The authentic PPase was purified from B subtilis
The authentic PPase was purified from B. subtilis as described previously , whereas the recombinant enzyme was purified to homogeneity from E. coli by a simplified procedure including only phenyl-Sepharose CL-4B column chromatography and DEAE-HPLC . About 26 mg of pure enzyme was obtained from 12 g of cell paste with a yield of 45%. To stabilize the enzyme, 1.5 mM MnCl2 was included in dub inhibitor at all purification steps. The purified enzyme gave a single band on SDS-PAGE gels stained with Coomassie brilliant blue . The N-terminal sequence of the authentic enzyme was determined by an automatic Edman degradation method with an Applied Biosystem 477A/120A gas-phase sequenator.
PPase activity was assayed at 37°C as described previously . The assay medium contained 1 mM PPi, 2 mM MgCl2 and 20 mM Tris-HCl buffer (pH 7.3). One unit of activity corresponds to 1 μmol of PPi converted per minute. Protein concentration was determined with a Pierce BCA protein assay kit, with BSA as standard.
Results and discussion Soluble PPase isolated from B. subtilis was purified to homogeneity and its N-terminal residues 1–39 were determined by automatic Edman degradation method giving the sequence: 1-M-E-K-I-L-I-F-G-H-Q-N-P-D-T-D-T-I-X-S-A-I-A-Y-A-D-L-K-N-K-L-G-F-N-A-E-P-V-R-L-39, where X denotes an unidentified residue. By searching the Swiss-Prot database, this sequence was found to be identical with an N-terminal gene-deduced sequence of a hypothetical 34-kDa protein encoded by a non-identified open reading frame located in the COTF-TETB intergenic region of B. subtilis. This open reading frame was amplified by PCR and its sequence was verified to be identical to that reported previously . By its size, the hypothetical protein is similar to B. subtilis PPase (34–36 kDa ). The PCR product was expressed under T7 promoter in E. coli, yielding transformants with 180-fold higher PPase activity than the host strain, clearly indicating that the open reading frame encodes a PPase. Like authentic B. subtilis PPase , the recombinant enzyme purified to homogeneity was activated by preincubation with Mn2+ and Co2+ (Fig. 1A), but not with Mg2+, Ca2+, Sr2+, Cd2+, Cu2+, Fe2+ or Ni2+. Full activation was obtained in 2 min of incubation. The activation was strongly dependent on metal ion concentration (Fig. 1A) and pH (Fig. 1B). Maximal activation was observed at pH 8.5 with 1.5 mM Mn2+ (Fig. 1). At pH>9, manganese hydroxide precipitated, limiting our ability to study the activation at high pH values. The specific activity of the enzyme at pH 7.3 was 323 and 9900 U/mg before and after the activation by 1.5 mM Mn2+, corresponding to the kcat values of 180 and 5500 s−1, respectively. Interestingly, the former kcat value is about the same as those for E. coli and yeast PPases 11, 12, whereas the latter is greater by an order of magnitude. According to its amino acid sequence, B. subtilis PPase is unique among known soluble PPases. There are 31 soluble PPase sequences currently available in the GenBank: prokaryotic PPases have 164–233 amino acid residues per subunit, whereas eukaryotic PPases have 211–310 residues per subunit. In this respect, B. subtilis PPase, with 309 amino acid residues per subunit, resembles eukaryotic PPases. Furthermore, B. subtilis PPase shows little sequence similarity to other soluble PPases, typical examples of which are shown in Fig. 2. Sequence identity between PPase of B. subtilis and E. coli is as low as 17%, whereas internal identity of the 21 available prokaryotic PPase sequences varies from 31% (Haemophilus influenzae vs. E. coli) to 61% (Legionella pneumophila vs. E. coli). Fan rom evolutionary point of view, it is interesting to note that Bacillus stearothermophilus, a close relative of B. subtilis, has a PPase which is very similar to other soluble PPases , but completely different from B. subtilis PPase (Fig. 2): identities of B. stearothermophilus PPase vs. E. coli PPase and B. subtilis PPase are 44 and 15%, respectively.