Effects of sulfhydryl regents on the activity of lambda Ser/Thr phosphoprotein phosphatase and inhibition of the enzyme by zinc ion.

Protein Engineering Design and Selection, Dec 1997

Sulfhydryl reagents, such as dithiothreitol (DTT), affected the activity of Ser/Thr phosphoprotein phosphatases. Addition of DTT to the assay buffer increased the affinity of lambda Ser/Thr phosphoprotein phosphatase (lambda-PPase) for its Mn2+ cofactor. On the other hand, the enzyme was found to be inactivated simply by dilution in Tris buffer. The inactivation could be completely prevented by the presence of DTT or Mn2+ in the buffer. Further studies showed that oxidation or reduction of cysteine residues in lambda-PPase may not be the cause of the change in the enzyme activity. Without exception, mutation of all cysteine residues in lambda-PPase to serine did not convert the enzyme into a thiol-insensitive mutant. By careful examination of the effects of different sulfhydryl reagents, metal ion cofactors and substrates on lambda-PPase, it was found that the role of sulfhydryl reagents was the chelation of small amounts of inhibitory metal ions, which were present in plastic laboratory ware, such as disposable cuvets and tubes, with prevention of the enzyme from inactivation. One of the main contaminants found in plastic cuvets was Zn2+, which is a potent inhibitor of lambda-PPase. The inhibition of lambda-PPase by Zn2+ was characterized. Pre-treatment of the enzyme (1-4 nM) with 1 microM of ZnCl2 almost completely inhibited the enzymatic activity in response to 2 mM Mn2+. However, no significant inhibition was found when the enzyme was added to the assay mixture containing 1 microM Zn2+ and 2 mM Mn2+ . This confirms the sensitivity of the holoenzyme to inhibitory metal ions in vitro. The kinetic analysis indicated that the inhibitory metal ion might compete with Mn2+ to bind to the active site of lambda-PPase. This was further supported by the mutation of metal cofactor binding amino acid residues of the enzyme. Mutants which have less affinity for Mn2+ are also less sensitive to Zn2+. Our results suggest that inhibitory metal ions may induce a different structural conformation for lambda-PPase.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

http://peds.oxfordjournals.org/content/10/12/1445.full.pdf

Effects of sulfhydryl regents on the activity of lambda Ser/Thr phosphoprotein phosphatase and inhibition of the enzyme by zinc ion.

Shaoqiu Zhuo 0 1 Jack E.Dixon 1 0 Present address: Chiron Corporation , 4560 Horton Street, Emeryville, CA 94608-2916 , USA 1 Department of Biological Chemistry, University of Michigan Medical School , Ann Arbor, MI 48109-0606 , USA 2To whom correspondence should be addressed Sulfhydryl reagents, such as dithiothreitol (DTT), affected the activity of Ser/Thr phosphoprotein phosphatases. Addition of DTT to the assay buffer increased the affinity of l Ser/Thr phosphoprotein phosphatase (l-PPase) for its Mn21 cofactor. On the other hand, the enzyme was found to be inactivated simply by dilution in Tris buffer. The inactivation could be completely prevented by the presence of DTT or Mn21 in the buffer. Further studies showed that oxidation or reduction of cysteine residues in l-PPase may not be the cause of the change in the enzyme activity. Without exception, mutation of all cysteine residues in l-PPase to serine did not convert the enzyme into a thiolinsensitive mutant. By careful examination of the effects of different sulfhydryl reagents, metal ion cofactors and substrates on l-PPase, it was found that the role of sulfhydryl reagents was the chelation of small amounts of inhibitory metal ions, which were present in plastic laboratory ware, such as disposable cuvets and tubes, with prevention of the enzyme from inactivation. One of the main contaminants found in plastic cuvets was Zn21, which is a potent inhibitor of l-PPase. The inhibition of l-PPase by Zn21 was characterized. Pre-treatment of the enzyme (1-4 nM) with 1 mM of ZnCl2 almost completely inhibited the enzymatic activity in response to 2 mM Mn21. However, no significant inhibition was found when the enzyme was added to the assay mixture containing 1 mM Zn21 and 2 mM Mn21. This confirms the sensitivity of the holoenzyme to inhibitory metal ions in vitro. The