Identification and Characterization of a Protein-tyrosine Phosphatase in Leishmania

Leishmania parasites are eukaryotic protozoans responsible for a variety of human diseases known as leishmaniasis, which ranges from skin lesions to fatal visceral infections. Leishmania is transmitted by the bite of an infected sandfly where it exists as promastigotes and, upon entry into a mammalian host, differentiates into amastigotes, which replicate exclusively in macro-phages. The biochemical pathways enabling Leishmania to differentiate and survive in the mammalian host are poorly defined. We have therefore examined the role of protein-tyrosine phosphorylation, which is essential in regulating cell function in higher eukaryotes. Using the recently completed Leishmania genome, we have identified and cloned a Leishmania protein-tyrosine phosphatase (PTP) gene (LPTP1) by virtue of its homology with the human protein-tyrosine phosphatase 1B gene (hPTP1B). The enzyme activity of recombinant LPTP1 was confirmed using a combination of PTP-specific substrates and inhibitors. We further demonstrate, by creating LPTP1 null mutants through gene targeting, that LPTP1 is necessary for survival as amastigotes in mice, but it is dispensable for survival as promastigotes in culture. Human PTPs, including the PTP1B enzyme, are actively pursued drug targets for a variety of diseases. The observations with the LPTP1 mutants in mice suggest that it may also represent a drug target against the mammalian amastigote stage. However, in silico structure analysis of LPTP1 revealed a striking similarity with hPTP1B in the active site suggesting that, although this is an attractive drug target, it may be difficult to develop an inhibitor specific for the Leishmania LPTP1.

and increases virulence in BALB/c mice (7). These observations argue that PTPs play a role in amastigote survival in the mammalian host and provide strong justification for the characterization of endogenous Leishmania PTP genes as detailed within.
In the present study, we describe the identification of a PTP gene from Leishmania major, Leishmania infantum, and L. donovani that has extensive sequence and corresponding structural homology with the human PTP1B gene product. These Leishmania genes have been designated LmPTP1, LiPTP1, and LdPTP1, respectively. The enzyme activity of LmPTP1 has been confirmed with relevant substrates and inhibitors. We have also developed LdPTP1 heterozygous and homozygous null mutant knock-out clones that proliferated in a similar fashion to wild type promastigotes in culture but were severely impaired with respect to survival as amastigotes in BALB/c mice. Through in silico structural analysis, we also show that the L. infantum PTP1 (LiPTP1) and human PTP1B (hPTP1B) shared remarkable structural conservation in the active site; however, notable differences outside this region are also present. This observation suggests that from a phenotypic perspective, the LPTP1 represents an attractive drug target. However, from a structural perspective, it may be difficult to develop a small molecule specific to the active site of LPTP1.

EXPERIMENTAL PROCEDURES
Parasite Cultures-The L. donovani 1S/Cl2D and L. major Friedlin V9 promastigotes were routinely cultured at pH 7.2 and 27°C in Medium 199 medium (Invitrogen) supplemented with 10% fetal bovine serum. L. donovani differentiation into amastigotes was performed by shifting to amastigote culture medium (37°C, pH 5.5 in RPMI 1640 plus 10% fetal bovine serum) overnight, which mimics the temperature and pH of the host macrophage phagolysosome.
