- Email: [email protected]
Rapid and Irreversible Inactivation of Protein Tyrosine Phosphatases PTP1B, CD45, and LAR by Peroxynitrite
Archives of Biochemistry and Biophysics Vol. 369, No. 2, September 15, pp. 197–207, 1999 Article ID abbi.1999.1374, available online at http://www.idealibrary.com on
Rapid and Irreversible Inactivation of Protein Tyrosine Phosphatases PTP1B, CD45, and LAR by Peroxynitrite 1 Ko Takakura,* Joseph S. Beckman,† ,‡ ,§ Lee Ann MacMillan-Crow, ¶ and John P. Crow† ,§ ,\ ,2 *Department of Anesthesiology and Reanimatology, Fukui Medical University, Fukui, Japan; and †Department of Anesthesiology, ‡Department of Cell and Molecular Biology, ¶Department of Surgery, \Department of Pharmacology/Toxicology, and §The Center for Free Radical Biology at UAB, University of Alabama at Birmingham, Birmingham, Alabama 35233
Received April 2, 1999, and in revised form July 1, 1999
Protein tyrosine phosphatases (PTPs) contain an essential thiol in the active site which may be susceptible to attack by nitric oxide-derived biological oxidants. We assessed the effects of peroxynitrite, nitric oxide, and S-nitrosoglutathione on the activity of three human tyrosine phosphatases in vitro. The receptor-like T-cell tyrosine phosphatase (CD45), the non-receptor-like tyrosine phosphatase PTP1B, and leukocyte–antigen-related (LAR) phosphatase were all irreversibly inactivated by peroxynitrite in less than 1 s with IC 50 values of <0.9 mM. PTP inactivation was also seen with equivalent concentrations of peroxynitrite generated by SIN-1, indicating that bolus peroxynitrite and cogeneration of superoxide and nitric oxide were equipotent. Rate constants for peroxynitrite-mediated PTP inactivation were determined by competition with cysteine and were among the fastest rates yet seen for reaction of peroxynitrite with any biological molecules. The bimolecular reaction rates for CD45, LAR, and PTP1B were 2.0 3 10 8, 2.3 3 10 7, and 2.2 3 10 7 M 21 s 21, respectively. Inactivation by peroxynitrite was essentially irreversible as incubation with dithiothreitol (DTT) restored less than 10% of the original phosphatase activity. Prolonged treatment with 0.4 mM DETA NONOate, which generated a steady-state concentration of 2 mM nitric oxide, was only slightly inhibitory. S-Nitrosoglutathione (1.0 mM) 1 This study was supported by grants from the Ministry of Education, Science, Sports, and Culture of Japan; The American Heart Association (LAMC); the National Institutes of Health (HL58209 and NS35871); and the Amyotrophic Lateral Sclerosis Association (J.S.B. and J.P.C.). 2 To whom correspondence should be addressed at Department of Pharmacology/Toxicology and Center for Free Radical Biology, University of Alabama at Birmingham, ZRB 906, 1900 University Boulevard, Birmingham, AL 35233. Fax: 205-934-7437. E-mail: [email protected].
0003-9861/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
inhibited PTPs by approximately 50% after 30 min and the inhibition was completely reversed by DTT. Nitrotyrosine immunoblots of peroxynitrite-treated PTP1B revealed that peroxynitrite completely inactivated PTP1B prior to the appearance of protein tyrosine nitration. Peroxynitrite anion is structurally similar to phosphate anion both in terms of molecular diameter and charge. Thus, the extreme vulnerability of these PTPs to peroxynitrite-mediated inactivation is consistent with attraction of peroxynitrite anion to the active site and subsequent oxidation of the essential thiolate. These findings suggest that any PTP possessing the CXXXXXR active-site sequence could potentially be inactivated by peroxynitrite in vivo resulting in a net increase in tyrosine phosphorylation and profound effects on phosphotyrosine-dependent signaling cascades. © 1999 Academic Press Key Words: peroxynitrite; phosphatase; nitrotyrosine; phosphotyrosine; nitration; nitric oxide; nitrosothiol; thiolate; carbon dioxide; rate constants.
Tyrosine phosphorylation plays a critical role in controlling cell growth, proliferation, and differentiation (1–3) and is regulated by a careful balance between the activities of tyrosine kinases and tyrosine phosphatases (4, 5). Although much work has focused on activation of protein tyrosine kinases, the ultimate physiological and pathological effects of phosphotyrosinemediated signaling are closely related to the duration of the phosphorylated state, which is highly regulated by specific tyrosine phosphatases (5). Tyrosine phosphatases can be divided into membrane-spanning receptor-like tyrosine phosphatases and soluble intracellular non-receptor-like tyrosine phosphatases (1, 2). The receptor-like tyrosine phos197
198
TAKAKURA ET AL.
phatases often contain two catalytic phosphatase domains and are controlled by binding of extracellular ligands (6). The non-receptor-like tyrosine phosphatases contain a single conserved phosphatase domain linked to variable sequences that modulate the activity and intracellular translocation of the enzyme. Both types of tyrosine phosphatases share the common active-site motif of containing cysteine and arginine, both of which are essential for enzyme activity (6). Following the attack on the substrate phosphotyrosine, the essential cysteine in the active site forms a covalent thiol–phosphate intermediate (7–9). The adjacent histidine and conserved proximal arginine lower the pK a of the cysteine from 9.5 to 4.7 (6) which activates it with respect to accepting the phosphate group. The positively charged arginine serves to attract the negatively charged phosphate moiety of phosphotyrosine. The phosphate group is oriented and stabilized by numerous hydrogen bonds to ONH x groups of surrounding residues (9). A conserved tyrosine in the protein tyrosine phosphatase (PTP) 3 active site appears to ring-stack with the aromatic ring of phosphotyrosine thereby confering specificity for phosphotyrosine over phosphoserine or phosphothreonine (6). Nitric oxide is an endogenous free radical formed in a variety of cell types by the enzymatic oxidation of L-arginine to citrulline and is implicated in diverse physiological processes including vasorelaxation, platelet inhibition, and neurotransmission (10). Nitric oxide and nitrogen oxides formed by the slow reaction of nitric oxide with oxygen have been suggested to modulate thiol-dependent enzymes through the formation of S-nitrosothiols (11) and subsequent oxidation to give disulfides (12). The highly reactive thiolate group in tyrosine phosphatases could also be a potential target for transnitrosation by S-nitrosothiols (12–14). The chemical reactivity of nitric oxide is greatly increased by its diffusion-limited reaction with superoxide (O 2•2) to form peroxynitrite (ONOO 2) (15). Peroxynitrite nitrates tyrosine residues in proteins to form 3-nitrotyrosine, which can potentially interfere with subsequent phosphorylation of the affected tyrosine (16 –19). In addition, peroxynitrite readily oxidizes both free and protein-bound sulfhydryls (20) and induces oxidative inactivation of other sulfhydryl-containing enzymes such as alcohol dehydrogenase (21), creatine kinase (22), and succinate dehydrogenase (23). Consequently, peroxynitrite could potentially inactivate tyrosine phosphatases by either oxidizing the thiolate or nitrating the tyrosine residue in the active site (24). 3 Abbreviations used: PTPs, protein tyrosine phosphatase; GSNO, S-nitrosoglutathione; LAR, leukocyte–antigen-related; pNPP, p-nitrophenyl phosphate; DCDHF, dichlorodihydrofluorescein; DCF, dichlorofluorescein; DTT, dithiothreitol; SOD, superoxide dismutase.
Hydrogen peroxide (H 2O 2) has been shown to inactivate PTPs containing the CXXXXXR active-site motif in vitro and in cells (25–29). However, inactivation by H 2O 2 often requires many minutes even at the relatively high concentrations of H 2O 2 typically used (29). Peroxynitrite oxidizes both free and protein-bound thiols up to 5000 times faster than H 2O 2 and thiols can react directly as peroxynitrite anion (20). Thus, thiols flanked by cationic amino acid residues, such as the CXXXXXR-containing PTPs, could preferentially be targeted by peroxynitrite. In this study, we compared the effects of nitric oxide, S-nitrosoglutathione (GSNO), and peroxynitrite on activities of three recombinant human PTPs: leukocyte– antigen-related tyrosine phosphatase (a receptor-like tyrosine phosphatase), T-cell tyrosine phosphatase (a non-receptor-like tyrosine phosphatase), and protein tyrosine phosphatase 1B (PTP1B). The tyrosine phosphatases were exceptionally susceptible to peroxynitrite but not to nitric oxide or GSNO. Moreover, the amount of peroxynitrite required for inhibition was essentially the same regardless of whether peroxynitrite was added from synthetic stocks or formed in situ via cogeneration of superoxide and nitric oxide. MATERIALS AND METHODS Human recombinant leukocyte antigen-related (LAR) protein tyrosine phosphatase (40-kDa molecular mass), human recombinant T-cell (CD45) protein tyrosine phosphatase (95-kDa molecular mass), and p-nitrophenyl phosphate (pNPP) were obtained from Calbiochem (San Diego, CA). Human recombinant protein tyrosine phosphatase 1B (PTP1B) (37-kDa molecular mass) was obtained from Biomol (Plymouth, PA). DETA NONOate and GSNO were from Alexis (San Diego, CA). All other reagents were from Sigma. Peroxynitrite was prepared using a quenched-flow reaction apparatus as described by Reed et al. (30) and Beckman et al. (31). An aqueous solution of 0.6 M sodium nitrite was rapidly mixed with an equal volume of 0.7 M hydrogen peroxide containing 0.6 M HCl and immediately quenched with the same volume of 1.5 M NaOH. Stock solutions were typically 150 mM in peroxynitrite. Working solutions (2 mM) were prepared by dilution of stocks into 0.1 M semiconductor grade NaOH and concentrations of all solutions were standardized spectrally (« 5 1700 M 21 cm 21). Working solutions were kept on ice and concentrated stock solutions were stored at 280°C. PTP assays and inactivation by peroxynitrite. Tyrosine phosphatase activity was measured using the synthetic substrate pNPP (1 mM) in 1.4 ml of 50 mM Tris buffer, pH 7.4, at 37°C (32). The increase in absorbance at 402 nm (associated with hydrolysis of pNPP to p-nitrophenol) was monitored in real time using stirred glass cuvettes mounted in the temperature-controlled cuvette holder of a Hewlett–Packard diode-array spectrophotometer. Concentration responses for inactivation by peroxynitrite were determined by adding peroxynitrite to rapidly stirred cuvettes at 37°C containing enzyme and Tris buffer. Substrate (pNPP) was added 30 s after peroxynitrite and activity was monitored for 3 min; rates were expressed as percentages of the untreated (control) phosphatase rate. Unless otherwise noted, the final concentration of the three PTPs in the cuvettes were as follows: 0.29 mg/ml or 3 nM enzyme monomer for CD45, 0.48 mg/ml or 12 nM enzyme monomer for LAR, and 0.36 mg/ml or 9.6 nM enzyme monomer for PTP1B. Determinations of activity versus pNPP substrate concentration revealed no increase in
INACTIVATION OF PROTEIN TYROSINE PHOSPHATASES BY PEROXYNITRITE
SCHEME 1
activity above 1 mM; thus, 1 mM was used for all activity assays unless otherwise noted. Determination of rate constants via competition assay. Because peroxynitrite decomposes so rapidly, it is not possible to determine an inactivation rate constant by measuring phosphatase activity over time as has been done with H 2O 2 (29). Therefore, bimolecular rate constants for reaction of peroxynitrite with tyrosine phosphatases were determined by means of a competition assay as previously described (21, 33, 34). Based on the concentration response for peroxynitrite-mediated inactivation, a concentration of peroxynitrite was chosen which would inhibit .90% but ,100% of the original PTP activity so that protection by cysteine could be precisely measured. Solutions of PTPs were prepared in Tris buffer and cysteine was added immediately before peroxynitrite. Remaining PTP activity was determined by adding 1 mM pNPP 30 s after peroxynitrite and monitoring hydrolysis at 402 nm for 3 min. Under the conditions used here, the possible fates for peroxynitrite include reaction with PTP, reaction with cysteine, and spontaneous rearrangement (decay) to nitrate (see Scheme 1). The rate constants for these three processes are k PTP, k cys, and k decay, respectively. The rate constant determination for the peroxynitrite/PTP reaction (k PTP) relies solely on the relative concentrations of competing targets (PTP and cysteine) for peroxynitrite and the known rate constant for the peroxynitrite/cysteine reaction. The rate of spontaneous decay of peroxynitrite (k decay) can be ignored because the fraction of peroxynitrite disappearing via decay is maximal at zero cysteine and decreases as the concentration of cysteine is increased; k decay would be relevant only if the fraction of peroxynitrite lost via decay somehow increased, thereby making less total peroxynitrite available to react with either PTP or cysteine. Cysteine reacts with peroxynitrite at a known rate of 5 3 10 3 M 21 s 21 at pH 7.4 and 37°C (20). Thus, by comparing the ability of different cysteine concentrations to prevent peroxynitrite-mediated inactivation of a fixed concentration of PTP enzyme monomer, the rate constant for the peroxynitrite/PTP reaction can be calculated (see Scheme 1) (21). Derivations of the equation describing the competition between cysteine and PTPs have been published previously (21, 34). Briefly, Eq. [1] was used to determine k PTP. 1 F@PTP# 5 @Cys#. ~1 2 F!k Cys k PTP
[1]
The term F describes the fraction of PTP inactivated by peroxynitrite. When F[PTP]/(1 2 F)k Cys is plotted as a function of cysteine concentration, the reciprocal of the slope gives k PTP in units of M 21 s 21 (21, 34, 35). As a further verification of our reported rate constants, we calculated them using an alternate method as described by Masumoto and Sies (36). This latter method involves plotting percent-
199
age remaining enzyme activity as a function of cysteine concentration and fitting data to an equation for first-order decay in order to determine the cysteine concentration providing 50% protection. At a concentration of cysteine which gives exactly 50% inactivation (or 50% inactivation), the rate constant for reaction of peroxynitrite with PTP is equal to the ratio of [PTP]/[cysteine] times the rate constant for the peroxynitrite/cysteine reaction. By definition, [PTP] would be one-half of the initial concentration and [cysteine] would be that concentration which provides 50% protection. k cysteine 5 5000 M 21 s 21 at pH 7.4 and 37°C. This latter method provided rate constants for the peroxynitrite/PTP reaction which were virtually identical to those obtained by the competition Eq. [1] shown above. Nitric oxide production from DETA NONOate and treatment of LAR phosphatase. The amount of NO generated by DETA NONOate was determined using an electrochemical probe (World Precision Instruments) mounted in a stirred, temperature-controlled vessel. A standard curve for probe response as a function of nitric oxide was prepared by adding small volumes of a saturated aqueous solution of nitric oxide (1.9 mM) to 1 ml of 50 mM Tris buffer, pH 7.4, as described previously (31). A known amount of DETA NONOate, which had been standardized spectrally (« 5 7940 M 21 cm 21 at 252 nm), was added to a 1-ml solution of Tris buffer, pH 7.4, at 37°C. DETA NONOate has a half-life of approximately 56 h under these conditions and releases 2 mol of nitric oxide per mole of DETA NONOate. At a concentration of 0.4 mM, DETA NONOate was found to produce a steady-state concentration of 2 mM nitric oxide. LAR phosphatase (0.12 mM enzyme monomer) was incubated for 3 h at 37°C with 0.4 mM DETA NONOate in 0.14 ml of 50 mM Tris buffer. After 3 h, the reaction solution was diluted to a final volume of 1.4 ml in 50 mM Tris buffer and assayed for activity after adding 1 mM pNPP. Determination of peroxynitrite generation from SIN-1. In simple buffered solutions, SIN-1 decomposes to produce equimolar amounts of superoxide and nitric oxide which combine quantitatively to form peroxynitrite (37, 38). Production of peroxynitrite from SIN-1 over time was quantified by measuring the oxidation of dichlorodihydrofluorescein (DCDHF) to dichlorofluorescein (DCF) spectrally (« 500 21 nm 5 59,500 M cm 21) (38) under the same conditions used for SIN-1 inactivation of PTPs (21). At pH 7.4 and 37°C, peroxynitrite production from SIN-1 decomposition reaches a steady state after approximately 3 min and remains essentially linear for up to 60 min (21, 38). Synthetic peroxynitrite oxidizes DCDHF to DCF with 38% efficiency, i.e., 1 mM peroxynitrite will produce 0.38 mM DCF (38). Thus, formation of DCF from SIN-1 was converted to total peroxynitrite production by dividing by 0.38. The amount of SIN-1-generated peroxynitrite required to inactivate PTPs was determined by comparing the time required for 50% PTP inactivation (corrected for spontaneous activity loss under turnover conditions) to the amount of peroxynitrite formed from SIN-1 during the same interval. Western blot analysis for protein nitration. Western blot analysis was performed as described previously (24, 39). Briefly, peroxynitrite was added to recombinant PTP1B in 60 ml of 50 mM Tris buffer, pH 7.4, while vortex mixing. PTP activity was immediately determined by adding 10-ml aliquots of reaction mixes to 1.4 ml standard activity assay mixes containing pNPP. The remaining 50 ml (2.5 mg) from each reaction was subjected to SDS–PAGE and transferred to nitrocellulose for nitrotyrosine immunoblotting.
