Chicken Anemia Virus VP2 Is a Novel Dual Specificity Protein Phosphatase*

The function of viral protein 2 (VP2) of the immunosuppressive circovirus chicken anemia virus (CAV) has not yet been established. We show that the CAV VP2 amino acid sequence has some similarity to a number of eukaryotic, receptor, protein-tyrosine phosphatase (PTPase) α proteins as well as to a cluster of human TT viruses within the Sanban group. To investigate if CAV VP2 functions as a PTPase, purified glutathione S-transferase (GST)-VP2 fusion protein was assayed for PTPase activity using the generalized peptide substrates ENDpYINASL and DADEpYLIPQQG (where pY represents phosphotyrosine), with free phosphate detected using the malachite green colorimetric assay. CAV GST-VP2 was shown to catalyze dephosphorylation of both substrates. CAV GST-VP2 PTPase activity for the ENDpYINASL substrate had a V max of 14,925 units/mg·min and a K m of 18.88 μm. Optimal activity was observed between pH 6 and 7, and activity was specifically inhibited by 0.01 mmorthovanadate. We also show that the ORF2 sequence of the CAV-related human virus TT-like minivirus (TLMV) possessed PTPase activity and steady state kinetics equivalent to CAV GST-VP2 when expressed as a GST fusion protein. To establish whether these viral proteins were dual specificity protein phosphatases, the CAV GST-VP2 and TLMV GST-ORF2 fusion proteins were also assayed for serine/threonine phosphatase (S/T PPase) activity using the generalized peptide substrate RRApTVA, with free phosphate detected using the malachite green colorimetric assay. Both CAV GST-VP2 and TLMV GST-ORF2 fusion proteins possessed S/T PPase activity, which was specifically inhibited by 50 mm sodium fluoride. CAV GST-VP2 exhibited S/T PPase activity with aV max of 28,600 units/mg·min and aK m of 76 μm. Mutagenesis of residue Cys95 to serine in CAV GST-VP2 abrogated both PTPase and S/T PPase activity, identifying it as the catalytic cysteine within the proposed signature motif. These studies thus show that the circoviruses CAV and TLMV encode dual specificity protein phosphatases (DSP) with an unusual signature motif that may play a role in intracellular signaling during viral replication. This is the first DSP gene to be identified in a small viral genome.

Chicken anemia virus (CAV), 1 a member of the family Circoviridae, causes severe immunosuppression, anemia, and thrombocytopenia (1). The Circoviridae include a number of very small plant and animal viruses that are characterized by the possession of a single-stranded, negative sense, circular DNA genome. Since there is minimal similarity between the genomic sequence and organization of CAV and the other characterized animal circoviruses, psittacine beak and feather disease virus and pigeon circovirus and porcine circoviruses 1 and 2, CAV has been reclassified within the floating genus Gyrovirus. TT viruses (TTV) have recently been identified in human hosts and other species as a heterogeneous cluster of singlestranded, negative sense, circular DNA viruses but have not yet been cultured in vitro. Sequence analysis of this group of viruses has demonstrated greatest overall homology to CAV, and Takahashi et al. have recently proposed the classification of the TTV, Sanban viruses, Yonban viruses, TT-like minivirus (TLMV), and CAV as the Paracircoviridae. However the phylogeny of this group of viruses remains an area of active revision (2). The highest sequence similarity between CAV and TTV is seen in the noncoding region and between ORF2 of TTV and VP2 of CAV. The high level of sequence conservation between CAV and TTV within these coding regions suggests that VP2 may play a critical role in viral replication.
CAV encodes only three proteins, with overlapping open reading frames in three frames. ORF3 encodes the 45-52-kDa capsid protein, VP1; ORF2 encodes the 11-13-kDa VP3 that has been shown to have apoptotic activity in transformed cell lines (1); and ORF1 encodes a 28-kDa nonstructural protein, VP2, of unknown function. VP2 is expressed at barely detectable levels during infection (3), and the low level of expression is consistent with a nonstructural, regulatory protein involved in viral replication and infection.
Preliminary comparisons of the CAV VP2 sequence with sequences available in the GenBank TM data base identified similarity to a number of eukaryotic receptor protein-tyrosine phosphatases (R-PTPases), particularly the human placental, rat, mouse, and chicken R-PTPase ␣ precursors. Reversible protein phosphorylation plays a crucial role in the regulation of cellular processes such as metabolism, gene regulation, cell cycle control, cytoskeletal organization, and cell adhesion. The PTPase family is highly diverse and includes the eukaryotic receptor-like transmembrane proteins and soluble cytosolic proteins, as well as bacterial PTPases, such as the YopH PTPase from pathogenic Yersinia species and a viral PTPase, VH1, found in Vaccinia virus (4).
The aim of this study was to investigate CAV VP2 and the homologous ORF2 of TT virus as novel viral PTPases and as serine/threonine protein phosphatases (S/T PPases). This investigation primarily stemmed from the finding of some sequence similarity between CAV VP2 and a number of PTPases. In order to establish the catalytic function of VP2, the protein was expressed as a fusion with glutathione S-transferase (GST), and the activity of the fusion protein was determined using universal oligopeptide PTPase substrates. The classification of the VP2 PTPase was assessed by examination of the kinetic parameters, the optimal pH for activity, and the sensitivity of the enzyme to orthovanadate. S/T PPase activity was similarly assessed using a universal oligopeptide substrate. Within the proposed signature motif of VP2, the cysteine 95 residue was mutagenized to serine, and protein phosphatase activity was assessed.

