence similarity or by experimental evidence. The domain structure of LmPPX and the inferred phylogenetic relationships of other apparently homologous proteins with similar domain structure are illustrated in Figs. 1, A and B, respectively.
Southern blotting was performed with the LmPPX gene as a probe. All restriction enzymes used gave single, strong bands that were distinct from one another, suggesting the presence of a single LmPPX gene in the L. major genome (data not shown).
Purification of Recombinant LmPPX-LmPPX was expressed in E. coli DH5␣ as a fusion protein with an N-terminal polyhistidine tail. Affinity chromatography on a nickel-chelated agarose column allowed protein purification. After 1-3 h of induction with 1 mM isopropyl-1-␤-D-galactopyranoside, the cells were disrupted, and the proteins were separated using SDS-polyacrylamide gels. The protein pattern of the crude extracts (lanes 1 and 2) and the eluted fractions from the successive purification steps (lanes 3-6) are shown in Fig. 2. LmPPX was purified under denaturing conditions from inclusion bodies (Fig. 2, lane 5, arrow) for antibody generation and from the soluble fraction of the cell lysate in its native form (Fig. 2, lane 6, arrow), for enzymatic characterization. Fig. 2, lane 2, shows that the expressed protein appears as a strong band with an approximate molecular mass of 48 kDa, which is similar to the predicted molecular mass (48 kDa). Other bands in lane 2 also seem to be overexpressed when compared with those in the sample taken before induction (lane 1). These bands co-purified with LmPPX even in the presence of protease inhibitors and seem to be either fragments originated from proteolysis of the recombinant protein (see lane 5) or aggregates that did not migrate properly in the gel due to their very high concentration in the sample in the case of the soluble protein (see lane 6). In fact, if this sample was concentrated, the only bands that could be seen were those of high molecular mass (data not shown), suggesting that this is probably an effect caused by aggregation.
Western Blot Analysis and Localization of LmPPX-Total homogenates of L. major and the recombinant protein were subjected to Western blot analysis with antibodies produced against the protein isolated from inclusion bodies and against the histidine tag (Fig. 3). Anti-LmPPX antibodies recognized a band of the expected size (48 kDa) as well as other low molecular weight bands (Fig. 3A, IS). No detectable bands were observed when the preimmune serum (Fig. 3A, PIS) was used at the same dilution. The protein bands were also recognized by antibodies against the polyhistidine tag (Fig. 3A, ␣-HIS), suggesting that the low molecular weight bands probably originated by proteolysis. Similar results were obtained when total homogenates of L. major were used (Fig. 3B).
The localization of LmPPX in promastigotes of L. major was determined by indirect immunofluorescence assays using affinity-purified anti-LmPPX antibodies (Fig. 4). The protein was localized in vacuoles and in the cell cytoplasm (Fig. 4, left). Co-localization studies were also done using antibodies against the vacuolar proton-translocating pyrophosphatase, an acidocalcisomal marker (Fig. 4, center). Using confocal microscopy we observed co-localization of the LmPPX and the vacuolar proton-translocating pyrophosphatase in acidocalcisomes (Fig.  4, right). No fluorescence was observed in control parasites incubated only in the presence of the secondary rhodaminelabeled goat anti-mouse or fluorescein-labeled goat anti-rabbit IgG (data not shown).
Reaction Requirements of the Recombinant Protein-The LmPPX was very unstable in solution after purification, loosing its activity after only a few hours of elution. Instability was also a problem found with other exopolyphosphatases (20,21). Stability increased when 300 mM NaCl and 250 mM sucrose were included in all steps of the purification procedure and 2 mM dithiothreitol was added after elution of the protein. Under these conditions the purified protein could be separated into aliquots and maintained at Ϫ80°C for several months without significant loss of its activity. As shown in Table I, the recombinant LmPPX preferentially hydrolyzed short chain polyP. The highest specific activity was found for tetrapolyphosphate (polyP 4 ), guanosine 5Ј-tetraphosphate (GP 4 ), and tripolyphosphate (polyP 3 ), which were shown to be the best substrates for the enzyme. No evidence of an increase in the affinity for the substrate with the increase of the phosphate chain was found. In fact, we were not able to detect any activity with long chain polyP (up to 200 residues) by measuring phosphate release. We therefore investigated the reaction products formed by LmPPX acting on long chain [ 32 P]polyP by 20% polyacrylamide, 7% urea gel electrophoresis, which is a more sensitive assay. Fig. 5 shows a time course of hydrolysis of [ 32 P]polyP 300 . Only after 90 min was it possible to detect some P i accumulation as a result of hydrolysis of the labeled substrate. Intermediary products could not be detected.
