The Purification and Characterization of Phosphonopyruvate Hydrolase, a Novel Carbon-Phosphorus Bond Cleavage Enzyme from Variovorax sp. Pal2*

Phosphonopyruvate hydrolase, a novel bacterial carbon-phosphorus bond cleavage enzyme, was purified to homogeneity by a series of chromatographic steps from cell extracts of a newly isolated environmental strain of Variovorax sp. Pal2. The enzyme was inducible in the presence of phosphonoalanine or phosphonopyruvate; unusually, its expression was independent of the phosphate status of the cell. The native enzyme had a molecular mass of 63 kDa with a subunit mass of 31.2 kDa. Activity of purified phosphonopyruvate hydrolase was Co2+-dependent and showed a pH optimum of 6.7–7.0. The enzyme had a Km of 0.53 mm for its sole substrate, phosphonopyruvate, and was inhibited by the analogues phosphonoformic acid, 3-phosphonopropionic acid, and hydroxymethylphosphonic acid. The nucleotide sequence of the phosphonopyruvate hydrolase structural gene indicated that it is a member of the phosphoenolpyruvate phosphomutase/isocitrate lyase superfamily with 41% identity at the amino acid level to the carbon-to-phosphorus bond-forming enzyme phosphoenolpyruvate phosphomutase from Tetrahymena pyriformis. Thus its apparently ancient evolutionary origins differ from those of each of the two carbon-phosphorus hydrolases that have been reported previously; phosphonoacetaldehyde hydrolase is a member of the haloacetate dehalogenase family, whereas phosphonoacetate hydrolase belongs to the alkaline phosphatase superfamily of zinc-dependent hydrolases. Phosphonopyruvate hydrolase is likely to be of considerable significance in global phosphorus cycling, because phosphonopyruvate is known to be a key intermediate in the formation of all naturally occurring compounds that contain the carbon-phosphorus bond.

Organophosphonates are a group of biogenic compounds characterized by the presence of a stable covalent carbon to phosphorus (C-P) 1 bond. Although they are believed to have originated before the earth's atmosphere became extensively oxygenated, C-P compounds have been isolated from a wide variety of extant life forms, and a significant percentage of biogenic phosphorus is believed to occur in the C-P linkage. Moreover a growing number of synthetic organophosphonates have found applications in industrial, agricultural, and domestic products and are ultimately disposed of to soils or natural waters. An understanding of C-P bond metabolism is thus a prerequisite to a fuller understanding of global biogeochemical phosphorus cycling.
Four distinct bacterial enzymes are known to carry out the C-P bond cleavage reaction that is central to organophosphonate mineralization. The C-P lyase complex(es) has a broad substrate specificity (1) and can act upon unsubstituted alkyl and aryl organophosphonates (general formula R-PO 3 H 2 ). The two C-P hydrolases that act on phosphonoacetaldehyde (OHC-CH 2 -PO 3 H 2 ) (2) and phosphonoacetate (HOOC-CH 2 -PO 3 H 2 ) (3) are essentially specific to their respective substrates. A fourth C-P bond-metabolizing enzyme, phosphoenolpyruvate phosphomutase (PEP mutase), catalyzes the intramolecular rearrangement and interconversion of 3-phosphonopyruvate (COOH-CO-CH 2 -PO 3 H 2 ) and phosphoenolpyruvate (COOH-C(ϭCH 2 )-O-PO 3 H 2 ) (4). Phosphonopyruvate formation by this route is a key step in the biosynthesis of all known natural products that contain the C-P bond (5).
