A pyrophosphatase regulating polyphosphate metabolism in acidocalcisomes is essential for Trypanosoma brucei virulence in mice.

We report the functional characterization of a soluble pyrophosphatase (TbVSP1), which localizes to acidocalcisomes, a vesicular acidic compartment of Trypanosoma brucei. Depending on the pH and the cofactors Mg(2+) or Zn(2+), both present in the compartment, the enzyme hydrolyzes either inorganic pyrophosphate (PP(i)) (k(cat) = 385 s(-1)) or tripolyP (polyP(3)) and polyphosphate (polyP) of 28 residues (polyP(28)) with k(cat) values of 52 and 3.5 s(-1), respectively. An unusual N-terminal domain of 160 amino acids, containing a putative calcium EF-hand-binding domain, is involved in protein oligomerization. Using double-stranded RNA interference methodology, we produced an inducible bloodstream form (BF) deficient in the TbVSP1 protein (BFiVSP1). The long-chain polyP levels of these mutants were reduced by 60%. Their phenotypes revealed a deficient polyP metabolism, as indicated by their defective response to phosphate starvation and hyposmotic stress. BFiVSP1 did not cause acute virulent infection in mice, demonstrating that TbVSP1 is essential for growth of bloodstream forms in the mammalian host.

Inorganic polyphosphate (polyP) 1 is a ubiquitous molecule formed by phosphate (P i ) residues linked by high energy phosphoanhydride bonds. Although polyP had been dismissed as a molecular fossil (1), the recent introduction of novel quantitative enzymatic analytical methods (2) has permitted reexamination of its biological functions. A study of the compartmentalization of polyP performed on various eukaryotic microorganisms has shown that each compartment contains its own type of polyP (3), and a correlation has been established between the basic biological processes occurring in each compartment and the structure and metabolism of polyP (3). Among the enzymes involved in the synthesis and utilization of polyP, exopolyphosphatases were considered as the central regulatory enzymes of its metabolism (4). In yeast, polyphosphatases differ from each other in their molecular masses, substrate specificity, and requirement for divalent cations (4). However, only one gene encoding an exopolyphosphatase has been identified so far (5). In eukaryotic organisms, genetic approaches have uncovered key roles for an endopolyphosphatase and an exopolyphophatase in polyP metabolism (6).
In trypanosomatids and many other protozoa causing human disease, an acidic electron-dense compartment has been shown to contain the major part of the cellular polyP. This compartment, initially named the acidocalcisome in T. brucei, was shown to be implicated in storage of cations and phosphate (7), regulation of intracellular pH (8,9), and osmoregulation (10,11). To better understand the relation between polyP metabolism and its cellular role, studies were performed on the polyP composition and on the characterization of enzymes involved in the metabolism of these compounds. Short-and long-chain polyP and PP i were shown to be present in the acidocalcisomes of T. brucei bloodstream forms (BF) (12,13). The turnover of the acidocalcisome PP i pool is low (14), suggesting that PP i could be a product of polyP metabolism.
In this work, we report the identification in acidocalcisomes of a novel type I soluble pyrophosphatase (TbVSP1). According to the divalent metal cofactor used, Mg 2ϩ or Zn 2ϩ , TbVSP1 specifically hydrolyzes either PP i or short-chain polyP. TbVSP1 was also shown to play a central role in the regulation of polyP metabolism in acidocalcisomes and to be essential for osmoregulation and virulence in mice.

EXPERIMENTAL PROCEDURES
Strains Used-T. brucei brucei bloodstream form host cell line 90-13, co-expressing the T7 RNA polymerase and the Tet repressor, was a gift from G. A. M. Cross (15). Cells were cultured in vitro in modified essential medium supplemented with 10% fetal calf serum (16). T. brucei strain 427 procyclic forms (PF) were cultured in SDM-79 medium (17) supplemented with 10% fetal calf serum.