Cloning, Sequencing, and Tagging of L. major PTP1-The 1.5-kb L. major DNA fragment homologous to the human PTP1B was identified in the L. major data base by BLAST search (Entry GeneDB LmjF36.5370; www.genedb.org). Based on the sequence obtained from the data base, two primers were designed to PCR amplify the LmPTP1 gene and to incorporate a sequence encoding a His tag at the N terminus, and the amplified product was then ligated into the mammalian expression vector pcDNA3. The primers used were HisF 5Ј-aagcttATGG-GCCATCATCATCATCATCATCATATGTGTGAAAAGC-AACTCAAGGAG-3Ј, which contained a HindIII site, and reverse R 5Ј-ggatccTTACACAAACGAAGGCGAGAAG-CGC-3Ј, which contained a BamHI site. The resulting Histagged LmPTP1 containing plasmid was used as template in a second PCR, where a new primer incorporated a 10-amino acid epitope tag from the L. donovani specific A2 protein that is recognized by an anti-A2 monoclonal antibody (8) A2-HisF 5Ј-aagcttATGCAGTCCGTTGGCCCGCTCTCCGTTGGCCC-GCATCATCATCATCATCATCAT-3Ј with a HindIII site and the same R 5Ј-ggatccTTACACAAACGAAGGCGAGAAG-CGC-3Ј with a BamHI site was used. The final amplified product was then cloned into the pcDNA3 vector and called pcDNA3-LmPTP1 as shown in Fig. 2A. As a positive control for subsequent experiments, the human PTP1B sequence was also PCR-amplified with the A2/His tag, using the oligonucleotide primers A2HisF (hPTP) 5Ј-ggatccATGCAGTCCGTTGGCC-CGCTCTCCGTTGGCCCGCATCATCATCATCATCATC-ATATGGAGATGGAAAAGGAGTTCGAG-3Ј with an EcoRI  site and the A2HisR (hPTP) 5Ј-gaattcCTATGTGTTGCTGT-TGAACAGGAAC-3Ј with a BamHI site. The L. donovani LdPTP1 gene was PCR-amplified from L. donovani 1S/Cl2D genomic DNA with the same primers used above for L. major and sequenced (GenBank TM ban-kit827438 DQ862810). In brief, the amplified fragment was cloned into the TOPO TA cloning vector (Invitrogen) and termed pLdPTP1-TOPO, and M13 reverse and M13 forward primers were used in sequencing reactions with the Mega-Bace500 (Molecular Dynamics of GE Healthcare). Sequencing reactions were performed by a DYEnamic ET TerminatorCycle sequencing kit with Thermo Sequenase II DNA polymerase, and post-reaction cleanup was achieved by extensive ethanol precipitation before adding formamide loading solution.
Transfection of Cos7 Cells and Purification of A2His-tagged LmPTP1 and hPTP1B-The Cos7 cell line was transiently transfected using Lipofectamine reagent according to the manufacturer's protocol (Invitrogen). Briefly, a 100-mm dish of Cos7 cells (ϳ8 ϫ 10 5 cells) was transfected with a total of 8 g of DNA (6 g of pcDNA3-LmPTP1 and 2 g of a ␤-galactosidase expression plasmid) in 20 l of Lipofectamine. The following day (ϳ20 h after transfection), the cells were washed with cold medium and lysed with Nonidet P-40 lysis buffer (150 mM NaCl, 1% Nonidet P-40, 20 mM Tris pH 8.0) with protease inhibitors (Roche Applied Science complete mixture tablets) on ice for 30 min. The cell lysates were centrifuged, and ␤-galactosidase assays were performed to determine the levels of transfection efficiency in each dish and to normalize the amount of protein used in each assay. His-bind resin (Novagen) was used for purification of the A2-His-tagged LmPTP1 and hPTP1B according to the manufacturer's protocol. Briefly, His-bind resin was activated in 1-binding buffer (8ϫ ϭ 4 mM NaCl, 160 mM Tris-HCl, 40 mM imidazole, pH 7.9) with 0.1% Nonidet P-40. 60 l of activated His-bind resin was added to 200 l of cell lysates and incubated with agitation for 3 h at 4°C. His-bind resin was thoroughly washed five times in 1.0 ml of 150 mM NaCl, 1%Nonidet P-40, 20 mM Tris, pH 8.0 buffer with protease inhibitors (Roche Applied Science complete mixture tablets), and half of the His-bind resin with the purified A2-His-tagged LmPTP1 or hPTP1B was used for activity assays, and the other half used for Western blot analysis with anti-A2 tag monoclonal antibodies.
Protein-tyrosine Phosphatase Activity and Inhibition Assays-The 4 p-nitrophenylphosphate (pNPP) assay was used for detection of total phosphatase activity as described previously (7). Briefly, 30 l of His-bind resin containing purified PTP proteins was washed once in 1.0 ml of 150 mM NaCl, 1% Nonidet P-40, 20 mM Tris, pH 8.0, Roche Applied Science complete protease inhibitor buffer and placed in a 96-well plate. 180 l of reaction buffer (50 mM Hepes, pH 7.5, 0.1% ␤-mercaptoethanol containing 10 mM fresh pNPP) was added to each well, and the plate was incubated at 37°C overnight. The plates were read at 405 nm.