kinetic analysis indicated that the inhibitory metal ion might compete with Mn21 to bind to the active site of l-PPase. This was further supported by the mutation of metal cofactor binding amino acid residues of the enzyme. Mutants which have less affinity for Mn21 are also less sensitive to Zn21. Our results suggest that inhibitory metal ions may induce a different structural conformation for l-PPase. - Ser/Thr protein phosphatases (PPases) are a family of phosphoesterases that catalyze phosphoprotein dephosphorylation, mediating a variety of biological processes such as metabolism, cell proliferation and differentiation, gene expression, transport, locomotion and memory (Cohen, 1989; Fischer and Krebs, 1990). The family consists of four main subclasses: type 1 (PPase-1), type 2A (PPase-2A), type 2B (PPase-2B or calcineurin) and type 2C (PPase-2C) (Ballou and Fischer, 1986; Cohen, 1991). The catalytic subunits of the enzymes are structurally related except for PPase-2C (Cohen, 1990). Biochemical studies of both PPase-1 and PPase-2A suggest that oxidation of sulfhydryl groups in the enzymes affects the activity (Ballou and Fischer, 1986). Modification of these two enzymes by reagents such as N-ethylmaleimide (NEM), pchloromercuribenzoate and oxidized glutathione (GSSG) inactivates the activity (Shimazu et al., 1978; Usami et al., 1980; Usui et al., 1983). When the sequences of PPase-1 and PPase-2A are compared, six cysteine residues are shown to be conserved (Zhang et al., 1993). However, when l -PPase, the phage enzyme, is included in the comparison, no invariant cysteines can be found (Cohen and Cohen, 1989). Furthermore, mutagenesis of the six cysteine residues in the catalytic subunit of rabbit muscle PPase-1 has no effect on the catalytic activity of the enzyme (Zhang et al., 1993). Nevertheless, the contribution of sulfhydryl groups in expression of enzymatic activity of Ser/Thr PPases has been unambiguously demonstrated (Ballou and Fischer, 1986). Although obscure, there appears to be a relationship between divalent metal ions and sulfhydryl residues in the activation of PPase-1 (Ballou and Fischer, 1986) and PPase-2A (Yan and Graves, 1982; SchacterNoiman and Chock, 1983; Villa-Moruzzi et al., 1984). The l protein phosphatase (l -PPase) from l bacteriophage is a member of the Ser/Thr protein phosphatase family (Cohen and Cohen, 1989). l -PPase has been shown to share a common catalytic structure and mechanism with the other members of the Ser/Thr PPase family (Zhuo et al., 1993, 1994). There are four cysteine residues in l -PPase. Although these cysteines are at different positions when compared with those in PPase-1 and PPase-2A, the enzyme also exhibits thiol agent-dependent activation in the hydrolysis of p-nitrophenyl phosphate (pNPP). We therefore sought to investigate further the nature of the activation of l -PPase by sulfhydryl reagents. Materials and methods All enzymes for the manipulation of recombinant DNA were from New England Biolabs. Other molecular biology reagents were supplied by Qiagen and Perkin-Elmer. pNPP was purchased from Fluka. All other chemicals were from Aldrich or Sigma. Oligonucleotides were synthesized by Bioservice. Bovine brain calcineurin (PPase-2B) and catalytic subunits of human PPase-1 g and bovine heart PPase-2A were purchased from Boehringer Mannheim. Site-directed mutagenesis Mutants D52N, R53A, R73A, E59Q and E77Q were generated as reported previously (Zhuo et al., 1994). Mutants C24S, C60S, C138S and C208S were generated by polymerase chain reaction (PCR) (Higuchi et al., 1988; Rashtchian et al., 1992). The following sets of primers were used for each specific mutation: 59-GCACGGAAGCTACACG and 59-CGTGTAGCTTCCGTGC (C24S); 59-CGTTGAAAGCCTGGAA and 59TTCCAGGCTTTCAACG (C60S); 59-TGTTATCAGCCACGCC and 59-GGCGTGGCTGATAACA (C138S); 59AGTGTTCAGCGGAAAC and 59-GTTTCCGCTGAACACT (C208S); 59-TTTGATCCATATGCGCTATTACGAA (N-terminal sequence with NdeI digestion site); and 59-TTTGAATTCTCATGCGCCTTCTCC (C-terminal sequence with EcoRI digestion site). The conditions of PCR were described previously (Zhuo et al., 1994). Mutant fragments were amplified with terminal primers paired with each specific mutation primer using wild-type pT77/l PP as template. The fragments were purified by low-melting agarose