RESULTS
Peroxynitrite-mediated inactivation of PTPs. Peroxynitrite inactivated T-cell tyrosine phosphatase (CD45), LAR tyrosine phosphatase, and PTP1B tyrosine phosphatase with IC 50 values of 0.3, 0.7, and 0.9 mM, respectively (Fig. 1a). Addition of a 100 mM peroxynitrite solution which had been allowed to decom-
200
TAKAKURA ET AL.
oxynitrite increased to 1.4 mM for CD45 and 5.8 mM for LAR tyrosine phosphatases but was slightly decreased (0.6 mM) with PTP1B (Fig. 1b). These results indicate that the peroxynitrite/carbon dioxide adduct is also capable of potently inactivating the PTPs, although subtle differences may exist with respect to recognition of the adduct compared with peroxynitrite anion. Peroxynitrite-mediated inactivation is largely irreversible. Previous studies have shown that the loss of PTP activity resulting from treatment with H 2O 2 was reversed by thiol reductants such as DTT (29, 44). Incubation of peroxynitrite-inactivated LAR phosphatase with 20 mM DTT for up to 1 h resulted in the recovery of only about 10% of the original activity (Fig. 2); similar limited activity recoveries were seen with CD45 and PTP1B phosphatases. LAR activity was inhibited to 20% of its untreated activity by 2 mM peroxynitrite. Addition of DTT to LAR phosphatase immediately prior to 2 mM peroxynitrite prevented inhibition in a dose-dependent manner (Fig. 2). No effect on phosphatase activity was seen when 100 mM peroxynitrite and 10 mM DTT were premixed immediately before addition to the LAR reaction mix (not shown), demonstrating that the protective effect of DTT was not due to any product of the peroxynitrite/DTT reaction. Rate constant determination. The rate of peroxynitrite-mediated inactivation of PTPs is too fast to permit use of PTP activity loss over time as a means of deter-
FIG. 1. Inactivation of protein tyrosine phosphatases by peroxynitrite and the effects of carbon dioxide. (a) PTPs were treated via bolus additions of peroxynitrite to rapidly stirred reaction cuvettes containing 50 mM Tris buffer, pH 7.4, at 37°C and assayed within 30 s using 1 mM pNPP (see Materials and Methods). Reaction cuvettes contained 3 nM CD45 PTP enzyme monomer, 12 nM LAR PTP enzyme monomer, or 9.6 nM PTP1B enzyme monomer and the indicated final concentrations of peroxynitrite. (b) Conditions were identical to those in (a) except that 25 mM sodium bicarbonate, which had been pH adjusted to 7.4, was added to the Tris buffer prior to adding peroxynitrite. PTP activity was not affected by the presence of bicarbonate.
pose in 50 mM Tris buffer, pH 7.4, had no effect on phosphatase activity. Likewise, peroxynitrite decomposition products nitrite (0.2 mM) and nitrate (0.2 mM) had no effect on phosphatase activity. Because carbon dioxide has been shown to react with and alter the reactivity of peroxynitrite (40 – 43), PTP inactivation by peroxynitrite was compared in the presence of a carbon dioxide-saturated (25 mM bicarbonate) Trisbuffered solution. The IC 50 for inactivation by per-
FIG. 2. Effects of DTT on peroxynitrite-mediated inactivation of PTPs. Reactions contained 12 nM LAR PTP in 50 mM Tris buffer, pH 7.4, and the indicated concentrations of DTT added prior to (■) 2 mM peroxynitrite; 1 mM pNPP was added immediately after peroxynitrite to determine the residual activity. The ability of DTT to reverse the activity loss induced by 2 mM peroxynitrite was determined by adding 10 mM DTT immediately after peroxynitrite (F) and incubating for 30 min prior to assaying for activity.
201
INACTIVATION OF PROTEIN TYROSINE PHOSPHATASES BY PEROXYNITRITE TABLE I
Rate Constants for Reaction of PTPs with Peroxynitrite
FIG. 3. Determination of rate constants for reaction of peroxynitrite with PTPs. Reaction mixtures contained the following concentrations of PTPs in 50 mM Tris buffer, pH 7.4, at 37°C: 0.29 mg/ml or 3 nM enzyme monomer for CD45, 0.96 mg/ml or 24 nM enzyme monomer for LAR, and 0.36 mg/ml or 9.6 nM enzyme monomer for PTP1B. Peroxynitrite concentrations which were found to inhibit .90 but ,100% of the activity were as follows: 3.8 mM peroxynitrite for CD45, 5 mM for PTP1B, and 10 mM for LAR. Cysteine, at the indicated concentrations, was added to solutions of PTPs in Tris buffer immediately before peroxynitrite. Remaining PTP activity was determined by adding 1 mM pNPP 30 s after peroxynitrite and monitoring hydrolysis at 402 nm for 3 min.
mining an inactivation rate constant. However, free cysteine reacts directly with peroxynitrite anion at a known rate (20). Thus, the rate of reaction between peroxynitrite and PTPs was determined by measuring protection as a function of increasing free cysteine concentrations (Fig. 3). By comparing the relative amounts of cysteine and enzyme monomer and the known rate constant for the cysteine/peroxynitrite reaction (EQ. [1]), rate constants for the PTP/peroxynitrite reaction were determined to be 2.0 3 10 8 M 21 s 21 for CD45 PTP, 2.3 3 10 7 M 21 s 21 for LAR PTP, and 2.2 3 10 7 M 21 s 21 for PTP1B (Table I). Effect of substrate on peroxynitrite-mediated inactivation. A concentration of peroxynitrite was chosen which would inactivate most but not all of the phosphatase activity so that effects due to substrate could be quantified. When 2 mM peroxynitrite was added to LAR in the absence of substrate (pNPP), only 20% of the original phosphatase activity remained (Fig. 4). Addition of 2 mM peroxynitrite to reaction mixtures containing LAR plus increasing amounts of substrate resulted in partial protection from inactivation. The protective effect saturated at approximately 2 mM pNPP which provided maximum protection of only 46% of the original activity (Fig. 4). Inactivation by SIN-1. Cogeneration of superoxide and nitric oxide by SIN-1 was used to compare the
Tyrosine phosphatase
Rate constant for reaction with peroxynitrite (M 21 s 21)
CD45 LAR PTP1B
2.0 3 10 8 2.3 3 10 7 2.2 3 10 7
Note. Each of the three PTPs was treated with a submaximal concentration of peroxynitrite at 37°C (Fig. 3) in the absence and presence of the indicated cysteine concentrations and immediately assayed for phosphatase activity (see Materials and Methods). Based on the known enzyme monomer concentration present ([PTP]), the biomolecular rate constant for peroxynitrite and cysteine (k SH), and the measured extent of protection from inactivation (F), the ordinate points were calculated and plotted versus cysteine concentration (EQ. [1]). The rate of reaction between peroxynitrite and phosphatase is determined by taking the inverse of the slope (1/k PTP).
effects of peroxynitrite formed in situ with bolus additions of synthetic peroxynitrite. SIN-1 releases both superoxide and nitric oxide at a constant rate leading to quantitative formation of peroxynitrite (37, 38). SIN-1 was added at 60 s to PTPs which were actively hydrolyzing pNPP and the time course for complete loss of phosphatase activity was monitored continuously (Fig. 5). In the presence of 0.1 mM SIN-1, the loss of LAR activity was considerably faster than the spontaneous activity loss that accompanies enzyme turnover. Complete inactivation oc-
FIG. 4. Protection from peroxynitrite-mediated inactivation by pNPP substrate. Reactions contained 12 nM LAR PTP in 50 mM Tris buffer, pH 7.4, and the indicated concentrations of pNPP added prior to 2 mM peroxynitrite. Immediately after peroxynitrite addition, more pNPP was added to give a final concentration of 3 mM in order to assay the remaining phosphatase activity.
202
TAKAKURA ET AL.
FIG. 5. Inhibition of LAR PTP by SIN-1 and protection by SOD. Reactions contained 24 nM LAR PTP in 50 mM Tris buffer, pH 7.4, and 1 mM pNPP at 37°C; pNPP hydrolysis to p-nitrophenol was monitored in real time at 402 nm. SIN-1 (0.1 mM) was added after 60 s to the reactions illustrated by broken lines. The reaction illustrated by the dotted line also contained 0.5 mg/ml bovine Cu,Zn SOD. SIN-1-mediated inactivation of PTP activity (dashed line) was taken as the time at which the absorbance change ceased—a time equal to 12 min of total SIN-1 exposure. The amount of peroxynitrite produced during this interval was determined under identical reaction conditions using dichlorodihydrofluorescein as the indicator of peroxynitrite production (see Materials and Methods). Neither SIN-1 nor its decomposition product SIN-1C absorbs at 402 nm.
curred approximately 840 s after SIN-1 addition (Fig. 5). The amount of peroxynitrite formed during this time was determined in a separate reaction using DCDHF oxidation as the measure of peroxynitrite production (21, 38). A comparison of the time required for 50% inactivation of LAR and the amount of peroxynitrite formed from SIN-1 during that in-
TABLE II
IC 50 Values for PTP Inhibition by SIN-1 Tyrosine phosphatase
IC 50 for inhibition by SIN-1 (mM)
CD45 LAR PTP1B
0.20 1.05 0.64
Note. PTP activity was monitored in real time at 37°C in the absence and presence of SIN-1 and the time required for 50% inactivation was determined (see Fig. 5 for illustration and Materials and Methods for details). Peroxynitrite formation from SIN-1 as a function of time was quantified in a separate reaction run under identical conditions. The amount of SIN-1-generated peroxynitrite required to inactivate PTPs was determined by comparing the time required for 50% PTP inactivation (corrected for spontaneous activity loss under turnover conditions) to the amount of peroxynitrite formed during the same interval.
terval produced an IC 50 equal to 1.05 mM peroxynitrite (Table II). Similar measurements for CD45 and PTP1B phosphatases produced IC 50 values of 0.2 and 0.64 mM, respectively. Addition of 0.5 mg/ml bovine copper, zinc superoxide dismutase (SOD), which scavenges superoxide and thereby inhibits peroxynitrite formation, completely protected the PTPs from SIN-1 inactivation (Fig. 5). Effects of nitric oxide and nitrosothiol donors. DETA NONOate releases nitric oxide continuously in a pH- and temperature-controlled manner. In a stirred vessel at pH 7.4 and 37°C, a steady-state concentration of 2 mM nitric oxide is obtained after 10 min at a concentration of 0.4 mM DETA NONOate (see Materials and Methods). Exposure of LAR to 0.4 mM DETA NONOate for 3 h resulted in a loss of only 12% of the original activity (Fig. 6a); CD45 and PTP1B were inhibited by less than 10% under the same conditions. Treatment with GSNO (0.1–1.0 mM) for 30 min inhibited LAR PTP activity in a concentration-dependent manner. The inhibitory effect of GSNO toward LAR PTP reached a maximum of 50% at a concentration of 1 mM (Fig. 6b). Unlike peroxynitrite-mediated inactivation, inhibition by GSNO was completely reversed by addition of 10 mM dithiothreitol.