The Malachite green phosphatase activity assay with the insulin receptor (IR) phosphopeptide is a more specific assay for hPTP1B, because the IR is a major substrate of this enzyme (9, 10). The assay was performed according to the manufacturer's instructions (Sigma; PTP101, nonradioactive phosphotyrosine phosphatase assay).
Two inhibitors, 1 mM of sodium orthovanadate (Na 3 VO 4 ), and 10 M of potassium bisperoxo(1,10-phenanthroline)oxovanadate (V) (bpV (phen)) were used to further confirm the activity of LmPTP1 and the hPTP1B as the control. bpV (phen) is a member of a class of potent and specific PTP inhibitors (11,12). For the inhibition assays, His-bind resin containing LmPTP1 and hPTP1B was washed five times with 1.0 ml of 150 mM NaCl, 1% Nonidet P-40, 20 mM Tris, pH 8.0, Roche Applied Science complete protease inhibitor buffer and then incubated with inhibitors for 1 h at 4°C. His-bind resin containing the PTPs was washed four times with 1.0 ml of 150 mM NaCl, 1% Nonidet P-40, 20 mM Tris, pH 8.0, Roche Applied Science complete protease inhibitor buffer and then assayed for phosphatase activity using pNPP as substrate, as detailed above.
Disruption of the LdPTP1 Genes from L. donovani-The L. donovani PTP1 gene disrupted strains were generated by homologous gene targeting as outlined in Fig. 4A. A Hin-dIII and XbaI fragment containing the L. donovani PTP1 gene from the pLdPTP1-TOPO plasmid described above was subcloned into a pBluescript vector. The resulting pBSLdPTP1 plasmid was digested with BclI to remove the 357-bp catalytic region of the LdPTP1 gene. The fragment containing the hygromycin resistance gene was removed from the pSPY hygromycin vector (13) with BamHI and BglII and inserted into the BclI site within the LdPTP1 sequence, generating the plasmid pBSLdPTP Hyg. The linear fragment containing the hygromycin gene and the LdPTP flanking sequences was then electroporated into L. donovani. Transfectants were initially selected, in the first round targeting, with 50 g/ml hygromycin to obtain the heterozygous single LdPTP1 knock-out mutant. The double knock-out homozygous null mutant for the LdPTP1 gene was achieved by increasing the hygromycin concentration to 200 g/ml in selection culture medium.
Complementation of the double knock-out null mutant with a plasmid containing the LdPTP1 gene was carried out as follows. The following oligonucleotide primers LdPTPF1 5Ј-ccc-aagcttTCACTTTTTGTTGCCCTTGGT with a HindIII site and the reverse LdPTPR1 5Ј-cgagatctCAGAGGTGCAGCCA- GTCATA with a BglII site were used to amplify a 3140-bp fragment from L. donovani genomic DNA that contained the LdPTP1 gene open reading frame including 655-bp upstream and 1000-bp downstream flanking sequences. The 3140-bp fragment was then inserted into HindIII and BamHI sites of plasmid pSPY-Neo (13) to generate the complementing plasmid pSPYNeoLdPTP1 as shown in Fig. 7A.
Southern Blotting-For Southern blot analysis, 10 g of Leishmania genomic DNA was digested with restriction enzymes PstI and SstI and separated in a 0.7% agarose gel. Hybridization and washing were performed as previously described (14). The LdPTP1 active domain encoding DNA (357-bp BclI fragment from nucleotides 677-1034) was used as a probe for DNA from the single (ϩ/Ϫ) and double knock-out (Ϫ/Ϫ) clones described above to demonstrate the disruption of the LdPTP1 gene (Fig. 4B). Southern blot using the hygromycin gene demonstrated specific targeting into the LdPTP1 gene (Fig. 4C). The L. major 1.5-kb LmPTP1 gene containing fragment was used in the Southern blot to confirm the presence of the episomal LdPTP1 added back to the double knock-out null clone Ld1PTP1 Ϫ/Ϫ (Fig. 7B).