gel electrophoresis. The excised bands were diluted in 10 volumes of water and 2.5 5.0 m l of each were used as the template for the second PCR. Each pair of mutant fragments was added to the reaction mixture and terminal primers were used to amplify the fulllength mutants. The products of the PCR were purified by phenolchloroform extraction and ethanol precipitation and then digested with NdeI and EcoRI. The digested fragments were ligated into a pT77 vector (Tabor and Richardson, 1985). DNA sequencing was performed on a Pharmacia-LKB ALF automated DNA sequencer using sequencing kits supplied by the manufacturer. Double mutant 2C-Mut (C60S/C138S) was generated by ligation of the NdeI/VspI-digested fragment from C60S mutant and the VspI/EcoRI-digested fragment from C138S mutant. The triple mutant 3C-Mut (C60S/C138S/ C208S) was generated from the NdeI/AfiIII fragment of the 2C-Mut mutant and the AfiIII/EcoRI fragment of the C208S mutant. The all cysteine-to-serine mutant 4C-Mut (C24S/C60S/ C138S/C208S) was generated by PCR with C24S primers using pT77/3C-Mut as template. Confirmed mutants were expressed and purified as described previously (Zhuo et al., 1994). Purified enzymes contained no endogenous Mn21 and Ni21 (typical samples: Ni21 ,0.14 m g per mg enzyme and Mn21 , 0.018 m g per mg enzyme) as assayed by inductively coupled plasma atomic emission spectrometry (Leeman Labs). Phosphatase activity assay Activity of l -PPase and the mutants was assayed as reported previously with pNPP as substrate (Zhuo et al., 1993). Typical assay buffer contained 50 mM TrisHCl, pH 7.67.8, with 2 mM MnCl2 and 20 mM pNPP. Kinetic data were processed by a statistical program written and described by Brooks (1992). Inductively coupled plasma mass spectrometry Samples were prepared by extracting each 1.5 ml disposable plastic (methacrylate) cuvet with 1 ml of Milli-Q water. A 20 ml volume of the extract were collected in a Falcon 50 ml conical tube which was extensively rinsed with Milli-Q water. Analyses were performed by Balazs Analytical Laboratory (Sunnyvale, CA). Activation of l -PPase by dithiothreitol (DTT) We found that DTT was able to accelerate the reaction rate of l -PPase in hydrolysis of the low molecular weight substrate pNPP. The acceleration was only observed at concentrations of Mn21 from 0.005 to 0.5 mM (Figure 1). With concentrations of Mn21 under 0.005 and above 1 mM, no effect of DTT on enzyme activity was observed. The activation by DTT was also dependent on the time that the enzyme had been stored, with fresh enzyme being more responsive to DTT. An Eadie Hofstee plot reveals that the enzyme has a slightly different Fig. 1. Activation of l -PPase by DTT. Purified metal ion-free l -PPase was assayed at 30C in 50 mM triethanolamineHCl, pH 7.8, containing 25 mM pNPP at various concentrations of MnCl2, with and without 10 mM DTT. For the inset, the data were plotted by the EadieHofstee method and subjected to linear regression analysis. Kd values were determined to be 4.5 m M with DTT and 16 m M without DTT. affinity for Mn21 in the presence and absence of DTT (Figure 1, inset). In the presence of DTT, the apparent Kd for Mn21 was found to be 4.5 m M; in the absence of DTT, the apparent Kd was 15.6 m M. At high concentrations of Mn21, DTT no longer affected the enzyme activity. Inactivation of l -PPase by dilution and the effects of Mn21, sulfhydryl agents, pNPP, glycerol and bovine serum albumin (BSA) on the inactivation Purified l -PPase is unstable at room temperature (t1/2 120 min, data not shown). However, enzyme solutions at concentrations of ~0.1 mg/ml or higher can be stabilized for up to 48 h at room temperature by the addition of 2040% glycerol (data not shown). When diluted into 50 mM Tris HCl buffer (final concentration ~30100 ng/ml) at pH 7.67.8 (the optimum catalytic pH for l -PPase when Mn21 is cofactor), the enzyme lost the pNPP activity in seconds (Figure 2A). The inactivation could be completely prevented by diluting the enzyme in the same buffer containing 200 m M Mn21 or 10 mM DTT. Figure 2B shows that different activities of l -PPase are obtained when the enzyme is added to the assay buffer in a different order. The buffer has to contain Mn21 or DTT before the enzyme can be added. If the enzyme is added to the buffer