Cloning of TbVSP1-To screen for genes encoding soluble pyrophosphatases (PPases) in T. brucei, the amino acid sequence of the yeast PPase (accession number 2781300) was used to search all available sequence databases using TBLASTN. This search yielded a 0.535-kb gene fragment (accession number AL496002), which closely resembled the region encompassed by nucleotides encoding amino acid residues 42-215 of the yeast PPase. In silico chromosome walking revealed two overlapping sequences (AQ951034 and AZ216908) encoding the N-terminal extremity of the T. brucei PPase. A genomic DNA fragment, PCR-amplified with the following set of primers, PPasePF1Nter (5Ј-GGGCGGCATATGAACAACACACACGGT-3Ј) and PPasePF1Cter (5Ј-GGCCGGCTCGAGCTTCTCAAATGT-3Ј), was used as a probe to screen a T. brucei cosmid library (18), generated into the c2X75 cosmid vector (19) as described previously (20). Hinc2 fragments of the isolated cosmids were subcloned into the pUC18 vector from Appligene and screened with the genomic DNA fragment PPasePF1Nter/PPase-PF1Cter. An Hinc2 fragment of 4000 bp was isolated and sequenced using the AmpliTaq DNA polymerase, as described by the manufacturer (ABI PRISM®, PerkinElmer Life Sciences). The complete gene sequence of the T. brucei PPase was obtained.
Cloning, Expression, and Purification of the N-terminal Domain of TbVSP1 (N-TbVSP1)-A 405-bp fragment comprising the N-terminal domain was generated by PCR. The 5Ј primer PPasePF1Nter contained an 18-nucleotide linker with a NdeI restriction site to facilitate subcloning and 5Ј adjacent N-terminal residues. The 3Ј primer, PPasePF1Cter included an 18-nucleotide linker with a XhoI site for cloning. The PCR product was inserted into the XhoI/NdeI sites of the pET23a plasmid (Novagen). The resulting recombinant protein was expressed in Escherichia coli BL21 (DE3) from Novagen according to the manufacturer's instructions. Cells were lysed, and the recombinant protein was purified as described previously (21).
Purification of the Recombinant N-TbVSP1, rec-TbVSP1 and ⌬N-Tb-VSP1, Immobilized Metal Affinity Chromatography-The recombinant proteins expressed in E. coli were purified from the His⅐Bind® column from Amersham Biosciences used according to the manufacturer's instructions (Novagen). Before performing the PPase assays, the enzyme was desalted in TrisDES buffer (50 mM Tris-HCl, pH 7.5, and 1 mM MgCl 2 ) on a PD10 column (Amersham Biosciences). The exclusion chromatography of the rec-TbVSP1 and ⌬N-TbVSP1 proteins was performed in a buffer containing 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM MgCl 2 using a Superdex 200 HR column from Amersham Biosciences.
Determination of Protein Concentration-Protein concentration was estimated by the method of Bradford (22) and by Coomassie Blue staining after SDS-PAGE. Bovine serum albumin was used as standard.
Production and Western Blot Analysis of Immune Serum against the Recombinant N-TbVSP1-Rabbits were first injected with 80 g of the N-TbVSP1 in complete Freund's adjuvant and with another 80 g in incomplete Freund's adjuvant 15 and 30 days later. Blood was collected before the first injection (preimmune serum) and 10, 13, and 15 days following the last injection. Western blotting was performed as described (23), with serum diluted 1 ⁄200 and incubated overnight at 4°C. The secondary antibody, a 1 ⁄5000 dilution of goat anti-mouse IgG conjugated to horseradish peroxidase (Sigma), was incubated for 2 h in PBS-Tweenmilk. Immunoreactive bands were revealed by washing in 50 mM Tris-HCl, pH 7.5, 20 mM NaCl, and revealed either by 0.5 mg/ml 3,3Ј-diaminobenzidine or by the ECL TM revelation kit (Amersham Biosciences).
PPase and Exopolyphosphatase Activity-0.020 -2 g of desalted recombinant PPase was incubated for 1-5 min at 25°C in a buffer containing 50 mM Tris-HCl, pH 7.5, 1 mM MgCl 2 , or 1 mM ZnCl 2 and the indicated concentration of sodium PP i , polyP 3 , or polyP 28 . The reaction was stopped, and released P i was quantified as described by Ref. 24. Acetate buffer was used from pH 4.5 to 5.0, MES buffer from 5.5 to 6.5, Tris-HCl from 7.0 to 8.8, and glycine NaOH at pH 9.5. Before use, polyP 28 was purified on G25 Sephadex.