EXPERIMENTAL PROCEDURES
Sequence Analysis-Protein sequences with homology to CAV VP2 were identified by searches of the GenBank TM data base using BLASTX (Basic Local Alignment Search Tool) via the NCBI interface (available on the World Wide Web at www.ncbi.nlm.nih.gov). Sequences identified by this method were then aligned to the CAV VP2 sequence using Macboxshade version 3.2 (R. Fusch and Glaxo Wellcome). A separate alignment was produced for both the R-PTPases and the TTV sequences.
Molecular Cloning of CAV Viral Protein 2-The Australian isolate of CAV, CAU269/7, was used in all experiments (GenBank TM accession number AF227982) (5). CAV ORF1 (encoding VP2) (GenBank TM accession number AAF34787.1) was amplified by PCR from the doublestranded replicative form of the CAV genome. Cellular DNA was purified from MDCC-MSB1 cells 48 h after infection with CAV using proteinase K and SDS lysis and phenol/chloroform extraction (6). Oligonucleotide forward primer CAV.1 (5Ј-CGGTCCGGATCCATGCACG-GAAACGGCGGACAAC-3Ј and reverse primer CAV.2 (5Ј-GGTTTGGA-ATTCTCACACTATACGTACCGGGGC-3Ј were synthesized to incorporate BamHI and EcoRI restriction endonuclease cleavage sites within their respective 5Ј-ends. A 100-l reaction mixture was prepared containing 300 M each dATP, dCTP, dGTP, and dTTP, 2 mM MgCl 2 , 0.2 M each primer, 10 l of 10ϫ TaqDNA polymerase buffer, 2 units of TaqDNA polymerase (Promega, Madison, WI), and 2 l of template DNA. The PCR was incubated at 95°C for 2 min, followed by 40 cycles of 96°C for 40 s, 60°C for 40 s, and 72°C for 40 s, with a final incubation at 72°C for 5 min. The PCR products were analyzed by agarose (1%) gel electrophoresis, and a band consistent with the expected size of 677 bp was excised and purified using a Qiaex II (Qiagen, Basel, Switzerland) gel extraction kit according to the manufacturer's instructions. The purified product was digested with BamHI and EcoRI and ligated to appropriately digested pGEX-4T-2 (Promega). E. coli strain DH5␣ was transformed with the ligated plasmid and cultured at 37°C on Luria-Bertani agar (LA) containing ampicillin at 50 g/ml. The cloned DNA was sequenced using a Taq Dye Deoxy Terminator Cycle Sequencing kit (PerkinElmer Life Sciences) using commercial sequencing primers (Promega) specific for pGEX-4T-2.
Molecular Cloning of TLMV Viral Protein 2-Purified nested PCR product extending from primer M-1360 to primer M-1363 (258 -886 bp) of TLMV strain CBD231 was kindly supplied by Dr. Shunji Mishiro (GenBank TM accession number AB026930) (2). TLMV ORF2 was amplified by PCR from the M-1360 to M-1363 template. Oligonucleotide forward primer TLMV.1 (5Ј-TTGGATCCATGAGCAGCTTTCTAACAC-CATC-3Ј and reverse primer TLMV.2 (5Ј-GGCGAATTCTTACCCATC-GTCTTCTTCGAAATC-3Ј were synthesized to incorporate BamHI and EcoRI cleavage sites within their respective 5Ј-ends. A 50-l reaction mixture was prepared containing 300 M each dATP, dCTP, dGTP, and dTTP, 2 mM MgCl 2 , 0.2 M each primer, 5 l of 10ϫ TaqDNA polymerase buffer, 1 unit of TaqDNA polymerase (Promega), and 1 l of template DNA. The PCR was incubated at 96°C for 2 min, followed by 40 cycles of 96°C for 40 s, 56°C for 40 s, and then 72°C for 40 s and a final incubation at 72°C for 5 min. Purification, digestion, and cloning of the 295-bp PCR product into the pGEM-T plasmid vector (Promega) were performed as described above for the CAV VP2 gene. The insert was subsequently subcloned into the pGEX-4T-2 plasmid vector, and the sequence of the insert was verified.
Mutagenesis of Cys 95 Residue in CAV-VP2-Overlap extension PCR (7) was used to introduce a mutation into the CAV VP2 gene sequence to change the cysteine residue at position 95 to serine (C95S). The PCR template was pCAU269/7 in the plasmid vector pGEX-4Z (Promega) (5). The oligonucleotide pair, positive sense (5Ј-cgctaagatcAgcaactgcg-3Ј) and negative sense (5Ј-cgcagttgcTgatcttagcgtg-3Ј), was synthesized to incorporate nucleotide substitutions encoding the amino acid alterations (shown in boldface type). Template DNA was prepared from E. coli DH5␣ containing pCAU269/7 cultured at 37°C in LB containing 50 g of ampicillin/ml. Plasmid was purified using a Qiagen Midi kit according to the manufacturer's instructions. The PCR was carried out in two stages. The first stage consisted of a set of two PCRs: one from the upstream flanking primer CAV.3 (5Ј-CTATCGAATTCCGAGTGGT-TACTAT-3Ј) to the negative sense mutagenesis primer and one from the downstream flanking primer CAV.4 (5Ј-TGCTCACGTATGTCAG-GTTC-3Ј) to the positive sense mutagenesis primer. In the second stage of mutagenesis, PCR products from the first pair of reactions acted as template for a PCR using only the flanking primers. The PCR product generated was bounded by the flanking sequences and incorporated in both strands the mutations introduced into the template in the first stage of PCR.
