Purification, Cloning, and Characterization of an Acidic Ectoprotein Phosphatase Differentially Expressed in the Infectious Bloodstream Form of Trypanosoma brucei*

We purified an ecto-phosphatase of 115 kDa (Try-AcP115) specifically expressed by bloodstream forms of Trypanosoma brucei . The corresponding gene coded for a 45-kDa protein potentially including a signal peptide, a membrane-spanning domain and an N-terminal domain containing 8 N -glycosylation sites. There was no significant sequence homology with other phosphatases. Antiserum to the Escherichia coli recombinant N-terminal domain, Petase7, recognized a protein of 55 kDa in Western blots after deglycosylation of the TryAcP115 protein by N -glycosidase F. Immunofluorescence and trypsin treatment of living parasites showed that TryAcP115 was localized to the surface of the parasite and that its N-terminal domain was oriented extracellularly. The recombinant N-terminal domains, expressed in E. coli and Leishmania amazonensis, harbored phosphatase activity against Tyr(P)-Raytide, Ser(P)-neurogranin, and ATP. The enzymatic properties of native TryAcP115 and the recombinant proteins for the substrate Tyr(P)-Raytide were virtually identical and included: (i) K m and V max values

Control levels of extracellular ATP and surface protein phosphorylation are important physiological regulatory mechanisms (1,2), but the correlation between the two in the regulation of cellular function has yet to be fully established. Extracellular ATP levels are under the control of ecto-ATPases and ecto-apyrases, which can be distinguished by the sub-strates they hydrolyze, NTPs 1 and both NTPs and NDPs, respectively (3). These ecto-enzymes may be involved in major cellular processes including adhesion, termination of purigenic signaling, and purine recycling (2).
Reversible phosphorylation of protein extracellular domains is controlled by a novel class of protein kinases, called ectoprotein kinases. This extracellular catalytic activity has been identified at the surface of many cell types. Ecto-kinases either reside on (4,5) or are shed from the cell surface (6 -8). Ectophosphatase activities have also been detected in some mammalian cells (9 -11) and in two parasitic protozoa Trypanosoma cruzi (12) and Trypanosoma brucei (13). Cellular activation, motility, growth, and differentiation are partially regulated by the combined action of ecto-kinases and ecto-phosphatases (6).
Because the proteins controlling extracellular phosphorylation and extracellular ATP levels are thought to play a central role in the modulation of cell growth (14,15), T-cell activation (5,10,16), and parasite-host interaction (17,18), identification of ecto kinases and phosphatases in a primitive parasite such as T. brucei would be of great interest. T. brucei is a digenetic parasitic protozoan with a complex life cycle characterized by proliferative and non-proliferative stages, accompanied by biochemical and morphological changes depending on the insect and the mammalian host (19). It also alternates between infectious and noninfectious stages, but the mechanisms that regulate these alternating periods of growth and infectivity are unknown. Immunosuppression, including a depressed T-cell response (20), is considered to be an essential element of the host-parasite relationships in the early stages of infection (21).
Previously we characterized differentially expressed, plasma membrane-associated tyrosine phosphatase activity in T. brucei (22). By monitoring tyrosine phosphatase activity, we have now purified and cloned a membrane protein phosphatase without significant sequence homology with other phosphatases. Kinetic values, substrate specificity, and modulator profiles were similar to those of acidic protein phosphatases. These characteristics and the extracellular localization of the catalytic domain suggest that this protein belongs to a new family of acidic ecto-phosphatases.

EXPERIMENTAL PROCEDURES
Trypanosomes and Leishmania-T. brucei brucei monomorphic clone MITat 1.4 (23,24) and clone AnTat 1 (25) were harvested at mid-log phase from infected rats and isolated by DEAE-cellulose chromatography (26). Procyclics were transformed from the MITat 1.4 clone and * This work was supported by the Centre National de la Recherche Scientifique (CNRS), Ministère de l'Enseignement Supérieur et de la Recherche, Groupement De Recherche Parasitologie (DGA-CNRS), Programme Recherche Fondamentale en Microbiologie Maladies Infectieuses et Parasitaires, and the Conseil Régional d'Aquitaine. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AF184925.
Preparation of Membrane Fractions-Purification from the total cell extract (F1) was performed as described by Seyfang and Duszenko (28). Briefly, F2 represented a ghost fraction obtained by hypotonic lysis, F3 was the F2 fraction stripped of peripheral and cytoskeletal proteins by EDTA/alkali treatment, and F4 was the octylthioglucoside-solubilized protein fraction. F4TS consisted of Triton X-100-solubilized proteins obtained by treating F3 for 30 min at 4°C with gentle agitation with 0.5% Triton X-100 and centrifugation for 1 h at 100,000 ϫ g. Protein concentrations were determined by the method of Bradford (29), 0.1% Triton X-100 was used to solubilize the proteins when necessary. Bovine serum albumin was used as a standard.