Similar results were found when [ 32 P]polyP 70 , [ 32 P]polyP 500 , or [ 32 P]polyP 700 were used (data not shown). These experiments suggest that the enzyme acts as an exoenzyme in a processive mode, releasing P i residues from the ends of the chain. The enzyme is not a general phosphatase. ATP, PP i , and p-nitro-   3. Western blot analysis. A, purified LmPPX (10 g) was subjected to SDS-PAGE on 10% polyacrylamide gel, transferred to nitrocellulose membrane, and probed with affinity-purified anti-LmPPX antibody (IS, 1:20,000), preimmune serum (PIS, 1:20,000), or anti-histidine-tag antibody (␣- His, 1:1,000). The asterisk shows the 48-kDa band corresponding to LmPPX. B, Western blot of L. major cell lysate (10 g) with LmPPXspecific polyclonal antibodies (IS) or preimmune serum (PIS) as described in A. phenylphosphate were not significantly hydrolyzed (data not shown). Because PP i does not act as a substrate for LmPPX, it qualifies as an end product in addition to P i . Despite the presence of a cleavage site for thrombin, the N-terminal fusion peptide was not removed by the enzyme (data not shown). Presumably, as occurs with the S. cerevisiae recombinant exopolyphosphatase (3), the cleavage site is inaccessible in this protein.
The pH optimum of the recombinant LmPPX was determined for polyP 3 , polyP 4 , GP 4 , and polyP 15 using imidazole buffer (for pH 6.0 -7.0) and Tris-HCl buffer (for pH 7.5-8.0). The use of MES buffer for the low pH values caused total inhibition of enzyme activity (data not shown). Fig. 6 shows the pH curve for the four substrates. For each substrate except polyP 3 the highest activity was found in the neutral range between pH 7.0 and 8.0. Hydrolysis of polyP 3 was more efficient at pH 6.5. At pH 6.0 the enzyme showed a significant reduction of its activity with all the substrates tested, ranging from 40 to 80% reduction as compared with its optimal activity.
The effect of divalent cations on the exopolyphosphatase activity is depicted in Table II. The recombinant LmPPX, like other exopolyphosphatases (18, 20 -25), required the presence of a divalent cation for activity. In the absence of divalent cations the activity was almost negligible. MgCl 2 at a concentration of 1 mM or above was the best activator using polyP 3 , polyP 4 , or polyP 15 as substrate. CoCl 2 enhanced the activity when polyP 3 , polyP 4 , or GP 4 was used as substrate but had little effect on polyP 15 hydrolysis. MnCl 2 was the best activator for GP 4 hydrolysis and had lower effect on polyP 3 and polyP 4 hydrolysis and no significant effect on the hydrolysis of polyP 15 .
Only a few reagents have been reported to inhibit exopolyphosphatases (18,20,23,24). We tested some of these compounds on the hydrolysis of polyP 4 . Some activation and/or inhibition of exopolyphosphatase activities was previously observed in the presence of increasing amounts of salts and amino acids (18,22,24,25). Our results show that the LmPPX is sensitive to high salt concentrations (Fig. 7A). Although in the case of polyP 4 the inhibitory effect of KCl was not so strong (Fig. 7A), for polyP 3 , polyP 15 , and GP 4 the activity was inhibited to an extent similar to that by NaCl in Fig. 7A (data not  shown). The effect of amino acids on the enzyme activity seems to depend on the amino acid tested, its concentration, and the exopolyphosphatase involved (18,25). LmPPX activity was stimulated by lysine and inhibited by arginine. The curves show a peak of stimulation/inhibition at a respective amino acid concentration of about 50 mM (Fig. 7B).