TbVSP1 and Ca 2ϩ kinetic constants were determined by preincubation of the enzyme with 100 M CaCl 2 . The enzyme was then diluted 100 times in the reaction buffer. Because polyP 3 hydrolysis produces P i and PP i , which could be hydrolyzed to P i , we used the equation developed for an enzyme with two substrates, as published by Zyryanov et al. (25), to calculate the Michaelis constants.
Expression of the Proteolipidic Subunit of the Vacuolar H ϩ -ATPase (V-H ϩ -ATPase) in T. brucei PF-The gene was amplified from AnTat genomic DNA with the two following primers: SC2 (5Ј-CTATATTTCG-AAATGCTAAGTGACGATACCTGTCAAC-3Ј) and SC1 (5Ј-CTCCGTG-GATCCTTAGTGATGGTGATGGTGATGAGAGCAACCACCCGTATA-GGA-3Ј) designed from the T. brucei V-H ϩ -ATPase proteolipidic subunit and cloned in the BamH1/HindIII restriction sites of the pLEW100 vector. Cells were then transfected with that construction and selected as described previously (23).
Electron Microscopy-For immunocytochemistry, procyclic trypomastigotes were washed with Dulbecco's PBS, fixed for 1 h in a solution containing 0.1% grade I glutaraldehyde, 4% freshly prepared formaldehyde, 0.8% picric acid, in 0.1 M cacodylate buffer, pH 7.2. Fixed parasites were washed with Dulbecco's PBS and dehydrated by successive incubations of 6 min with increasing concentrations of ethanol (10, 25, 50, 75, 95, and 100%) at Ϫ20°C. Samples were embedded in Unicryl at 4°C by incubation with 1:1 ethanol/Unicryl for 1 h and 100% Unicryl for 1, 16, and 8 h. Embedded samples were polymerized under UV irradiation at Ϫ20°C for 48 h. Thin sections were collected on 300 mesh nickel grids and blocked for 60 min with PBS containing 0.1% Tween 20 and 0.5% cold fish gelatin (PBS-TW-FG). Grids were incubated for 3 h with a mix of a rabbit polyclonal antibody against TbVSP1 (1:50) and a mouse anti-His monoclonal antibody (1:50), diluted in PBS-TW-FG. After washing in PBS-TW-FG, grids were incubated for 1 h with a mixture of a 10-nm gold-conjugated goat anti-rabbit antibody (1:75) and a 20-nm gold-conjugated goat anti-mouse antibody (1:20), diluted in PBS-TW-FG. After incubations, grids were washed with 5 ml of PBS and with 5 ml of distilled water before being stained with uranyl acetate and lead citrate. Routine and immunocytochemistry samples were observed in a Hitachi H 600 electron microscope.
Double-stranded RNA Expression and Trypanosome Transfection-The inducible T7 RNA polymerase-based protein expression system developed by Wirtz et al. (15) (kindly provided by G. A. M. Cross) and the p2T7 vector (kindly provided by D. LaCount) (26) were used in this study. DNA fragments corresponding to the coding regions of TbVSP1 from nucleotides 1-462 were PCR-amplified. The 462-bp fragment was amplified using the following set of primers: DB/PP1 (5Ј-CGGGGCAA-GCTTATGAACAACACACACGG-3Ј) and DB/PP2 (5Ј-GCCCCGGAATT-CCATCGAGAAACATTTCG-3Ј). After precloning in the pGEMt vector (Promega), the resulting fragment was cloned between the HindIII/ SacII restriction sites of p2T7 generating the p2T7PPase construction. For stable transfection of bloodstream forms, cells were harvested from a log phase culture, washed once in ZPFMG (ZPFM (Zimmerman Post Fusion Medium) and 55 mM glucose), and resuspended in ZPMG (containing 2 mM ATP and 5 mM glutathione) at a density of 2 ϫ 10 7 cells/ml (27,28). 2 ϫ 10 7 cells were electroporated with 10 g of DNA at 1600 V, infinite resistance, 40 microfarads using the Eurogentech cellject machine. Surviving cells were then seeded in a microtiter plate as described by Wirtz et al. (27), and selection was applied the following day by adding phleomycin to a final concentration of 0.5 g/ml. The genetically modified strain was named BFiTbVSP1 and induced for 4 days with 1 g/ml of tetracycline. Several clones with identical phenotypes were obtained.