before other components, 98% of the activity is lost. Addition of 10 mM DTT after the enzyme is diluted can only rescue part of the activity. Continuous incubation of the inactivated enzyme with D-cysteine resulted in gradual recovery of the enzyme activity (Figure 2C). The effects of other sulfhydryl reagents were also examined (Figure 3A). Both Dand L-cysteine and DTT give the best protection with an apparent Kd in the micromolar concentration range. The Fig. 2. (A) Inactivation of l -PPase by dilution in TrisHCl buffer. A 1 m l volume of l -PPase (0.072 mg/ml) in 40% glycerol was diluted into 0.95 ml of 50 mM TrisHCl (pH 7.8) and mixed at 28C. At the time indicated, the activity was assayed by addition of pNPP and MnCl2 to final concentrations of 30 mM and 200 m M, respectively (j). For the control (d), the enzyme was incubated in the same TrisHCl buffer under the same conditions without dilution. (B) Protection of l -PPase from dilution inactivation by Mn21 and DTT. The enzyme activity was assayed in 50 mM TrisHCl, pH 7.8, containing 30 mM pNPP, 50 m M MnCl2 and 10 mM DTT. The orders of addition of components were as follows: I, DTT, l -PPase, Mn21 and pNPP; II, Mn21, l -PPase, DTT and pNPP; III, l -PPase, Mn21, pNPP and DTT; IV, l -PPase, Mn21 and pNPP. The time interval between the addition of components was 515 s. (C) Reactivation of diluted l -PPase by D-cysteine. l -PPase was inactivated by dilution in TrisHCl buffer for 1 min and then incubated with 1 mM D-cysteine. 25 mM pNPP and 2 mM MnCl2 were added at the time indicated to assess the enzyme activity. The activity of the untreated enzyme was set to 100%. Fig. 3. Protection of l -PPase from dilute inactivation. (A) Protection by sulfhydryl reagents. Assays were carried out in 1.0 ml of 50 mM TrisHCl, pH 7.62, containing 25 mM pNPP, 2 mM MnCl2 and 10 mM of sulfhydryl reagents. A 50 ng amount of l -PPase was incubated in 1 ml of TrisHCl buffer containing sulfhydryl reagents as indicated for 1 min before pNPP, Mn21 and additional sulfhydryl reagents were added. The activity of the enzyme was compared with that of the unincubated l -PPase. (B) Protection by pNPP. Assays were carried out in 1 ml of 50 mM TrisHCl, pH 7.62, containing 25 mM pNPP and 2 mM MnCl2. A 50 ng amount of l -PPase was incubated in 1.0 ml of TrisHCl buffer containing pNPP or NaCl as indicated for 1 min before the activity was determined. (C) Protection by BSA. Enzyme was assayed as in (A) without sulfhydryl reagents. A 50 ng amount of l -PPase was incubated in 1 ml of Tris HCl buffer containing BSA as indicated for 1 min before the substrate and MnCl2 were added and activity was assayed. apparent Kd of glutathione is in the low millimolar concentration range. b -Mercaptoethanol is ~5001000-fold less effective than cysteine. The effects of the sulfhydryl reagents in reactivation of the inactivated l -PPase are cysteine . DTT . glutathione . b -mercaptoethanol (results not shown). The results indicate that the structures of sulfhydryl reagents are important for the mechanism of function. The inactivation of l -PPase in Tris buffer can also be prevented by binding of substrate. pNPP slightly protects l -PPase from inactivation by dilution in 50 mM Tris buffer in a dose-dependent manner (Figure 3B) with an apparent Kd . 30 mM, which is much higher than the Km for hydrolysis. The protection is not due to the change of the ionic strength of the buffer because no effect is observed when NaCl is added up to 120 mM. We further examined the effects of glycerol and protein concentrations on the stability of l -PPase. Glycerol did not protect l -PPase from the dilution inactivation at concentrations up to 40% (v/v) (data not shown). BSA has frequently been used to stabilize enzyme activity by preventing protein denaturation at the liquidair interface. BSA protects the activity of l -PPase upon dilution with an apparent Kd of about Fig. 4. Effects of D-cysteine on the activity of 4C-Mut mutant. (A) Enzyme was assayed in 1 ml of 50 mM TrisHCl, pH 7.62, containing 25 mM pNPP and different concentrations of MnCl2 in the presence and absence of 1 mM D-cysteine. (B) Assays were carried out in 1.0 ml of 50 mM TrisHCl, pH 7.62, containing 25 mM pNPP, 2 mM MnCl2 and 1 mM D-cysteine. Different activities were obtained when the enzyme was preincubated with 1 mM D-cysteine in 1.0 ml of the TrisHCl buffer for different times as indicated. 