One of the most widely distributed of such biogenic C-P compounds is 2-amino-3-phosphonopropionic acid, commonly called phosphonoalanine (HOOC-CH(NH 2 )-CH 2 -PO 3 H 2 ); it is formed through the transamination of phosphonopyruvate by many lower organisms, such as the sea anemone Zoanthus sociatus (6) and the protozoan Tetrahymena pyriformis (7). Our earlier studies on its biodegradation showed that 47 of 100 randomly chosen environmental bacterial isolates had the ability to utilize phosphonoalanine as sole phosphorus source for growth (8). Subsequently Ternan et al. (9) reported the isolation of an environmental Burkholderia cepacia strain capable of growth on phosphonoalanine as sole source of carbon, nitrogen, and phosphorus. Cell extracts of the isolate contained a previously unknown P i -independent activity designated phosphonopyruvate hydrolase (PPH) that catalyzed the hydrolytic cleavage of the C-P bond of phosphonopyruvate to yield pyruvate and P i (Fig. 1).
We now describe the purification and properties of PPH from a soil isolate of a Variovorax sp. and the sequence analysis of the structural gene that encodes it. Intriguingly, the enzyme shares a common ancestry with PEP mutase, which is responsible for phosphonopyruvate biosynthesis (see above); the interaction of the two activities is thus likely to play an important regulatory role in global phosphorus cycling.

EXPERIMENTAL PROCEDURES
Microorganism and Culture Conditions-Samples from soils, natural water bodies, and waste treatment installations in Northern Ireland were incubated in mineral salts medium (8) with D,L-phosphonoalanine (5 mM) as sole carbon, nitrogen, and phosphorus source for 5-7 days. Cultures were plated on the same medium solidified with agar, single colonies were isolated, and those that grew most rapidly were chosen for further study.
Cells of Variovorax sp. Pal2 were routinely grown in batch culture at 30°C on an orbital shaker at 100 rpm in mineral salts medium (8) supplemented with 3 g pyruvate/liter and 2 g gluconate/liter as carbon sources, 5 mM NH 4 Cl as nitrogen source, and 0.5 mM phosphonoalanine as phosphorus source. Cells were harvested in mid-log phase by centrifugation, washed twice in 20 mM HEPES-KOH buffer, pH 7.0, and stored at Ϫ20°C.
Synthesis of 3-Phosphonopyruvate-The trilithium salt of phosphonopyruvate was prepared according to the method of Sparkes et al. (10) by Chemical Synthesis Services, Belfast, Northern Ireland via chemical transamination of phosphonoalanine (Sigma) in the presence of Cu 2ϩ and pyridine catalysts with glyoxylate as the amino group acceptor. The synthesis yielded a pale yellow solid, which was characterized by 31 P and 1 H NMR spectroscopy; this showed the material to be 3-phosphonopyruvate with a purity of 95% by 1 H NMR and high pressure liquid chromatography.
Enzyme Assay-PPH activity was assayed at 37°C by measuring the amount of phosphate liberated from phosphonopyruvate. The assay mixture (1 ml) contained 100 mol HEPES-KOH buffer, pH 7.0, 5 mol phosphonopyruvate, and 5 mol CoCl 2 . The reaction was initiated by the addition of 25 l of enzyme solution (0.1-0.3 g of protein/ml) and terminated after 10 min by the addition of 200 l of 3 M trichloroacetic acid. The mixture was centrifuged, and the phosphate released was measured by the method of Fiske and Subbarrow (11). Activities were expressed as units of 1.0 mol of phosphate released per min (mean of three replicates). When required, pyruvate release was measured using a lactate dehydrogenase-based pyruvate test kit (Sigma).
Protein Assay-Protein concentrations were determined by the BCA (bicinchoninic acid) method with bovine albumin as standard (Pierce).