Analysis of PolyP-PolyP levels were determined as described by Ruiz et al. (29) using the recombinant exopolyphosphatase (rPPX1) from Saccharomyces cerevisiae. BFiVSP1 bloodstream forms were induced or not with 1 g/ml tetracycline during 4 days. Then, 2 ϫ 10 7 cells were washed twice with Dulbecco's PBS, and long-and short-chain polyP were quantified.

RESULTS AND DISCUSSION
Gene Cloning, Classification, and Sequence Analysis of T. brucei VSP1-The nucleotide sequence revealed an open reading frame of 1245 bp, which encodes a protein of 415 amino acids with a relative molecular mass of 47.3 kDa and a calculated pI of 5.79 (Fig. 1). A Southern blot analysis indicated the presence of a muticopy gene family (data not shown). While this manuscript was in preparation, a search in the T. brucei genome project databases (www.genedb.org) using the TbVSP1 coding sequence indicated the existence on chromosome X and XI of at least three copies of the TbVSP1. The sequence on chromosome X encodes for the pyrophosphatase we worked with. Sequences on chromosomes XI are isoforms. One sequence diverged on 11 amino acids ( Sequence comparison analysis (www.ncbi.nlm.nih.gov/ BLAST) revealed that the TbVSP1 belongs to the family I PPase. The 17 polar amino acids present in the active site of yeast PPase (30) are conserved in the TbVSP1 protein (Fig. 1). With respect to size and primary structure, type I PPases have polyP Metabolism as Potential Pharmacological Target been divided into three families (30). The best identity was obtained with the animal/fungal (Ic) subfamily (Fig. 1). In particular, the TbVSP1 contains three of the four insertions characteristic of the animal/fungal subfamily (Fig. 1). In addition, the TbVSP1 has a large N-terminal extension domain of 160 amino acids (Fig. 1). A search of conserved domains performed by INTERPRO (www.ebi.ac.uk/interpro/scan.html) revealed the presence of a calcium-binding type II EF-hand domain (IPR002048) in this N-terminal region. This analysis suggested that the TbVSP1 might possess specific functions or a particular location. Therefore, in a first step toward the characterization of the TbVSP1, we analyzed its expression and its subcellular localization.
Expression and Localization of TbVSP1-Immunoblotting with a specific rabbit antiserum directed against the N-terminal domain (amino acids 1-160) of TbVSP1 indicated that expression levels by the mammalian host bloodstream and the insect procyclic forms were similar ( Fig. 2A). To determine the localization of the TbVSP1, immunoelectron microscopy was performed on thin sections of PF parasites using the rabbit anti-N-TbVSP1 antiserum. The results obtained (Fig. 2B) indicated that TbVSP1 was located in large vesicles containing electron-dense material, a characteristic feature of acidocalcisomes (7). Co-localization studies were performed on procyclic forms expressing a His tag subunit c of the vacuolar H ϩ -ATPase using an anti-His monoclonal antibody (Fig. 2B). This co-localization confirmed the presence of TbVSP1 protein in acidocalcisomes (31,32). No reaction was detected when the primary antibodies were omitted (data not shown). To further analyze the subcellular localization of TbVSP1, we prepared acidocalcisomes from procyclic forms from a high density Per-coll fraction of trypanosomes as described previously (9). TbVSP1 was detected by Western blots in this high density fraction (Fig. 2C), which was shown to contain the acidocalcisomal protein marker V-H ϩ -PPase (9). We estimated at around 10-fold the enrichment of the acidocalcisome fraction in TbVSP1. Although some membrane-associated PPases have been shown to localize in mitochondria (33) and chloroplast thylakoid (34) previously, none has been localized to a vacuolar compartment before.