5 mg/ml, which is too high for its expected function (Figure 3C). However, the effect of BSA could be greatly enhanced when 40% glycerol was included (Figure 3C). This is not due to the reduction of dielectric constant of the solvent, since no enhancing effects were observed when 20% propan-2-ol was used. Mutagenesis of cysteine residues in l -PPase The effects of sulfhydryl reagents on enzyme activity have been reported previously for PPase-1 (Ballou et al., 1985). When the catalytic subunit of PPase-1 was purified in the absence of disulfide reducing reagents, it was inactive and unresponsive to Mn21 treatment. Its reactivation by disulfide reducing reagents was time (t1/2 1 min), temperature (30C), concentration (50 mM DTT) and pH dependent. Maximum activity was obtained when the enzyme was incubated with Mn21 together with the disulfide reducing agents. It was assumed that the sulfhydryl groups of the enzyme were involved in this activity expression (Ballou et al., 1985). Interestingly, the recently published three-dimensional structure of the catalytic subunit of PPase-1 determined by x-ray crystallography indicated that Cys127 in the crystals of the bacterially expressed enzyme was oxidized into a terminal sulfinyl or sulfonyl group (Goldberg et al., 1995). To study further the mechanism of cysteine residues in the activation of l -PPase, we constructed l -PPase mutants with cysteine to serine conversion and expressed the enzymes in Escherichia coli. To our surprise, all of these mutants still responded to the dilution inactivation of l -PPase in Tris buffer and the protection by sulfhydryl agents (results not shown). Characterization of a mutant of l -PPase, in which all cysteine residues were mutated to serine (4C-Mut), showed that this mutant was similar to the wild-type enzyme in its response to D-cysteine (Figure 4). The results argue that the effects of cysteine are not due to the reduction of disulfide bonds or oxidized forms of cysteine in the protein. It is also unlikely that the role of cysteine is to prevent the enzyme from oxidation or modification by free radicals, since ascorbic acid, NADH, NADPH and catalase do not have any effects on the protection of the enzyme from the dilution inactivation (results not shown). SMethyl-L-cysteine, L-cystine and L-serine were also found to have no or only very small effects on the activation or inactivation of the enzyme (results not shown). Protection of l -PPase activity from dilution inactivation by EDTA and identification of the inhibitory source Our results clearly indicate that the inactivation of l -PPase in Tris buffer is not caused by oxidation or disulfide bond formation among the cysteine residues in the protein. Denaturation of the protein structure due to dilution is also unlikely because protection by BSA requires very high concentrations. By careful examination of the sulfhydryl reagents which protect the enzyme from the inactivation, it can be seen that the best protection is offered by the best metal ion chelators, not the best reducing agents. Therefore, it is very possible that the effect of sulfhydryl reagents on l -PPase is to chelate an inhibitory metal ion which competes with Mn21 to bind to the catalytic site of l -PPase. If that is the case, other non-thiol chelating agents should also be able to protect l -PPase from the inactivation. Indeed, Figure 5 shows that the enzyme activity can be completely protected by the inclusion of as little as 2.5 m M EDTA in the incubation buffer. This confirms that the dilution inactivation of l -PPase is due to contamination by inhibitory metal ions in the dilution buffer. To identify the Fig. 5. Protection of l -PPase from dilution inactivation by EDTA. The enzyme activity was assayed in 50 mM TrisHCl, pH 7.7, containing 20 mM pNPP, 2 mM MnCl2 and 10 mM DTT. I: l -PPase was pre-incubated with buffer only in the assay cuvet for 1 min before Mn21, pNPP and DTT were added. II: No pre-incubation. III: l -PPase was pre-incubated for 1 min with 2.5 m M EDTA before Mn21, pNPP and DTT were added. source of the contamination, the dilution buffer was circulated in an iminodiacetic acid (IDA) or AG-MP-50 ion-exchange