Although divalent cations are a requirement for optimum activity, some of them, when tested in the presence of MgCl 2 , act as strong inhibitors of this activity, probably by competing with Mg 2ϩ for specific sites (20,21). This same effect was observed with the recombinant LmPPX (Fig. 8A). CuSO 4 and ZnSO 4 proved to be the most efficient, totally inhibiting the activity at very low concentrations. CaCl 2 was also a good inhibitor, although higher concentrations were required for almost total inhibition of the activity (ϳ90%). The recombinant enzyme was insensitive to o-vanadate and molybdate at concentrations up to 1 mM. Heparin, a good inhibitor for other well characterized exopolyphosphatases (23,25,26), was also effective against LmPPX (Fig. 8B).
Distribution of PolyP in L. major-LmPPX is mainly located in acidocalcisomes (Fig. 4), where the vast accumulation of polyP in L. major promastigotes occurs (27). Levels of 55.9 Ϯ 5.6 and 21.4 Ϯ 3.0 mM (in terms of P i residues and calculated taking into account the cell volume indicated under "Experimental Procedures") in chains of about 700 -800-residues-long and in chains of less than 50-residues-long, respectively, were found in L. major promastigotes. Cell fractionation studies showed that most of the cellular polyP (ϳ96 -98%) was located in the pellet fractions, whereas only 2.05 and 3.76% of the total amount of long chain and short chain polyP, respectively, was in the cytosolic fraction (data not shown). DISCUSSION We report here that a gene, LmPPX, present in the L. major genome, encodes a functional exopolyphosphatase. The open reading frame corresponding to LmPPX encodes a protein of 388 amino acids and a molecular mass of 48 kDa (Fig. 1). The LmPPX protein is a member of a family of proteins characterized by the presence of a DHH domain in the N terminus of the sequence. The DHH domain was originally identified as common to a novel family of predicted phosphoesterases including the Drosophila prune protein and bacterial RecJ exonuclease (28). The additional presence of the DHHA2 locates LmPPX to "subfamily 2" of the DHH family, which includes sequences from eubacteria, Archaea, and eukaryotes (Fig. 1). The consensus sequence for the DHHA2 domain is often found in tandem with the DHH domain, but its function is not known. With the exception of the well characterized exopolyphosphatase from S. cerevisiae, there has been no biochemical examination of the members of the subfamily 2 of the DHH domain family.
LmPPX was overexpressed in a bacterial host, and the addition of an N-terminal histidine tag to the recombinant protein allowed efficient purification by affinity chromatography (Fig. 2). As occurred with the S. cerevisiae recombinant exopolyphosphatase (3), the N-terminal appendix probably interacts with the protein since it could not be removed by thrombin, despite possessing the cleavage site for this protease. The recombinant enzyme was similar to S. cerevisiae exopolyphosphatase 1 with respect to its divalent cation requirement, optimum pH, and sensitivity to salts, amino acids, and heparin (Figs. 6 -8), but in contrast to that enzyme (3), LmPPX acts on polyPs of short chain lengths with higher rates and affinity ( Table I). Hydrolysis of long chain polyP was negligible. As with most known exopolyphosphatases, LmPPX is a processive enzyme releasing P i residues from the ends of the chain, with no intermediate products detected upon polyP hydrolysis (Fig. 5). The lack of a signal sequence suggests that it does not enter the secretory pathway. Its cytosolic localization suggests that it is synthesized in free polysomes and then transported into the acidocalcisomes. Confocal immunofluorescence microscopy us- ing affinity-purified antibodies against the recombinant enzyme confirmed its cytosolic and acidocalcisomal localization (Fig. 4). LmPPX is the first gene functionally related to polyP metabolism identified from a eukaryotic cell different from yeast.