Substrate Specificity of TbVSP1-In the presence of Mg 2ϩ , PPases display nearly absolute PP i specificity. However, this specificity is lost when transition metal ions such as Zn 2ϩ , Mn 2ϩ , or Co 2ϩ are used as cofactors (25). Because acidocalcisomes contain high levels of Mg 2ϩ , Zn 2 , and Ca 2ϩ (35), we determined the Michaelis-Menten constants of the TbVSP1 for the three substrates polyP 28 , polyP 3 , and PP i , in the presence of Mg 2ϩ , Zn 2ϩ , and Ca 2ϩ . With 1 mM Mg 2ϩ , TbVSP1 specifically hydolyzed PP i with an optimum pH of 7.5 and a k cat value of 385 s Ϫ1 (Fig. 3A). The replacement of Mg 2ϩ for Zn 2ϩ in the reaction buffer resulted in the hydrolysis of polyP 3 and polyP 28 , in addition to PP i . However, the optimum pH for hydrolysis shifted to a more acidic pH of 6.5 with k cat values of 52, 3.5, and 56 s Ϫ1 , respectively (Fig. 3, B and C). This effect of specificity change was discussed previously by Zyryanov et al. (25) and attributed to the fact that the Zn 2ϩ transition metal ion allowed a more favorable positioning of the terminal phosphate group for catalysis (25). Ca 2ϩ -PPi, conversely, is an efficient competitive inhibitor of Mg 2ϩ -PPi hydrolysis. TbVSP1 and polyP 28 hydrolysis. In conclusion, TbVSP1 substrate specificity in acidocalcisomes is regulated in a complex manner by the pH and Ca 2ϩ , Zn 2ϩ , and Mg 2ϩ content of the compartment.

Role of the N-terminal Extension in the Oligomerization and Enzymatic
Activity of TbVSP1-The sequence that directs proteins to acidocalcisomes has not been identified. To analyze the potential implication of the N-terminal domain for acidocalcisome protein targeting, we constructed a chimeric protein containing this N-terminal domain and the green fluorescent protein. The resulting fusion protein expressed in T. brucei procyclic forms was not addressed to acidocalcisomes (data not shown). However, the main roles attributed to the EF-hand domain are the regulation of cellular activity, buffering/transporting Ca 2ϩ , or anchoring protein complexes (36,37). Therefore, to further investigate the role of the N-terminal extension, we deleted it from a mutant protein (⌬N-TbVSP1). The resulting kinetic constants of ⌬N-TbVSP1 for PP i hydrolysis and the Ca 2ϩ inhibition constant did not vary (13 M), suggesting that the EF-hand domain of this N-terminal region was not implicated in PP i hydrolysis or Ca 2ϩ regulation. Nevertheless, whereas size analysis performed by exclusion chromatography indicated that ⌬N-TbVSP1 principally existed as a dimer (Fig.  4), TbVSP1, in the presence of 5 mM MgCl 2 , was shown to mainly exist as an hexamer. Because of the type II Ca 2ϩ EFhand domain, we tested the effect of Ca 2ϩ on the enzyme oligomerization state. Exclusion chromatography showed that addition of 100 M CaCl 2 favored the formation of large complexes of around 600,000 Da (Fig. 4.). Large complexes did not show different kinetic constants for PP i (k cat ϭ 324 s Ϫ1 ), polyP 3 (k cat ϭ 33.5 s Ϫ1 ), and polyP 28 (k cat ϭ 3.7 s Ϫ1 ). In conclusion, Ca 2ϩ could regulate the oligomerization state of the enzyme by its binding to the N-terminal region. This oligomerization could be a mechanism for association of TbVSP1 with the membrane compartment (38) in which polyP is present (7).
Implication of the TbVSP1 in the Regulation of PolyP Metabolism in Acidocalcisomes-We used RNA interference to produce bloodstream TbVSP1 mutants (BFiVSP1). After 4 days of double-stranded RNA expression, TbVSP1 protein completely disappeared (Fig. 5A). Concomitantly, the level of short-and long-chain polyP greatly decreased. In particular, the steady state level of long-chain polyP underwent a 60% reduction (Fig. 5B).