column (Bio-Rad) for several hours to eliminate or reduce metal ion contamination. However, only a small decrease in the inactivation effects was observed (results not shown). Further experiments indicated that the main source of the contamination was the disposable methacrylate or polystyrene cuvets, possibly due to the residues from metal molds or the leaching of metal fillings, since a much smaller inactivation effect was found if glass cuvets were used (results not shown). Analysis of a Milli-Q water extract of the plastic cuvets by inductively coupled plasma mass spectrometry (ICP-MS) scanning (68 common metals) revealed that the cuvets contained metal ions which could be recovered at nanomolar to micromolar concentrations (Table I). Among these metal ions, zinc was shown to be a potent inhibitor of l -PPase (Zhuo et al., 1993). Since the enzyme concentrations were 14 nM under the assay conditions, it is reasonable to conclude that these trace metal ions cause the inactivation of l -PPase in the incubation. Characterization of inhibition of l -PPase by Zn21 Many d-block transition metal ions have been shown to inhibit l -PPase (Zhuo et al., 1993). Sc21, Yb31, Cu21, Zn21 and Hg21 were found to be most potent. Zn21 also inhibited the Mn21-stimulated activity of PPase-1 and PPase-2B (Chernoff et al., 1984; Chan et al., 1986; Pallen and Wang, 1986). In the plastic cuvet, the Zn21 concentration was found to be as high as 0.6 m M in 1 ml of assay buffer (Table I). To understand the roles of metal ions in the catalysis and inhibition, we further studied the inhibition of l -PPase and other PPases by Zn21. Figure 6A shows that hydrolysis of pNPP by l -PPase in the presence of 2 mM Mn21 is almost completely inhibited by 30 m M of Zn21. Addition of DTT or b -mercaptoethanol 17 230 0.82 5.7 5.6 21 4.3 4.4 100 830 37 5.0 Twenty plastic cuvets (Fisher 14-385-938) were extracted with 1 ml of Milli-Q water each. The extracts were combined and diluted to 100 ml before being subjected to analysis. The experiments were performed by Balazs Analytical Laboratory (Sunnyvale, CA). Only elements with concentration 0.01 m M are presented. prevented the inhibition. However, when the enzyme was preincubated with Zn21 at a concentration as low as 1 m M before addition of Mn21, the activity was almost completely suppressed (Table II). The enzyme no longer responded to 2 mM Mn21 treatment. Incubation with DTT slowly and partially produced a recovery of the activity (results not shown). This is similar to the dilution inactivation. Table II shows the inactivation of four different PPases by pre-incubation with 1 m M Zn21. As a comparison, no significant inhibition was found when the enzymes were added to the mixture of 1 m M Zn21 and 2 mM Mn21. This suggests that once Zn21 is bound in the pre-incubation, it does not exchange with Mn21 in the assay buffer. On the other hand, pre-incubation of l -PPase with Mn21 did not prevent enzyme inhibition by Zn21. Addition of 20 m M Zn21 to the dephosphorylation reaction activated by 2 mM Mn21 immediately reduced the rate of the catalysis (Figure 6B). The inhibition of l -PPase by Zn21 was demonstrated to be competitive with Mn21 (Figure 7A) with an apparent Ki of ~0.6 m M at pH 7.8. This suggests that Zn21 and Mn21 may bind to the same sites on the active center. Similar conclusions can be drawn from study of the inhibition of l -PPase mutants by Zn21. Ligand residues of l -PPase which are involved in chelation of metal ions have been identified by mutagenesis (Zhuo et al., 1994) and confirmed by the crystal structures of PPase-1 and PPase-2B (Egloff et al., 1995; Goldberg et al., 1995; Griffith et al., 1995). Mutants which affected binding of Mn21 were also found to be less inhibited by Zn21 except E77Q (Table III). However, Glu77 does not directly chelate metal ions according to the crystal structures of PPase-1 and PPase-2B (Egloff et al., 1995; Goldberg et al., 1995; Griffith et al., 1995). Mutation of this residue favors the binding of Zn21 and decreases the affinity for Mn21. The inhibition of l -PPase by Zn21 is also competitive with pNPP (Figure 7B) with an apparent Ki of 12 m M. Discussion Thiol reagents were found to modulate the activities of many Ser/Thr phosphoprotein phosphatases (Ballou and Fischer, 