INACTIVATION OF PROTEIN TYROSINE PHOSPHATASES BY PEROXYNITRITE
203
first appeared (Fig. 7, lane 6). These results reveal that inactivation is not correlated with nitration of proteinbound tyrosine. DISCUSSION
Peroxynitrite readily nitrates tyrosine residues in proteins, and peptides containing nitrotyrosine are resistant to phosphorylation in vitro (16, 42, 45), suggesting that peroxynitrite could alter phosphotyrosine-mediated signaling. Peroxynitrite has been shown to affect signal transduction in cultured cells and to induce apoptosis (46 – 48). However, tyrosine phosphorylation of proteins is typically increased rather than inhibited when cells are treated with peroxynitrite (49 –51). This led us to consider whether tyrosine phosphatases might be inactivated by peroxynitrite. The reaction rates between peroxynitrite anion and PTPs, particularly CD45, are among the fastest rates ever reported for peroxynitrite reacting with biological molecules. One possible explanation for the rapid rates emerges from an examination of the active site of PTP1B (9) and a comparison between binding of phosphate anion and peroxynitrite anion (Fig. 8). Orientation of the phosphate group of phosphotyrosine is determined by hydrogen bonding (Fig. 8, dotted lines) of the phosphate oxygens to ONH x groups of the residues (216 –221) surrounding the active-site cysteine 215 of PTP1B (9). It is reasonable to propose that peroxynitrite anion reacts with cys 215 faster than with other cysteine residues for three reasons: (i) charge attraction by the essential arginine 221 cation occurs, (ii) per-
FIG. 6. Effects of DETA NONOate and GSNO on PTP activity. (a) LAR PTP was incubated with DETA NONOate for 3 h as described under Materials and Methods, diluted 10-fold, and assayed for activity via addition of pNPP. (b) LAR PTP (12 nM enzyme monomer) in 50 mM Tris buffer, pH 7.4, at 37°C was incubated with the indicated concentrations of S-nitrosoglutathione for 30 min and immediately assayed by the addition of 1 mM pNPP (F). After 30 min incubation with 1 mM GSNO, 10 mM DTT was added to the LAR PTP reaction solution and activity was added after 30 min (■).
Peroxynitrite-mediated tyrosine nitration of PTPs. PTPs possess a conserved tyrosine residue near the active site which appears to confer specificity for phosphotyrosine over phosphoserine or phosphothreonine (6). Although peroxynitrite is capable of nitrating tyrosine residues in proteins, nitration does not appear to play a role in PTP inactivation (Fig. 7). Reaction mixtures containing 50 mg/ml (1.33 mM enzyme monomer) PTP1B were treated with increasing concentrations of peroxynitrite (5–160 mM) and immediately assayed for activity and nitrotyrosine content via antinitrotyrosine immunoblot analysis (Fig. 7). PTP1B was inactivated by 95% at the point at which nitrotyrosine
FIG. 7. Comparison of PTP1B activity with protein nitration via nitrotyrosine immunoblot. Peroxynitrite (0, 5, 10, 20, 40, 80, and 160 mM) was added to recombinant PTP1B (50 mg/ml or 1.33 mM enzyme monomer) in 60 ml of 50 mM Tris buffer, pH 7.4, while vortex mixing. PTP activity was immediately determined by adding 10-ml aliquots of reaction solutions to 1.4-ml standard activity assay mixes containing pNPP (see Materials and Methods). The remaining 50 ml (2.5 mg protein) from each reaction was subjected to SDS–PAGE and nitrotyrosine immunoblot as described under Materials and Methods.
204
TAKAKURA ET AL.
FIG. 8. Active-site residues of PTP1B: Illustration of relative molecular size and hydrogen bonding between phosphate and peroxynitrite anions. The active-site residues 216 –221 and the cysteine 215–phosphate intermediate (structure at left) were taken from the X-ray crystal structure of recombinant human PTP1B (Brookhaven Database Entry 1A5Y) (9). The entire X-ray crystal structure of PTP1B was imported into the molecular visualization program Quantum Cache 3.0 (Oxford Molecular Group, Eugene, OR) and edited to show only the active-site residues (minus noninvolved hydrogens) without altering the original X-ray coordinates. The structure on the right is identical to that at left except that cis peroxynitrite anion has been superimposed over the dotted spheres of active-site cysteine designated with the yellow C and the bound phosphate (labeled P). The structure of cis peroxynitrite anion is based on several earlier physical– chemical studies (56, 68) and is in agreement with recent X-ray crystal data (69). Peroxynitrite anion was superimposed onto the active site of PTP1B (structure at right) and rotated through all planes to determine potential alignments with the known coordinates of the phosphate oxygen atoms (note two asterisks). The structure at right illustrates just one of several such alignments wherein two of the oxygen atoms of peroxynitrite precisely superimposed over two of the phosphate oxygen atoms. Hydrogen bonding interactions between enzyme ONH x groups and the cysteinebound phosphate are those proposed by Pannifer et al. (9). Hydrogen bonding interactions with peroxynitrite anion are proposed as an additional explanation for the fast reaction rate constants.
oxynitrite and phosphate are very similar in size and interatomic distance as evidenced by the precise alignment of two oxygen atoms of cis peroxynitrite anion with two of the phosphate oxygens (Fig. 8, asterisks), and (iii) every atom of peroxynitrite is close enough to hydrogen bond with ONH groups of the surrounding residues. Figure 8 shows just one of several potential orientations of peroxynitrite anion, any one of which is compatible with numerous hydrogen bonding interactions and subsequent oxidation of cys 215. Fast reaction rates suggest that PTPs could be targeted for peroxynitrite-mediated inactivation even in the presence of much higher concentrations of endogenous scavengers like glutathione. For example, the peroxynitrite reaction rate with CD45 is 2.0 3 10 8 M 21 s 21 compared to 5 3 10 3 M 21 s 21 for glutathione, a difference of 40,000. Thus, 25 nM CD45 could be inac-
tivated by peroxynitrite even in the presence of 1 mM glutathione. Both the in vivo concentration of carbon dioxide and its rate of reaction with peroxynitrite are quite similar to those of glutathione (52). However, in the case of carbon dioxide, the reaction is catalytic and generates an adduct which is only slightly less inhibitory toward PTPs (Fig. 1) and thiols in general (43). The potential for peroxynitrite-mediated inactivation of PTPs in vivo becomes all the more relevant when considering that it is irreversible. Remarkably low concentrations (,0.9 mM) of peroxynitrite irreversibly inactivated three human tyrosine phosphatases, whereas continuous generation of micromolar concentrations of nitric oxide over periods of up to 3 h did not. Because the half-lives of oxidants vary, relative toxicities of reactive species must be assessed by comparing the concentration over time
INACTIVATION OF PROTEIN TYROSINE PHOSPHATASES BY PEROXYNITRITE
(area under the curve or total exposure) rather than by initial concentration (53, 54). Peroxynitrite decomposes with a half-life of approximately 1.0 s at pH 7.4 and 37°C (55) and produced 50% inhibition of PTPs at concentrations ranging from 0.3 to 0.9 mM. Thus, total peroxynitrite exposure equaled 0.3– 0.9 mM 3 s. In contrast, exposure to nitric oxide generated from DETA NONOate was 21,600 mM 3 s (2 mM 3 10,800 s)—an exposure which had little effect on PTP activity. Exposure to GSNO was 180,000 mM 3 s (1000 mM 3 1800 s) and inhibited PTP activity by 50%, but inhibition was completely reversible. A simple comparison reveals that peroxynitrite is at least 100,000 times more inhibitory to PTPs than is nitric oxide or GSNO. Several reports indicate that H 2O 2 can inhibit PTPs (25–29). Unlike peroxynitrite, inactivation by H 2O 2 appears to be completely reversible by thiol reductants (29). Extrapolation of data from Denu and Tanner (29) reveals that 0.9 mM H 2O 2 inactivated PTP1B by 50% within approximately 17 s 4 or a total H 2O 2 exposure of 15,300 mM 3 s (900 mM 3 17 s). Thus, peroxynitrite is roughly 17,000-fold more inhibitory to PTP1B than is H 2O 2. Formation of peroxynitrite in vivo is most likely to occur via the radical–radical combination of superoxide and nitric oxide (56, 57). A recent report suggested that cogeneration of superoxide and nitric oxide was both qualitatively and quantitatively different from peroxynitrite with respect to modification of target molecules (58). With regard to PTP inhibition, we found precisely the opposite—namely that slow, continuous cogeneration of superoxide and nitric oxide with SIN-1 inhibited 50% of PTP activity at virtually identical concentrations of peroxynitrite. The equivalence of SIN-1 and bolus peroxynitrite has also been seen previously with respect to inactivation of the yeast alcohol dehydrogenase (21). Recent experiments designed to directly compare synthetic (bolus) peroxynitrite to rapid mixing of nitric oxide with potassium superoxide have also revealed their equivalency with regard to the yield of nitrotyrosine (C. Reiter and J. S. Beckman, unpublished observations). SOD scavenges the superoxide produced during SIN-1 decomposition and thereby converts SIN-1 predominantly into a nitric oxide donor (59). The ability of SOD to completely prevent PTP inactivation indicates that the effects of SIN-1 are mediated by peroxynitrite and provides additional evidence that nitric oxide alone is not inhibitory. The protein tyrosine phosphatases have an unusually acidic cysteine in the active site with a pK a esti4 The time required for 50% inhibition of PTP1B by 0.9 mM H 2O 2 was determined from Fig. 1 of the paper by Denu and Tanner (29) which reveals the hyperbolic loss of activity over a 100-s interval at pH 7 and 30°C.
205
mated to be around 4.7 (6). Thus, this cysteine exists as the thiolate anion at physiological pH, making it susceptible to oxidation. The active-site cysteine functions as a nucleophile to accept the OPO 3 moiety from phosphotyrosine (7, 8) thereby generating a phosphocysteine intermediate (9). Site-directed mutagenesis has confirmed that this cysteine residue is essential for catalysis in both receptor-like and non-receptor-like protein tyrosine phosphatases (60, 61). Based on previous studies which unequivocally identified the activesite cysteine as the target of H 2O 2-mediated inactivation (29, 62), it is reasonable to suggest that oxidation of this cysteine thiolate is also the mechanism of peroxynitrite-mediated inactivation. It is clear that peroxynitrite-mediated inactivation is not dependent on tyrosine nitration since activity loss occurred at peroxynitrite concentrations below those at which nitration became apparent (Fig. 7). Carbon dioxide forms an adduct with peroxynitrite which is more efficient at nitrating tyrosine than at oxidizing thiols (43). In the presence of carbon dioxide, a fivefold higher concentration of peroxynitrite was required to produce the same inhibition with at least two of the PTPs. This finding is consistent with inhibition via thiolate oxidation and not tyrosine nitration. Treatment with DTT after peroxynitrite-mediated inhibition of PTPs resulted in less than 10% recovery of activity. This contrasts with inactivation by H 2O 2 which is completely reversed by thiol reductants (29). Thiol oxidation may result in the formation of sulfenic acid (OSOH) or in direct disulfide formation, either of which can be reduced back to OSH with reductants such as DTT (20). Therefore, the inability of DTT to restore activity to peroxynitrite-treated PTPs suggests that the active-site thiol is oxidized to sulfinic acid (OSO 2H) or even to sulfonic acid (OSO 3H) which cannot be reduced by the low-molecular-weight antioxidants present in cells such as glutathione and ascorbate (34, 63). Because DTT was unable to restore activity, its ability to prevent peroxynitrite-mediated inhibition (when added first) is most likely related to direct scavenging of peroxynitrite. Relatively high amounts of GNSO partially inhibited phosphatase activity, but the inhibition was rapidly reversed by low-molecular-weight thiols such as DTT and mercaptoethanol. Partial, reversible inhibition by the nitrosothiol GSNO, but not by nitric oxide itself, is consistent with a mechanism involving direct nitrosation of the active-site thiolate by GSNO. Transnitrosation reactions involving transfer of NO from a nitrosothiol to a reduced thiol (OSH) have been shown to occur quite readily in in vitro systems (12–14). Rates of transnitrosation increase with pH indicating that thiolate anions are better acceptors of the nitrosating species (11, 64). At physiological pH, the active-site cysteine (pK a 5 4.7) (6) of PTPs would exist in the ionized
206
TAKAKURA ET AL.