In the first stage, a 100-l reaction mixture was prepared containing 300 M each dATP, dCTP, dGTP, and dTTP, 2 mM MgSO 4 , 0.2 M of each primer, 10 l of 10ϫ Platinum Pfu TaqDNA polymerase buffer, 2 units of Platinum Pfu TaqDNA polymerase (Promega), and 1 l of template DNA. The reaction was incubated at 95°C for 2 min, followed by 30 cycles of 95°C for 40 s, 60°C for 60 s, and 68°C for 40 s, with a final incubation at 68°C for 5 min. First stage template was removed by digestion with DpnI (Roche Molecular Biochemicals). The PCR products were analyzed by agarose (1%) gel electrophoresis, and the bands corresponding to the stage one products were excised and purified using a Qiaex II gel extraction kit according to the manufacturer's instructions.
For the second stage PCR, a 100-l reaction mixture was prepared containing 300 M each dATP, dCTP, dGTP, and dTTP, 2 mM MgSO 4 , 0.2 M each primer, 10 l of 10ϫ Platinum Pfu TaqDNA polymerase buffer, 2 units of Platinum Pfu TaqDNA polymerase, and 1 l of each template DNA. The PCR was incubated at 95°C for 2 min, followed by one cycle of 95°C for 40 s, 57°C for 90 s, and 68°C for 40 s and then 15 cycles of 95°C for 40 s, 57°C for 60 s, and 68°C for 40 s, with a final incubation at 68°C for 5 min.
The PCR products were digested with StuI (New England Biolabs, Beverly, MA) and BsmI (New England Biolabs) and analyzed by agarose (1%) gel electrophoresis. A band of the expected size of 357 bp was excised and purified using a Qiaex II gel extraction kit according to the manufacturer's instructions. The VP2-pGEX-4T-2 clone was similarly digested with StuI and BsmI to remove the region of 357 bp to be replaced with the mutant sequence, and a band of the expected size of 5260 bp was purified from a 1% agarose gel. The digested PCR product was then ligated to the digested VP2-pGEX-4T-2 backbone. E. coli DH5␣ was transformed by electroporation with the ligated plasmid and cultured at 37°C on LA containing ampicillin at 50 g/ml. Plasmid was purified from selected clones using a Qiagen Midi kit according to the manufacturer's instructions. Clones were screened for the presence of insert by PCR using the forward primer CAV.2 and a reverse primer overlapping the VP2 pGEX multicloning site (5Ј-ctctagaggatcctcacactatacgtaccggggc-3Ј). The cloned DNA was sequenced using a Taq Dye Deoxy Terminator Cycle Sequencing kit using the above primers.
Protein Expression and Purification-CAV VP2 was produced as a C-terminal fusion with glutathione S-transferase. Briefly, 1-liter cultures of E. coli DH5␣ possessing the CAV VP2 pGEX-4T-2 construct were cultured in LB containing ampicillin at 50 g/ml. Expression was induced by adding isopropyl-␤-D-thiogalactopyranoside to a final concentration of 1 mM when the culture reached an optical density of 0.6 at 600 nm, and the culture was incubated an additional 1 h prior to harvest. Bacteria were recovered by centrifugation at 6000 ϫ g for 30 min, and the pellets were washed twice in phosphate-buffered saline. The cells were resuspended in 25 ml of phosphate-buffered saline containing 0.3 M EDTA, 200 mg of lysozyme, and 100 g of phenylmethylsulfonyl fluoride (Sigma-Aldrich)/ml and lysed by 10-s bursts of sonication at low frequency. The lysate was solubilized in 0.1% Triton X-100 and incubated a further 10 min at 4°C, and the cellular debris was removed by centrifugation at 10,000 ϫ g for 30 min. The fusion protein was affinity-purified using glutathione-Sepharose resin (Promega) following the manufacturer's protocol. The eluate was extensively dialyzed against a buffer containing 137 mM NaCl, 2.7 mM KCl, and 25 mM Tris-HCl, pH 7.4 (TBS). CAV GST-VP2 mutagenized to contain serine at position 95 and negative control glutathione S-transferase were purified from E. coli DH5␣ transformed with either the C95S mu-tagenized VP2 pGEX-4T-2 clone or pGEX-4T-2, respectively, following the same method as was used to purify the CAV GST-VP2 fusion. The GST-TLMV ORF2 fusion protein was purified from E. coli DH5␣ transformed with the TLMV ORF2 pGEX 4T-2 clone following the method used to purify the CAV GST-VP2 fusion, except that protein expression was induced for only 30 min.
Purified proteins were separated by electrophoresis in 12.5% SDSpolyacrylamide gels and stained with Coomassie Brilliant Blue (8). Proteins were electrotransferred to a polyvinylidene difluoride membrane (Immobilon; Millipore Corp., Bedford, MA). Western blots of CAV GST-VP2 and GST were probed with a rabbit polyclonal antiserum raised against GST, at a dilution of 1:500, followed by a secondary swine anti-rabbit horseradish peroxidase conjugate (Dako Australia, Botany, Australia) diluted 1:1000 and developed with 3,3Ј-diaminobenzidine substrate (Sigma Fast DAB) (Sigma-Aldrich) according to the manufacturer's instructions. A second Western blot was probed with pooled chicken serum from flocks positive to CAV by enzyme-linked immunosorbent assay (IDEXX, Westbrook, ME), followed by rabbit antichicken-horseradish peroxidase conjugate at a dilution of 1:500, and developed with 3,3Ј-diaminobenzidine substrate. Protein concentration was quantified using the Bradford assay (Bio-Rad) with a bovine serum albumin (Sigma-Aldrich) standard.