Columns and Instruments-DEAE-Sepharose fast flow (Amersham Pharmacia Biotech, Saclay, France), was used for ion exchange chromatography. O-Phospho-L-Tyrosine (O-Tyr(P)) immobilized on crosslinked 4% beaded agarose from Sigma was used for affinity chromatography. The chromatographic system used throughout this study was the FPLC workstation from Amersham Pharmacia Biotech. All buffers contained a protease inhibitor mixture with final concentrations of 1 M chemostatin, 1 M leupeptin, 1 M pepstatin, and 10 M phenylmethylsulfonyl fluoride.
Ion Exchange Chromatography-The F3 fraction was equilibrated in buffer A1 (40 mM MES, pH 5.6, 0.5% Triton X-100, 20 mM NaCl, 1 mM DTT, 1 mM EDTA, protease inhibitor mixture) and loaded on the DEAE fast flow column from Amersham Pharmacia Biotech pre-equilibrated in buffer A1. The column was then washed with buffer A1 and the breakthrough collected. Elution was performed by the discontinuous step gradient method. Fractions were collected at 70 and 100% buffer B1 (40 mM MES, pH 5.6, 0.5% Triton X-100, 200 mM NaCl, 1 mM DTT, 1 mM EDTA, protease inhibitor mixture). Proteins were eluted at each step and the major phosphotyrosine phosphatase (PTPase) activity was identified at the 70% buffer B1 step. Flow rate was 2 ml/min for each step.
Affinity Chromatography-The 70% eluted DEAE fraction was diluted and equilibrated with pre-equilibration buffer A2 (50 mM sodium acetate, pH 4, 0.5% Triton X-100, 20 mM NaCl, 1 mM DTT, 1 mM EDTA, protease inhibitor mixture and adsorbed onto a buffer A2 pre-equilibrated O-Tyr(P)-agarose affinity column. The column was washed with buffer A2 and the breakthrough collected. The enzyme was eluted with a linear salt gradient from 20 to 500 mM NaCl in buffer B2 (50 mM sodium acetate, pH 4, 0.5% Triton X-100, 600 mM NaCl, 1 mM DTT, 1 mM EDTA, protease inhibitor mixture) in a total volume of 20 ml. A major protein peak with PTPase activity was eluted at 150 mM salt. Flow rate was 1 ml/min for each step. Petase7, the recombinant protein expressed in Escherichia coli was purified over the same column and conditions with the following buffer compositions: buffer A3 (50 mM sodium acetate, pH 4, 20 mM NaCl, 15 mM ␤-mercaptoethanol, protease inhibitors); buffer B3 (50 mM sodium acetate, pH 4, 500 mM NaCl, 15 mM ␤-mercaptoethanol protease inhibitor mixture). Before loading, the dialyzed His⅐Bind TM eluted fraction was equilibrated in buffer A3.
Immobilized Metal Affinity Chromatography-The His⅐Bind TM column from Novagen was used according to the manufacturer's instructions. Before loading, the Leishmania culture medium was equilibrated in 1ϫ binding buffer. The eluted fractions were dialyzed overnight against 2 liters of 30 mM Tris-HCl, pH 7.2, 15 mM ␤-mercaptoethanol. Concentration of pTEX/AcP protein was estimated by Western blot using Petase7 as a control.
Phosphatase and Reconstitution Assays-Determination of PTPase activity, optimum pH, the inhibitor profile, and reconstitution assays were conducted as described previously (30). All activity assays were carried out for 1 h at 30°C in 50 mM acetate buffer, pH 4, with 100 nM Tyr(P)-Raytide or 50 mM p-nitrophenyl phosphate (pNPP). ATPase activity was followed by measuring the release of phosphate according to Mitsui et al. (31) at pH 6.9 using 1 mM [␥-32 P]ATP (55,000 cpm/nmol). Before the assay, protein extract buffer was exchanged by PEM buffer (0.1 M Pipes, 1 mM MgCl 2 , 1 mM EGTA, pH 6.9) using a PD10 column from Amersham Pharmacia Biotech.
The amplified and cloned MEX2/P3CaAS fragment was used as a probe to screen a T. brucei genomic DNA library (33) generated into the c2X75 cosmid vector (34) as described previously (35). KpnI fragments of the isolated I.3 cosmid were subcloned in the pUC18 vector from Appligene and screened with the cDNA fragment MEX2/P3CaAS. A KpnI fragment of 5000 base pairs was isolated and sequenced using the AmpliTaq DNA polymerase, as described by the manufacturer (ABI PRISM TM , Perkin Elmer).
