INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Prevotella intermedia, an anaerobic, gram-negative, rod-shaped bacterium, has been reported to be associated with advanced adult periodontal disease (4, 39), acute necrotizing ulcerative gingivitis (21), and pregnancy gingivitis (16, 32). Although the exact role of oral microbes in the etiology of periodontal disease remains unknown, they have been shown to produce a variety of potential virulence factors, including high phosphatase activity (38). To date, alkaline phosphatases (ALPases) of both P. intermedia and Porphyromonas gingivalis have been characterized (3, 48). A neutral phosphatase gene was also isolated from the oral spirochete Treponema denticola, which is associated with chronic periodontal disease (14). In addition, acid phosphatases (ACPases) are widely found in oral gram-negative bacteria (22), although only limited information regarding their role in periodontal disease or in the bacterial life cycle is currently available. For this reason, we have initiated a study of ACPase activities of the oral pathogens. P. intermedia was selected as the bacterium of choice, since ACPase activities of P. intermedia strains isolated from active sites in patients were significantly higher than those from healthy subjects (23), suggesting a possible clinical correlation. While studying the properties of the ACPase from P. intermedia, we found that purified enzyme exhibited substantial activities against O-phospho-L-tyrosine as well as phosphotyrosine-containing peptides, compounds that act as substrates for phosphotyrosyl phosphatases (PTPases).

In general, protein tyrosine phosphorylation is associated with alterations in receptor activity, cellular proliferation, and modulation of the cell cycle (18). Eukaryotic PTPases have been shown to constitute a family of enzymes that contain a number of conserved motifs, such as HCXAGXXR. However, prokaryotic PTPases appear to constitute novel families that are devoid of these motifs but that contain unique signature motifs of their own. For instance, small, acidic PTPases contain N-terminally located, highly conserved active sites (FVCXGNICRSPXAEAXF) (20). Moreover, PiALP, an ALPase from P. intermedia, appeared to represent a new family of alkaline PTPases that showed significant homology with the predicted primary structures of PhoD from Zymomonas mobilis and PhoV from Synechococcus spp., although PTPase activities of PhoD and PhoV were not investigated (3).

In the present study, we have purified an ACPase containing a PTPase activity, designated PiACP, from P. intermedia ATCC 25611 and isolated and sequenced its corresponding gene. The gene encoded a protein exhibiting significant homology to the class A bacterial ACPases. This is the first report that one of the class A bacterial ACPases possesses PTPase activity. Furthermore, preliminary mutational analysis of the enzyme was carried out to ascertain the essential role of conserved amino acid residues in enzymatic activity.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strain and growth conditions. P. intermedia ATCC 25611 was obtained from the American Type Culture Collection and grown anaerobically (5% [vol/vol] CO₂, 10% [vol/vol] H₂, 85% [vol/vol] N₂) at 37°C in brain heart infusion broth (Difco) supplemented with yeast extract (0.5%), hemin (5 µg/ml), and menadione (0.5 µg/ml). Escherichia coli JM109 and E. coli BL21(DE3) and BL21(DE3)/pLysS were used in subcloning and expression experiments. All E. coli strains were grown on Luria-Bertani (LB) agar plates or in LB broth in the presence of appropriate antibiotics (ampicillin, 50 µg/ml; chloramphenicol, 34 µg/ml).

Enzyme assays. Standard phosphatase reaction mixtures contained 50 mM sodium acetate (pH 4.9), 50 mM NaCl, 5% glycerol, 1.0 to 10 ng of enzyme (depending on activity), and p-nitrophenylphosphate (pNPP) (10 mM) as a substrate. The pNPP reactions were carried out in 0.5 ml, terminated by the addition of 0.5 ml of 1 N NaOH, and quantified by measuring absorbance at 420 nm.