form which should be more susceptible than other protein thiols to nitrosation by GSNO. It is quite possible that previous reports of PTP inactivation in cells by compounds such as S-nitroso-N-acetylpenicillamine (SNAP) or SIN-1 (65, 66) could have been related to transnitrosation or peroxynitrite-mediated inactivation, respectively. Pretreatment with the PTP substrate pNPP afforded some protection from peroxynitrite-mediated inactivation, providing additional support for active-site thiol oxidation as the mechanism of inactivation. Studies which have assessed PTP inactivation by H 2O 2 have reported that saturating concentrations of pNPP provided complete protection (29). This difference can be explained in terms of the much faster reaction rate of peroxynitrite with PTPs. That is to say, transient occupation of the active site by a substrate during catalytic turnover will block a slow interaction such as H 2O 2 much more effectively than a fast one such as peroxynitrite. If pNPP were acting as a direct scavenger of peroxynitrite, then protection should have been linear with pNPP concentration, which was not observed. Recent demonstration that H 2O 2 is capable of reversibly inactivating PTPs both in vitro and in cells (25–29, 44) has ignited interest in the concept of oxidative regulation of tyrosine phosphatase activity and, by extension, in oxidative regulation of phosphotyrosine-mediated cell signaling. In that regard, it is especially noteworthy that the rates of PTP inactivation by peroxynitrite are many orders of magnitude faster than inactivation by H 2O 2. More importantly, inactivation by peroxynitrite is not reversed by thiol reductants. Mallozzi et al. (67) reported an initial decrease in total PTP activity in human erythrocytes treated with low concentrations of peroxynitrite followed by a time-dependent recovery of PTP activity; similar findings have been reported with human platelets (50). These results suggest that at least some cell types may have mechanisms for restoring activity to oxidized PTPs beyond simple low-molecular-weight reductants such as glutathione; the possibility of enzymatic reduction of oxidized PTPs warrants further investigation. The ability of peroxynitrite to inactivate PTPs much faster and at much lower concentrations than H 2O 2 argues that peroxynitrite may be the more likely biological mediator of PTP inactivation and subsequent enhancement of tyrosine phosphorylation under conditions of oxidative stress. Physiological concentrations of nitric oxide did not inhibit PTPs and inhibition of PTPs by the nitrosothiol GSNO was slow even at a concentration of 1 mM. Moreover, the rapid reversal of GSNO-mediated inhibition by DTT suggests that nitrosothiols would be very poor inhibitors in the presence of low-molecular-weight thiols in vivo. Thus, in situations where nitric oxide synthase inhibitors pre-
vent the loss of PTP activity in cells or tissues, it is reasonable to implicate peroxynitrite. It is also reasonable to conclude that endogenous formation of peroxynitrite and subsequent inactivation of PTPs may profoundly alter cellular physiology by enhancing protein tyrosine phosphorylation. REFERENCES 1. Fischer, E. H., Charbonneau, H., and Tonks, N. K. (1991) Science 253, 401– 406. 2. Pot, D. A., and Dixon, J. E. (1992) Biochim. Biophys. Acta 1136, 35– 43. 3. Walton, K. M., and Dixon, J. E. (1993) Annu. Rev. Biochem. 62, 101–120. 4. Sun, H., and Tonks, N. K. (1994) Trends Biochem. Sci. 19, 480 – 485. 5. Inagaki, N., Ito, M., Nakano, T., and Inagaki, M. (1994) Trends Biochem. Sci. 19, 448 – 452. 6. Fauman, E. B., and Saper, M. A. (1996) Trends Biochem. Sci. 21, 413– 417. 7. Guan, K. L., and Dixon, J. E. (1991) J. Biol. Chem. 266, 17026 – 17030. 8. Denu, J. M., Lohse, D. L., Vijayalakshmi, J., Saper, M. A., and Dixon, J. E. (1996) Proc. Natl. Acad. Sci. USA 93, 2493–2498. 9. Pannifer, A. D. B., Flint, A. J., Tonks, N. K., and Barford, D. (1998) J. Biol. Chem. 273, 10454 –10462. 10. Moncada, S., and Higgs, E. A. (1991) Eur. J. Clin. Invest. 21, 361–374. 11. Stamler, J. S. (1995) Curr. Top. Micro. Immunol. 196, 19 –36. 12. Singh, R. J., Hogg, N., Joseph, J., and Kalyanaraman, B. (1996) J. Biol. Chem. 271, 18596 –18603. 13. Liu, Z., Rudd, M. A., Freedman, J. E., and Loscalzo, J. (1998) J. Pharm. Exp. Ther. 284, 526 –534. 14. Rossi, R., Lusini, L., Giannerini, F., Giustarini, D., Lungarella, G., and di Simplicio, P. (1997) Anal. Biochem. 254, 215–220. 15. Kissner, R., Nauser, T., Bugnon, P., Lye, P. G., and Koppenol, W. H. (1998) Chem. Res. Toxicol. 11, 557. 16. Martin, B. L., Wu, D., Jakes, S., and Graves, D. J. (1990) J. Biol. Chem. 265, 7108 –7111. 17. Gow, A. J., Duran, D., Malcolm, S., and Ischiropoulos, H. (1996) FEBS Lett. 385, 63– 66. 18. Berlett, B. S., Friguet, B., Yim, M. B., Chock, P. B., and Stadtman, E. R. (1996) Proc. Natl. Acad. Sci. USA 93, 1776 –1780. 19. Li, X. H., Desarno, P., Song, L., Beckman, J. S., and Jope, R. S. (1998) Biochem. J. 331(2), 599 – 606. 20. Radi, R., Beckman, J. S., Bush, K. M., and Freeman, B. A. (1991) J. Biol. Chem. 266, 4244 – 4250. 21. Crow, J. P., Beckman, J. S., and Mccord, J. M. (1995) Biochemistry 34, 3544 –3552. 22. Konorev, E. A., Hogg, N., and Kalyanaraman, B. (1998) FEBS Lett. 427, 171–174. 23. Rubbo, H., Denicola, A., and Radi, R. (1994) Arch. Biochem. Biophys. 308, 96 –102. 24. MacMillan-Crow, L. A., Crow, J. P., and Thompson, J. A. (1998) Biochemistry 37, 1613–1622. 25. Hadari, Y. R., Geiger, B., Nadiv, O., Sabanay, I., Roberts, C. T., Jr., LeRoith, D., and Zick, Y. (1993) Mol. Cell. Endocrinol. 97, 9 –17. 26. Sullivan, S. G., Chiu, D. T., Errasfa, M., Wang, J. M., Qi, J. S., and Stern, A. (1994) Free Radical Biol. Med. 16, 399 – 403.
INACTIVATION OF PROTEIN TYROSINE PHOSPHATASES BY PEROXYNITRITE 27. Natarajan, V., Scribner, W. M., Alhassani, M., and Vepa, S. (1998) Environ. Health Perspect. 106(5), 1205–1212. 28. Carbajal, J. M., and Schaeffer, R. C. (1998) Biochem. Biophys. Res. Commun. 249, 461– 466. 29. Denu, J. M., and Tanner, K. G. (1998) Biochemistry 37, 5633– 5642. 30. Reed, J. W., Ho, H. H., and Jolly, W. L. (1974) J. Am. Chem. Soc. 96, 1248 –1249. 31. Beckman, J. S., Chen, J., Ischiropoulos, H., and Crow, J. P. (1994) Methods Enzymol. 233, 229 –240. 32. Tonks, N. K., Charbonneau, H., Diltz, C. D., Fischer, E. H., and Walsh, K. A. (1988) Biochemistry 27, 8695– 8701. 33. Castro, L., Rodriguez, M., and Radi, R. (1994) J. Biol. Chem. 269, 29409 –29415. 34. Radi, R. (1996) Methods Enzymol. 269, 354 –366. 35. Souza, J. M., and Radi, R. (1998) Arch. Biochem. Biophys. 360, 187–194. 36. Masumoto, H., and Sies, H. (1996) Chem. Res. Toxicol. 9, 1057– 1062. 37. Hogg, N., Darley-Usmar, V. M., Wilson, M. T., and Moncada, S. (1992) Biochem. J. 281, 419 – 424. 38. Crow, J. P. (1997) Nitric Oxide 1, 145–157. 39. MacMillan-Crow, L. A., Crow, J. P., Kerby, J. D., Beckman, J. S., and Thompson, J. A. (1996) Proc. Natl. Acad. Sci. USA 93, 11853–11858. 40. Lymar, S. V., and Hurst, J. K. (1996) Chem. Res. Toxicol. 9, 845– 850. 41. Lymar, S. V., Jiang, Q., and Hurst, J. K. (1996) Biochemistry 35, 7855–7861. 42. Gow, A., Duran, D., Thom, S. R., and Ischiropoulos, H. (1996) Arch. Biochem. Biophys. 333, 42– 48. 43. Zhang, H., Squadrito, G. L., Uppu, R. M., Lemercier, J. N., Cueto, R., and Pryor, W. A. (1997) Arch. Biochem. Biophys. 339, 183–189. 44. Lee, S. R., Kwon, K. S., Kim, S. R., and Rhee, S. G. (1998) J. Biol. Chem. 273, 15366 –15372. 45. Kong, S. K., Yim, M. B., Stadtman, E. R., and Chock, P. B. (1996) Proc. Natl. Acad. Sci. USA 93, 3377–3382. 46. Estevez, A. G., Radi, R., Barbeito, L., Shin, J. T., Thompson, J. A., and Beckman, J. S. (1995) J. Neurochem. 65, 1543–1550. 47. Spear, N., Estevez, A. G., Barbeito, L., Beckman, J. S., and Johnson, G. V. (1997) J. Neurochem. 69, 53–59. 48. Estevez, A. G., Spear, N., Manuel, S. M., Radi, R., Henderson, C. E., Barbeito, L., and Beckman, J. S. (1998) J. Neurosci. 18, 923–931.
207
49. Hogg, N., Joseph, J., and Kalyanaraman, B. (1994) Arch. Biochem. Biophys. 314, 153–158. 50. Mondoro, T. H., Shafer, B. C., and Vostal, J. G. (1997) Free Radical Biol. Med. 22, 1055–1063. 51. Spear, N., Estevez, A. G., Johnson, G. V. W., Bredesen, D. E., Thompson, J. A., and Beckman, J. S. (1998) Arch. Biochem. Biophys. 356, 41– 45. 52. Denicola, A., Freeman, B. A., Trujillo, M., and Radi, R. (1996) Arch. Biochem. Biophys. 333, 49 –58. 53. Zhu, L., Gunn, C., and Beckman, J. S. (1992) Arch. Biochem. Biophys. 298, 452– 457. 54. Beckman, J. S., and Koppenol, W. H. (1996) Am. J. Physiol. 271, C1424 –C1437. 55. Beckman, J. S., Beckman, T. W., Chen, J., Marshall, P. A., and Freeman, B. A. (1990) Proc. Natl. Acad. Sci. USA 87, 1620 –1624. 56. Koppenol, W. H., Moreno, J. J., Pryor, W. A., Ischiropoulos, H., and Beckman, J. S. (1992) Chem. Res. Toxicol. 5, 834 – 842. 57. Beckman, J. S., and Crow, J. P. (1993) Biochem. Soc. Trans. 21, 330 –334. 58. Pfeiffer, S., and Mayer, B. (1998) J. Biol. Chem. 273, 27280 – 27285. 59. Haddad, I. Y., Crow, J. P., Hu, P., Ye, Y., Beckman, J., and Matalon, S. (1994) Am. J. Physiol. 267, L242–L249. 60. Streuli, M., Krueger, N. X., Tsai, A. Y., and Saito, H. (1989) Proc. Natl. Acad. Sci. USA 86, 8698 – 8702. 61. Streuli, M., Krueger, N. X., Thai, T., Tang, M., and Saito, H. (1990) EMBO J. 9, 2399 –2407. 62. Degnore, J. P., Konig, S., Barrett, W. C., Chock, P. B., and Fales, H. M. (1998) Rapid Commun. Mass Spectrom. 12, 1457–1462. 63. Quijano, C., Alvarez, B., Gatti, R. M., Augusto, O., and Radi, R. (1997) Biochem. J. 322, 167–173. 64. Keshive, M., Singh, S., Wishnok, J. S., Tannenbaum, S. R., and Deen, W. M. (1996) Chem. Res. Toxicol. 9, 988 –993. 65. Estrada, C., Gomez, C., Martin-Nieto, J., De Frutos, T., Jimenez, A., and Villalobo, A. (1997) Biochem. J. 326, 369 –376. 66. Caselli, A., Chiarugi, P., Camici, G., Manao, G., and Ramponi, G. (1995) FEBS Lett. 374, 249 –252. 67. Mallozzi, C., Distasi, A. M. M., and Minetti, M. (1997) FASEB J. 11, 1281–1290. 68. Hamilton, T. P., Tsai, H. H., and Beckman, J. S. (1996) Methods Enzymol. 269, 329 –346. 69. Worle, M., Latal, P., Kissner, R., Nesper, R., and Koppenol, W. H. (1999) Chem. Res. Toxicol. 12, 305–307.
