Diesterase activity and substrate binding in purple acid phosphatases.

Purple acid phosphatases (PAPs) are dinuclear monoesterases1,2 that are structurally diverse but with a conserved set of ligands to the Fe3+-M2+ metal centers (M2+ ) Fe, Zn, or Mn).3 The identity of the nucleophile and the substrate binding mode have been matters of controversy (Figure 1).1,3-5 In one proposal, substrate binds to the divalent metal and is attacked by a terminal Fe3+-bound hydroxide. In an alternative mechanism, substrate coordinates first to the divalent metal and then forms a bridging complex, followed by nucleophilic attack by the μ-hydroxide. We report that the PAPs from pig and kidney bean catalyze the hydrolysis of diesters if the second ester group is small, and the kinetics of the reaction with methyl p-nitrophenylphosphate suggest that a terminal-bound hydroxide is the nucleophile for the diester, followed by attack of the bridging hydroxide upon the resulting monoester without release into solution. The PAPs are reported to lack the ability to hydrolyze diesters, based on the lack of activity with bis(p-nitrophenyl)phosphate (BpNPP).6 In contrast, model complexes designed to mimic the PAP metal center hydrolyze diesters, while monoesterase activity has not been reported.7 This suggests that the inability of PAP to hydrolyze BpNPP might result from steric considerations. Thus, we tested the PAPs from pig (also termed uteroferrin) and from red kidney bean using two diesters bearing smaller second ester groups, methyl pNPP (MpNPP) and ethyl pNPP (EpNPP). While BpNPP gave negligible turnover, consistent with a previous report,6 the other two diesters, particularly MpNPP, are substrates although binding is poor. The pH dependence for kcat and kcat/Km of the reactions with MpNPP was followed by monitoring release of p-nitrophenol (see Supporting Information, Figure S1). The pH optima for monoesterase and diesterase reactions were the same; data in Table 1 were obtained at the optimal pH of 4.9 for pig PAP and pH 6.25 for kbPAP (see Supporting Information for EpNPP and BpNPP data). At its pH optimum, the kcat for MpNPP hydrolysis by pig PAP is close to that for the monoester pNPP,5,8 while for kbPAP, the MpNPP kcat is about half that for pNPP. The high diester Km values presumably result from steric requirements of the second ester group.14 For comparison, Km values of aryl and benzyl monoesters range from 1.25 to 5.8 mM.5 Inorganic phosphate is a competitive inhibitor of the reactions of both PAPs with MpNPP, with Ki values close to the literature values for the reactions with pNPP (Table 2). The reaction of MpNPP is assumed to first yield p-nitrophenol and methyl phosphate, on the basis of the large difference in the pKa values of the two potential leaving groups and the observation that PAP catalysis shows a significant dependence on the leaving group (âlg ) -0.4 and -0.6 for pig and kbPAP, respectively).5 Methyl phosphate (MP), the presumed product of the diester reaction, is a substrate, with kcat values of 17 and 20 s-1 for the pig and kbPAPs respectively, with surprisingly high Km values (Table 1). MP is a weak competitive inhibitor of the PAP-catalyzed hydrolysis of pNPP and of MpNPP, with a Ki of ∼100 mM (Table 2). These Ki values confirm the unexpectedly weak binding of MP. As noted above, previously characterized substrates with low millimolar Km values are aryl and benzyl esters. It may be that the PAP active site region recognizes the aromatic moiety in these substrates. Studies of PAP inhibitors have shown that aromatic groups are important in achieving tight binding.8 In the MpNPP reactions, the rates of formation of p-nitrophenol and of inorganic phosphate decreased significantly after only a few percent of the diester was consumed (see Figure 2 and Supporting Information). Addition of pNPP immediately resulted in rapid production of p-nitrophenol, showing that this was not due to loss of enzymatic activity (see Supporting Information for details). The low Ki for inorganic phosphate and high Km of the diester leads to the expectation of significant product inhibition. This was confirmed by an experiment in which enzyme was preincubated with an amount of inorganic phosphate corresponding to that produced after the first 5% of reaction. Addition of MpNPP substrate gave rise to the same slow rate observed after ∼5% had been turned over in the experiments without preincubation with phosphate. † Utah State University. § University of Queensland. Figure 1. Schematic diagram of two potential binding modes of a phosphate ester to the metal center of PAPs. In A, the nucleophile is the Fe(III)coordinated hydroxide; in B, it is a bridging μ-hydroxide.