Purification of PiACP. Unless otherwise mentioned, all the procedures were performed at 0 to 4°C. P. intermedia ATCC 25611 cells (20 g) were harvested from 8 liters of culture by centrifugation at 10,000 × g for 20 min. The cells were washed twice with 0.1 M Tris-HCl buffer (pH 7.5) containing 0.15 M NaCl, suspended in 50 mM Tris-HCl (pH 7.5) containing 5% glycerol (buffer A) and 50 mM NaCl, and then broken by sonic disruption with 1-min pulses of a sonic disrupter. To the resulting lysate, Triton X-114 was added to a final concentration of 1%, and the mixture was stirred for 90 min at 4°C and then centrifuged at 100,000 × g for 60 min. The supernatant fraction was collected as the crude enzyme extract and applied to a DEAE Bio-Gel (Bio-Rad Laboratories, Hercules, Calif.) column (6 by 20 cm) equilibrated with buffer A containing 50 mM NaCl. The column was washed with the equilibration buffer until no protein was detected in the effluent by measurement of A₂₈₀. The flowthrough (unbound) fraction was concentrated by filtration through a Centricon-10 ultrafiltration device (Millipore, Bedford, Mass.) and applied to a carboxymethyl Bio-Gel (Bio-Rad) column (2 by 40 cm) equilibrated with buffer A containing 50 mM NaCl. The flowthrough fraction was concentrated as described above and applied to a phosphocellulose (Sigma, St. Louis, Mo.) column (2 by 20 cm) equilibrated with buffer A containing 50 mM NaCl. The flowthrough fraction was dialyzed against buffer A containing 0.5 M NaCl and loaded onto a phenyl-Sepharose CL-4B (Pharmacia Fine Chemicals, Uppsala, Sweden) column (1 by 7 cm) equilibrated with 0.5 M NaCl in buffer A. The enzyme was eluted with 50 mM Tris-HCl (pH 7.5) containing 50% glycerol. The active fractions were pooled and concentrated as described above. The purified enzyme was frozen and stored in small portions at 80°C.

Dephosphorylation of phosphotyrosine proteins in A431 lysate. A 2-µl aliquot of purified PiACP was incubated with 10 µl of lysate of the human epidermoid carcinoma cell line A431 that contained phosphotyrosine EGFR (Upstate Biotechnology) in the presence of 50 mM sodium acetate (pH 4.9)-50 mM NaCl-5% glycerol at 37°C for 1, 3, or 15 h. Proteins were resolved in 10% polyacrylamide gels containing sodium dodecyl sulfate (SDS) (17), followed by transfer to Immobilon-P membranes (Millipore) for immunoblot (Western) analysis by using RC20H antibodies (1:1,000) and the ECL method (Amersham). In multiple trials, 1 to 5 min of exposure to X-ray films was sufficient to visualize bands.

Determination of partial amino acid sequences. The purified samples were subjected to SDS-polyacrylamide gel electrophoresis (PAGE), transferred electrophoretically to Immobilon-P membranes as described above, and then stained with Coomassie blue R-250 (Sigma). The stained bands were excised, and the absorbed proteins were sequenced with automatic gas phase sequencer HP G1005A (Hewlett-Packard, Palo Alto, Calif.) by Takara Shuzo (Tokyo, Japan). Another sample was digested with Staphylococcus aureus V8 protease (Sigma). The peptides generated were resolved by SDS-PAGE (18% polyacrylamide), electrophoretically transferred to Immobilon-P membranes, and stained with Coomassie blue R-250. Portions of the membranes containing the two most visually prominent bands were excised and subjected to automatic gas phase sequencing.

Oligonucleotides and DNA amplification by PCR. For amplification of DNA fragments encoding the partial amino acid sequences from the purified PiACP, two synthetic oligonucleotide primers were designed on the basis of the amino acid sequences of two internal regions of the purified enzyme. The primers TA42 (5'-TNYTNCCNACNCCNCCNCAR-3') and TA43 (5'-GTRTGNCCNSWNGGRTANSWNCC-3') (N denotes complete degeneracy) were synthesized by Takara Shuzo. Amplification of the DNA fragments was carried out with these primers in standard PCR buffer that included 2.5 mM MgCl₂, 200 µM (each) dATP, dTTP, dCTP, and dGTP, and 2 U of Taq polymerase (Promega, Madison, Wis.). The thermal cycle parameters were 94°C for 1 min (denaturation), 34°C for 2 min (annealing) and 72°C for 1 min 10 s (extension), for 35 cycles. In addition, time delays of 2 min at 94 and 72°C were incorporated at the beginning and end, respectively. The PCR product was purified by use of a gel extraction kit (Qiagen) and cloned into the pGEM-T vector (Promega). Sequencing the resulting clones led to the identification of one encoding the anticipated amino acid sequence of PiACP. The DNA insert of this clone was then used as a probe for further screening to obtain the full-length PiACP clone.