Rapid and Irreversible Inactivation of Protein Tyrosine Phosphatases PTP1B, CD45, and LAR by Peroxynitrite 1 Ko Takakura,* Joseph S. Beckman,† ,‡ ,§ Lee Ann MacMillan-Crow, ¶ and John P. Crow† ,§ ,\ ,2 *Department of Anesthesiology and Reanimatology, Fukui Medical University, Fukui, Japan; and †Department of Anesthesiology, ‡Department of Cell and Molecular Biology, ¶Department of Surgery, \Department of Pharmacology/Toxicology, and §The Center for Free Radical Biology at UAB, University of Alabama at Birmingham, Birmingham, Alabama 35233
Received April 2, 1999, and in revised form July 1, 1999
Protein tyrosine phosphatases (PTPs) contain an essential thiol in the active site which may be susceptible to attack by nitric oxide-derived biological oxidants. We assessed the effects of peroxynitrite, nitric oxide, and S-nitrosoglutathione on the activity of three human tyrosine phosphatases in vitro. The receptor-like T-cell tyrosine phosphatase (CD45), the non-receptor-like tyrosine phosphatase PTP1B, and leukocyte–antigen-related (LAR) phosphatase were all irreversibly inactivated by peroxynitrite in less than 1 s with IC 50 values of <0.9 mM. PTP inactivation was also seen with equivalent concentrations of peroxynitrite generated by SIN-1, indicating that bolus peroxynitrite and cogeneration of superoxide and nitric oxide were equipotent. Rate constants for peroxynitrite-mediated PTP inactivation were determined by competition with cysteine and were among the fastest rates yet seen for reaction of peroxynitrite with any biological molecules. The bimolecular reaction rates for CD45, LAR, and PTP1B were 2.0 3 10 8, 2.3 3 10 7, and 2.2 3 10 7 M 21 s 21, respectively. Inactivation by peroxynitrite was essentially irreversible as incubation with dithiothreitol (DTT) restored less than 10% of the original phosphatase activity. Prolonged treatment with 0.4 mM DETA NONOate, which generated a steady-state concentration of 2 mM nitric oxide, was only slightly inhibitory. S-Nitrosoglutathione (1.0 mM) 1 This study was supported by grants from the Ministry of Education, Science, Sports, and Culture of Japan; The American Heart Association (LAMC); the National Institutes of Health (HL58209 and NS35871); and the Amyotrophic Lateral Sclerosis Association (J.S.B. and J.P.C.). 2 To whom correspondence should be addressed at Department of Pharmacology/Toxicology and Center for Free Radical Biology, University of Alabama at Birmingham, ZRB 906, 1900 University Boulevard, Birmingham, AL 35233. Fax: 205-934-7437. E-mail: [email protected].
0003-9861/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
inhibited PTPs by approximately 50% after 30 min and the inhibition was completely reversed by DTT. Nitrotyrosine immunoblots of peroxynitrite-treated PTP1B revealed that peroxynitrite completely inactivated PTP1B prior to the appearance of protein tyrosine nitration. Peroxynitrite anion is structurally similar to phosphate anion both in terms of molecular diameter and charge. Thus, the extreme vulnerability of these PTPs to peroxynitrite-mediated inactivation is consistent with attraction of peroxynitrite anion to the active site and subsequent oxidation of the essential thiolate. These findings suggest that any PTP possessing the CXXXXXR active-site sequence could potentially be inactivated by peroxynitrite in vivo resulting in a net increase in tyrosine phosphorylation and profound effects on phosphotyrosine-dependent signaling cascades. © 1999 Academic Press Key Words: peroxynitrite; phosphatase; nitrotyrosine; phosphotyrosine; nitration; nitric oxide; nitrosothiol; thiolate; carbon dioxide; rate constants.
Tyrosine phosphorylation plays a critical role in controlling cell growth, proliferation, and differentiation (1–3) and is regulated by a careful balance between the activities of tyrosine kinases and tyrosine phosphatases (4, 5). Although much work has focused on activation of protein tyrosine kinases, the ultimate physiological and pathological effects of phosphotyrosinemediated signaling are closely related to the duration of the phosphorylated state, which is highly regulated by specific tyrosine phosphatases (5). Tyrosine phosphatases can be divided into membrane-spanning receptor-like tyrosine phosphatases and soluble intracellular non-receptor-like tyrosine phosphatases (1, 2). The receptor-like tyrosine phos197
198
TAKAKURA ET AL.
phatases often contain two catalytic phosphatase domains and are controlled by binding of extracellular ligands (6). The non-receptor-like tyrosine phosphatases contain a single conserved phosphatase domain linked to variable sequences that modulate the activity and intracellular translocation of the enzyme. Both types of tyrosine phosphatases share the common active-site motif of containing cysteine and arginine, both of which are essential for enzyme activity (6). Following the attack on the substrate phosphotyrosine, the essential cysteine in the active site forms a covalent thiol–phosphate intermediate (7–9). The adjacent histidine and conserved proximal arginine lower the pK a of the cysteine from 9.5 to 4.7 (6) which activates it with respect to accepting the phosphate group. The positively charged arginine serves to attract the negatively charged phosphate moiety of phosphotyrosine. The phosphate group is oriented and stabilized by numerous hydrogen bonds to ONH x groups of surrounding residues (9). A conserved tyrosine in the protein tyrosine phosphatase (PTP) 3 active site appears to ring-stack with the aromatic ring of phosphotyrosine thereby confering specificity for phosphotyrosine over phosphoserine or phosphothreonine (6). Nitric oxide is an endogenous free radical formed in a variety of cell types by the enzymatic oxidation of L-arginine to citrulline and is implicated in diverse physiological processes including vasorelaxation, platelet inhibition, and neurotransmission (10). Nitric oxide and nitrogen oxides formed by the slow reaction of nitric oxide with oxygen have been suggested to modulate thiol-dependent enzymes through the formation of S-nitrosothiols (11) and subsequent oxidation to give disulfides (12). The highly reactive thiolate group in tyrosine phosphatases could also be a potential target for transnitrosation by S-nitrosothiols (12–14). The chemical reactivity of nitric oxide is greatly increased by its diffusion-limited reaction with superoxide (O 2•2) to form peroxynitrite (ONOO 2) (15). Peroxynitrite nitrates tyrosine residues in proteins to form 3-nitrotyrosine, which can potentially interfere with subsequent phosphorylation of the affected tyrosine (16 –19). In addition, peroxynitrite readily oxidizes both free and protein-bound sulfhydryls (20) and induces oxidative inactivation of other sulfhydryl-containing enzymes such as alcohol dehydrogenase (21), creatine kinase (22), and succinate dehydrogenase (23). Consequently, peroxynitrite could potentially inactivate tyrosine phosphatases by either oxidizing the thiolate or nitrating the tyrosine residue in the active site (24). 3 Abbreviations used: PTPs, protein tyrosine phosphatase; GSNO, S-nitrosoglutathione; LAR, leukocyte–antigen-related; pNPP, p-nitrophenyl phosphate; DCDHF, dichlorodihydrofluorescein; DCF, dichlorofluorescein; DTT, dithiothreitol; SOD, superoxide dismutase.
Hydrogen peroxide (H 2O 2) has been shown to inactivate PTPs containing the CXXXXXR active-site motif in vitro and in cells (25–29). However, inactivation by H 2O 2 often requires many minutes even at the relatively high concentrations of H 2O 2 typically used (29). Peroxynitrite oxidizes both free and protein-bound thiols up to 5000 times faster than H 2O 2 and thiols can react directly as peroxynitrite anion (20). Thus, thiols flanked by cationic amino acid residues, such as the CXXXXXR-containing PTPs, could preferentially be targeted by peroxynitrite. In this study, we compared the effects of nitric oxide, S-nitrosoglutathione (GSNO), and peroxynitrite on activities of three recombinant human PTPs: leukocyte– antigen-related tyrosine phosphatase (a receptor-like tyrosine phosphatase), T-cell tyrosine phosphatase (a non-receptor-like tyrosine phosphatase), and protein tyrosine phosphatase 1B (PTP1B). The tyrosine phosphatases were exceptionally susceptible to peroxynitrite but not to nitric oxide or GSNO. Moreover, the amount of peroxynitrite required for inhibition was essentially the same regardless of whether peroxynitrite was added from synthetic stocks or formed in situ via cogeneration of superoxide and nitric oxide. MATERIALS AND METHODS Human recombinant leukocyte antigen-related (LAR) protein tyrosine phosphatase (40-kDa molecular mass), human recombinant T-cell (CD45) protein tyrosine phosphatase (95-kDa molecular mass), and p-nitrophenyl phosphate (pNPP) were obtained from Calbiochem (San Diego, CA). Human recombinant protein tyrosine phosphatase 1B (PTP1B) (37-kDa molecular mass) was obtained from Biomol (Plymouth, PA). DETA NONOate and GSNO were from Alexis (San Diego, CA). All other reagents were from Sigma. Peroxynitrite was prepared using a quenched-flow reaction apparatus as described by Reed et al. (30) and Beckman et al. (31). An aqueous solution of 0.6 M sodium nitrite was rapidly mixed with an equal volume of 0.7 M hydrogen peroxide containing 0.6 M HCl and immediately quenched with the same volume of 1.5 M NaOH. Stock solutions were typically 150 mM in peroxynitrite. Working solutions (2 mM) were prepared by dilution of stocks into 0.1 M semiconductor grade NaOH and concentrations of all solutions were standardized spectrally (« 5 1700 M 21 cm 21). Working solutions were kept on ice and concentrated stock solutions were stored at 280°C. PTP assays and inactivation by peroxynitrite. Tyrosine phosphatase activity was measured using the synthetic substrate pNPP (1 mM) in 1.4 ml of 50 mM Tris buffer, pH 7.4, at 37°C (32). The increase in absorbance at 402 nm (associated with hydrolysis of pNPP to p-nitrophenol) was monitored in real time using stirred glass cuvettes mounted in the temperature-controlled cuvette holder of a Hewlett–Packard diode-array spectrophotometer. Concentration responses for inactivation by peroxynitrite were determined by adding peroxynitrite to rapidly stirred cuvettes at 37°C containing enzyme and Tris buffer. Substrate (pNPP) was added 30 s after peroxynitrite and activity was monitored for 3 min; rates were expressed as percentages of the untreated (control) phosphatase rate. Unless otherwise noted, the final concentration of the three PTPs in the cuvettes were as follows: 0.29 mg/ml or 3 nM enzyme monomer for CD45, 0.48 mg/ml or 12 nM enzyme monomer for LAR, and 0.36 mg/ml or 9.6 nM enzyme monomer for PTP1B. Determinations of activity versus pNPP substrate concentration revealed no increase in
INACTIVATION OF PROTEIN TYROSINE PHOSPHATASES BY PEROXYNITRITE
SCHEME 1
activity above 1 mM; thus, 1 mM was used for all activity assays unless otherwise noted. Determination of rate constants via competition assay. Because peroxynitrite decomposes so rapidly, it is not possible to determine an inactivation rate constant by measuring phosphatase activity over time as has been done with H 2O 2 (29). Therefore, bimolecular rate constants for reaction of peroxynitrite with tyrosine phosphatases were determined by means of a competition assay as previously described (21, 33, 34). Based on the concentration response for peroxynitrite-mediated inactivation, a concentration of peroxynitrite was chosen which would inhibit .90% but ,100% of the original PTP activity so that protection by cysteine could be precisely measured. Solutions of PTPs were prepared in Tris buffer and cysteine was added immediately before peroxynitrite. Remaining PTP activity was determined by adding 1 mM pNPP 30 s after peroxynitrite and monitoring hydrolysis at 402 nm for 3 min. Under the conditions used here, the possible fates for peroxynitrite include reaction with PTP, reaction with cysteine, and spontaneous rearrangement (decay) to nitrate (see Scheme 1). The rate constants for these three processes are k PTP, k cys, and k decay, respectively. The rate constant determination for the peroxynitrite/PTP reaction (k PTP) relies solely on the relative concentrations of competing targets (PTP and cysteine) for peroxynitrite and the known rate constant for the peroxynitrite/cysteine reaction. The rate of spontaneous decay of peroxynitrite (k decay) can be ignored because the fraction of peroxynitrite disappearing via decay is maximal at zero cysteine and decreases as the concentration of cysteine is increased; k decay would be relevant only if the fraction of peroxynitrite lost via decay somehow increased, thereby making less total peroxynitrite available to react with either PTP or cysteine. Cysteine reacts with peroxynitrite at a known rate of 5 3 10 3 M 21 s 21 at pH 7.4 and 37°C (20). Thus, by comparing the ability of different cysteine concentrations to prevent peroxynitrite-mediated inactivation of a fixed concentration of PTP enzyme monomer, the rate constant for the peroxynitrite/PTP reaction can be calculated (see Scheme 1) (21). Derivations of the equation describing the competition between cysteine and PTPs have been published previously (21, 34). Briefly, Eq. [1] was used to determine k PTP. 1 F@PTP# 5 @Cys#. ~1 2 F!k Cys k PTP
[1]
The term F describes the fraction of PTP inactivated by peroxynitrite. When F[PTP]/(1 2 F)k Cys is plotted as a function of cysteine concentration, the reciprocal of the slope gives k PTP in units of M 21 s 21 (21, 34, 35). As a further verification of our reported rate constants, we calculated them using an alternate method as described by Masumoto and Sies (36). This latter method involves plotting percent-
199
age remaining enzyme activity as a function of cysteine concentration and fitting data to an equation for first-order decay in order to determine the cysteine concentration providing 50% protection. At a concentration of cysteine which gives exactly 50% inactivation (or 50% inactivation), the rate constant for reaction of peroxynitrite with PTP is equal to the ratio of [PTP]/[cysteine] times the rate constant for the peroxynitrite/cysteine reaction. By definition, [PTP] would be one-half of the initial concentration and [cysteine] would be that concentration which provides 50% protection. k cysteine 5 5000 M 21 s 21 at pH 7.4 and 37°C. This latter method provided rate constants for the peroxynitrite/PTP reaction which were virtually identical to those obtained by the competition Eq. [1] shown above. Nitric oxide production from DETA NONOate and treatment of LAR phosphatase. The amount of NO generated by DETA NONOate was determined using an electrochemical probe (World Precision Instruments) mounted in a stirred, temperature-controlled vessel. A standard curve for probe response as a function of nitric oxide was prepared by adding small volumes of a saturated aqueous solution of nitric oxide (1.9 mM) to 1 ml of 50 mM Tris buffer, pH 7.4, as described previously (31). A known amount of DETA NONOate, which had been standardized spectrally (« 5 7940 M 21 cm 21 at 252 nm), was added to a 1-ml solution of Tris buffer, pH 7.4, at 37°C. DETA NONOate has a half-life of approximately 56 h under these conditions and releases 2 mol of nitric oxide per mole of DETA NONOate. At a concentration of 0.4 mM, DETA NONOate was found to produce a steady-state concentration of 2 mM nitric oxide. LAR phosphatase (0.12 mM enzyme monomer) was incubated for 3 h at 37°C with 0.4 mM DETA NONOate in 0.14 ml of 50 mM Tris buffer. After 3 h, the reaction solution was diluted to a final volume of 1.4 ml in 50 mM Tris buffer and assayed for activity after adding 1 mM pNPP. Determination of peroxynitrite generation from SIN-1. In simple buffered solutions, SIN-1 decomposes to produce equimolar amounts of superoxide and nitric oxide which combine quantitatively to form peroxynitrite (37, 38). Production of peroxynitrite from SIN-1 over time was quantified by measuring the oxidation of dichlorodihydrofluorescein (DCDHF) to dichlorofluorescein (DCF) spectrally (« 500 21 nm 5 59,500 M cm 21) (38) under the same conditions used for SIN-1 inactivation of PTPs (21). At pH 7.4 and 37°C, peroxynitrite production from SIN-1 decomposition reaches a steady state after approximately 3 min and remains essentially linear for up to 60 min (21, 38). Synthetic peroxynitrite oxidizes DCDHF to DCF with 38% efficiency, i.e., 1 mM peroxynitrite will produce 0.38 mM DCF (38). Thus, formation of DCF from SIN-1 was converted to total peroxynitrite production by dividing by 0.38. The amount of SIN-1-generated peroxynitrite required to inactivate PTPs was determined by comparing the time required for 50% PTP inactivation (corrected for spontaneous activity loss under turnover conditions) to the amount of peroxynitrite formed from SIN-1 during the same interval. Western blot analysis for protein nitration. Western blot analysis was performed as described previously (24, 39). Briefly, peroxynitrite was added to recombinant PTP1B in 60 ml of 50 mM Tris buffer, pH 7.4, while vortex mixing. PTP activity was immediately determined by adding 10-ml aliquots of reaction mixes to 1.4 ml standard activity assay mixes containing pNPP. The remaining 50 ml (2.5 mg) from each reaction was subjected to SDS–PAGE and transferred to nitrocellulose for nitrotyrosine immunoblotting.