Purification of Phosphonopyruvate Hydrolase-During enzyme puri-fication all column chromatography steps, each lasting for no more than 2.5 h, were performed at room temperature. Collected fractions were removed to ice. For preparation of crude extract, the washed cells (ϳ15 g wet weight from 5 liters of the culture) were suspended in 65 ml of 20 mM HEPES-KOH buffer, pH 7.0, and disrupted by sonication on ice for 6 min (30-s sonication followed by 2-min cooling) at 16 kHz. The cellular debris was removed by centrifugation (25,000 ϫ g for 30 min at 4°C), and the supernatant was immediately used in the next step. For ammonium sulfate fractionation, the cell-free supernatant was fractionated in two steps by addition of solid ammonium sulfate at 20°C and stirring for 1 h. The precipitate obtained between 2.1 and 3.1 M was collected by centrifugation (10,000 ϫ g for 30 min at 4°C) and dissolved in 20 mM HEPES-KOH buffer, pH 7.0. For hydrophobic interaction chromatography, the protein fraction obtained by salt precipitation was applied to a phenyl-Sepharose HP column (5.0 ϫ 4.0 cm) (Amersham Biosciences) equilibrated with 50 mM HEPES-KOH buffer, pH 7.0, containing 1 M (NH 4 ) 2 SO 4 . The column was washed with 3 bed volumes of 0.5 M (NH 4 ) 2 SO 4 in 50 mM HEPES-KOH buffer, pH 7.0. The proteins were eluted at 8 ml/min first with 0.1 M (NH 4 ) 2 SO 4 in 50 mM HEPES-KOH buffer, pH 7.0 (200 ml), and then the enzyme with the buffer alone (160 ml). Fractions containing active enzyme were pooled, concentrated, and desalted with 20 mM HEPES-KOH buffer, pH 7.0, by ultrafiltration through a Vivaspin 15 concentrator (10,000 molecular weight cut-off) (Sartorius). For anion-exchange chromatography (column 1), the enzyme solution was applied to a Q-Sepharose HP anion-exchange column (1.6 ϫ 4.8 cm) (Amersham Biosciences) equilibrated with 20 mM HEPES-KOH buffer, pH 7.0. Proteins were eluted at 2.5 ml/min, applying an initial isocratic step in the equilibration buffer (27 ml) followed by a linear gradient of 0 -0.2 M NaCl in the same buffer (80 ml). Fractions containing the enzyme were pooled and dialyzed against 25 mM HEPES-KOH buffer, pH 7.0, containing 0.5 M NaCl. For immobilized metal ion affinity chromatography, the enzyme solution (1.1 ml) was applied to a chelating-Sepharose column (1.0 ϫ 7.5 cm) (Amersham Biosciences) charged with Cu 2ϩ and equilibrated with 25 mM HEPES-KOH buffer, pH 7.0, containing 0.5 M NaCl. The column was washed with the same buffer (20 ml), and the bound proteins were eluted at 2 ml/min with a linear gradient of 0 -0.5 M NH 4 Cl in 25 mM HEPES-KOH buffer, pH 7.0, containing 0.5 M NaCl. The active fractions were pooled, concentrated, and desalted with 20 mM HEPES-KOH buffer, pH 7.8, by ultrafiltration as before. For anion-exchange chromatography (column 2), the sample (1.1 ml) was applied to a Q-Sepharose HP column (1.0 ϫ 4.5 cm) equilibrated with 20 mM HEPES-KOH buffer, pH 7.8. The column was washed with the same buffer (10 ml), and the bound proteins were removed with a linear gradient of 0 -0.2 M NaCl in the same buffer at a flow rate of 1 ml/min. PPH was eluted in a single major peak at about 0.19 M NaCl. The active fractions were pooled, dialyzed, and concentrated by ultra filtration as before and stored in aliquots with 25% glycerol at Ϫ20°C.
Gel Electrophoresis-SDS-PAGE used precast 8 -16% Tris/glycine or NuPAGE BisTris gels in an X cell 11 mini-cell with Mark 12 molecular mass standards (Invitrogen). Gels were stained with EZBlue gel staining reagent (Sigma) or the SilverQuest kit (Invitrogen). Electrophoresis under non-denaturing conditions used precast 8 -16% Tris/glycine polyacrylamide gels (Invitrogen). Enzyme activity was located by immersion of the gel for 15 min at 37°C in the phosphonopyruvate hydrolase assay mixture and then detecting the phosphate released with the Fiske and Subbarrow reagent (see above); the enzyme's activity was manifested by a blue band. Protein was detected using the EZBlue reagent (Sigma).