Production and Western Blot Analysis of Immune Serum against the Petase7 Protein-A rabbit was injected with 150 g of Petase7 in complete Freund's adjuvant and with 100 g in incomplete 15 and 30 days later. Blood was collected before the first injection (preimmune serum) and 10, 13, and 15 days following the last injection. Affinity purified proteins in sample buffer (2.2% SDS, 50 mM DTT, 90 mM Tris-HCl, pH 6.8, and 10% glycerol, mass/volume) were boiled for 5 min and subjected to 10% polyacrylamide gel electrophoresis. Proteins were transferred to polyvinylidene difluoride (Immobilon P, Millipore) membranes by semi-dry blotting (36). Filters were blocked for 15 min with PBS/Tween/milk (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na 2 HPO 4 , 1.4 mM KH 2 PO 4 , 0,005% Tween 20, 5% milk), incubated overnight at 4°C with a 1/500 dilution of serum against Petase7 in PBS/Tween/milk and incubated for 2 h with either 1/1000 goat anti-rabbit IgG conjugated to horseradish peroxidase (Sanofi-Pasteur) in PBS/Tween/milk or 1/2000 goat anti-rabbit IgG conjugated to alkaline phosphatase (Bio-Rad) in TBS/Tween/milk. Immunoreactive bands were revealed by washing in 50 mM Tris-HCl, pH 7.5, 20 mM NaCl and a solution containing 0.05% H 2 O 2 and 2.8 mM 4-chloro-1-naphtol for the peroxidase conjugate, and according to Sambrook et al. (33) for the alkaline phosphatase conjugate (Pasteur Sanofi).
Immunofluorescence-Bloodstream form trypanosomes (AnTat) were obtained at peak parasitemia from infected mice, and fixed on ice for 2 min in 0.2% formaldehyde. The formaldehyde was neutralized for 10 min with 0.1 M glycine at room temperature. After centrifugation and resuspension in PBS, trypanosomes were transferred to a microscope slide and treated with anti-Petase7 diluted (1/500) in PBS containing 0.1% bovine serum albumin for 30 min at room temperature followed by the fluorescein isothiocyanate-labeled secondary antibody (Pasteur Sanofi) diluted (1/1000) in PBS, 0.1% bovine serum albumin, 0.001% Evans blue for an additional 30 min. After washing, slides containing the treated trypanosomes were mounted with anti-fade Vectashield (Vector Laboratories).
For Leishmania, 10 6 promastigotes were transferred to circular glass poly-L-lysine-treated (30 min. in 10 g/ml) coverslips and placed inside a 24-well plate. After centrifugation at 130 ϫ g for 10 min, cells were fixed with 2% paraformaldehyde for 1 h and washed twice with PBS before incubation for 1 h at room temperature with anti-Petase7 serum, diluted 1000-fold in PBS containing 10% mouse serum. After three washes of 10 min each in PBS, fluorescein-conjugated goat anti-rabbit IgG diluted 2000-fold in PBS with 10% goat serum was added as a secondary antibody. Photomicrographs were taken through a Zeiss UV microscope and images analyzed by the use of a camera (Photometrics) with Iplab software (Sigma Analytics) and Adobe photoshop 4.0 on a Macintosh 7100/80 computer.

Cloning, Expression, and Purification of the N-terminal Domain of
TryAcP115 in E. coli-A 1-kilobase fragment comprising the N-terminal region lacking the putative signal peptide was generated by PCR. A 5-kilobase KpnI genomic fragment from AnTat 1 subcloned in pBlue-Script (Stratagene) was used as a template in the PCR reaction. The 5Ј primer (5Ј-GGCAAACATATGGAGTCGAGCAGCGATGCGCAA-3Ј) contained a 12-nucleotide linker with a NdeI restriction site to facilitate subcloning and 7 adjacent N-terminal residues (ESSSDAQ). The 3Ј primer, (5Ј-AACAAGGGATCCTTACTGATGCTGATGCTGATGCTCG-GGTATCGTTAACGG-3Ј), included a 12-nucleotide linker with a BamHI site for cloning, a stop codon, codons for 6-histidine tag residues to allow binding to the His⅐Bind TM column, and codons for 6 C-terminal residues (PLTIPE). The PCR product was inserted into a NdeI/BamHI sites of the pET3a plasmid (Novagen) to create the Petase7 construction. The mutated recombinant gene coding for the mutated protein PetaseM4 was obtained by using the site-directed mutagenesis kit from CLONTECH, Petase7 as DNA matrix and the following mutated oligonucleotides MUT1 (ACCGGCACTCCGGAGAAGTCAAG) and MUT2 (GAGTCAACACGGCGACACTGATGT). The resulting recombinant Petase7 and PetM4 proteins were expressed in E. coli BL21 (DE3) from Novagen according to the manufacturer's instructions. Cells were lysed in 1ϫ binding buffer, containing the protease inhibitor mixture, by three steps of freezing and thawing and brief sonication. The lysate was centrifuged for 30 min at 10,000 ϫ g and the resulting supernatant applied to a Ni 2ϩ His⅐Bind TM column.