Construction of a genomic library and screening. Genomic DNA from P. intermedia ATCC 25611 was isolated as previously described (1). Standard procedures for recombinant DNA manipulations were carried out as described by Sambrook et al. (34). For construction of the genomic library, P. intermedia chromosomal DNA was digested with HincII and HindIII and was ligated to pBluescript SK(+) and KS(+), respectively, that had been cut with the corresponding restriction enzymes and treated with ALPase. The oligonucleotide probe (~400 bp of PCR product) was labeled with digoxigenin by using a labeling kit (Boehringer Mannheim, Indianapolis, Ind.) according to the manufacturer's instructions. Colony hybridization was performed as previously described (1). Positive clones were detected with the reagents and procedures of the same kit.

DNA sequencing. Plasmid DNA of the pBluescript clones obtained above was prepared with the Wizard Miniprep system (Promega) and sequenced by the dideoxy method of Sanger et al. (35) with a dye terminator sequencing kit (Applied Biosystems) together with a synthetic oligonucleotide primer. The sequence was determined with an Applied Biosystems model 373S automated DNA sequencer. The nucleotide sequences were analyzed with the computer software package DNA Strider, version 1.2 (24). Amino acid homology searches and comparisons were done with GENETYX-Mac software (Software Development Co., Ltd., Tokyo, Japan) and BLAST network services of DDBJ. Sequence alignments were optimized with the CLUSTAL W program (45).

Expression of PiACP gene in E. coli. The putative PiACP gene open reading frame (ORF) (792 bp) was subcloned in the T7-based bacterial expression plasmid pET3a as follows. The gene was amplified by the following primers (corresponding to the 5' and 3' ends of the gene, respectively): 5'-GGAGTTGCATATGACAAAAAAGACTTTACTTGTCGG-3' and 5'-GGAGTGGATCCTTAGTTTGCTGCCTTGAAAGTG-3' (the NdeI and BamHI sites, respectively, are underlined). The PCR product was restricted with NdeI and BamHI and cloned into the same sites of pET3a as described previously (25). The resulting clone, pET3a-PiACP, was confirmed by DNA sequencing and introduced into E. coli BL21(DE3)/pLysS. Growth of the transformant, induction with isopropyl-1-thio--D-galactopyranoside (IPTG), and lysis with lysozyme were carried out essentially as described previously (2). SDS-PAGE analysis of proteins was performed as described by Laemmli (17) with an 18% acrylamide (acrylamide/bisacrylamide ratio, 30:0.4) gel. Site-directed mutagenesis and deletion of the cloned PiACP gene were performed by the PCR-based "megaprimer" method as described previously (5).

Purification of recombinant PiACP. Identical procedures were used for the purification of wild-type and three mutant recombinant PiACPs. In brief, the total extract of 1 g of induced E. coli BL21(DE3)/pLysS cells containing the pET3a-PiACP plasmid was prepared as described above. After growth for 24 h at 37°C with vigorous shaking, the cells were harvested by centrifugation, resuspended in lysis buffer (50 mM Tris-HCl [pH 7.5], 50 mM NaCl, 2 mg of lysozyme per ml), and lysed by sonication in ice. The lysate was centrifuged at 100,000 × g for 20 min. The pellet, which contained nearly all of the expressed PiACP protein, was dissolved in 6 ml of buffer A containing 6 M guanidine hydrochloride with the aid of a homogenizer under cold conditions and incubated in ice for another 1 h. The solution was centrifuged at 100,000 × g for 20 min. The supernatant was dialyzed against 2 liters of buffer A containing 50 mM NaCl with two changes. The dialyzed material was loaded onto a phosphocellulose column (1 by 6 cm) equilibrated with buffer A containing 50 mM NaCl. An NaCl gradient in buffer A was used to subsequently elute the enzyme at an NaCl concentration of about 200 mM. The phosphocellulose fraction, already highly pure, was concentrated by Centricon-10 ultrafiltration and further purified by gel filtration chromatography through TSK gel G3000SWXL, from which it was eluted as an apparent monomer (data not shown).