Size Exclusion Chromatography-This was performed using an Ultraspherogel SEC 3,000 column (Beckman Coulter) equilibrated with 20 mM sodium phosphate buffer, pH 7.2, containing 0.25 M NaCl, at a flow rate of 1 ml/min.
Protein Sequencing-Approximately 50 g of dried enzyme was dissolved in trifluoroacetic acid (35 l) and water (15 l), and a small crystal of CNBr was added. After incubation at room temperature for 24 h the sample was lyophilized and dissolved in 10 l of SDS-PAGE sample buffer. Pure enzyme and its CNBr fragments were separated in NuPAGE BisTris gels (Invitrogen) and blotted by the semi-dry method onto polyvinylidene difluoride membranes (ProBlot; Applied Biosystems) using 10 mM CAPS buffer, pH 11.0, containing 10% methanol. Enzyme Stability-The effect of temperature on the stability of PPH was determined by exposure of the reaction mixture without substrate at temperatures from 30 to 75°C for 10 min prior to performing the assay.
Effect of Temperature-PPH activity was assayed at different temperatures over the range from 0 to 60°C.
Effect of Metal Ions-The effect of metal ions on purified enzyme that had been dialyzed was tested by the addition of EDTA (1 mM) for 1 h at 0°C to the reaction mixture and then incubating different metal ions (5 mM final concentration) for 5 min at 37°C prior to measuring activity. The activity in the dialyzed preparation was 61 units/mg, and after EDTA treatment this fell to 30 units/mg; all activities were compared with the latter value.
Determination of Kinetic Parameters-V max and K m of the purified enzyme were estimated for phosphonopyruvate by using concentrations ranging from 0.05 to 25.0 mM. Activity was measured at 37°C as described above.
Cloning and Analysis of the pphA Gene-General DNA isolation and manipulation techniques were carried out following published protocols (12). Restriction endonuclease digestions and ligations with T4 DNA ligase were performed as recommended by the enzymes' suppliers (Amersham Biosciences and Sigma). Plasmid pGEM-T Easy (Promega) was used for cloning of the PCR fragments. Plasmids pUC18 and pUC19 containing inserts from the pphA gene region and PCR fragments were used as templates for the CEQ DTCS Dye Terminator Cycle sequencing kit (Beckman Coulter). DNA sequences were obtained using an automatic DNA sequencer (CEQ2000; Beckman Coulter). The nucleotide sequences of both strands were determined. Initial computer analysis of the sequences was performed using the DNASIS V2.6 (Hitachi) software package. The alignment of sequences was performed by using ClustalW (13) with parameters set at default values.
On the basis of the N-terminal amino acid sequence of the purified enzyme, a forward oligonucleotide primer was designed: 5Ј-TTYAC-NGCNATGGCNGCNCACAA-3Ј (JQ102). A reverse primer was designed using the sequence of the internal 17.4-kDa polypeptide: 5Ј-TTNGGRAANGTYTTRTCYTCCAT-3Ј (JQ111). A 300-bp fragment of the pphA gene was amplified by PCR, using total Variovorax sp. Pal2 genomic DNA and primers JQ102 and JQ111. The reaction was carried out in a volume of 25 l with concentrations of deoxynucleotide triphosphates at 200 M and the JQ102 and JQ111 primers at 0.15 M each. The following temperature profile was used: initial denaturing at 95°C for 1 min and then 30 cycles of 95°C for 30s, 60°C for 30s, and 70°C for 30s. Subsequently the 300 bp pphA gene fragment was used as a probe for Southern hybridization.