Cloning, Expression, and Purification of the N-terminal Domain of TryAcP115 in Leishmania-To ensure the presence of the signal peptide, the N-terminal region was PCR amplified with the following primers designed with the same characteristics as for E. coli: L115PSNter (CTAGCCGGGATCCATCTCGCGTACTATTGTG) and L115Cter (CCG-G CCG AAT TCTTAGTGATGGTGATGGTGATGGTGATGGTGATGGT-GATGCTCGGGTATCGTTAACGG). The fragment was cloned in the pTEX vector (gift of Dr. J. Kelly (37)) at EcoRI/BamHI sites giving the pTEX/AcP construction. 10 8 exponentially growing cells (L. amazonensis LV79) were washed and resuspended in 200 l of cold electroporation buffer (21 mM Hepes, pH 7.4, 137 mM NaCl, 5 mM KCl, 0.7 mM NaH 2 PO 4 , 6 mM glucose). 50 g of plasmid DNA was added and the cells electroporated at 0°C in 2-mm cuvettes with a Cellject apparatus set on 450 V, 74 W, 600 microfarads, single pulse mode. Cells were transferred to 10 ml of culture medium and 24 h later G418 (10 g/ml) was added. For purification, 50 ml of the culture supernatant was applied to a 50-ml Q fast flow column 1ϫ binding buffer (20 mM Tris, pH 7.9, 500 mM NaCl, 6 mM imidazole). The nonretained fraction was applied to the His⅐Bind TM column and the pTEX/AcP protein eluted as described above.

RESULTS
Purification of the 115-kDa Acidic Phosphatase (TryAcP115) from T. brucei-Since the detergent n-octylthioglucoside used in our earlier study (30) solubilized, but inactivated, the phosphatase activity, we tested the effect of 0.5% Triton X-100. The resulting PTPase activity in the Triton X-100-solubilized fraction (F4TS) shared the characteristics of the original TbPTPase activity of the F4 membrane fraction (30). It was sensitive to 200 M vanadate and 1 mM sodium fluoride, was more active at pH 4 than pH 7, and had a K m value of 10 nM for Tyr(P)-Raytide. Next we devised a two-step chromatographic purification procedure (Fig. 1). Chromatography on DEAE fast flow at   (Fig. 1, A and B) removed VSG (60 kDa), the major surface protein of the F4TS fraction and two other abundant proteins of 70 and 30 kDa (Fig. 1B). Dephosphorylating activities were recovered (Table I) both in the breakthrough (BT-DEAE) and in the 150 mM NaCl eluate (150EL-DEAE). The specific activity of the 150EL-DEAE on Tyr(P)-Raytide was 32 (Ϯ18) pmol/min/mg, which represented a purification of 14 (Ϯ4)-fold with a yield of 89% (Ϯ10). Purification of the DEAE eluted activity (Fig. 1C) on an O-phospho-L-tyrosine affinity column resulted in a large increase in the tyrosine phosphatase specific activity of the eluate (400EL-AFF) to 206 (Ϯ100) pmol/ min/mg, which represented a total purification factor of 90 (Ϯ20)-fold and a yield of 44% (Ϯ20) ( Table I). The values of the purification parameters were similar when measured by either a specific (Tyr(P)-Raytide), or a nonspecific (pNPP) substrate, suggesting that the isolated activity dephosphorylated both Tyr(P)-Raytide and pNPP. The optimum activity of the affinity eluted fraction for Tyr(P)-Raytide was at pH 4. Affinity eluted PTPase activity catalyzed the removal of phosphate from Tyr(P)-Raytide with a K m value of 15 nM. These values (not shown) were similar to those estimated for F4 and F4TS. Examination of the 400EL-AFF fraction by SDS-PAGE under reducing conditions and by Coomassie staining (Fig. 1D) showed the presence of a major polypeptide of 115-kDa and two minor polypeptides of 66 and 30 kDa. The 30-kDa polypeptide was not well stained by Coomassie. Protein-reconstitution assays (Fig. 1E) indicated that the PTPase activity was in the 115-kDa polypeptide, which we named 115 acidic phosphatase protein (TryAcP115).