RESULTS
Peroxynitrite-mediated inactivation of PTPs. Peroxynitrite inactivated T-cell tyrosine phosphatase (CD45), LAR tyrosine phosphatase, and PTP1B tyrosine phosphatase with IC 50 values of 0.3, 0.7, and 0.9 mM, respectively (Fig. 1a). Addition of a 100 mM peroxynitrite solution which had been allowed to decom-
200
TAKAKURA ET AL.
oxynitrite increased to 1.4 mM for CD45 and 5.8 mM for LAR tyrosine phosphatases but was slightly decreased (0.6 mM) with PTP1B (Fig. 1b). These results indicate that the peroxynitrite/carbon dioxide adduct is also capable of potently inactivating the PTPs, although subtle differences may exist with respect to recognition of the adduct compared with peroxynitrite anion. Peroxynitrite-mediated inactivation is largely irreversible. Previous studies have shown that the loss of PTP activity resulting from treatment with H 2O 2 was reversed by thiol reductants such as DTT (29, 44). Incubation of peroxynitrite-inactivated LAR phosphatase with 20 mM DTT for up to 1 h resulted in the recovery of only about 10% of the original activity (Fig. 2); similar limited activity recoveries were seen with CD45 and PTP1B phosphatases. LAR activity was inhibited to 20% of its untreated activity by 2 mM peroxynitrite. Addition of DTT to LAR phosphatase immediately prior to 2 mM peroxynitrite prevented inhibition in a dose-dependent manner (Fig. 2). No effect on phosphatase activity was seen when 100 mM peroxynitrite and 10 mM DTT were premixed immediately before addition to the LAR reaction mix (not shown), demonstrating that the protective effect of DTT was not due to any product of the peroxynitrite/DTT reaction. Rate constant determination. The rate of peroxynitrite-mediated inactivation of PTPs is too fast to permit use of PTP activity loss over time as a means of deter-
FIG. 1. Inactivation of protein tyrosine phosphatases by peroxynitrite and the effects of carbon dioxide. (a) PTPs were treated via bolus additions of peroxynitrite to rapidly stirred reaction cuvettes containing 50 mM Tris buffer, pH 7.4, at 37°C and assayed within 30 s using 1 mM pNPP (see Materials and Methods). Reaction cuvettes contained 3 nM CD45 PTP enzyme monomer, 12 nM LAR PTP enzyme monomer, or 9.6 nM PTP1B enzyme monomer and the indicated final concentrations of peroxynitrite. (b) Conditions were identical to those in (a) except that 25 mM sodium bicarbonate, which had been pH adjusted to 7.4, was added to the Tris buffer prior to adding peroxynitrite. PTP activity was not affected by the presence of bicarbonate.
pose in 50 mM Tris buffer, pH 7.4, had no effect on phosphatase activity. Likewise, peroxynitrite decomposition products nitrite (0.2 mM) and nitrate (0.2 mM) had no effect on phosphatase activity. Because carbon dioxide has been shown to react with and alter the reactivity of peroxynitrite (40 – 43), PTP inactivation by peroxynitrite was compared in the presence of a carbon dioxide-saturated (25 mM bicarbonate) Trisbuffered solution. The IC 50 for inactivation by per-
FIG. 2. Effects of DTT on peroxynitrite-mediated inactivation of PTPs. Reactions contained 12 nM LAR PTP in 50 mM Tris buffer, pH 7.4, and the indicated concentrations of DTT added prior to (■) 2 mM peroxynitrite; 1 mM pNPP was added immediately after peroxynitrite to determine the residual activity. The ability of DTT to reverse the activity loss induced by 2 mM peroxynitrite was determined by adding 10 mM DTT immediately after peroxynitrite (F) and incubating for 30 min prior to assaying for activity.
201
INACTIVATION OF PROTEIN TYROSINE PHOSPHATASES BY PEROXYNITRITE TABLE I
Rate Constants for Reaction of PTPs with Peroxynitrite
FIG. 3. Determination of rate constants for reaction of peroxynitrite with PTPs. Reaction mixtures contained the following concentrations of PTPs in 50 mM Tris buffer, pH 7.4, at 37°C: 0.29 mg/ml or 3 nM enzyme monomer for CD45, 0.96 mg/ml or 24 nM enzyme monomer for LAR, and 0.36 mg/ml or 9.6 nM enzyme monomer for PTP1B. Peroxynitrite concentrations which were found to inhibit .90 but ,100% of the activity were as follows: 3.8 mM peroxynitrite for CD45, 5 mM for PTP1B, and 10 mM for LAR. Cysteine, at the indicated concentrations, was added to solutions of PTPs in Tris buffer immediately before peroxynitrite. Remaining PTP activity was determined by adding 1 mM pNPP 30 s after peroxynitrite and monitoring hydrolysis at 402 nm for 3 min.
mining an inactivation rate constant. However, free cysteine reacts directly with peroxynitrite anion at a known rate (20). Thus, the rate of reaction between peroxynitrite and PTPs was determined by measuring protection as a function of increasing free cysteine concentrations (Fig. 3). By comparing the relative amounts of cysteine and enzyme monomer and the known rate constant for the cysteine/peroxynitrite reaction (EQ. [1]), rate constants for the PTP/peroxynitrite reaction were determined to be 2.0 3 10 8 M 21 s 21 for CD45 PTP, 2.3 3 10 7 M 21 s 21 for LAR PTP, and 2.2 3 10 7 M 21 s 21 for PTP1B (Table I). Effect of substrate on peroxynitrite-mediated inactivation. A concentration of peroxynitrite was chosen which would inactivate most but not all of the phosphatase activity so that effects due to substrate could be quantified. When 2 mM peroxynitrite was added to LAR in the absence of substrate (pNPP), only 20% of the original phosphatase activity remained (Fig. 4). Addition of 2 mM peroxynitrite to reaction mixtures containing LAR plus increasing amounts of substrate resulted in partial protection from inactivation. The protective effect saturated at approximately 2 mM pNPP which provided maximum protection of only 46% of the original activity (Fig. 4). Inactivation by SIN-1. Cogeneration of superoxide and nitric oxide by SIN-1 was used to compare the
Tyrosine phosphatase
Rate constant for reaction with peroxynitrite (M 21 s 21)
CD45 LAR PTP1B
2.0 3 10 8 2.3 3 10 7 2.2 3 10 7
Note. Each of the three PTPs was treated with a submaximal concentration of peroxynitrite at 37°C (Fig. 3) in the absence and presence of the indicated cysteine concentrations and immediately assayed for phosphatase activity (see Materials and Methods). Based on the known enzyme monomer concentration present ([PTP]), the biomolecular rate constant for peroxynitrite and cysteine (k SH), and the measured extent of protection from inactivation (F), the ordinate points were calculated and plotted versus cysteine concentration (EQ. [1]). The rate of reaction between peroxynitrite and phosphatase is determined by taking the inverse of the slope (1/k PTP).
effects of peroxynitrite formed in situ with bolus additions of synthetic peroxynitrite. SIN-1 releases both superoxide and nitric oxide at a constant rate leading to quantitative formation of peroxynitrite (37, 38). SIN-1 was added at 60 s to PTPs which were actively hydrolyzing pNPP and the time course for complete loss of phosphatase activity was monitored continuously (Fig. 5). In the presence of 0.1 mM SIN-1, the loss of LAR activity was considerably faster than the spontaneous activity loss that accompanies enzyme turnover. Complete inactivation oc-
FIG. 4. Protection from peroxynitrite-mediated inactivation by pNPP substrate. Reactions contained 12 nM LAR PTP in 50 mM Tris buffer, pH 7.4, and the indicated concentrations of pNPP added prior to 2 mM peroxynitrite. Immediately after peroxynitrite addition, more pNPP was added to give a final concentration of 3 mM in order to assay the remaining phosphatase activity.
202
TAKAKURA ET AL.
FIG. 5. Inhibition of LAR PTP by SIN-1 and protection by SOD. Reactions contained 24 nM LAR PTP in 50 mM Tris buffer, pH 7.4, and 1 mM pNPP at 37°C; pNPP hydrolysis to p-nitrophenol was monitored in real time at 402 nm. SIN-1 (0.1 mM) was added after 60 s to the reactions illustrated by broken lines. The reaction illustrated by the dotted line also contained 0.5 mg/ml bovine Cu,Zn SOD. SIN-1-mediated inactivation of PTP activity (dashed line) was taken as the time at which the absorbance change ceased—a time equal to 12 min of total SIN-1 exposure. The amount of peroxynitrite produced during this interval was determined under identical reaction conditions using dichlorodihydrofluorescein as the indicator of peroxynitrite production (see Materials and Methods). Neither SIN-1 nor its decomposition product SIN-1C absorbs at 402 nm.
curred approximately 840 s after SIN-1 addition (Fig. 5). The amount of peroxynitrite formed during this time was determined in a separate reaction using DCDHF oxidation as the measure of peroxynitrite production (21, 38). A comparison of the time required for 50% inactivation of LAR and the amount of peroxynitrite formed from SIN-1 during that in-
TABLE II
IC 50 Values for PTP Inhibition by SIN-1 Tyrosine phosphatase
IC 50 for inhibition by SIN-1 (mM)
CD45 LAR PTP1B
0.20 1.05 0.64
Note. PTP activity was monitored in real time at 37°C in the absence and presence of SIN-1 and the time required for 50% inactivation was determined (see Fig. 5 for illustration and Materials and Methods for details). Peroxynitrite formation from SIN-1 as a function of time was quantified in a separate reaction run under identical conditions. The amount of SIN-1-generated peroxynitrite required to inactivate PTPs was determined by comparing the time required for 50% PTP inactivation (corrected for spontaneous activity loss under turnover conditions) to the amount of peroxynitrite formed during the same interval.
terval produced an IC 50 equal to 1.05 mM peroxynitrite (Table II). Similar measurements for CD45 and PTP1B phosphatases produced IC 50 values of 0.2 and 0.64 mM, respectively. Addition of 0.5 mg/ml bovine copper, zinc superoxide dismutase (SOD), which scavenges superoxide and thereby inhibits peroxynitrite formation, completely protected the PTPs from SIN-1 inactivation (Fig. 5). Effects of nitric oxide and nitrosothiol donors. DETA NONOate releases nitric oxide continuously in a pH- and temperature-controlled manner. In a stirred vessel at pH 7.4 and 37°C, a steady-state concentration of 2 mM nitric oxide is obtained after 10 min at a concentration of 0.4 mM DETA NONOate (see Materials and Methods). Exposure of LAR to 0.4 mM DETA NONOate for 3 h resulted in a loss of only 12% of the original activity (Fig. 6a); CD45 and PTP1B were inhibited by less than 10% under the same conditions. Treatment with GSNO (0.1–1.0 mM) for 30 min inhibited LAR PTP activity in a concentration-dependent manner. The inhibitory effect of GSNO toward LAR PTP reached a maximum of 50% at a concentration of 1 mM (Fig. 6b). Unlike peroxynitrite-mediated inactivation, inhibition by GSNO was completely reversed by addition of 10 mM dithiothreitol.