Protein-tyrosine Phosphatase Assay-Protein-tyrosine phosphatase activity was assayed using the protein-tyrosine phosphatase assay system (Promega) according to the manufacturer's instructions. Briefly, the assays were performed in 50-l volumes in a microtiter plate using the generalized PTPase substrates ENDpYINASL and DADEpYLIPQQG (where pY represents phosphotyrosine) and assay buffer containing 25 mM Tris-HCl (pH 7.4), 50 mM NaCl, 2 mM EDTA, 5 mM dithiothreitol (New England Biolabs), 0.01% Brij 35 (New England Biolabs), and 1 mg of bovine serum albumin (New England Biolabs)/ml. The reactions were started by the addition of 5 g of either CAV GST-VP2, GST, CAV GST-VP2 containing the C95S mutation, or TLMV GST-ORF2 or 2 units of the positive control T cell proteintyrosine phosphatase (TC-PTP; New England Biolabs) in assay buffer. Control reactions were assayed with either ENDpYINASL substrate alone, DADEpYLIPQQG substrate alone, CAV GST-VP2 without substrate, CAV GST-VP2 containing the C95S mutation without substrate, TLMV GST-ORF2 without substrate, TC-PTP without substrate, or assay buffer with neither enzyme nor substrate. A phosphate standard curve was derived using a supplied phosphate standard (Promega). The reactions were terminated by the addition of the malachite green detection reagent according to the manufacturer's instructions (Promega).
PTPase Kinetic and Inhibition Studies-The generalized proteintyrosine phosphatase substrate described by Daum et al. (9) was used in assays of enzyme kinetics. The phosphopeptide sequence was H-Glu-Asn-Asp-Tyr(PO 3 H 2 )-Ile-Asn-Ala-Ser-Leu-OH. Briefly, the nonapeptide was assembled manually in the solid phase using Fmoc chemistry (10). All chemicals for use in peptide synthesis were of analytical grade. Fmoc-protected amino acid residues (Auspep, Melbourne, Australia) were used for synthesis. The residues used were Fmoc-L-Leu-OH, Fmoc-L-Ser(t-butyl)-OH, Fmoc-L-Ala-OH, Fmoc-L-Asn(trityl)-OH, Fmoc-L-Ile- The support resin PAC-PEG-PS (capacity 0.18 mmol/g) (Perspective Biosystems, Herts, UK) was used for the synthesis. The amino acids were activated by incubation with equimolar quantities of O-benzotriazole-N,N,NЈ,NЈ-tetramethyluronium hexafluorophosphate (Auspep) and 1-hydroxybenzotriazole (Auspep) and two equivalents of diisopropylethylamine (Auspep). The coupling reaction was carried out for 60 min and followed by the trinitrobenzene sulfonic acid test described by Hancock et al. (11). The Fmoc groups were removed after each coupling reaction by washing in 2.5% 1,8-diazabicyclo- [5.4.0]undec-7-ene. For the coupling of residues 4 -9, each cycle was repeated twice. The side chain protective groups were removed, and the peptide was cleaved from the resin by treatment with 88% trifluoroacetic acid, 5% phenol, and 2% tri-isopropylsilane (Sigma-Aldrich) in water. The crude peptide was precipitated in cold diethyl ether prior to purification by reverse-phase high performance liquid chromatography over a Vydac C4 semipreparative column in 0.1% trifluoroacetic acid and eluted with a 2%/min gradient of acetonitrile. The identity of the peptide was confirmed by mass spectroscopy.
The method for the protein-tyrosine phosphatase kinetic analyses was adapted from the method of Tonks et al. (12). The following reaction conditions were used for all assays performed unless otherwise stated. An assay buffer (AB) was prepared with 50 mM Tris-HCl (pH 7 at 25°C), 1 mM EDTA, 50 mM 2-mercaptoethanol, and 1% (w/v) bovine serum albumin. A second buffer (TB) was prepared with 50 mM Tris-HCl (pH 7 at 25°C) and 0.01% (w/v) Brij 35 (Sigma-Aldrich). All reactions were carried out in a total volume of 200 l in a microtiter plate. Fifteen nanomoles of synthesized phosphopeptide substrate at 1 mM in AB buffer was added to each triplicate reaction mixture composed of a 1:1 mixture of AB and TB. The reactions were started by the addition of 9 g of either CAV GST-VP2 or GST in TBS. Control reactions were assayed with substrate alone or with CAV GST-VP2 or GST without substrate. Reactions were incubated with shaking at room temperature for 0, 1, 2, 3, 4, 5, or 10 min and terminated by the addition of malachite green reagent. All assays were repeated on at least three occasions, and the average activity was plotted for each time point. Activity was adjusted by a factor of 0.52 to account for the contribution to mass of the 24-kDa GST fusion tag and expressed as nmol of catalyzed substrate/g of enzyme. The GST-TLMV ORF2 protein was assayed for end point kinetics using the reaction conditions described for CAV GST-VP2. Reactions were repeated four times.