Nucleotide sequence accession number. The GenBank accession number for the nucleotide sequence of the PiACP gene reported in this paper is AB017537.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Purification of PiACP. The purification of PiACP from Triton X-114 extract is summarized in Table 1. The enzyme was purified 194-fold with a final specific activity of 132.1 U/mg, and the overall yield of the activity was 23.5%. The molecular mass of native enzyme from P. intermedia was estimated by gel filtration to be about 28 kDa. On an SDS-PAGE gel, ~4 µg of purified PiACP was analyzed; as shown (Fig. 1, lane B), the preparation contains a single major band with a molecular mass of ~30 kDa, with no single contaminant being more than 5% of the total protein. These results suggested that the enzyme was a monomer of an ~30-kDa polypeptide.

View larger version (25K): [in this window] [in a new window]   FIG. 1.   SDS-PAGE (16% polyacrylamide) analysis of the purified PiACP. Lane A, molecular mass markers; lane B, purified PiACP (4 µg). The positions and molecular masses of standard proteins are indicated at the left.

Enzymatic activity of purified PiACP. The optimal temperature and pH for PiACP were determined in vitro by using pNPP as a substrate; essentially similar results were obtained by using a phosphorylated peptide (EGFR) (data not shown). The purified enzyme had the highest activity at around 46°C and was inactive at 60°C. The optimum pH value for the activity was approximately 4.9. To characterize PiACP further, the effect of inhibitors on the activity of the enzyme was investigated. As shown in Table 2, sodium molybdate and sodium orthovanadate, which are known inhibitors of PTPases (18), and sodium fluoride were found to inhibit the activity of PiACP. On the other hand, the activity was not inhibited by okadaic acid and microcystin-LR, which are known to be specific inhibitors of serine/threonine protein phosphatases of the PP1 and PP2A families (37). Sulfhydryl-modifying reagents, such as N-ethylmaleimide, iodoacetic acid, and phenylarsine oxide (PAO), which are known to inhibit PTPases of eukaryotic origin, had no significant effect. None of the divalent metal ions stimulated the activity, while only Cu²⁺ and Zn²⁺ inhibited the activity. Like PhoN of Salmonella typhimurium, PiACP was also resistant to the ACPase inhibitors EDTA and tartrate (15). The purified PiACP dephosphorylated p-Tyr strongly and p-Ser and p-Thr weakly. In addition, it dephosphorylated the EGFR phosphopeptide containing p-Tyr (40). A commercial phosphopeptide, H₂N-T-S-T-E-P-Q-Y(P)-Q-P-G-E-N-L-COOH, containing the C-terminal phosphorylation site of c-src, was also efficiently dephosphorylated by PiACP (data not shown). As shown in Fig. 2, the K_m for purified PiACP with EGFR as a substrate was determined to be 0.83 mM (pH 4.9; 37°C) and the V_max was 8.44 µmol/min · mg of enzyme. The K_m and V_max for purified PiACP with pNPP as the substrate were determined to be 0.24 mM (pH 4.9; 37°C) and 32.46 µmol/min · mg, respectively. These values were similar to those for the rat liver low-molecular-weight PTPases with the same substrate (40). Furthermore, the enzyme strongly dephosphorylated a 170-kDa protein (molecular size corresponds to that of EGFR) in A431 cell lysate which was phosphorylated on tyrosine (Fig. 3).

View larger version (15K): [in this window] [in a new window]   FIG. 2.   Substrate dependence curves for PiACP. Purified PiACP was assayed at 37°C in 50 mM sodium acetate (pH 4.9) buffer with a synthetic peptide corresponding to human EGFR, either A-E-N-A-E-Y(P)-L-R-V [0.04 to 0.4 mM; Y(P), phosphotyrosine] or pNPP (0.25 to 10 mM). The data were plotted by the Lineweaver-Burk method and subjected to linear regression. The Km and Vmax values for EGFR were determined to be 0.83 ± 0.035 mM and 8.44 ± 0.24 µmol/min · mg, respectively. Km and Vmax values for pNPP were determined to be 0.24 ± 0.009 mM and 32.46 ± 0.53 µmol/min · mg, respectively.