1986, and references therein). Similar effects were also observed for l -PPase. It was unclear how the oxidation of the cysteine residues in the enzymes changed its activity since the Fig. 6. Inhibition of l -PPase by ZnCl2. (A) The enzyme was assayed in 50 mM TrisHCl, pH 7.8, containing 2 mM MnCl2 and 20 mM pNPP at 30C in the absence (1) or presence of 10 mM DTT (m) or b -mercaptoethanol (d). (B) 20 m M (final concentration) ZnCl2 was added to the dephosphorylation reaction mixture of 20 mU of l -PPase, 20 mM pNPP and 2 mM Mn21 in 50 mM TrisHCl, pH 7.7, at the time indicated. The broken line is the control reaction with the addition of buffer only. The dashed line represents the activity when l -PPase was added to the assay mixture containing 20 m M ZnCl2. 22 5 2.5 4 For co-incubation, enzymes were incubated in 50 mM TrisHCl, pH 7.65, treated with metal ion chelating resin for 1 min at 25C. 20 mM pNPP, 2 mM MnCl2 and 1 m M ZnSO4 (final concentrations) were added to start the reaction. For pre-incubation, enzymes were incubated with 1 m M ZnSO4 in 50 mM TrisHCl, pH 7.65, for 1 min at 25C. 2 mM MnCl2 and 20 mM pNPP were added to start the reaction. Assays were carried out as described in Experimental. aOne unit is the amount of enzyme which catalyzes the conversion of 1 m mol of pNPP to p-nitrophenol and inorganic phosphate in 1 min at 30C. bOwing to the enzyme storage conditions, the assay buffer contained 0.5 m M of EGTA. conditions used to reactivate the enzymes were unusual (Ballou and Fischer, 1986, and references therein). Modification with thiol-reactive compounds did inactivate the enzyme (Ballou and Fischer, 1986, and references therein; S.Zhuo, unpublished results). However, mutagenesis of the cysteine residues did not affect the catalysis (Zhang et al., 1993). The effects of thiol reagents also depended on the concentrations of metal ion cofactor. All these observations were contradictory and confusing. Our discovery of the sensitivity of l -PPase to inhibitory metal ions explains well the relationship between thiol reagents and metal ion cofactors. The role of thiol reagents is to chelate the inhibitory metal ions and prevent the inhibition. At high concentrations of Mn21 or Ni21, since the ratio of the cofactor to the inhibitory ion is so high, the effects of thiol reagents greatly decrease. At least for l -PPase, contamination by inhibitory metal ions is the main cause of the decrease in enzymatic activity. Addition of metal ion chelators to the enzyme stock solution will stabilize the activity. Another interesting question is why and how metal ions such as zinc inhibit PPases. Both the kinetic data and the mutagenesis results indicate that Zn21 binds to the same site on l -PPase as Mn21. If the catalytic role of the metal ions in PPases is an electrostatic interaction, it should be insensitive to the nature of the metals (Goldberg et al., 1995). It has been suggested that metal ions make a phosphate ester more susceptible to nucleophilic attack and stabilize the transition state or intermediate (Goldberg et al., 1995). This model has some difficulty in explaining the observation that many metal ions, such as Zn21, Hg21 and Cu21, which possibly bind to the same sites as Mn21, are potent inhibitors of PPases (Cohen, 1991; Zhuo et al., 1993). Cu21 and Zn21 have much higher Lewis-acid strength than Ni21 and Mn21 (Frau sto da Silva and Williams, 1991) and should, therefore, be more powerful in catalysis. What is common between Ni21 and Mn21 is that their preferred geometry is octahedral (Frau sto da Silva and Williams, 1991). The other octahedral ions include Co21, Co31, Fe31 and Cr31. Co21 is also an activator of PPase-1, PPase-2A and PPase-2B (reviewed by Cohen, 1991) although it activates l -PPase much less (Zhuo et al., 1993). Fe31 has been shown to be present in the active sites of the crystal structures of PPase-2B (Goldberg et al., 1995). The preferred geometry for Zn21 is tetrahedral, which creates less ligand coordination than octahedral ions (Frau sto da Silva and Williams, 1991). Therefore, the function of these metal ions may result from a combination of their charge, geometry and ionic radii. The ionic radii for Ni21 and Mn21 are about 70 Fig. 7. Kinetic characterization of inhibition of l -PPase by ZnCl2. (A) Competitive