Infection of BALB/c Mice and Recovery of Amastigotes-Female BALB/c mice (Charles River) weighing 20 -25 g (n ϭ four mice/group) were injected via tail vein with 1.5 ϫ 10 8 late log phase promastigotes in 100 l of phosphate-buffered saline, as described previously (15). After 4 weeks of infection, the mice were examined for L. donovani parasite burden by counting the number of amastigotes in the Giemsa-stained liver imprints. Liver parasite burden, expressed as Leishman-Donovan units, was calculated by multiplying the number of amastigotes/1000 cell nuclei ϫ liver weight (g). Spleen parasite burden were determined by limiting dilution in 96-well plates as previously detailed (16).
Proliferation in Culture-Parasite growth was evaluated by determination of the optical density at 600 nm of diluted cultures (starting from 10 6 cell/ml) grown in 96-well plates from days 0 -8 (promastigotes) or 0 -5 (amastigotes).
LiPTP1 Catalytic Domain Modeling-The alignment between the L. infantum PTP1 amino acid sequence and human PTP1B obtained with ClustalW ( Fig. 1) in conjunction with the PTP1B crystal structure (Protein Data Bank code 1SUG) were used in Modeler 8 (version 2, default configuration) (17) to create the L. infantum PTP1 homology model. A 20-residue stretch of amino acids present in the Leishmania PTP1 sequence that was outside of the enzyme active site was not present in the human enzyme sequence and, as a result, was not modeled (residues 32-51). In addition, the construct used to elucidate the human enzyme structure did not contain any residues beyond position 319. Therefore, these two regions were removed from the Leishmania homology model.

RESULTS
We began this study by performing a BLAST search of the L. major data base (Ref. 18; www.genedb.org) for sequences that could represent PTP genes by virtue of their homology with the human PTP1B gene. In total, nine potential PTP genes were identified in the Leishmania genome. The one with the greatest identity with human PTP1B was LmjF36.5370 located on chro-mosome 36, which we have designated LmPTP1. The L. infantum PTP1 (LiPTP1) sequence was also identified in this manner from the L. infantum data base (LinJ36.5860; www.genedb.org). Based on these sequences, we designed PCR primers to amplify, clone, and sequence the L. donovani LdPTP1 homolog (Gen-Bank TM bankit827438 DQ862810) as detailed under "Experimental Procedures." Alignment comparison of the various Leishmania PTP1 and hPTP1B proteins revealed they share ϳ40% sequence identity, including a number of important conserved amino acid residues within the hPTP1B signature catalytic domain ((I/V)HCXXGXXR(S/T/G)), which contains the essential cysteine and arginine residues required for enzyme activity (boxed region in Fig. 1) (reviewed in Ref. 4). In addition to the conserved catalytic domain, LPTP1s share relevant accessory motifs with hPTP1B including the adjacent signature WPD and the Q residues (boxed region indicated with bold type in Fig. 1), which play a role in maintaining the conformation of the active site. The LPTP1s and hPTP1B also share a proline-rich region from amino acids 325-340 that are responsible for SH3 domain protein-protein type interactions and cellular localization in hPTP1B (4). Southern blot analysis of genomic DNA from L. major demonstrated that LmPTP1 was a single copy gene in the haploid genome. 6 Included in Fig. 1 is the closest PTP sequence homolog from Saccharomyces cerevisiae, which is more divergent from hPTP1B than is the LPTP1s.  Although the Leishmania PTP1 sequences shown in Fig. 1 suggest that they encode for a hPTP1B homolog, it was necessary to validate this experimentally using relevant enzyme substrates and inhibitors. We designed primers to amplify, clone, and insert the LmPTP1 gene into a eukaryotic expression vec-tor (pcDNA3) for expression in transfected simian Cos7 cells. To detect and partially purify the LmPTP1 from Cos7 cells, a 10-amino acid epitope tag derived from the L. donovani A2 protein, followed by a His 7 tag encoding sequence, were inserted at the 5Ј end of the LmPTP1 gene ( Fig. 2A). In this FIGURE 3. LmPTP enzymatic activity and inhibition following expression in Cos7 cells. LmPTP1 and hPTP1B purified on the His-bindா resin were assayed on PTP substrates: pNPP (A) and IR phosphopeptide (B) as indicated. Also shown is a Western blot confirming equal levels of PTP protein used for each substrate enzyme assay. For the enzyme inhibition assays, LmPTP1 and hPTP1B purified on the His 7 tag resin were assayed in the absence (Ϫ) or presence (ϩ) of the PTP inhibitors; Na 3 VO 4 (C) and bpV (phen) (D) as indicated. Also shown is a Western blot confirming equal levels of PTP protein used in the presence (ϩ) or the absence (Ϫ) of the inhibitors in each assay. Note than both inhibitors impaired the enzyme activity to levels similar to that derived from the control transfected cells (control pcDNA3). The values reported are the average means of three independent experiments, and the results in bar graphs are the means Ϯ S.E. Microsoft Excel was used to calculate the Student's test. *, p Յ 0.05; **, p Յ 0.01, statistical difference from control. manner, the Cos7 cell expressing A2-His 7 -tagged LmPTP1 could be detected with anti-A2 monoclonal antibodies (8) and partially purified from cell lysates by affinity chromatography with His-bind resin. We also generated the same construct using the human hPTP1B gene as a positive control for subsequent comparison.