INACTIVATION OF PROTEIN TYROSINE PHOSPHATASES BY PEROXYNITRITE
203
first appeared (Fig. 7, lane 6). These results reveal that inactivation is not correlated with nitration of proteinbound tyrosine. DISCUSSION
Peroxynitrite readily nitrates tyrosine residues in proteins, and peptides containing nitrotyrosine are resistant to phosphorylation in vitro (16, 42, 45), suggesting that peroxynitrite could alter phosphotyrosine-mediated signaling. Peroxynitrite has been shown to affect signal transduction in cultured cells and to induce apoptosis (46 – 48). However, tyrosine phosphorylation of proteins is typically increased rather than inhibited when cells are treated with peroxynitrite (49 –51). This led us to consider whether tyrosine phosphatases might be inactivated by peroxynitrite. The reaction rates between peroxynitrite anion and PTPs, particularly CD45, are among the fastest rates ever reported for peroxynitrite reacting with biological molecules. One possible explanation for the rapid rates emerges from an examination of the active site of PTP1B (9) and a comparison between binding of phosphate anion and peroxynitrite anion (Fig. 8). Orientation of the phosphate group of phosphotyrosine is determined by hydrogen bonding (Fig. 8, dotted lines) of the phosphate oxygens to ONH x groups of the residues (216 –221) surrounding the active-site cysteine 215 of PTP1B (9). It is reasonable to propose that peroxynitrite anion reacts with cys 215 faster than with other cysteine residues for three reasons: (i) charge attraction by the essential arginine 221 cation occurs, (ii) per-
FIG. 6. Effects of DETA NONOate and GSNO on PTP activity. (a) LAR PTP was incubated with DETA NONOate for 3 h as described under Materials and Methods, diluted 10-fold, and assayed for activity via addition of pNPP. (b) LAR PTP (12 nM enzyme monomer) in 50 mM Tris buffer, pH 7.4, at 37°C was incubated with the indicated concentrations of S-nitrosoglutathione for 30 min and immediately assayed by the addition of 1 mM pNPP (F). After 30 min incubation with 1 mM GSNO, 10 mM DTT was added to the LAR PTP reaction solution and activity was added after 30 min (■).
Peroxynitrite-mediated tyrosine nitration of PTPs. PTPs possess a conserved tyrosine residue near the active site which appears to confer specificity for phosphotyrosine over phosphoserine or phosphothreonine (6). Although peroxynitrite is capable of nitrating tyrosine residues in proteins, nitration does not appear to play a role in PTP inactivation (Fig. 7). Reaction mixtures containing 50 mg/ml (1.33 mM enzyme monomer) PTP1B were treated with increasing concentrations of peroxynitrite (5–160 mM) and immediately assayed for activity and nitrotyrosine content via antinitrotyrosine immunoblot analysis (Fig. 7). PTP1B was inactivated by 95% at the point at which nitrotyrosine
FIG. 7. Comparison of PTP1B activity with protein nitration via nitrotyrosine immunoblot. Peroxynitrite (0, 5, 10, 20, 40, 80, and 160 mM) was added to recombinant PTP1B (50 mg/ml or 1.33 mM enzyme monomer) in 60 ml of 50 mM Tris buffer, pH 7.4, while vortex mixing. PTP activity was immediately determined by adding 10-ml aliquots of reaction solutions to 1.4-ml standard activity assay mixes containing pNPP (see Materials and Methods). The remaining 50 ml (2.5 mg protein) from each reaction was subjected to SDS–PAGE and nitrotyrosine immunoblot as described under Materials and Methods.
204
TAKAKURA ET AL.
FIG. 8. Active-site residues of PTP1B: Illustration of relative molecular size and hydrogen bonding between phosphate and peroxynitrite anions. The active-site residues 216 –221 and the cysteine 215–phosphate intermediate (structure at left) were taken from the X-ray crystal structure of recombinant human PTP1B (Brookhaven Database Entry 1A5Y) (9). The entire X-ray crystal structure of PTP1B was imported into the molecular visualization program Quantum Cache 3.0 (Oxford Molecular Group, Eugene, OR) and edited to show only the active-site residues (minus noninvolved hydrogens) without altering the original X-ray coordinates. The structure on the right is identical to that at left except that cis peroxynitrite anion has been superimposed over the dotted spheres of active-site cysteine designated with the yellow C and the bound phosphate (labeled P). The structure of cis peroxynitrite anion is based on several earlier physical– chemical studies (56, 68) and is in agreement with recent X-ray crystal data (69). Peroxynitrite anion was superimposed onto the active site of PTP1B (structure at right) and rotated through all planes to determine potential alignments with the known coordinates of the phosphate oxygen atoms (note two asterisks). The structure at right illustrates just one of several such alignments wherein two of the oxygen atoms of peroxynitrite precisely superimposed over two of the phosphate oxygen atoms. Hydrogen bonding interactions between enzyme ONH x groups and the cysteinebound phosphate are those proposed by Pannifer et al. (9). Hydrogen bonding interactions with peroxynitrite anion are proposed as an additional explanation for the fast reaction rate constants.
oxynitrite and phosphate are very similar in size and interatomic distance as evidenced by the precise alignment of two oxygen atoms of cis peroxynitrite anion with two of the phosphate oxygens (Fig. 8, asterisks), and (iii) every atom of peroxynitrite is close enough to hydrogen bond with ONH groups of the surrounding residues. Figure 8 shows just one of several potential orientations of peroxynitrite anion, any one of which is compatible with numerous hydrogen bonding interactions and subsequent oxidation of cys 215. Fast reaction rates suggest that PTPs could be targeted for peroxynitrite-mediated inactivation even in the presence of much higher concentrations of endogenous scavengers like glutathione. For example, the peroxynitrite reaction rate with CD45 is 2.0 3 10 8 M 21 s 21 compared to 5 3 10 3 M 21 s 21 for glutathione, a difference of 40,000. Thus, 25 nM CD45 could be inac-
tivated by peroxynitrite even in the presence of 1 mM glutathione. Both the in vivo concentration of carbon dioxide and its rate of reaction with peroxynitrite are quite similar to those of glutathione (52). However, in the case of carbon dioxide, the reaction is catalytic and generates an adduct which is only slightly less inhibitory toward PTPs (Fig. 1) and thiols in general (43). The potential for peroxynitrite-mediated inactivation of PTPs in vivo becomes all the more relevant when considering that it is irreversible. Remarkably low concentrations (,0.9 mM) of peroxynitrite irreversibly inactivated three human tyrosine phosphatases, whereas continuous generation of micromolar concentrations of nitric oxide over periods of up to 3 h did not. Because the half-lives of oxidants vary, relative toxicities of reactive species must be assessed by comparing the concentration over time
INACTIVATION OF PROTEIN TYROSINE PHOSPHATASES BY PEROXYNITRITE
(area under the curve or total exposure) rather than by initial concentration (53, 54). Peroxynitrite decomposes with a half-life of approximately 1.0 s at pH 7.4 and 37°C (55) and produced 50% inhibition of PTPs at concentrations ranging from 0.3 to 0.9 mM. Thus, total peroxynitrite exposure equaled 0.3– 0.9 mM 3 s. In contrast, exposure to nitric oxide generated from DETA NONOate was 21,600 mM 3 s (2 mM 3 10,800 s)—an exposure which had little effect on PTP activity. Exposure to GSNO was 180,000 mM 3 s (1000 mM 3 1800 s) and inhibited PTP activity by 50%, but inhibition was completely reversible. A simple comparison reveals that peroxynitrite is at least 100,000 times more inhibitory to PTPs than is nitric oxide or GSNO. Several reports indicate that H 2O 2 can inhibit PTPs (25–29). Unlike peroxynitrite, inactivation by H 2O 2 appears to be completely reversible by thiol reductants (29). Extrapolation of data from Denu and Tanner (29) reveals that 0.9 mM H 2O 2 inactivated PTP1B by 50% within approximately 17 s 4 or a total H 2O 2 exposure of 15,300 mM 3 s (900 mM 3 17 s). Thus, peroxynitrite is roughly 17,000-fold more inhibitory to PTP1B than is H 2O 2. Formation of peroxynitrite in vivo is most likely to occur via the radical–radical combination of superoxide and nitric oxide (56, 57). A recent report suggested that cogeneration of superoxide and nitric oxide was both qualitatively and quantitatively different from peroxynitrite with respect to modification of target molecules (58). With regard to PTP inhibition, we found precisely the opposite—namely that slow, continuous cogeneration of superoxide and nitric oxide with SIN-1 inhibited 50% of PTP activity at virtually identical concentrations of peroxynitrite. The equivalence of SIN-1 and bolus peroxynitrite has also been seen previously with respect to inactivation of the yeast alcohol dehydrogenase (21). Recent experiments designed to directly compare synthetic (bolus) peroxynitrite to rapid mixing of nitric oxide with potassium superoxide have also revealed their equivalency with regard to the yield of nitrotyrosine (C. Reiter and J. S. Beckman, unpublished observations). SOD scavenges the superoxide produced during SIN-1 decomposition and thereby converts SIN-1 predominantly into a nitric oxide donor (59). The ability of SOD to completely prevent PTP inactivation indicates that the effects of SIN-1 are mediated by peroxynitrite and provides additional evidence that nitric oxide alone is not inhibitory. The protein tyrosine phosphatases have an unusually acidic cysteine in the active site with a pK a esti4 The time required for 50% inhibition of PTP1B by 0.9 mM H 2O 2 was determined from Fig. 1 of the paper by Denu and Tanner (29) which reveals the hyperbolic loss of activity over a 100-s interval at pH 7 and 30°C.
205
mated to be around 4.7 (6). Thus, this cysteine exists as the thiolate anion at physiological pH, making it susceptible to oxidation. The active-site cysteine functions as a nucleophile to accept the OPO 3 moiety from phosphotyrosine (7, 8) thereby generating a phosphocysteine intermediate (9). Site-directed mutagenesis has confirmed that this cysteine residue is essential for catalysis in both receptor-like and non-receptor-like protein tyrosine phosphatases (60, 61). Based on previous studies which unequivocally identified the activesite cysteine as the target of H 2O 2-mediated inactivation (29, 62), it is reasonable to suggest that oxidation of this cysteine thiolate is also the mechanism of peroxynitrite-mediated inactivation. It is clear that peroxynitrite-mediated inactivation is not dependent on tyrosine nitration since activity loss occurred at peroxynitrite concentrations below those at which nitration became apparent (Fig. 7). Carbon dioxide forms an adduct with peroxynitrite which is more efficient at nitrating tyrosine than at oxidizing thiols (43). In the presence of carbon dioxide, a fivefold higher concentration of peroxynitrite was required to produce the same inhibition with at least two of the PTPs. This finding is consistent with inhibition via thiolate oxidation and not tyrosine nitration. Treatment with DTT after peroxynitrite-mediated inhibition of PTPs resulted in less than 10% recovery of activity. This contrasts with inactivation by H 2O 2 which is completely reversed by thiol reductants (29). Thiol oxidation may result in the formation of sulfenic acid (OSOH) or in direct disulfide formation, either of which can be reduced back to OSH with reductants such as DTT (20). Therefore, the inability of DTT to restore activity to peroxynitrite-treated PTPs suggests that the active-site thiol is oxidized to sulfinic acid (OSO 2H) or even to sulfonic acid (OSO 3H) which cannot be reduced by the low-molecular-weight antioxidants present in cells such as glutathione and ascorbate (34, 63). Because DTT was unable to restore activity, its ability to prevent peroxynitrite-mediated inhibition (when added first) is most likely related to direct scavenging of peroxynitrite. Relatively high amounts of GNSO partially inhibited phosphatase activity, but the inhibition was rapidly reversed by low-molecular-weight thiols such as DTT and mercaptoethanol. Partial, reversible inhibition by the nitrosothiol GSNO, but not by nitric oxide itself, is consistent with a mechanism involving direct nitrosation of the active-site thiolate by GSNO. Transnitrosation reactions involving transfer of NO from a nitrosothiol to a reduced thiol (OSH) have been shown to occur quite readily in in vitro systems (12–14). Rates of transnitrosation increase with pH indicating that thiolate anions are better acceptors of the nitrosating species (11, 64). At physiological pH, the active-site cysteine (pK a 5 4.7) (6) of PTPs would exist in the ionized
206
TAKAKURA ET AL.