Sequencing and Cloning of TryAcP115-5 g of TryAcP115 protein were obtained from 1.5 ϫ 10 11 bloodstream form trypanosomes. The protein was further purified by 10% reducing SDS-PAGE. Microsequencing of peptides generated by Endo-K cleavage and high performance liquid chromatography purification yielded 2 peptide sequences: 1) RVGY-FGSQVISLSGR; and 2) KAFDEWESK ( Fig. 2A). Degenerated primers and miniexon primers were designed according to the peptide sequences and the T. brucei spliced leader sequence, respectively. DNA fragments of 330, 760, and 430 base pairs were amplified by RT-PCR with, respectively, the following primer associations, P1CaAS/Mex2, P3CaAS/Mex2, and P1CaS/P3CaAS, and cloned. The nucleotide sequence started as expected at the 5Ј end by the miniexon sequence characteristic of trypanosomal mRNA. Moreover, since the 760-base pair fragment coded for the amino acid sequence obtained by microsequencing of peptide 1 and since the three DNA fragment sequences shared identical overlapping regions, we concluded that the 3 amplified fragments resulted from the amplification of the same mRNA. Screening of 2 ϫ 10 5 clones from the T. brucei cosmid library with the 760-base pair fragment as a probe at high stringency yielded 5 colonies. A KpnI/KpnI fragment of 5000 base pairs containing the entire gene was sub-cloned and sequenced. This fragment showed an open reading frame of 401 amino acids predicting for a protein of 45 kDa with a calculated pI of 5.62. The algorithm of Kyte and Doolittle (38) revealed two hydrophobic regions between residues 1 and 22 and residues 340 and 362 ( Fig. 2A). The N-terminal region, which is 81% homologous to the first 16 residues of the peptide signal of the ephrin receptor tyrosine kinase family (RPTK) (39), may correspond to a signal peptide (Fig. 2B). The Cterminal hydrophobic region is likely to be a membrane-spanning ␣-helix. The amino acid sequence also predicts eight potential N-glycosylation sites at Asn residues ( Fig. 2A), which may explain the discrepancy between the 115-kDa apparent molecular mass of the purified protein as determined by SDS-PAGE and the molecular mass, 45.3-kDa, as determined from the amino acid sequence. A search of sequence data bases with Bisance service (40) and sequence alignments with Match-Box service (41) revealed no protein with statistically significant similarities to the tryAcP115 gene sequence.
Genomic Southern Blot and mRNA Expression Analysis of T. brucei AnTat tryAcP115 Gene-After digestion of genomic DNA with several restriction enzymes, Southern blot patterns indicated (data not shown) that these trypanosomal genomes harbor a single copy gene of tryAcP115.
Northern blots probed with tryAcP115 revealed a 2-kilobase messenger RNA only in bloodstream forms (data not shown). This is in accord with the bloodstream form specific PTPase activity that we characterized previously (30).
Expression, Glycosylation, and Localization of the Try-AcP115 Protein-An histidine-tagged recombinant protein, Petase7, containing the N-terminal part of TryAcP115 deleted from the signal peptide ( Fig. 2A) was expressed in E. coli, purified by Ni 2ϩ His⅐Bind TM and affinity O-Tyr(P)-agarose chromatography and used to immunize rabbits. Western blots confirmed the differential expression of the 115 acidic protein phosphatase. It (Fig. 3A) was recognized by the rabbit antiserum in fraction F3 of bloodstream trypanosomes, but not in the equivalent procyclic trypanosome fraction. Several questions raised by the sequence analysis including the degree of glycosylation, the membrane localization, and the orientation of TryAcP115 were also addressed. The antiserum reacted in immunoblots with a high molecular mass band of 115-kDa (Fig.  3, A and B) which shifted to 55-kDa after N-glycosidase F treatment (Fig. 3B, lane F4 dg) showing that the TryAcP115 protein (45.3 kDa estimated molecular mass) was highly glycosylated. A minor band of about 66 kDa (Fig. 3B, lanes F2, F3,  and F4) was also recognized and may represent a partially glycosylated product of TryAcP115 or cleaved peptide. Enrichment of the TryAcP115 protein content during membrane fraction preparation was accompanied by enrichment of total plasma membrane protein in the successive fractionation steps (Fig. 3B,  lanes F2, F3, and F4) (30). Under these conditions TryAcP115 was not detectable in the total cell lysate fraction (F1). a F4TS, BT, and EL correspond to the triton-solubilized fraction, the nonretained and the salt eluted fractions from the different columns, respectively. DEAE and AFF indicate the ion exchange, and O-Tyr(P) columns, respectively.
b The resulting yield measured by activity against Tyr(P)-Raytide of the (BT ϩ 150EL) DEAE fast flow column fractions is larger than 100% indicating the probable removal of an inhibitor from the F4TS fraction.