Expression of the A2-His 7 tagged LmPTP1 and hPTP1B genes in Cos7 cells was analyzed by Western blot analysis with anti-A2 monoclonal antibodies 24 h after transfection. As shown in Fig. 2B, the hPTP1B and LmPTP1 proteins were detectable at the predicted molecular masses of 50 and 55 kDa, respectively (lanes 2 and 3). The control empty pcDNA3 vector did not produce bands (lane 1), confirming the specificity of the Western blot for A2-tagged PTPs. The A2-His 7 -tagged proteins were subsequently partially purified from the transfected Cos7 cell lysate using His-bind resin and washed extensively, followed by Western blot analysis with anti-A2 monoclonal antibodies. As shown in Fig. 2B  (lanes 5 and 6), the washed His tagbinding resin contained approximately equal amounts of hPTP1B and LmPTP1. These data confirmed that it was possible to express, detect, and partially purify similar levels of hPTP1B and LmPTP1 from transfected Cos7 cells.
We next determined whether we could detect PTP enzyme activity in the washed His-bind resin containing the extracted hPTP1B and LmPTP1 from the transfected Cos7 cells. Two protein phosphatase substrates were used for these assays including pNPP and a specific tyrosine phosphatase substrate, IR phosphopeptide (Fig. 3, A and B). For this assay, the hPTP1B and LmPTP1 expression constructs were transfected into Cos7 cells, purified on His-Tag resin, assayed for activity, and further subjected to parallel Western blot analysis to confirm similar levels of hPTP1B and LmPTP in each assay. As shown in Fig. 3 (A and B), both LmPTP1 and hPTP1B enzyme activities were detected on the His-bind resin with the pNPP and IR phosphopeptide substrates when compared with the control (His tag-binding resin from control pcDNA3-transfected cells).
To confirm that the activity detected in the samples shown was due to PTP activity, we determined whether it was possible to specifically inhibit this activity using the protein phosphatase inhibitor Na 3 VO 4 and a more specific protein-tyrosine phosphatase inhibitor bpV (phen) (12,19). As shown in Fig. 3 (C and D), both the LmPTP1 and hPTP1B were inhibited with Na 3 VO 4 and bpV (phen) (white bars). Accompanying Western blots confirmed there were similar levels of LmPTP1 and hPTP1B assayed in the presence (ϩ, white bars) and absence (Ϫ, black bars) of these inhibitors. Importantly, the inhibitors did not affect the background activity observed on His tag-binding resin from the control vec- A hygromycin selectable marker gene was inserted into a BclI site of the pBsLdPTP1 construct containing the LdPTP1 gene, resulting in the construct pBsLdPTP1-Hyg, which was then linearized and transfected into L. donovani promastigotes. For the single gene targeting (heterozygous LdPTP1 knock-out), recombinant parasites were selected in 50 g/ml hygromycin, and for the double gene targeting (homozygous LdPTP1 knockout) recombinant parasites were selected in 200 g/ml hygromycin. B, Southern blot analysis of genomic DNA derived from the wild type (ϩ/ϩ), the heterozygous (ϩ/Ϫ), and homozygous (Ϫ/Ϫ) gene targeted clones following digestion with PstI or SstI. The probe consisted of the 357-bp fragment (see A) derived from the LdPTP1 active site encoding region. C, Southern blot analysis as in B above with a probe consisting of the Hyg R gene. Note that these Southern blots confirm the removal of the 357-bp fragment containing the active site of the LdPTP1 gene in the homozygous (Ϫ/Ϫ) gene targeted clones and the Hyg R gene targeted specifically into the LdPTP1 gene in all clones tested. tor transfected cells (pcDNA3 control). Taken together, these data confirm the bioinformatic prediction that the LmPTP1 gene encodes for a protein-tyrosine phosphatase enzyme that, when assayed under these conditions, had a level of activity similar to that of hPTP1B.