View larger version (52K): [in this window] [in a new window]   FIG. 3.   Dephosphorylation of tyrosine-phosphorylated proteins in A431 lysate by PiACP. A 2-µl aliquot of purified PiACP was incubated with 10 µl of lysate of the human epidermoid carcinoma cell line A431 for 1 h (lane b), 3 h (lane c), or 15 h (lane d). Lane a, control. Proteins were resolved by SDS-PAGE and transferred to Immobilon-P membranes for immunoanalysis as described in Materials and Methods. The molecular size corresponding to EGFR (170 kDa) is indicated. Molecular mass markers are at the left.

Partial amino acid sequence of PiACP. The N-terminal amino acid sequence of the purified PiACP polypeptide, determined by Edman degradation, was Lys-Lys-Ile-Lys-Asp-Ala-Arg-Thr-Asn-Pro-Asp-Leu-Tyr-Tyr-Leu-Gln-Asp-Gly. Two internal sequences obtained from N-terminal sequencing of V8 protease fragments were Leu-Leu-Pro-Thr-Pro-Pro-Gln-Pro-Gly-Ser-Ile-Gln-Phe-Leu-Tyr-Asp-Glu-Ala-Gln-Tyr and Leu-Ser-Thr-Asn-Gly-Ser-Tyr-Pro-Ser-Gly-His-Thr-Ala-Ile-Gly-Trp-Ala-Thr-Ala-Leu. Interestingly, these two internal amino acid sequences were found to contain the conservative motifs in class A bacterial ACPases, thus providing preliminary evidence that the primary structure of PiACP may be similar to those of other ACPases, a conclusion further supported by sequence analysis of the complete gene as described below.

Cloning of PiACP. On the basis of the internal amino acid sequences of PiACP, oligonucleotide primers were synthesized and PCR was performed to obtain a partial clone of the enzyme gene as described in Materials and Methods, which was then used to screen the P. intermedia genomic DNA library to identify the full-length clone. Southern blot analysis indicated that P. intermedia genomic DNA may not produce convenient restriction fragments that will contain a complete structural gene for PiACP. Since the partial region for the PiACP gene contained HincII and HindIII sites, the upstream region of the PiACP gene was screened from a HincII library and the downstream region was screened from a HindIII library of P. intermedia genomic DNA. Two positive clones from the HincII library and one from the HindIII library were independently obtained from a total of 2,000 transformants by colony hybridization. Southern blot analysis indicated that the two HincII clones contained identical inserts. The results of Southern blotting as well as restriction mapping also indicated that the HincII and HindIII clones contained overlapping DNA fragments as expected (data not shown). The clones from HincII and HindIII libraries were designated pAC1 and pAC2, respectively.

Sequence analysis of the phosphatase gene. In order to identify and characterize the gene for PiACP, both HindIII and HincII clones were sequenced and the gene sequences were assembled and analyzed. Since the coding region was sequenced and confirmed for both strands by using overlapping fragments, sequencing errors were ruled out. This total sequence contained one complete ORF and two incomplete ORFs. Furthermore, the N-terminal and two internal sequences described earlier could be identified in a long ORF (Fig. 4). The PiACP gene contained 792 bp coding for a putative polypeptide of 264 amino acids with a calculated molecular mass of 29,164 Da and an estimated pI of 8.39. A potential ribosome-binding site was not identified since little information is known about P. intermedia gene control at present. A sequence with hyphenated dyad symmetry with the potential to form a stem-loop structure in the RNA was located downstream from the translational termination codon TAA, suggesting that this sequence may act as a transcriptional terminator. The overall G+C content of the gene (46.9%) agrees with that estimated for the chromosomal DNA from P. intermedia strains (41 to 44%) (36). The first 20 amino acid residues appeared to have a hydrophobic region characteristic of a signal sequence including a potential cleavage site (between Ala-20 and Gln-21), as defined by the criteria of von Heijne (47). This indicates that PiACP may be secreted across the cytoplasmic membrane by using the signal peptide.