To clone the complete pphA gene, hybridization experiments were set up with SacI, SalI, SmaI, and PstI digests of total Variovorax sp. Pal2 DNA. DNA fragments were transferred to Hybond-Nϩ membrane (Amersham Biosciences) by Southern blotting, and the membranes were then treated according to the instructions of the manufacturer. Hybridization probes were labeled using the Gene Images Random Prime Labeling and Detection system (Amersham Biosciences). Hybridization was carried out at 64°C, and detection was performed according to the manufacturer's instructions. Based on the results of these experiments libraries of 3-4-kbp SacI and 2.5-3.5-kbp SmaI fragments of Variovorax sp. Pal2 DNA were constructed in pUC18 and pUC19 vectors, respectively. Clones containing fragments with the pphA gene were identified by colony hybridization with the same fluorescein labeled 300-bp fragment probe.
16 S rRNA Gene Sequencing Analysis-An almost complete 16 S rRNA gene from isolate Pal2 was amplified by PCR with the universal primers described by Pascual et al. (14). Sequencing of the 16 S rRNA gene and its analysis were carried out as described previously (15).

RESULTS
Isolation of Variovorax sp. Strain Pal2-A program of enrichment culture on D,L-phosphonoalanine was carried out using samples from local soils, natural water bodies, and waste treatment installations. Bacterial strains capable of metabolizing the compound as sole carbon, nitrogen, and phosphorus source were obtained from all six sites sampled; the isolates were found to utilize only the L-enantiomer of phosphonoalanine. A small number of isolates were also obtained that were capable of growing on phosphonopyruvate, the transamination product of phosphonoalanine, as the sole carbon and phosphorus source. Enzyme assays carried out on crude extracts of the cultures demonstrated that each strain contained PPH activity, which was inducible in the presence of phosphonoalanine or phosphonopyruvate or both. Induction of the enzyme was not repressed in any isolate by the presence of phosphate. The bacterial strain, designated Pal2, that grew most quickly on phosphonoalanine was chosen for further study; it was identified by 16 S rDNA sequence as a Variovorax sp.
Purification of PPH-Initial studies on cell extracts from Variovorax sp. Pal2 revealed a 20% loss in activity on storage at room temperature for 24 h. It was therefore necessary to minimize exposure to higher temperatures and all chromatography fractions were collected, on ice. A five-step procedure was developed to obtain pure enzyme; this employed a combination of salt precipitation and chromatography using hydrophobic interaction, immobilized metal affinity, and anion exchange (Table I). The PPH was purified 189-fold (with a 19% yield) and had a specific activity of 163 units/mg of protein. Enzyme prepared in this way migrated as a single protein band of 31.5 kDa on silver-stained SDS-containing electrophoresis gels (Fig.  2a). When non-denaturing Tris/glycine polyacrylamide gels were used the PPH activity was found to coincide with the single band observed on Coomassie Blue staining (Fig. 2b).
The Structure of PPH-The enzyme had a molecular mass of 31.187 kDa determined by mass spectrometry and 63 kDa by size exclusion chromatography under non-denaturing conditions. This indicates that PPH is a dimer. The pI of the purified protein was 5.35. The N-terminal sequences of the complete polypeptide and two internal fragments generated by CNBr (20.4 and 17.4 kDa) were determined (Table II).
The Stability of PPH-The effect of heat on the enzyme showed that activity was lost after an exposure of 10 min to 50°C. The stability of pure enzyme during storage was increased by inclusion of 25% glycerol, and after 6 months of storage at Ϫ20°C the loss in activity was ϳ20%.
The Effects of pH, Temperature, and Metal Ions on Activity-The influence of pH on the reaction was examined over the pH range from 3.5 to 10.0, and an optimum of about 7.0 was found. The effect of temperature was also determined, and 37°C was selected for routine use.