Once the Leishmania PTP1 gene had been identified and characterized as detailed above, it was necessary to determine its role in the life cycle of the parasite by developing null PTP1 mutants. We performed single and double knock-outs of the catalytic domain of the PTP1 gene from L. donovani and characterized the resulting mutant parasite phenotype. L. donovani was chosen for this analysis because it can be cultured in vitro as both promastigotes and amastigotes, whereas L. major can only be cultured as promastigotes. Additionally, it has been previously shown that overexpression of a transfected hPTP1B gene in L. donovani resulted in increased virulence in the amastigote stage (7). Because Leishmania are diploid organisms, the two alleles of the catalytic domain of the LdPTP1 gene were targeted for deletion as summarized in Fig. 4A. The BclI restriction enzyme fragment containing the LdPTP1 catalytic domain (nucleotides 677-1034; Fig. 1) was replaced with a BamHI-BglII fragment containing the selectable marker gene conferring hygromycin (Hyg) resistance. The resulting plasmid, termed pBsLdPTP1hyg, was linearized and targeted into the LdPTP1 site of the L. donovani genome by transfection, and transformants were selected for hygromycin resistance. The first round of gene targeting (heterozygous deletion) was carried out using 50 g of hygromycin selection, and the second round (homozygous deletion) was carried out using 200 g of hygromycin. In this manner, both LdPTP1 alleles could be targeted with one selectable marker. Cultures were then subjected to serial dilution to isolate individual LdPTP1 knock-out clones.
Southern blot analyses were performed to confirm the heterozygous and homozygous targeted deletion of the catalytic domain of the LdPTP1 gene in the cloned L. donovani mutant cultures. For this analysis, the 357-bp BclI restriction enzyme fragment, which was deleted by gene targeting (Fig. 4A), was used as the hybridization probe. As shown in Fig. 4B, the BclI LdPTP1 gene fragment encoding the catalytic domain was eliminated from the L. donovani clones as indicated by the absence of the band containing this sequence in the null mutant double knock out (Ϫ/Ϫ) cultures and further by a 50% reduction in the single knock out clone (ϩ/Ϫ), as compared with wild type L. donovani (ϩ/ϩ) culture. PCR analysis of the LdPTP1 gene in these mutant clones confirmed the Southern blot data showing deletion of the catalytic domain (data not shown). We  further confirmed the accurate replacement of the Hygromycin resistance gene specifically into the LdPTP1 locus on the same clones by performing Southern blot analysis with a probe specific for the hygromycin resistance gene. As shown in Fig. 4C, the hygromycin resistance gene was only present in the specific site for the LdPTP1 gene. Taken together, these Southern blot analyses demonstrated that the active site region of the LdPTP1 gene was specifically and completely disrupted in the null mutant clones, thereby confirming their suitability for subsequent phenotypic analysis.
Initially, the phenotype of the LdPTP1 null mutants was compared with that of the wild type L. donovani using well established in vitro axenic culture protocols for promastigotes and amastigotes. As observed in Fig. 5A, the two individual homozygous LdPTP1 null mutant clones (Ld1PTP1 Ϫ/Ϫ , Ld2PTP1 Ϫ/Ϫ ), cultured under promastigote conditions (26°C, pH 7.2), proliferated at a slightly slower rate compared with wild type L. donovani (L.dWT). The heterozygous single knock-out clone (LdPTP ϩ/Ϫ ) proliferated at a similar rate to the null mutant clones. Under amastigote culture conditions (37°C, pH 5.5), the homozygous null mutants (Ld1PTP1 Ϫ/Ϫ and Ld2PTP1 Ϫ/Ϫ ) also demonstrated a slight reduction in proliferation compared with the single knock-out clone (LdPTP ϩ/Ϫ ) and the wild type culture (LdWT) (Fig. 5B). With respect to promastigote morphology, there did not appear to be any difference between the null mutant clones and the wild type cultures as shown in Fig. 5C.