form which should be more susceptible than other protein thiols to nitrosation by GSNO. It is quite possible that previous reports of PTP inactivation in cells by compounds such as S-nitroso-N-acetylpenicillamine (SNAP) or SIN-1 (65, 66) could have been related to transnitrosation or peroxynitrite-mediated inactivation, respectively. Pretreatment with the PTP substrate pNPP afforded some protection from peroxynitrite-mediated inactivation, providing additional support for active-site thiol oxidation as the mechanism of inactivation. Studies which have assessed PTP inactivation by H 2O 2 have reported that saturating concentrations of pNPP provided complete protection (29). This difference can be explained in terms of the much faster reaction rate of peroxynitrite with PTPs. That is to say, transient occupation of the active site by a substrate during catalytic turnover will block a slow interaction such as H 2O 2 much more effectively than a fast one such as peroxynitrite. If pNPP were acting as a direct scavenger of peroxynitrite, then protection should have been linear with pNPP concentration, which was not observed. Recent demonstration that H 2O 2 is capable of reversibly inactivating PTPs both in vitro and in cells (25–29, 44) has ignited interest in the concept of oxidative regulation of tyrosine phosphatase activity and, by extension, in oxidative regulation of phosphotyrosine-mediated cell signaling. In that regard, it is especially noteworthy that the rates of PTP inactivation by peroxynitrite are many orders of magnitude faster than inactivation by H 2O 2. More importantly, inactivation by peroxynitrite is not reversed by thiol reductants. Mallozzi et al. (67) reported an initial decrease in total PTP activity in human erythrocytes treated with low concentrations of peroxynitrite followed by a time-dependent recovery of PTP activity; similar findings have been reported with human platelets (50). These results suggest that at least some cell types may have mechanisms for restoring activity to oxidized PTPs beyond simple low-molecular-weight reductants such as glutathione; the possibility of enzymatic reduction of oxidized PTPs warrants further investigation. The ability of peroxynitrite to inactivate PTPs much faster and at much lower concentrations than H 2O 2 argues that peroxynitrite may be the more likely biological mediator of PTP inactivation and subsequent enhancement of tyrosine phosphorylation under conditions of oxidative stress. Physiological concentrations of nitric oxide did not inhibit PTPs and inhibition of PTPs by the nitrosothiol GSNO was slow even at a concentration of 1 mM. Moreover, the rapid reversal of GSNO-mediated inhibition by DTT suggests that nitrosothiols would be very poor inhibitors in the presence of low-molecular-weight thiols in vivo. Thus, in situations where nitric oxide synthase inhibitors pre-
vent the loss of PTP activity in cells or tissues, it is reasonable to implicate peroxynitrite. It is also reasonable to conclude that endogenous formation of peroxynitrite and subsequent inactivation of PTPs may profoundly alter cellular physiology by enhancing protein tyrosine phosphorylation. REFERENCES 1. Fischer, E. H., Charbonneau, H., and Tonks, N. K. (1991) Science 253, 401– 406. 2. Pot, D. A., and Dixon, J. E. (1992) Biochim. Biophys. Acta 1136, 35– 43. 3. Walton, K. M., and Dixon, J. E. (1993) Annu. Rev. Biochem. 62, 101–120. 4. Sun, H., and Tonks, N. K. (1994) Trends Biochem. Sci. 19, 480 – 485. 5. Inagaki, N., Ito, M., Nakano, T., and Inagaki, M. (1994) Trends Biochem. Sci. 19, 448 – 452. 6. Fauman, E. B., and Saper, M. A. (1996) Trends Biochem. Sci. 21, 413– 417. 7. Guan, K. L., and Dixon, J. E. (1991) J. Biol. Chem. 266, 17026 – 17030. 8. Denu, J. M., Lohse, D. L., Vijayalakshmi, J., Saper, M. A., and Dixon, J. E. (1996) Proc. Natl. Acad. Sci. USA 93, 2493–2498. 9. Pannifer, A. D. B., Flint, A. J., Tonks, N. K., and Barford, D. (1998) J. Biol. Chem. 273, 10454 –10462. 10. Moncada, S., and Higgs, E. A. (1991) Eur. J. Clin. Invest. 21, 361–374. 11. Stamler, J. S. (1995) Curr. Top. Micro. Immunol. 196, 19 –36. 12. Singh, R. J., Hogg, N., Joseph, J., and Kalyanaraman, B. (1996) J. Biol. Chem. 271, 18596 –18603. 13. Liu, Z., Rudd, M. A., Freedman, J. E., and Loscalzo, J. (1998) J. Pharm. Exp. Ther. 284, 526 –534. 14. Rossi, R., Lusini, L., Giannerini, F., Giustarini, D., Lungarella, G., and di Simplicio, P. (1997) Anal. Biochem. 254, 215–220. 15. Kissner, R., Nauser, T., Bugnon, P., Lye, P. G., and Koppenol, W. H. (1998) Chem. Res. Toxicol. 11, 557. 16. Martin, B. L., Wu, D., Jakes, S., and Graves, D. J. (1990) J. Biol. Chem. 265, 7108 –7111. 17. Gow, A. J., Duran, D., Malcolm, S., and Ischiropoulos, H. (1996) FEBS Lett. 385, 63– 66. 18. Berlett, B. S., Friguet, B., Yim, M. B., Chock, P. B., and Stadtman, E. R. (1996) Proc. Natl. Acad. Sci. USA 93, 1776 –1780. 19. Li, X. H., Desarno, P., Song, L., Beckman, J. S., and Jope, R. S. (1998) Biochem. J. 331(2), 599 – 606. 20. Radi, R., Beckman, J. S., Bush, K. M., and Freeman, B. A. (1991) J. Biol. Chem. 266, 4244 – 4250. 21. Crow, J. P., Beckman, J. S., and Mccord, J. M. (1995) Biochemistry 34, 3544 –3552. 22. Konorev, E. A., Hogg, N., and Kalyanaraman, B. (1998) FEBS Lett. 427, 171–174. 23. Rubbo, H., Denicola, A., and Radi, R. (1994) Arch. Biochem. Biophys. 308, 96 –102. 24. MacMillan-Crow, L. A., Crow, J. P., and Thompson, J. A. (1998) Biochemistry 37, 1613–1622. 25. Hadari, Y. R., Geiger, B., Nadiv, O., Sabanay, I., Roberts, C. T., Jr., LeRoith, D., and Zick, Y. (1993) Mol. Cell. Endocrinol. 97, 9 –17. 26. Sullivan, S. G., Chiu, D. T., Errasfa, M., Wang, J. M., Qi, J. S., and Stern, A. (1994) Free Radical Biol. Med. 16, 399 – 403.
INACTIVATION OF PROTEIN TYROSINE PHOSPHATASES BY PEROXYNITRITE 27. Natarajan, V., Scribner, W. M., Alhassani, M., and Vepa, S. (1998) Environ. Health Perspect. 106(5), 1205–1212. 28. Carbajal, J. M., and Schaeffer, R. C. (1998) Biochem. Biophys. Res. Commun. 249, 461– 466. 29. Denu, J. M., and Tanner, K. G. (1998) Biochemistry 37, 5633– 5642. 30. Reed, J. W., Ho, H. H., and Jolly, W. L. (1974) J. Am. Chem. Soc. 96, 1248 –1249. 31. Beckman, J. S., Chen, J., Ischiropoulos, H., and Crow, J. P. (1994) Methods Enzymol. 233, 229 –240. 32. Tonks, N. K., Charbonneau, H., Diltz, C. D., Fischer, E. H., and Walsh, K. A. (1988) Biochemistry 27, 8695– 8701. 33. Castro, L., Rodriguez, M., and Radi, R. (1994) J. Biol. Chem. 269, 29409 –29415. 34. Radi, R. (1996) Methods Enzymol. 269, 354 –366. 35. Souza, J. M., and Radi, R. (1998) Arch. Biochem. Biophys. 360, 187–194. 36. Masumoto, H., and Sies, H. (1996) Chem. Res. Toxicol. 9, 1057– 1062. 37. Hogg, N., Darley-Usmar, V. M., Wilson, M. T., and Moncada, S. (1992) Biochem. J. 281, 419 – 424. 38. Crow, J. P. (1997) Nitric Oxide 1, 145–157. 39. MacMillan-Crow, L. A., Crow, J. P., Kerby, J. D., Beckman, J. S., and Thompson, J. A. (1996) Proc. Natl. Acad. Sci. USA 93, 11853–11858. 40. Lymar, S. V., and Hurst, J. K. (1996) Chem. Res. Toxicol. 9, 845– 850. 41. Lymar, S. V., Jiang, Q., and Hurst, J. K. (1996) Biochemistry 35, 7855–7861. 42. Gow, A., Duran, D., Thom, S. R., and Ischiropoulos, H. (1996) Arch. Biochem. Biophys. 333, 42– 48. 43. Zhang, H., Squadrito, G. L., Uppu, R. M., Lemercier, J. N., Cueto, R., and Pryor, W. A. (1997) Arch. Biochem. Biophys. 339, 183–189. 44. Lee, S. R., Kwon, K. S., Kim, S. R., and Rhee, S. G. (1998) J. Biol. Chem. 273, 15366 –15372. 45. Kong, S. K., Yim, M. B., Stadtman, E. R., and Chock, P. B. (1996) Proc. Natl. Acad. Sci. USA 93, 3377–3382. 46. Estevez, A. G., Radi, R., Barbeito, L., Shin, J. T., Thompson, J. A., and Beckman, J. S. (1995) J. Neurochem. 65, 1543–1550. 47. Spear, N., Estevez, A. G., Barbeito, L., Beckman, J. S., and Johnson, G. V. (1997) J. Neurochem. 69, 53–59. 48. Estevez, A. G., Spear, N., Manuel, S. M., Radi, R., Henderson, C. E., Barbeito, L., and Beckman, J. S. (1998) J. Neurosci. 18, 923–931.
207
49. Hogg, N., Joseph, J., and Kalyanaraman, B. (1994) Arch. Biochem. Biophys. 314, 153–158. 50. Mondoro, T. H., Shafer, B. C., and Vostal, J. G. (1997) Free Radical Biol. Med. 22, 1055–1063. 51. Spear, N., Estevez, A. G., Johnson, G. V. W., Bredesen, D. E., Thompson, J. A., and Beckman, J. S. (1998) Arch. Biochem. Biophys. 356, 41– 45. 52. Denicola, A., Freeman, B. A., Trujillo, M., and Radi, R. (1996) Arch. Biochem. Biophys. 333, 49 –58. 53. Zhu, L., Gunn, C., and Beckman, J. S. (1992) Arch. Biochem. Biophys. 298, 452– 457. 54. Beckman, J. S., and Koppenol, W. H. (1996) Am. J. Physiol. 271, C1424 –C1437. 55. Beckman, J. S., Beckman, T. W., Chen, J., Marshall, P. A., and Freeman, B. A. (1990) Proc. Natl. Acad. Sci. USA 87, 1620 –1624. 56. Koppenol, W. H., Moreno, J. J., Pryor, W. A., Ischiropoulos, H., and Beckman, J. S. (1992) Chem. Res. Toxicol. 5, 834 – 842. 57. Beckman, J. S., and Crow, J. P. (1993) Biochem. Soc. Trans. 21, 330 –334. 58. Pfeiffer, S., and Mayer, B. (1998) J. Biol. Chem. 273, 27280 – 27285. 59. Haddad, I. Y., Crow, J. P., Hu, P., Ye, Y., Beckman, J., and Matalon, S. (1994) Am. J. Physiol. 267, L242–L249. 60. Streuli, M., Krueger, N. X., Tsai, A. Y., and Saito, H. (1989) Proc. Natl. Acad. Sci. USA 86, 8698 – 8702. 61. Streuli, M., Krueger, N. X., Thai, T., Tang, M., and Saito, H. (1990) EMBO J. 9, 2399 –2407. 62. Degnore, J. P., Konig, S., Barrett, W. C., Chock, P. B., and Fales, H. M. (1998) Rapid Commun. Mass Spectrom. 12, 1457–1462. 63. Quijano, C., Alvarez, B., Gatti, R. M., Augusto, O., and Radi, R. (1997) Biochem. J. 322, 167–173. 64. Keshive, M., Singh, S., Wishnok, J. S., Tannenbaum, S. R., and Deen, W. M. (1996) Chem. Res. Toxicol. 9, 988 –993. 65. Estrada, C., Gomez, C., Martin-Nieto, J., De Frutos, T., Jimenez, A., and Villalobo, A. (1997) Biochem. J. 326, 369 –376. 66. Caselli, A., Chiarugi, P., Camici, G., Manao, G., and Ramponi, G. (1995) FEBS Lett. 374, 249 –252. 67. Mallozzi, C., Distasi, A. M. M., and Minetti, M. (1997) FASEB J. 11, 1281–1290. 68. Hamilton, T. P., Tsai, H. H., and Beckman, J. S. (1996) Methods Enzymol. 269, 329 –346. 69. Worle, M., Latal, P., Kissner, R., Nesper, R., and Koppenol, W. H. (1999) Chem. Res. Toxicol. 12, 305–307.