Bloodstream Form-Specific Acidic Ecto-phosphatase of T. brucei
Indirect immunofluorescence showed that the antiserum reacted specifically with surface elements of the parasite body (Fig. 4). Interestingly, the protein appeared to be unequally distributed along the cell body. Regions with higher fluorescence density indicated a patchwork organization of small protein aggregates and larger fluorescent regions. Subsequently, we attempted to determine the orientation of TryAcP115 across the membrane. Although the N terminus was glycosylated (Fig.  3B) and a signal peptide had been observed (Fig. 2), we had no definitive evidence that the N-terminal domain was extracellular. In order to study this orientation further, we digested surface proteins of living trypanosomes with trypsin for 5, 10, and 15 min (Fig. 3C). During the treatment and 1 h after blocking, the motility of the cells was unaffected, and the parasites retained their elongated shape. Analysis of these proteins by SDS-PAGE, Coomassie staining, and Western blotting with the same rabbit antiserum showed that a 30-kDa peptide was released from the TryAcP115 protein (Fig. 3C). Control trypanosomes incubated without trypsin (Fig. 3C)   ; lane pet7, 0.5 g of affinity purified Petase7. Proteins were separated by reducing 12% SDS-PAGE, blotted to polyvinylidene difluoride probed with rabbit antiserum (1/500) directed against Petase7, and developed with a goat peroxidase-conjugated second antibody (1/1000) from Pasteur Sanofi. C, the N-terminal region of TryAcP115 is extracellularly located. Surface proteins from 3 ϫ 10 7 trypanosomes were digested with 80 g of trypsin (ϩ) in 1 ml of TD buffer (20 mM Na 2 HPO 4 , 2 mM NaH 2 PO 4 , 1 mM MgCl 2 , 80 mM NaCl, 5 mM KCl, 20 mM glucose, pH 7.4) for 5, 10, and 15 min. As a control the same amount of trypanosomes were incubated under the same conditions for 10 and 30 min without trypsin (Ϫ). After digestion 100 g of trypsin inhibitor (Serva) was added and trypanosomes centrifuged for 10 min at 600 ϫ g. The collected supernatant was centrifuged for 45 min at 100,000 ϫ g. Trifluoroacetic acid (1% final) was added to the resulting supernatant, dialyzed overnight against water, lyophilized, and resuspended in water. Proteins were separated on reducing 12% SDS-PAGE, blotted to polyvinylidene difluoride probed with rabbit antiserum (1/100) directed against Petase7, and revealed by treatment with goat phosphatase-conjugated second antibody (1/2000). that the tryAcP115 gene encoded a protein with intrinsic PT-Pase activity, we examined the recombinant proteins Petase7 and pTEX/AcP (entire N-terminal domain expressed in L. amazonensis) for phosphatase activity (Fig. 5). Petase7 overexpression and purification by a two-step procedure were analyzed by Coomassie staining (Fig. 5A). PTEX/AcP expression was examined by immunofluorescence (Fig. 5, D and E). A 60-kDa protein in the culture supernatant of Leishmania was identified, after immobilized metal affinity chromatography purification, by probing Western blots with the Petase7 antiserum (Fig. 5F). Phosphatase activity against pNPP was undetectable in the corresponding purified fractions of E. coli and Leishmania controls expressing non-related proteins. Tyr(P)-Raytide was dephosphorylated by Petase7 with an optimum pH of 4 (Fig. 5B). The specific activity was estimated to 3.6 (Ϯ2.6) pmol/min/mg. However, the specific activity of pTEX/AcP against the same substrate (200 pmol/min/mg) was much higher (Fig. 5G). This large difference in specific activity could be partially due to the presence of protein inactivated by improper folding in E. coli. As shown in Fig. 5 (see legend), only 5-20% of the His⅐Tag eluted protein was retained on the O-Tyr(P) affinity column. Although the specific activity of Petase7 for Tyr(P)-Raytide increased after this purification step, it only reached an average value of 16 (Ϯ13) pmol/min/mg. Interestingly the specific activity of PetaseM4, a mutated Petase7 enzyme, with Cys-166 and -185 replaced by serine and glycine residues, respectively, was reduced to 0.2 (Ϯ0.1). These cysteine residues were initially chosen because they were part (the arginine residue was absent) of the hallmark of the PTPase family (42). In contrast, the K m values of Petase7 and pTEX/AcP of about 15 nM were similar to TryAcP115 (Fig. 5, C and G) Inhibitor Profiles of Petase7 Protein-By protein reconstitution (30) we have previously characterized a 115-kDa peptide in the F4 fraction as a protein-tyrosine phosphatase sensitive to vanadate, tartrate, and sodium fluoride. A modulator profile similar to the F4TS fraction was obtained with TryAcP115, Petase7, and pTEX/AcP. Among the usual phosphatase inhibitors tested, vanadate, bVp(phen), molybdate, sodium fluoride, and tartrate inhibited the phosphatase activity with IC 50 values of 3, 2, 4, 100, and 250 M, respectively. None of these preparations were affected by 150 nM okadaic acid, an inhibitor of type-2A and type 1 protein Ser/Thr phosphatases, nor by 1 mM tetramisole, which inhibits alkaline phosphatases. The divalent cations, Ca 2ϩ , Cu 2ϩ , Zn 2ϩ , and Ni 2ϩ tested at 1 mM inhibited the phosphatase activity by about 50% while Mg 2ϩ stimulated it (Table II).