The most stringent assay for L. donovani virulence is its ability to survive in the visceral organs in a mammalian host. We therefore compared the ability of the LdPTP1 null mutant clones and the parental wild type L. donovani to survive in the liver 4 weeks following injection in the tail vein of BALB/c mice. As shown in Fig. 6, the two LdPTP1 null mutant clones (Ld1PTP1 Ϫ/Ϫ and Ld2PTP1 Ϫ/Ϫ ) displayed significantly reduced virulence compared with the wild type parasites as indicated by both the number of amastigotes per nuclei in liver imprints (upper panel) and by calculating the Leishman-Donovan units determined by multiplying the level of infection by liver weight (lower panel). Both LdPTP1 null mutant clones displayed the same phenotype with an approximately 80 -90% reduction in virulence as determined by their ability to survive in the liver. The single knock-out clone LdPTP1 ϩ/Ϫ showed a clear intermediary reduction in virulence, consistent with reduced expression of LdPTP1. These results argue that the LdPTP1 gene plays a significant role in parasite survival in the mammalian host.
We next determined whether it was possible to restore the virulence of the null mutant clone (Ld1PTP1 Ϫ/Ϫ ) by comple-mentation with the wild type LdPTP1 gene. This was performed by introducing the LdPTP1 gene including its flanking regulatory sequences into a plasmid (Fig. 7A). The LdPTP1 containing plasmid (pSPYneoLdPTP1) was transfected into the null mutant clone (Ld1PTP1 Ϫ/Ϫ ), and transformants were selected in G418 and cloned by limiting dilution. For subsequent infections in mice, the wild type L. donovani promastigotes that had been in culture for several months and used to generate the null mutants characterized in Fig. 6 were no longer available. Nevertheless, the heterozygous single knock-out clone LdPTP1 ϩ/Ϫ also represented a relevant control because it had retained one wild type PTP1 allele (Fig. 4B), was more virulent than the null mutant clones (Fig. 6), and had undergone the same selection procedure as the null mutant clone, Ld1PTP1 Ϫ/Ϫ . Southern blot analysis with the entire LdPTP1 gene probe confirmed that the pSPYneoLdPTP1 plasmid had been successfully introduced into the null mutant clone (Fig.  7B, lanes 5 and 6).
We compared the virulence of the mutant Ld1PTP1 Ϫ/Ϫ containing the complementary pSPYneoLdPTP1 plasmid to the parental null mutant Ld1PTP1 Ϫ/Ϫ and the heterozygous mutant (LdPTP1 ϩ/Ϫ ), which retained one intact endogenous LdPTP1 allele. In addition, a newly established L. donovani culture was used for comparison. As shown in Fig. 7 (C and D), adding back the plasmid-derived LdPTP1 gene to the Ld1PTP1 Ϫ/Ϫ null mutant increased virulence in the liver (lanes 4) in comparison with the parental Ld1PTP1 Ϫ/Ϫ null mutant (lanes 3). The Ld1PTP1 Ϫ/Ϫ parasites containing the add back LdPTP1 was similar in virulence to the heterozygous LdPTP1 ϩ/Ϫ clone (Fig. 7, C and D, lanes 2), which retained one functional endogenous allele for the LdPTP1 gene. It is also noteworthy that the infection levels shown in Fig. 7 (C and D), for both the heterozygous LdPTP1 ϩ/Ϫ (lanes 2) and null mutant Ld1PTP1 Ϫ/Ϫ (lanes 3) were almost identical to the previous experiment shown in Fig. 6. As expected, the newly thawed culture of wild type L. donovani (Fig. 7, C and D, lanes 1) was considerably more virulent than the previous L. donovani culture that was used to generate the original LdPTP1 null mutants and which had been maintained in culture for several months (Fig. 6, A and B, lanes 1).