As the hydrolases described for other phosphonate com-pounds were shown to be metalloproteins (2,3), the effect of EDTA treatment and metal ion addition to the dialyzed enzyme were investigated (Table III). Exposure to EDTA reduced activity by about 50%. Co 2ϩ and, to some extent, Ni 2ϩ and Mg 2ϩ , was found to restore PPH activity. Cs 2ϩ and Ca 2ϩ had no effect, whereas Mn 2ϩ and Cu 2ϩ reduced the activity by 41 and 45%, respectively. These results suggest the enzyme is dependent on Co 2ϩ , and this ion was included in the assay mixtures.
Kinetics and Specificity-The stoichiometry of the catalyzed reaction was confirmed by the equimolar release of pyruvate and phosphate from phosphonopyruvate with the purified enzyme. The ability of the enzyme to catalyze the intramolecular rearrangement of phosphonopyruvate to phosphoenolpyruvate was examined using the coupled ADP-pyruvate kinase-NADHlactate dehydrogenase system of Nakashita et al. (17); this revealed a rate of phosphoenolpyruvate formation some 1000fold lower than that of pyruvate and P i . Phosphoenolpyruvate did not itself serve as a substrate for phosphonopyruvate hydrolase. In addition, some small phosphatase substrates (glycerophosphate, phospho-L-serine, and phosphoglycolic acid) were examined as possible PPH substrates; none were hydrolyzed.
Of 20 phosphonate compounds tested activity was detected only with phosphonopyruvate. The possible inhibitory effects of the other 18 compounds on the purified enzyme were also studied (Table IV). It was found that only 3-phosphonopropionic acid (17%), hydroxymethylphosphonic acid (25%), and phosphonoformic acid (73%) caused significant inhibition (at 5 mM). As in the case of the phosphonoacetate hydrolase (3), P i did not inhibit the reaction. The enzyme had an apparent K m of 0.53 mM for its sole substrate, phosphonopyruvate, and a V max of 202 units per mg of protein.
Nucleotide Sequence Analysis of the pphA Gene-N-terminal amino acid sequences of PPH and its fragments produced by CNBr fragmentation were used to design primers for analysis of the PPH (pphA) locus (Table II). PCR reactions with JQ102 and JQ111 primers yielded a fragment of 300 bp, which was cloned into pGEM-T Easy vector and analyzed. This fragment was subsequently used as a hybridization probe for detection of pphA fragments and clones.
Two clones containing the pphA gene region were identified. These were pPPH5 (which contained a 3.4-kb SacI fragment) and pUPH11 (containing a 3.0-kb SmaI fragment). The nucleotide sequence of the pphA gene was determined. This gene encodes a putative PPH protein of 290 amino acids with an estimated molecular mass of 31,177 Da. This is in good correspondence with the molecular mass obtained for the purified PPH by mass spectrometry. The N-terminal sequences of the PPH protein and the two of its CNBr fragments corresponded to translated amino acid sequences of pphA. The start codon of the pphA gene was unequivocally identified by comparison with the PPH N-terminal amino acid sequence. To further    (18) and homology analyses showed that the PPH protein has similarities with a number of members of the PEP mutase superfamily. The highest scores were with PEP mutases from T. pyriformis (19) (41% amino acid sequence identity) and Mytilus edulis (20, 21) (40% identity). A slightly lower level of identity (38%) was found with the bacterial PEP mutase from Streptomyces viridochromogenes. Even less pronounced similarities were found between PPH and 2-methylisocitrate lyase from Salmonella typhimurium (32% identity) and isocitrate lyase of E. coli (20% identity) (Fig. 3). Alignment of the PPH sequence with several proteins belonging to the PEP mutase superfamily is presented in Fig. 3. The conserved residues in various groups of the superfamily and the active site residues of M. edulis PEP mutase are shown as described by Dunaway-Mariano and co-workers (20). Although there are only moderate levels of similarity between PPH and the PEP mutase superfamily proteins, there is a good conservation of amino acid residues between PPH and PEP mutases in several regions (Fig. 3, positions 72-99, 152-165, 188 -198, 230 -240, 278 -295, and 355-360). It is particularly notable that most of the amino acid residues that comprise the active site of M. edulis PEP mutase (20) are conserved in PPH from Variovorax sp. Pal2.