To test the substrate of TryAcP115 and the recombinant proteins, we first examined three sets of phosphorylated substrates (phosphorylated nucleotides, phosphorylated sugars, and tyrosine phosphatase substrates) as inhibitors of Petase7 phosphotyrosyl phosphatase activity. ATP at only 1 mM inhibited significantly (50%), while 5 mM pNPP were necessary to inhibit equally. The serine-phosphorylated substrate, Kemptide, was a poor inhibitor (20% at 1 mM). Other phosphorylated sugars, O-phospho-L-tyrosine, and other nucleotides failed to inhibit significantly. In a second step, compounds which inhibited the phosphotyrosyl phosphatase activity of TryAcP115 and the recombinant proteins were tested as substrates for their phosphatase activity. We used the pTEX/AcP protein in these assays because it exhibited the highest specific activity. As shown in Table III, while Ser(P)-Kemptide was not a substrate for pTEX/AcP, the recombinant enzyme dephosphorylated the nonspecific substrate pNPP, ATP, and the serine-phosphorylated peptide, neurogranin, with K m and V max values, respectively, of 26 mM, 1.8 mol/min/mg; 250 M, 120 nmol/min/mg; and 200 nM, 200 pmol/min/mg, respectively. In contrast, pNPP and Ser(P)-neurogranin were poor substrates for Petase7, but it dephosphorylated ATP with kinetic constant values of 200 M and 24 nmol/min/mg for K m and V max , respectively. This ATPase activity was magnesium dependent and inhibited by 1 mM azide, 5 mM ADP, and 1 mM NTPs at 50, 80, and 50%, respectively. DISCUSSION We have purifed, cloned, and characterized an acidic ectophosphatase. Throughout the purification process the phosphotyrosyl peptide phosphatase activity and pNPP dephosphorylation cofractionated with a constant relative activity ratio (Table I). This tight association of phosphatase activities suggested that they were mediated by the same enzyme. Because the O-phospho-L-tyrosine ligand was a poor substrate for Try-AcP115 (Table II), the O-Tyr(P)-agarose column retained the phosphatase without significantly dephosphorylating it. After the affinity purified active fraction was eluted with a salt gradient (Fig. 1), Coomassie and silver stains revealed three polypeptides of 115, 66, and 30 kDa. The 115-kDa polypeptide was identified by membrane reconstitution activity as the protein harboring the tyrosine phosphatase activity (Fig. 1E). The 30-kDa protein and its genomic sequence were not homologous with known phosphatases. The 66-kDa protein was recognized by antiserum to the N-terminal region of TryAcP115 suggesting that it was differently processed or partially degraded (Fig.  3B). The recombinant pTEX/AcP protein of 60-kDa expressed in L. amazonensis may correspond to this 66-kDa protein.
Sequence analysis of several T. brucei subspecies revealed that tryAcP115 is invariant and ubiquitous in the Trypanozoon subgenus. Discrepancy between the apparent (115 kDa) and predicted (45 kDa) molecular mass, as well as the apparent mass (55 kDa) after N-glycosidase F (Fig. 3B) treatment confirmed modification by N-linked glycans. The existence of 8 potential sites of N-glycosylation within the N-terminal domain ( Fig. 2A) corroborated this conclusion. The calculated isoelectric point of 5.62 may explain the slower migration on SDS gels of the deglycosylated protein and the expressed protein in E. coli (Petase7) (Fig. 3B).