To further examine the ability of the added back LdPTP1 gene to restore virulence to the null mutant, we compared infection levels in the spleen. As shown in Fig. 7E, adding back the plasmid derived LdPTP1 gene to the Ld1PTP1 Ϫ/Ϫ null mutant resulted in increased spleen parasite levels (lane 4) compared with the parental Ld1PTP1 Ϫ/Ϫ null mutant (lane 3) and was similar in virulence to the heterozygous single allele knock-out clone (ϩ/Ϫ, lane 2). Taken together, these infection results in the mouse liver and spleen demonstrate that adding back the LdPTP1 gene on a plasmid partially complemented the virulent phenotype in the null mutant Ld1PTP1 Ϫ/Ϫ clone, thus confirming the importance for LdPTP1 in amastigote survival in the mammalian host.
The preceding observations provide an argument that inhibition of Leishmania PTP1 in amastigotes may have therapeutic potential and therefore may represent a drug target. With this in mind, in silico homology modeling was performed to compare the three-dimensional structures of the L. infantum PTP with the hPTP1B (Fig. 8A). This was performed to determine whether there were significant differences in the active sites between these enzymes that could be exploited to develop specific small molecule inhibitors. The structure of the complex between a tyrosine-phosphorylated peptide substrate (sequence etdy(Ptr)rkggkgll) and human PTP (HEPTP, Protein Data Bank code 1G1G), was used to model the position of the peptide ligand into the L. infantum homology via superposition. Enzyme residues within eight angstroms of the ligand were used to perform the superposition. The hydrogen bonding pattern seen in the substrate peptide-human HEPTP complex between the phosphate group and the backbone of the protein appears to be very well conserved in the L. infantum PTP1 homology model, and a high level of structural conservation between the human and Leishmania enzymes is concentrated in and around the active site (Fig. 8, A and B). Thus, it seems likely that the active sites of the two proteins are similar, although there are significant differences in residues further away from the active site, including residues 32-51 (Fig. 8, A and  B). Two differences close to the active site are worth mentioning. First, although the L. infantum enzyme has a proline residue at position 67, the human enzyme has an arginine. Second, residue 205 in the L. infantum sequence is a glutamine, whereas in all human tyrosine phosphatases there is a phenylalanine residue at this position. These differences affect the hydrophobicity of the substrate-binding pocket and thus, potentially, the preferred substrate. This altered specificity may provide scope for designing an inhibitor that binds more tightly to the L. infantum enzyme than to the human homologs.

DISCUSSION
Among the most important signaling mechanisms in eukaryotic cells are those that involve protein-tyrosine phosphorylation and dephosphorylation (4). It was therefore important to identify a prototype PTP-related gene in the Leishmania genome, determine its role in the parasite life cycle, and determine its potential as a drug target. We were particularly interested in identifying a Leishmania PTP1B-type gene because we had previously observed that overexpression of the human PTP1B increased the virulence of L. donovani and appeared to mediate certain aspects of differentiation toward the amastigote stage (7). The present study confirmed that Leishmania parasites do have a functional PTP1B-like gene, which we termed LPTP1. Deletion of the L. donovani PTP1 gene resulted in attenuated amastigotes in BALB/c mice but did not significantly impair promas-tigote survival in culture. These results argue that Leishmania PTP1 plays a significant role in biochemical pathways associated with amastigote survival in the mammalian host. A, the homology model is shown as a cartoon representation, highlighting ␣-helices, ␤-strands, and loops. The color scheme used is according to the level of conservation at each residue position between the L. infantum sequence and the human PTP1B sequence with red representing identical residues, orange representing conserved substitutions, green representing semi-conserved substitutions, turquoise representing positional conservation, and blue representing no conservation. Those stretches of sequence outside the active site (residues 32-51 and 319 -493) that could not be modeled, and therefore were not included, are represented schematically by ovals, and their size is indicative of the number of residues missing from the model. The oval representing the residues 32-51 that are present in the Leishmania sequence but not the human sequence is colored blue, and the oval that represents residues 319 -493 that are present in both enzyme sequences but were not present in the structure of the human enzyme is colored mauve. B, using the identical orientation used in A, a molecular surface representation is shown, which is colored using the same scheme used in A. The active site is shown occupied by the monophosphorylated peptide substrate from the human PTP1B-peptide complex (Protein Data Bank code 1G1G), which is represented by a stick model colored by atom type as follows: white, carbon; blue, nitrogen; red, oxygen; yellow, phosphorous. Also indicated are residues proline 67 and glutamine 205, which may affect substrate preference.