LmPPX preference for a particular substrate depended on the divalent cation present in the reaction mixture. PolyP 3 hydrolysis could be stimulated to the same extent by Mg 2ϩ or Co 2ϩ and less by Mn 2ϩ . Mg 2ϩ was the best activator for polyP 4 and polyP 15 hydrolysis, whereas Mn 2ϩ was the best activator of GP 4 hydrolysis. The optimum pH varied according to the substrate used.
Our results also indicated the presence of high levels of short and long chain polyP in L. major promastigotes, confirming a previous report (29). The vast bulk of polyP in L. major is located in the acidocalcisomes (27), where the LmPPX is local- ized. Because the enzyme has an extremely low activity in the hydrolysis of long chain polyP, it is possible that either an endopolyphosphatase or a different exopolyphosphatase is also present in acidocalcisomes. In this regard, endopolyphosphatases that act on long chain polyP, generating tripolyphosphate, have been detected in several eukaryotes including the protist Giardia lamblia (30). Interestingly, the yeast endopolyphosphatase is localized in vacuoles (31). We cannot rule out the possibility that this low activity is due to an impurity in the preparation because the purification gel shows that other bands co-purify with the recombinant protein. However, the low molecular weight bands are also recognized by the antibody against the histidine tag (Fig. 3A), suggesting that they belong to the same protein that could have been degraded by proteolysis.
The cytosolic localization of the exopolyphosphatase suggests that if short chain polyP is located, at least transiently, in the promastigote cytosol, the enzyme would be able to hydrolyze it. As a polyanion, polyP is able to inhibit a number of enzymes (2), and a low cytosolic level should be required to prevent these toxic effects.
Crucial for the accumulation of polyP in trypanosomatid acidocalcisomes is the presence of an active vacuolar H ϩ -pyrophosphatase, which acidifies the acidocalcisomal lumen and provides a pH gradient across the acidocalcisomal membrane (8). The pH gradient in turn is used for the accumulation and/or retention of Ca 2ϩ , Na ϩ , and probably polyP (5,15,16). The rapid decrease of acidocalcisomal polyP in the absence of the pH gradient or after the addition of ionophores (8) suggests an involvement of the gradient either in biosynthesis of polyP or in preventing it from being degraded by polyphosphatases. LmPPX has an alkaline optimum pH and is inhibited by calcium, whereas under physiological conditions acidocalcisomes are acidic and possess high calcium content. Alkalinization and calcium release would contribute to the rapid disappearance of polyP after the acidocalcisomal pH gradient is lost by the action of ionophores and alkalinizing agents (8).
Few exopolyphosphatases have been shown to preferentially hydrolyze short chain polyP in other cells. S. cerevisiae exopolyphosphatase 1, which is homologous to LmPPX, showed an increase in the K m of more that 4 orders of magnitude when the chain length was decreased from 250 up to 3 phosphate residues/molecule, although all the polyPs of S. cerevisiae, remarkably heterogeneous in length, were substrates for S. cerevisiae exopolyphosphatase 1 (3). Although a 28-kDa exopolyphosphatase of S. cerevisiae was also shown to cleave short chain polyP, the K m value for the polyP substrates markedly increased with the decrease in the chain length of the polymer (21). The E. coli exopolyphosphatase (22) and the mammalian intestinal alkaline phosphatase (32) also preferentially cleave long chain polyP. The cell envelope polyphosphatase (20) and the soluble mitochondrial exopolyphosphatase (24, 25) from S. cerevisiae were also shown to hydrolyze short chain polyP but with lower affinity than for long chain polyP. The mechanism that enables an exopolyphosphatase to distinguish the ends of a long chain from those of a short one is intriguing (21), although multiple binding sites on distant portions of E. coli exopolyphosphatase were determined to be responsible for the polymer length recognition of the enzyme (33).
In conclusion, our results indicate that LmPPX is unique in that it preferentially hydrolyzes short chain polyPs, which are abundantly present in the parasite. The lack of a similar enzyme in other eukaryotic cells suggests that it could constitute an attractive target for chemotherapy.