View larger version (69K): [in this window] [in a new window]   FIG. 4.   Nucleotide and predicted amino acid sequences of PiACP. The underlined areas indicate areas of the protein that were previously analyzed by amino acid sequencing. The predicted site of proteolytic cleavage of the putative signal sequence is indicated by an arrowhead. The two convergent arrows indicate repeat sequences (details in the text). RBS?, potential ribosome-binding site.

View larger version (85K): [in this window] [in a new window]   FIG. 5.   Comparison of the deduced amino acid sequence of PiACP with those of the class A bacterial ACPases. ACP-Sfl, ACPase from S. flexneri; ACP-Pst, ACPase from P. stuartii; ACP-Mm, ACPase from M. morganii; Apy-Sfl, apyrase from S. flexneri; ACP-Sty, ACPase from S. typhimurium; ACP-Zm, ACPase from Z. mobilis. Identical residues are indicated by an asterisk; conservative amino acid substitutions are indicated by a colon or dot.

Expression of recombinant PiACP in E. coli. To ascertain that the PiACP gene indeed codes for a phosphatase, full-length PiACP was overexpressed in E. coli. Upon induction of E. coli BL21(DE3)/pLysS containing pET3a-PiACP with IPTG, a polypeptide with an M_r of ~30,000 was produced (Fig. 6), in good agreement with the calculated molecular mass of the predicted amino acid sequence (29,164 Da). Recombinant PiACP was purified as described in Materials and Methods. During purification, the ACPase activity, with either pNPP or EGFR as the substrate, always cochromatographed with the 30-kDa protein. The final preparation (Fig. 6, lane 2) was judged at least 90% pure. Bacterial extract obtained from uninduced cells contained very little ACPase activity, while the activity of an extract of E. coli BL21(DE3)/pLysS containing the pET3a vector was undetectable (data not shown).

View larger version (63K): [in this window] [in a new window]   FIG. 6.   Overexpression of recombinant PiACP in E. coli. The ORFs for PiACP and its mutants were cloned and expressed in pET3a, as described in Materials and Methods. BL21(DE3)/pLysS cells containing the following recombinant PiACP clones were induced with (lane +) or without (lane ; control) IPTG (0.5 mM), and total extracts (lanes  and +), the denaturant-dissolved fraction (lane 1), or the purified fraction (lane 2; gel filtration fraction) were analyzed by SDS-PAGE followed by staining with Coomassie blue. Lanes , +, 1, and 2, wild-type PiACP; lane 3, GSYPSGHT mutant; lane 4, H170Q mutant; lane 5, H209T mutant; lane M, protein standards as in Fig. 1.

View larger version (20K): [in this window] [in a new window]   FIG. 7.   Chemical modification of recombinant PiACP in E. coli. The residual activity was measured at 37°C (pH 4.9) with pNPP as the substrate after the following treatments: PAO (10 mM [] and 20 mM []), DEPC (1 mM [] and 5 mM []), and phenylglyoxal (1 mM [] and 5 mM []).