The release of free phosphate into solution was detected by the malachite green colorimetric assay (13). Briefly, stock malachite green solution was made by the slow addition of 60 ml of concentrated sulfuric acid to 300 ml of water followed by cooling to room temperature, and then 0.44 g of malachite green (Fisher) was added. Immediately before use, the colorimetric reagent was made from 10 ml of stock malachite green, 3% (w/v) (NH 4 ) 3 MoO 3 (Sigma-Aldrich) and 0.15% Tween 20 (Sigma-Aldrich).
Fifty microliters of the colorimetric reagent was added to the 200-l reaction volumes in each well of a microtiter plate and allowed to equilibrate for 20 min at room temperature. Absorbance was read at 620 nm, and phosphate release was calibrated against a phosphate standard curve.
For each phosphatase assay, a phosphate standard curve was prepared for phosphate at 0, 2.5, 5, 7.5, 10, 15, 20, 25, 30, 40, and 45 nmol. Phosphate dilutions were prepared in triplicate, in 200-l volumes in a microtiter plate, in a 1:1 ratio of AB and TB buffer. For each reaction, 50 l of the colorimetric reagent was added and allowed to equilibrate for 20 min at room temperature. Absorbance was then measured at 620 nm.
PTPase assays were performed with 0, 2.5, 5, 7.5, 10, 15, 20, 25, 30, or 40 nmol of substrate. Reactions were incubated for 1 min, with all other reaction conditions as described above. For each substrate, concentration activity was measured in at least six replicate reactions, and the S.E. activity was calculated for each concentration. V max and K m estimates were derived by linear regression analysis from a double reciprocal plot, and the S.E. and p value were calculated for the constant 1/V max and the coefficient K m /V max . Stock 1 mM Na 3 VO 3 was made in distilled water and adjusted to pH 10 with sulfuric acid. Once dissolved, a 0.3 mM solution of Na 3 VO 3 was made in AB and TB buffer at a ratio of 1:1:1 and adjusted to pH 7. PTPase assays were performed with 10 nmol of substrate and 9 g of CAV GST-VP2 and were performed in triplicate for each concentration of inhibitor. Inhibition studies were conducted at 0.1, 0.01, and 0.001 mM of Na 3 VO 3 . All other reaction conditions were as described above.
Enzyme pH Optimum-Triplicate reactions were set up at pH 4, 5, 6, 7, 8, and 9. Assays were performed with 10 nmol of substrate and 9 g of CAV GST-VP2 with all other reaction conditions as described above.
Prior to the addition of the malachite green reagent, the pH was neutralized to pH 7 with either sulfuric acid or sodium hydroxide.
Reactions were incubated for 1 min, with all other reaction conditions as described above. V max and K m estimates were derived by linear regression analysis from a double reciprocal plot.
The S/T PPase inhibitor NaF was made at 250 mM in distilled water, and S/T PPase assays were performed in triplicate with 50 mM NaF, 200 M RRApTVA, and 5 g of either CAV GST-VP2 or TLMV GST-ORF2, with all other reaction conditions as described above.

CAV VP2 Possesses a PTPase-like Domain-A data base
search for protein sequences with similarity to CAV VP2 identified some sequence similarity to R-PTPase ␣ proteins of human, rat, mouse, and chicken origin. The region of sequence similarity extended over residues 48 -187 of VP2, and the same region of VP2 was similar in all cases. CAV VP2 similarity was highest to the WPD loop flanking the P-loop in all R-PTPase homologues (Fig. 1). The P-loop contains the catalytic site and signature motif. The P and WPD loops are segments of random coil lying within a catalytic cleft formed by five ␤ sheets (14). Sequence similarity over the region of overlap between CAV VP2 and all the R-PTPase ␣ sequences over this restricted region of 139 residues was 55%, and identity was 30 -32%. From the studies of R-PTPases, the consensus signature motif has been defined as (I/V)HCXAGXGR(S/T). The cysteine residue is critical in binding the phosphate, and the arginine coordinates the phosphotyrosine in the catalytic cleft. A minimal signature motif has been defined for the PTPase superfamily as CXXXXXR. This definition is based on the subgroup of low M r , dual specificity PTPases that lack overall sequence similarity to the conserved PTPase domain but contain this minimal signature motif. The proposed catalytic motif in CAV VP2 is ICNCGQFRK, encoded by amino acid residues 94 -102. The alignment in Fig. 1 illustrates the overall sequence similarity between CAV VP2 and the catalytic subunit of the R-PTPases over an extended region. However, the proposed CAV VP2 signature motif diverges from the highly conserved consensus motif seen in R-PTPases and contains a novel configuration, including a second cysteine within the motif and a second basic residue (lysine) adjacent to the arginine of the motif.
A second cluster of sequences, comprising the Sanban group of TTV, was identified as related to CAV VP2. In all Sanban viral sequences, the region of similarity extended from residue 46 to 118 of the amino acid sequence encoded by the CAV VP2 gene. The related region in CAV VP2 was the same as that for the R-PTPases, but different residues were shared (Fig. 1b). Similarity between CAV VP2 and the Sanban viral sequences was 53-55%, and identity was 22-26%. As identified previously, TTV, Sanban, Yonban, and TLMV viruses have in common with CAV the sequence WX 7 HX 3 CXCX 5 H in ORF2 (2). This sequence corresponds to the 5Ј-end of the PTPase signature motif we have proposed. However, as is evident from the alignment in Fig. 1b, the similarity between the Sanban isolates and CAV VP2 is more extensive.