The primary sequence of TryAcP115 predicted the presence of a N-terminal signal peptide, a large extracellular N-terminal domain, a C-terminal transmembrane domain, and a short C-terminal intracellular domain of 30 amino acids (Fig. 2). The N-terminal extremity contains the three structural features characteristic of a signal peptide (43,44): (i) a positively charged N-terminal amino acid (n region); (ii) a central hydrophobic domain (h region) of 16 amino acids (7 residues is the FIG. 4. Subcellular localization of TryAcP115 by immunofluorescence. Immunofluorescence of bloodstream form trypanosomes (MITat). The rabbit antiserum directed against the N-terminal region of TryAcP115 (1/200) was used as first antibody and goat anti-rabbit fluorescein-conjugated as the second antibody. No signal was observed when the antiserum was used in the same conditions on procyclic forms. minimum required); and (iii) a COOH-terminal polar domain containing a glycine residue (c region). This c region is about 4 to 6 amino acids long and typically contains a proline or a glycine residue. During translocation the signal peptide is cleaved by signal peptidase with specificity due partly to the structural features of the signal peptide. The sequence homology of TryAcP115 with signal peptides of RPTK of the ephrin family (Fig. 2B), may suggest cleavage by a similar endopeptidase (45). Although the recombinant L. amazonensis protein pTEX/AcP, containing this potential signal peptide was purified from the culture supernatant, it failed to identify similarities of the peptidase similar to that of the ephrin family, since the same protein deleted of the signal peptide was also recovered from the supernatant (data not shown).
In order to study the localization, membrane orientation, and phosphatase activity of the TryAcP115 protein, its N-terminal domain was cloned and expressed in E. coli. The recombinant protein, Petase7, was purified by His⅐Tag and affinity columns (Fig. 5A). Indirect immunofluorescence with antiserum to Petase7 confirmed the surface localization of the TryAcP115 protein (Fig. 4). It was unequally spread over the cell surface and organized in patchwork patterns of larger stained areas.
The extracellular location of its N-terminal region was verified when Western blots of trypsin-treated living trypanosomes (Fig. 3C) revealed a partially degraded N-terminal domain. Since no TryAcP115 protein was released without trypsin treatment, the presence of the soluble degraded protein in the Western blot is not due to secretion. This result strengthened our conclusion that the N-terminal domain of TryAcP115 protein was oriented extracellularly. The short C-terminal cytoplasmic tail was not homologous to any other intracellular domain. Furthermore, it contained neither a tyrosine residue nor a Leu-Ile motif, which are essential for endocytosis and lysosome targeting (46,47) . The importance of C-terminal domains in protein localization suggests that the intracellular domain of TryAcP115 may have a pivotal role in the distribution of this protein over the parasite body, but its lack of homology with other intracellular domains prohibits further speculation about its function.
To determine if the phosphatase activities demonstrated by the F4 and F4TS fractions and by the affinity eluted Try-AcP115, Petase7, and pTEX/AcP proteins were due to the same protein, we compared their kinetic constants, modulator effect profiles, and substrate specificity. First, we observed that the FIG. 5. Expression, purification, and characterization of the phosphatase activity of the purified Petase7, PetaseM4, and pTEX/ AcP recombinant proteins. A, protein samples obtained during the production and purification steps were analyzed by electrophoresis on 12% reducing SDS-polyacrylamide gels and visualized by Coomassie Blue staining. Lane T, 4 g of total lysate proteins (6.3 mg total) from BL21 expressing Petase7; lane Ni, 0,5 g of eluted protein (0.35 mg total) from the Ni 2ϩ His⅐Bind column; lane Aff, 0,5 g of Petase7 protein (0.07 mg total) recovered from the affinity column O-Tyr(P). B, effect of pH on Petase7 tyrosine phosphatase activity. 2 g of Petase7 protein eluted from the His⅐Bind column were incubated for 60 min with 100 nM Tyr(P)-Raytide at different pH values as described under "Experimental Procedures." C, and G, kinetics of Petase7 and pTEX/AcP dephosphorylation activities. Tyrosine phosphatase activity was measured at concentrations of 1-200 nM Tyr(P)-Raytide. 2 g of Petase7 (⅐) and 30 ng of pTEX/AcP (Ⅺ) eluted from the Ni 2ϩ His⅐Bind column were assayed for 1 h. Double-reciprocal transformation (inset) revealed an apparent K m value of 15 nM. These data represent typical results of several experiments. D and E, nontransformed (D) and pTEX115 transformed (E) L. amazonensis were tested for TryAcP115 expression by immunofluorescence using the rabbit antiserum anti-Petase7 (1/2000) as first antibody and goat anti-rabbit conjugated to fluorescein as second antibody. F, Western blot analysis of the the Ni 2ϩ His⅐Bind eluted proteins from L. amazonensis culture supernatant. Proteins were loaded on a 12% reducing SDS-polyacrylamide and revealed after transfer with the rabbit anti-Petase7 serum used as first antibody (1/200). Bloodstream Form-Specific Acidic Ecto-phosphatase of T. brucei