Moreno et al. (12) have reported that short-chain polyP of an average chain length of 3.3 phosphates and PP i are the most abundant phosphorus-containing compounds in acidocalcisomes. The short-chain polyP concentration in T. brucei PF has been previously reported to be around 50 mM (9). Fig. 5B shows that the concentration in BF is around 600 M. The respective K m values for PP i and polyP 3 hydrolysis in the presence of Mg 2ϩ and Zn 2ϩ were estimated as 19 and 373 M (Fig. 3). Interestingly, the K m for polyP 28 , which is present in a much lower concentration in acidocalcisomes, was estimated as 14 M (Fig. 3). Moreover, K m values for the same substrates and cofactors calculated for the hexameric and larger complexes were similar. Therefore, according to the K m values of TbVSP1, it appears that the TbVSP1 would work at V max to produce P i from PP i and short-chain polyP. However, before it could hydrolyze PP i , TbVSP1 would require a pH increase and a calcium concentration decrease in acidocalcisomes.
This conclusion suggests a central role for the enzyme in the regulation of polyP metabolism. In S. cerevisiae, polyP is hydrolyzed through the combined action of an endopolyphosphatase and an exopolyphosphatase (39), which hydrolyze long-and short-chain polyP to polyP 3 and PP i . In Leishmania   FIG. 2. Expression and subcellular localization of TbVSP1 As shown in A, BF and PF homogenates containing 2 g of protein were subjected to SDS-polyacrylamide gel electrophoresis on 12% polyacrylamide gels and transferred to Immobilon membranes. Lanes were probed with antibodies against N-TbVSP1 and the 50 kDa TbVSP1 protein was revealed with ECL. B, immunoelectron microscopy of T. brucei procyclic forms with anti-TbVSP1 and anti-V-H ϩ -ATPase. Labeling of TbVSP1 is revealed by the 10-nm gold beads, and labeling of the V-H ϩ -ATPase is revealed by the 20-nm gold beads. The electrondense material observed close to the membrane is typical of acidocalcisomes (Ac). Controls without primary antibodies did not give labeling. Bar, 56 nm. C, immunoblots of a total cell homogenate (lane Tch, 20 g of protein) and an acidocalcisomal fraction (lanes Acf1 and Acf2, 3 g of protein). Lanes Tch and Acf1 were probed with antibodies against N-TbVSP1. In a second run, as a control, lane Tch was probed with a monoclonal antibody directed against aldolase. Lane Acf2 was probed with antibodies directed against the V-H ϩ -PPase (9). These immunoblots were revealed with 3,3Ј-diaminobenzidine. As a control, we have shown previously that the acidocalcisome fraction did not contain ␣-tubulin (9) .   FIG. 3. Effect of substrate concentration on TbVSP1 activity. TbVSP1 was measured as described under "Experimental Procedures" over a range of pH between 4.5 and 9.5 (insets) and in the presence of different concentrations of PPi, polyP 3 , and polyP 28 . The activity in the presence of PPi was measured in buffer containing Mg 2ϩ (A), whereas the activity in the presence of polyP 3 and polyP 28 was measured in buffer containing Zn 2ϩ (B and C). major (40) and in T. brucei (41), genes encoding for exopolyphosphatases have been sequenced, but in T. brucei, we do not have evidence for its location in acidocalcisomes. This exopolyphosphatase is active at pH 7.5 and is inhibited by CaCl 2 (40,41). To avoid polyP 3 accumulation in the cell compartment where polyP hydrolysis is occurring and to produce Pi, polyP 3 should be further hydrolyzed. We hypothesize that polyP 3 at acidic pH in the presence of Zn 2ϩ is hydrolyzed by the TbVSP1 and that, following compartment alkalinization and the release of H ϩ and Ca 2ϩ , the exopolyphosphatase and TbVSP1 could drive the reaction by producing P i directly from polyP 3 and PP i within acidocalcisomes. The produced P i could be either exported or further utilized for polyP synthesis.
Acidocalcisomes have been shown to be involved in several biological processes after partial alkalinization (9,29). Alkalinization creates environmental conditions that activate TbVSP1 and the previously reported exopolyphosphatase activity (9,29,32,40). Accordingly, TbVSP1 might be essential for polyP hydrolysis to occur during these processes. We therefore tested the BF mutant cells under different physiological conditions that have been shown to require polyP hydrolysis.