There was no significant similarity identified by alignments of CAV VP2 or TTV VP2 sequences with known S/T PPases.

FIG. 1. Sequence alignments of CAV VP2 with similar sequences. R-PTPase homologues (a) and TT virus sequences (b)
were aligned to the CAV VP2 amino acid sequence using ECLUSTALW (WebANGIS; available on the World Wide Web at www.angis.org.au) and displayed graphically using Macboxshade version 3.5 (R. Fuchs and Glaxo Wellcome). Residues identical to the CAV sequence are illustrated by capital letters, and lowercase type is used for similar residues. Nonhomologous residues are represented by dashes, and spaces are used for residues present in one sequence but not in the aligned sequence. a, row 1, CAU269/7 VP2 (AF227982); row 2, chicken protein-tyrosine phosphatase ␣ (Z32749); row 3, human R-PTPase ␣ (PP18433); row 4, rat R-PTPase ␣ (Q03348); row 5, mouse R-PTPase ␣ (P18052); row 6, human R-PTPase ␣ (17011300A); row 7, human placental protein-tyrosine phosphatase (CAA38065). b, row 1, CAU269/7 VP2 (AF227982), residues 48 -187; row 2, TTV Sanban sequence AB024366; row 3, TTV Sanban sequence AB024365; row 4, TTV Sanban sequence AB024355; row 5, TTV Sanban sequence AB024356. The CAV VP2 signature motif is illustrated as boxed text above the relevant sequence. tigate the possibility that CAV VP2 functioned as a PTPase, a recombinant form of the protein was synthesized and tested in vitro for PTPase activity. For this, CAV ORF1 (encoding VP2) was amplified from the CAU269/7 Australian isolate of CAV. This isolate is equivalent in pathogenicity and infectivity to other described isolates of CAV (5). The PCR product was cloned into the pGEX 4T-2 vector, and CAV VP2 was produced as a recombinant fusion protein with GST, and the fusion protein was purified by glutathione affinity chromatography. A band of 58 kDa corresponding to the CAV GST-VP2 fusion protein was identified by SDS-PAGE from affinity-purified eluate (Fig. 2). The protein band reacted specifically with antiserum raised against GST and also with pooled serum from chickens seropositive for CAV. Dialyzed CAV GST-VP2 was readily soluble in Tris-buffered saline and was used directly in PTPase assays. Based on the sequence similarity between CAV GST-VP2 and TLMV GST-ORF2, the TLMV protein was also expressed in order to investigate its PTPase activity.
PTPase activity was demonstrated for the reaction of 2 units of the positive control TC-PTP with both ENDpYINASL and DADEpYLIPQQG, and no activity was detected for reactions containing either ENDpYINASL substrate alone, DADEpYLIPQQG substrate alone, CAV GST-VP2 without substrate, CAV GST-VP2 containing the C95S mutation without substrate, TLMV GST-ORF2 without substrate, TC-PTP without substrate, or assay buffer with neither enzyme nor substrate. Both CAV GST-VP2 and TLMV GST-ORF2 were shown to have PTPase activity using both ENDpYINASL and DADEpYLIPQQG. Steady state activities for the reaction of 5 g of CAV GST-VP2 with ENDpYINASL and DADEpYLIPQQG were 100 and 208%, respectively, of those seen for 2 units of TC-PTP. Steady state activities for the reaction of 5 g of TLMV GST-ORF2 with ENDpYINASL and DADEpYLIPQQG were 37 and 0.04%, respectively, of those seen for 2 units of TC-PTP.
PTPase Kinetic and Inhibition Studies-Peptide substrate was synthesized on a solid support using standard Fmoc chemistry. Cyclical additions of amino acids following the phosphotyrosine residue were duplicated to counter potential steric hindrance to coupling by the large phosphate group. A single peak consistent with a pure phosphate nonapeptide was seen on analytical reverse-phase high performance liquid chroma-tography, and the formula weight was confirmed as 1116.3 by mass spectroscopy.
A standard curve of absorbance at 620 nm as a function of phosphate concentration was established for the assay conditions. The sensitivity of the malachite green colorimetric detection was 2.5 nmol of phosphate, and the relationship between log [P i ] and absorbance at 620 nm was linear over the range of 0 -45 nmol of phosphate. Precipitation of phosphomolybdate was seen for concentrations of phosphate greater than 45 nmol.
The rate of CAV GST-VP2 activity was determined from a time course study (Fig. 3a). V 0 was measured at 1 min in all subsequent reactions, since this time point was within the linear region of the activity curve in the time course study. CAV GST-VP2 V 0 relative to control GST protein is shown in Fig. 3b. PTPase activity of CAV GST-VP2 displayed Michaelis-Menten kinetics, and the relationship between V 0 and [S] was given by 1/[V 0 ] ϭ (1.265).1/[S] ϩ 0.067. The Lineweaver-Burk double reciprocal plot for CAV GST-VP2 is shown in Fig. 3c. From the Lineweaver-Burk plot, 1/V max was found by linear regression to be 0.067 Ϯ 0.0137 (p Ͻ 0.0001), and K m /V max was found to be 1.265 Ϯ 0.1085 (p Ͻ 0.0001). Based on these results, V max was estimated to be 14,925 units/mg⅐min, and K m was estimated to be 18.88 M. All assays were repeated three times using two different preparations of CAV GST-VP2, and reactions at each substrate concentration were repeated at least four times.