inhibition of l -PPase by Zn21 towards MnCl2. The assays were carried out in 50 mM TrisHCl, pH 7.8, with 20 mM pNPP as substrate in the presence of 0 (1), 0.5 (m), 2 (d) and 4 (.) m M ZnCl2 at 30C. The results are presented as LineweaverBurk plots. Ki for ZnCl2 is calculated to be 0.6 6 0.2 m M. (B) Competitive inhibition of l -PPase towards pNPP. The enzyme was assayed as in (A) with addition of 2 mM MnCl2 in the presence of 0 (1), 4 (.), 8 (j), 11 (r) and 20 (m) m M ZnCl2. pNPP concentrations range from 1 to 100 mM. The results are presented as Lineweaver-Burk plots. Ki for ZnCl2 is calculated to be 12 6 4 m M. and 80 pm, respectively (Frau sto da Silva and Williams, 1991). The latter has fewer 3d electrons (five) and binds to l -PPase more tightly (Zhuo et al., 1994) than the former (eight 3d electrons). Zn21 does not have any reactive 3d shell. Inhibition of PPases by Zn21 could be explained by the fact that Zn21 has a different geometry and ligand coordination. The importance of the geometry of ligand coordination can also be seen by examination of the l -PPase mutants. Mutation of the metal ion chelating ligands of l -PPase does not drastically decrease the binding affinity of Mn21 (Zhuo et al., 1994). The destruction of the enzyme activity must be due to the distortion of the coordination geometry of the metal ions. The inhibition by Zn21 also is competitive with the artificial substrate, pNPP. This double competitive inhibition could suggest that the metal cofactor might participate in substrate binding and Zn21 inhibits the enzyme by preventing the substrate binding. Also, binding of pNPP was shown to protect the enzyme from the inactivation by pre-incubation with Zn21. However, it could be the result of chelation of the metal ion by the phosphate substrate although monophosphate has very low affinity for metal ions. This theory requires the scrutiny of further experiments. Under the experimental conditions, it seems that l -PPase does not discriminate between the cofactors and the inhibitors. The Kd (4.5 m M) of Mn21 and the Ki (0.6 m M) of Zn21 are only 7.5-fold different. However, if the holoenzyme is exposed to Zn21 before Mn21, it is switched to a conformation which no longer responds to Mn21. The crystal structures of PPase-1 and PPase-2B reveal two metal ions bound at the active site, which is analogous to that in the DNA exonucleases and other phosphotransferases (Goldberg et al., 1995; Griffith et al., 1995). One hypothesis is that if both metal ion binding sites on the PPases are occupied by Zn21, the enzyme will change its conformation and no longer respond to Mn21. Incubation with metal ion chelators will slowly switch the conformation and reactivate the enzyme. If the enzyme is exposed to Zn21 and Mn21 together, then the chance for binding of Zn21 to both sites is much lower because of the concentration ratio and, hence, the inhibitory effect is much smaller. Zn21 is required for catalytic activity for some alkaline and acid phosphatases. It forms a binuclear Fe31Zn21 active center in kidney bean purple acid phosphatases (Strater et al., 1995). The Ser/Thr phosphoprotein phosphatases also have similar active centers (Egloff et al., 1995; Goldberg et al., 1995; Griffith et al., 1995). Furthermore, in the crystal structure of PPase-2B, one of the enzyme-bound metal ions was identified to be Zn21 (Griffith et al., 1995). The concentration of Zn21 to inactivate PPases in vitro is in the physiological range. This raises the possibility that Zn21 could be a regulator for PPases in vivo. Further physiological and cell biological studies may reveal the importance of Zn21 in the activity of PPases. Acknowledgments The authors are grateful to Professor Jack Kyte (University of California, San Diego) for consultation and advice. They also thank Drs Frank Masiarz, Lawrence Cousens and Frank Marcus (Chiron Corporation) for critical reading and helpful discussion of the manuscript.


This is a preview of a remote PDF: http://peds.oxfordjournals.org/content/10/12/1445.full.pdf

S Zhuo, J E Dixon. Effects of sulfhydryl regents on the activity of lambda Ser/Thr phosphoprotein phosphatase and inhibition of the enzyme by zinc ion., Protein Engineering Design and Selection, 1997, 1445-1452, DOI: 10.1093/protein/10.12.1445