Phenotypic Analysis of TbVSP1 BF Mutants (BFiVSP1), pH and Ca 2ϩ Homeostasis and Survival in a Deprived Phosphate Medium and under Osmotic Stress-In contrast to what we reported previously for V-H ϩ -PPase mutated parasites (9), BFiVSP1 recovered as well as wild type cells from pH stresses (acidic or basic, data not shown). Moreover, Ca 2ϩ release in-duced by nigericin, NH 4 Cl, or ionomycin was not altered in the induced BFiVSP1 (data not shown), whereas that amount was significantly reduced in the V-H ϩ -PPase mutants (9). Because the polyP content was reduced by 90% in V-H ϩ -PPase mutants, we suggest that the remaining polyP in the BFiVSP1 mutants was sufficient for these homeostatic mechanisms to take place.
It has been established that polyP serves as a P i reservoir for cellular metabolism and growth. We showed that neither induced nor non-induced BFiVSP1 mutant cells failed to overcome the effects of an incubation in a phosphate-depleted medium (data not shown). However, after 45 min of incubation in low PO 4 -containing medium, the percent of cell survival is Surviving cells were diluted 20 times in complete modified essential medium and counted 48 h later. The figure shows the percentage of surviving cells as compared with untreated cells. As shown in B, hyposmotic treatment was done by dilution of the culture medium with distilled water (1:1) (squares, n ϭ 4). Control cells were cultured in iso-osmotic medium (triangles, n ϭ 4) obtained by adding 60 mM NaCl to the previously diluted medium. Cells were then counted every day. Black and dashed lines correspond to induced and non-induced cells, respectively. C, growth of T. brucei in mice infected with induced (black line, n ϭ 8), non-induced (dashed line, n ϭ 5) BFiVSP1, or non-transfected cells (gray line, n ϭ 5, wild type (wt)). Swiss mice were intraperitoneally inoculated with 2 ϫ 10 4 cultured BF parasites. For induction, doxycycline (200 g/ml) and 5% sucrose were added to the drinking water beginning 3 days before infection. Parasites were counted each day. For each set of data, a typical result is shown. about 30% for non-induced BFiVSP1 parasites and 1% for those induced (Fig. 6A). This observation indicated that induced parasites were more sensitive to a low PO 4 -containing medium. Ruiz et al. (29) have demonstrated that polyP hydrolysis was induced by hyposmotic stress. LeFurgey et al. (10) also confirmed that acidocalcisomes play a physiological role in osmoregulation. Induced BFiVSP1 forms were therefore tested for growth under hyposmotic growth conditions. Induced mutated cells did not develop under hyposmotic stress, whereas noninduced cells started to develop normally after 30 h (Fig. 6B). As a control, we found that the addition of NaCl to a final concentration of 120 mM (isoosmotic conditions) was required for normal growth of induced BFiVSP1. This cellular response is not immediate since we noted a 20-h latent period before growth of the non-induced BFiVSP1. In conclusion, these results suggested that the absence of TbVSP1 largely reduced the ability of acidocalcisomes to hydrolyze polyP and recover from hyposmotic stress.
TbVSP1 Is Essential for T. brucei Virulence in Mice-First, we examined whether induced BFiVSP1 grew normally in isoosmotic modified essential medium. Induced BFiVSP1 started to grow at about 120 h and then declined in numbers to an undetectable level for at least 10 days (Fig. 6C). In the absence of either tetracycline induction or doxycline treatment, BFiVSP1 parasites killed the mice after 100 h of infection. After 100 h, there were 40-fold fewer induced BFiVSP1 parasites in the blood than in the controls (non-induced and wild type) (Fig. 6C). Taken together, these results indicate that the absence of TbVSP1 decreases acute virulence of T. brucei in mice such that animals survive an otherwise lethal infection with this strain.
These data confirmed that TbVSP1 plays a key role in polyP metabolism and might be an alternative to exopolyphosphatase activity in eukaryotic cells. Because TbVSP1 was essential for osmoregulation and the establishment of full virulence in mice, it could be considered as a potential pharmacological target. We also have identified a gene encoding this enzyme in Leishmania amazonensis (accession number AL354096). Moreover, bisphosphonates, which are known as good trypanocides (42,43), are potent inhibitors of pyrophosphatase activities (44). Since these molecules accumulate on polyphosphate (45), it would be of great interest to evaluate bisphosphonate accumulation in this compartment and to test them for inhibitory effects on TbVSP1 enzymatic activity.