Mutagenesis of recombinant PiACP. The invariant His and Arg residues are shared among class A bacterial ACPases, a neutral phosphatase from T. denticola, and lipid phosphatases and G6Pases (42). In G6Pases, the mutation of two conserved His residues resulted in the loss of activity (19, 29). In this study, to further confirm the role of His residues, we mutated specific His residues in recombinant PiACP and investigated the effect of the mutations on enzyme activity. A comparison between PiACP and class A bacterial ACPases revealed two conserved regions containing His residues, one of which is the peptide stretch GSYPSGHT (residues 164 to 171) of PiACP. A mutant clone in which the nucleotides corresponding to this stretch were deleted by in vitro mutagenesis was constructed. In addition, two mutants were constructed by site-specific mutagenesis of pET3a-PiACP DNA by PCR, viz., H170Q and H209T mutants. Figure 6 shows that the GSYPSGHT deletion mutant (lane 3) expressed a protein that migrated faster than the wild type (lane 2), indicating a molecular mass difference of ~0.8 kDa, whereas proteins expressed by H170Q (lane 4) and H209T (lane 5) mutant clones were ~2 kDa bigger than the wild type. The three mutant proteins were purified to near homogeneity, and their phosphatase activities were tested by using pNPP as the substrate. All three mutants were completely devoid of activity (data not shown). These results not only confirm the 30-kDa protein as the phosphatase but in addition suggest an essential role for these specific amino acid residues in enzyme activity.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study, we isolated, cloned and analyzed a novel ACPase, PiACP, in molecular detail and discovered that it contained PTPase activity. Phosphatase activities and gene sequences in several periodontopathogenic bacteria, including P. intermedia, have been reported (3, 14, 48). However, limited information about these potentially important enzymes is available. In this study, comparison of the sequence of PiACP with the sequences of other proteins clearly indicated that PiACP was closely related to the class A bacterial ACPase family (6, 15, 30, 43, 46). Class A bacterial ACPases are further classified into two major subgroups: classes A1 and A2 (44). Class A1 enzymes are resistant to fluoride and contain a slightly smaller polypeptide component (~25 kDa, such as PhoC of M. morganii), while class A2 enzymes are inhibited by fluoride and contain a slightly larger polypeptide component (~27 kDa, such as PhoN of S. typhimurium). Very recently, a new subclass, viz., class A3, has been recognized (33); this class consists of monomeric enzymes that are inhibited by fluoride, O-vanadate, and various divalent cations, including Cu²⁺ and Zn²⁺. An example of a class A3 phosphatase is the apyrase of S. flexneri, whose activity is strongly inhibited by 1 mM O-vanadate, 5 mM sodium fluoride, 5 mM Zn²⁺, and 5 mM Cu²⁺ to less than 5% of its initial value, although it was resistant to a high concentration of EDTA (20 mM) (6, 7). Based on these diagnostic criteria, we propose that PiACP may be classified as a class A3 enzyme rather than a class A2 enzyme.

Stukey and Carman (42) have recently recognized a conserved sequence motif, KX₆RP-(X_12-54)-PSGH-(X_31-54)-SRX₅HX₃D, that is shared among class A bacterial ACPases, a neutral phosphatase from T. denticola, and mammalian phosphatases (such as G6Pases and lipid phosphatases), although their overall amino acid identities were low. Of these phosphatases, G6Pases have been well investigated by structure-function analysis, which has shown that two His residues participate in the catalytic mechanism and that His-176 could act as the phosphoryl acceptor in catalysis (19, 29). Curiously, two His residues were also shown to be essential for activity in our mutational studies of PiACP (i.e., H170 and H209), although the exact role of each His residue was not determined. On the other hand, sequence alignment analysis also revealed the presence of two Arg residues conserved between PiACP and G6Pase: Arg-142 and Arg-203 in PiACP and Arg-83 and Arg-170 in G6Pase, respectively. Structure-function studies suggest that Arg-83 in G6Pase is involved in stabilizing the phosphoryl enzyme intermediate formed during catalysis (19). A detailed mutational analysis of specific PiACP residues, including Arg residues, will be needed to define their roles in the PiACP catalytic mechanism. This is in progress.

A phylogenetic tree depicting the evolutionary relationships among the primary structures of the above-described enzymes is shown in Fig. 8. In particular, PiACP was found to be closely related to the enzyme of the enterobacteria, such as M. morganii and S. flexneri. Interestingly, it has been suggested that the phoN gene of S. typhimurium, a well-studied enterobacterium, was acquired by lateral transmission from another species, since the overall base composition of phoN was different from that of the Salmonella chromosome (11). However, this does not appear to be the case with the PiACP gene, because the G+C content of the PiACP gene (46.9%) is almost consistent with the overall G+C content (41 to 44%) of P. intermedia (36). These observations raise interesting questions about the evolution and origin of these phosphatases, since an ancestral relatedness between P. intermedia and several enterobacteria, such as Morganella species, seems to exist. In this group of phosphatase genes, only the product of the PiACP gene has thus far been shown to contain in vitro PTPase activity (this paper). It will be interesting to see if other ACPases of this group also exhibit PTPase activity.