DISCUSSION
Phosphonopyruvate is produced biogenically through the rearrangement of the high energy phosphate ester phosphoenolpyruvate by PEP mutase, the major C-P bond-forming enzyme identified to date. Thus the compound is central to the formation of all natural phosphonates. In some organisms phosphonopyruvate may subsequently be aminated to phosphonoalanine (22); however, the initial step in all of the major biosynthetic pathways of phosphonate natural products is believed to be its decarboxylation to phosphonoacetaldehyde by phosphonopyruvate decarboxylase (22). Because the equilib-rium of the reaction carried out by PEP mutase lies strongly toward phosphoenolpyruvate this decarboxylation serves to drive the reaction in the direction of C-P bond formation (23). Consistent with this suggestion is the apparent translational coupling of the expression of the PEP mutase and phosphonopyruvate decarboxylase genes observed in the phosphinothricin tripeptide-producing strain S. viridochromogenes (24) and the bialaphos-producing Streptomyces hygroscopicus (19).
The existence of PPH, a novel C-P bond cleavage enzyme of apparently unique substrate specificity whose purification and properties we now report, clearly indicates the possibility of an alternative metabolic fate for phosphonopyruvate; indeed the activity of this enzyme has now been demonstrated by us in cell extracts of a range of environmental bacteria capable of the utilization of phosphonoalanine as sole carbon, nitrogen, and phosphorus source or of phosphonopyruvate itself as sole carbon and phosphorus source (results not shown).  PPH is the third substrate-specific bacterial phosphonohydrolase capable of heterolytic C-P bond cleavage to be described to date. In its requirement for a divalent metal ion it has similarities to many phosphoryl transferring enzymes including the C-P hydrolases specific to phosphonoacetate (3) and phosphonoacetaldehyde (25); the former is zinc-dependent whereas in the latter magnesium is a cofactor in a mechanism that involves both an active site nucleophile and formation of a Schiff-base intermediate (25). PPH is also similar to the other two hydrolases in its homodimeric structure. However, the subunit sizes are different; the phosphonoacetate hydrolase is 30 kDa whereas the phosphonoacetaldehyde hydrolase is 40 kDa.
Analysis of the corresponding sequences of the three enzymes suggests, however, that they have very different ancestries. Thus whereas phosphonoacetaldehyde hydrolase belongs to the haloacid dehalogenase family that includes several phosphohydrolases (26), and phosphonoacetate hydrolase is a member of the alkaline phosphatase superfamily of zinc-dependent hydrolases (27), PPH shows strong homology (Fig. 3) to the members of a recently identified ␣/␤-barrel enzyme superfamily (20) that contains isocitrate lyase and PEP mutase.
Unsurprisingly, in view of their common substrate, PPH showed highest levels of amino acid sequence identity with the PEP mutases; in particular the active site residues identified in the PEP mutase of the marine mussel M. edulis (28) are highly conserved in PPH, as they are in other members of the superfamily (20). These amino acids include the Lys-120 and Asp-58 of the M. edulis enzyme that have been implicated in phosphoryl group activation and transfer (20). Significantly, too, PPH displays a very weak ability to convert phosphonopyruvate to PEP (however it is unable to release P i from the latter).
The fact that the highest degree of homology was found between the PPH of Variovorax sp. Pal2 and the PEP mutases from the lower eukaryotes M. edulis and T. pyriformis would appear to confirm the ancient origin of both enzymes. Further studies should reveal the precise way in which their two active sites have diverged to support their differing roles in phosphonate metabolism. Another intriguing question concerns the interaction of the two activities when they occur in the same cell, whereas the metabolic relationship of each with phosphonopyruvate decarboxylase also requires clarification.