A number of other descriptive features have been defined for the PTPase family (15). As a family, they are resistant to inhibition of activity by EDTA and display an optimal activity within the range of pH 5.5-7. Protein-tyrosine phosphatase activity was measured for CAV at varying reaction pH. The optimal CAV GST-VP2 PTPase activity was found to be in the range of pH 6 -7 (Table I). As for other PTPases, EDTA was required in the reaction buffer for activity. The low M r dual specificity PTPases are characterized by activity extending over a pH range of 4 -7, whereas the high M r PTPases have optimal activity only at neutral pH (16). CAV VP2 has a low M r , but its activity was intermediate between those of the high and low M r PTPases, and the pH range for optimal activity was similar to that seen with high M r PTPases such as the R-PTPases. The CAV PTPase therefore has a combination of biochemical properties that are unique within the PTPase family.
PTPases catalyze the removal of phosphate from phosphotyrosine via a cysteinyl-phosphate intermediate formed with the active cysteine in the catalytic cleft. The mechanism of catalysis is unique to the PTPase family, as is inhibition of activity by low concentrations of orthovanadate (15). Orthovanadate is a structural analogue of phosphate and as such competitively inhibits the cysteinyl-phosphate intermediate. The inhibitory effect of sodium orthovanadate on CAV GST-VP2 PTPase activity is shown in Table II. Orthovanadate concentrations of 0.001, 0.01, and 0.1 mM completely inhibited PTPase activity by GST-VP2. Inhibition by orthovanadate concentrations as low as 0.001 mM indicated a catalytic mechanism in common with the PTPase family.
CAV VP2 and TLMV ORF2 Have S/T PPase Activity-S/T PPase activity was demonstrated for the reaction of 1 unit of the positive control PPase-2B with the substrate RRApTVA, and no activity was detected for reactions containing either RRApTVA substrate alone, CAV GST-VP2 without substrate, CAV GST-VP2 containing the C95S mutation without substrate, TLMV GST-ORF2 without substrate, PPase-2B without substrate, or assay buffer with neither enzyme nor substrate. Both CAV GST-VP2 and TLMV GST-ORF2 were shown to have S/T PPase activity using the generalized substrate RRApTVA. Steady state activity for the reaction of 5 g of CAV GST-VP2 with RRApTVA was 97% of that seen for 1 unit of PPase-2B. Steady state activity for the reaction of 5 g of TLMV GST-ORF2 with RRApTVA was 49% of that seen for 1 unit of PPase-2B.
V 0 for S/T PPase activity of CAV GST-VP2 was measured

DISCUSSION
This work is the first to define a function for CAV VP2 and TLMV ORF2 and has established that these proteins are novel dual specificity phosphatases. PTPases are defined by their capacity to remove phosphate specifically from phosphotyrosine residues in phosphoprotein substrates. The kinetic studies clearly demonstrated that CAV VP2 has protein-tyrosine phosphatase activity. Under the assay conditions described, CAV VP2 has a V max of 14,925 units/mg⅐min and a K m of 18.88 M. The V max of PTPases ranges from 1,000 units/mg⅐min at the lower end for CD45 from human spleen to 20,000 units/ mg⅐min at the upper end for PTP1B. In general, PTPases of low molecular mass tend to have higher specific activity than the high molecular mass PTPases (12). Although CAV VP2 is a protein of low molecular mass, its PTPase activity is intermediate between those characteristic of high and low molecular mass PTPases. PTPases have been found to vary in their specific activity for different complex protein substrates, but the generalized peptide substrates ENDpYINASL and DADEpYLIPQQG used in these assays can be used by a wide range of PTPases, allowing comparison of kinetic parameters across the family.
Data base searches identified a number of eukaryotic receptor PTPases (R-PTPases) with 30 -32% identity and 55% similarity to CAV VP2 over a region of 139 residues. This group of R-PTPases themselves have significant similarity to each other. However, the proposed circovirus signature motif is similar to a region upstream from the defined catalytic motif of the eukaryotic proteins. This region of random coil is designated the WPD loop and, with the catalytic P-loop, is located in the catalytic cleft formed by five ␤ sheets. Alignments of CAV VP2 with sequences of dual specificity and multiple specificity phosphatases did not identify significant similarities. However, CAV VP2 and TLMV ORF2 were clearly shown to have S/T PPase activity, and this activity was inhibited by 50 mM NaF, an inhibitor of S/T PPases but not of PTPases.
A minimal signature motif with the configuration CXXXXXR is highly conserved in all PTPases. PTPases catalyze the removal of phosphate from phosphotyrosine via a cysteinyl-phosphate intermediate formed with the active cysteine in the signature motif. A proposed signature motif ICNCGQFRKH from residue 94 to 103 was defined for CAV VP2. Mutagenesis of C95S in the CAV GST-VP2 fusion protein abrogated both PTPase and S/T PPase activities. The abrogation of tyrosine and serine/threonine protein phosphatase functions by the C95S mutation identified this residue as essential to the catalytic mechanism and confirms the identity of the signature motif proposed on the basis of features of known PTPases.