View larger version (12K): [in this window] [in a new window]   FIG. 8.   Phylogenetic tree of the class A bacterial ACPase family and other phosphatases. The unrooted tree was constructed with NjPlot of the CLUSTAL W software package. All of these phosphatases contain the motif KXXXXXXRP-(X12-54)-PSGH-(X31-54)-SRXXXXXHXXXD as described in the text. Abbreviations for phosphatases are as follows: E. coli phosphatidylglycerol phosphate phosphatase, PGP-Ec; Haemophilus influenzae phosphatidylglycerol phosphate phosphatase, PGP-Hi; T. denticola neutral phosphatase, NP-Td; rat G6Pase, G6P-rat; and human G6Pase, G6P-human. The other abbreviations are as defined in the legend for Fig. 5.

To date, several prokaryotic PTPases have been genetically and biochemically identified. For instance, IphP from the cyanobacterium Nostoc commune UTEX 584 has been cloned and characterized (31). The enzyme displayed both protein phosphoserine and PTPase activities, thus showing significant similarity with VH1 (13). Another bacterial PTPase previously described, YopH, was found to be an important virulence determinant in Yersinia spp. (12). These PTPases contain the highly conserved motif HCXAGXXR (both Cys and Arg are essential for activity) in the active domains of the enzymes (41). On the other hand, a PTPase without this motif, viz., small, acidic PTPase, has recently been identified (20) and has been found to contain another motif, FVCXGNICRSPXAEAXF, near the N terminus. Upon examination of the complete sequences of class A bacterial ACPases including PiACP, however, we were not able to find either the HCXAGXXR or the FVCXGNICRSPXAEAXF motifs. Class A bacterial ACPases also shared no overall sequence with the prokaryotic PTPases described above. Nevertheless, PiACP was inhibited by sodium orthovanadate and sodium molybdate, known inhibitors of PTPases. Furthermore, it did exhibit substantial activity against protein tyrosine phosphates, such as those present in growth factor receptors containing EGFR. These results suggest that one can consider the class A bacterial ACPase group including PiACP as constituting a novel family of acidic PTPases significantly different from other known PTPases. It is also tempting to speculate that the class A bacterial ACPase family belongs to a novel widespread PTPase family containing several mammalian phosphatases, such as G6Pases.

In our study, the effects by Cys modifier PAO on PiACP activity were weaker than those of the His modifier DEPC and the Arg modifier phenylglyoxal (Fig. 7). Considering that catalysis by all PTPases reported to date has been shown to proceed through the formation of a covalent phosphorus- and sulfur-containing intermediate involving Cys, the mechanism of PiACP activity, as well as that of class A bacterial ACPases, must be different from that of the known PTPases.

It is currently not known why the H170Q and H209T mutant proteins migrated more slowly than the wild type in the SDS-PAGE gel and exhibited an apparent molecular mass of 32 kDa, which is 2 kDa higher than that of the wild-type enzyme. However, because this size difference corresponds approximately to the molecular weight of the PiACP signal peptide, it seemed possible that the replacement of His-170 or His-209 had an effect on the processing of the PiACP precursor and that this somehow resulted in the secretion of unprocessed enzyme, as demonstrated for the ACPase enzyme of E. coli (27).

Finally, the biological function of PiACP is currently unknown. In view of the pathogenesis of periodontal disease, the ability of PiACP to dephosphorylate the phosphopeptide corresponding to EGFR may play a significant role. A clinical isolate of P. intermedia was recently found to invade a human oral epithelial cell line (8). Polypeptide growth factors, such as EGF, are biological mediators of cellular functions, such as differentiation, motility, and matrix synthesis (10), and have been shown to play important roles in the regenerative response (28). Considering that EGFRs were expressed at high levels on the cell surfaces of basal cell layers of the gingival epithelium (26), it can be speculated that PiACP released from this bacterium inhibits EGF promotion of cell proliferation. Additional studies are needed to ascertain the exact nature and physiological function of PiACP.