In dual specificity protein phosphatases, the essential cysteine for the catalysis of phosphotyrosine substrates is also the catalytic cysteine for the catalysis of phosphothreonine or phosphoserine substrates. Crystallographic studies indicate that the principle structural difference between tyrosine specific phosphatases, such as PTP1B and YOP5, and dual specificity phosphatases, such as VHR, is the overall depth of the active site pocket (17). VHR has a relatively shallow catalytic cleft that can accommodate both the short Thr(P) and Ser(P) moieties as well as the longer Tyr(P) moieties. In contrast, the catalytic cleft of tyrosine-specific phosphatases is deeper and can only accommodate the longer Tyr(P) moiety. The modeling of CAV VP2 by threading of the sequence onto the mouse PTPA sequence identified the VP2 signature motif as located within a random loop that is designated the WPD loop in mouse PTPA. The location of the loop in the cleft is shallower than the second loop of random coil, the P loop, in which the signature motif of mouse PTPA is located. The predicted shallowness of the location of the signature motif in the three-dimensional model is therefore consistent with VP2 having dual specificity phosphatase activity.
Of additional interest is the finding of significant similarity over the same region between CAV VP2 and the Sanban subgroup of TTV. Our results clearly identify TLMV ORF2 as a dual specificity protein phosphatase. The demonstration of the phosphatase activity of TLMV ORF2 is indicative of a common strategy for infection and replication in CAV and TTV. This is consistent with the similarity found in genome organization and the sequence similarity between TTV and CAV.
The current designation of open reading frames in TTV is based on sequence analysis alone, since these viruses have not yet been grown in culture and the viral protein expression profiles have not been characterized, although recently the protein encoded by ORF3 of TTV has been shown to be a phosphoprotein (18). Numerous studies have identified a significant proportion of asymptomatic individuals as positive for TTV DNA using PCR (19 -27). Based on these PCR results, and a lack of evidence to support any specific association between TTV infection and disease, it has been suggested that the virus may produce only asymptomatic infection. However, with the identification of close functional similarity between CAV VP2 and TLMV ORF2, there is perhaps reason to suggest that there may be further similarities between TTV and CAV infection strategies. Whereas TTV ORF2 activity was shown to be approximately equivalent to CAV VP2 for the substrate ENDpYINASL, its activity was markedly lower for the substrate DADEpYLIPQQG. These results indicate differences between the two viral proteins in substrate affinities, and this requires further investigation in the context of activity during cellular infection. The influence of variation in ORF2 of TTV on PTPase function may be of further interest in deciphering the significance of the heterogeneity in the TTV cluster and could indicate the potential for variation in pathogenicity between different viruses.
Protein phosphatases are known to function in the regulation of mitogenesis, gene transcription, signal transduction, cell-cell interactions, and cellular differentiation and in cytokine responses of lymphocytes (28 -38). CAV infection of Tlymphocyte and hemocytoblast populations of chickens leads to profound immunosuppression and anemia. It will therefore be of interest to investigate VP2 protein phosphatase activity in infected cells and infected chicken lymphocyte populations. VP2 protein phosphatase activity during infection could be expected to induce regulatory changes in infected lymphocyte populations. CAV is highly dependent on host function for completion of its replication cycle, and it is possible that a capacity to regulate the cell cycle may be a critical viral function. Alternatively, the protein phosphatase activity may be primarily a requirement for viral replication rather than cellular regulation, and any cellular effects may be secondary to this function.
All previous accounts of virus-encoded regulatory proteins have involved viruses with large genome sizes and extensive coding capacity. It has been suggested that only viruses with a large genome can maintain cell regulatory proteins in addition to critical viral structural and replicative proteins. This includes the previously described VH1 PTPase from Vaccinia virus and VH1-like protein analogues of poxviruses (39 -41). Our finding is therefore unusual in that CAV has an extremely small genome size (2.3 kb) and only three viral proteins expressed from overlapping reading frames. To the best of our knowledge, only two other examples of virally encoded PTPases have been described, 1) the VH1 protein and VH1-like proteins of poxviruses and 2) BVP PTPase, which has been identified recently in a baculovirus, Autographa californica nuclear polyhedrosis virus (41). Both of these enzymes are found in enveloped viruses with complex, double-stranded DNA genomes (39 -41). The BVP and VH1 PTPases share sequence similarity with each other and with the cell cycle regulatory protein Cdc25. However, sequence similarity was not identified between CAV VP2 and BVP or VH1, suggesting a distinction between these viral enzymes. Both BVP and VH1 have additional substrate phosphatase activities. BVP exhibits modest PTPase activity and a high level of RNA triphosphatase and RNA diphosphatase activity. The serine phosphatase activity of VH1 protein is of greater magnitude than the PTPase activity (39 -41). The BVP1 gene is not essential, but BVP1 gene deletants grow to 50% of the titer of wild type virus and are inefficient in the formation of viral occlusions. When a VH1 allele containing a dominant selectable marker is introduced into vaccinia virus, only viruses with an intact VH1 gene can be recovered, suggesting that VH1 is essential (42). The VP2 gene, as one of only three expressed by CAV, is most likely to be essential to viral replication. CAV VP2 PTPase is only the third viral PTPase to be described and is the only PTPase described in a small virus.
During Vaccinia virus infection, the VH1 protein blocks interferon ␥ signaling, thereby enabling evasion of the immune response to virus infection (4). This role of the VH1 DSP, the only viral phosphatase with a characterized in vivo function, does highlight the potential for virus-encoded phosphatases to be involved in mechanisms of immune evasion and virus persistence. Whereas the function of the circovirus protein phosphatase during infection remains to be elucidated, the characterization of a biochemical function for CAV VP2 provides a focal point for further examination of the pathogenesis of CAV infection.