Inhibition of Protein-tyrosine Phosphatases by Mild Oxidative Stresses Is Dependent on S-Nitrosylation*

Previous studies have shown that a Ca2+-dependent nitric-oxide synthase (NOS) is activated as part of a cellular response to low doses of ionizing radiation. Genetic and pharmacological inhibitor studies linked this NO signaling to the radiation-induced activation of ERK1/2. Herein, a mechanism for the radiation-induced activation of Tyr phosphorylation-dependent pathways (e.g. ERK1/2) involving the inhibition of protein-Tyr phosphatases (PTPs) by S-nitrosylation is tested. The basis for this mechanism resides in the redox-sensitive active site Cys in PTPs. These studies also examined oxidative stress induced by low concentrations of H2O2. S-Nitrosylation of total cellular PTP and immunopurified SHP-1 and SHP-2 was detected as protection of PTP enzymatic activity from alkylation by N-ethylmaleimide and reversal by ascorbate. Both radiation and H2O2 protected PTP activity from alkylation by a mechanism reversible by ascorbate and inhibited by NOS inhibitors or expression of a dominant negative mutant of NOS-1. Radiation and H2O2 stimulated a transient increase in cytoplasmic free [Ca2+]. Radiation, H2O2, and the Ca2+ ionophore, ionomycin, also stimulated NOS activity, and this was associated with an enhanced S-nitrosylation of the active site Cys453 determined by isolation of S-nitrosylated wild type but not active site Cys453 → Ser SHP-1 mutant by the “biotin-switch” method. Thus, one consequence of oxidative stimulation of NO generation is S-nitrosylation and inhibition of PTPs critical in cellular signal transduction pathways. These results support the conclusion that a mild oxidative signal is converted to a nitrosative one due to the better redox signaling properties of NO.

Phosphotyrosine phosphatases (PTPs) 1 compose a superfamily of phosphatases that hydrolyze phospho-Tyr residues in proteins critically involved in several cell-signaling pathways (1)(2)(3). Their active sites are characterized by the consensus sequence (I/V)HCXAGXXR(S/T). Mutation of the active site Cys results in a catalytically inactive, dominant negative enzyme (3). The redox state of the active site Cys has been shown to be critical modulator of PTP activity (4). For example, the reversible oxidation of the active site Cys to sulfenic acid by ROS such as H 2 O 2 inhibits PTP activity (4). A prolonged exposure to high concentrations of H 2 O 2 can irreversibly oxidize sulfenic acid to sulfinic acid (5). Several investigators have also provided evidence suggesting that ligand-stimulated generation of ROS activates growth factor receptor Tyr kinase-dependent signal transduction pathways by inhibiting counteracting PTPs (1, 2, 6 -12). Key evidence in these studies has included the use of fluorescent dyes to qualitatively measure the ROS generated. However, these dyes neither identify the ROS nor distinguish between ROS and RNS (reviewed in Ref. 12).
High cellular concentrations of ROS are toxic and are not generated under normal physiological conditions (13,14). This toxicity, the lack of target selectivity by ROS, and a cell's complement of ROS scavenging mechanisms would appear to preclude ROS as physiological signal-transducing molecules (12). An alternative mechanism for redox modulation of cellular signal transduction pathways proposes that the oxidative signal is converted to a nitrosative signal (11,12,15,16). In contrast to ROS, cellular NO synthesis is tightly regulated by cellular NO buffering of Ca 2ϩ -dependent NOS and by catabolic mechanisms (12,14,17). This is best exemplified by the catalytic requirements of these enzymes including homodimerization, three cosubstrates (NADPH, L-arginine, and O 2 ), and five cofactors (calmodulin, heme, FAD, FMN, and tetrahydrobiopterin) (17). In addition, NOS activities are modulated by phosphorylation and protein-protein interactions (e.g. see Refs. 12,18,19). In contrast to ROS, physiologically relevant RNS are not toxic at the intracellular concentrations generated by Ca 2ϩactivated constitutive NOS. At these concentrations, RNS are relatively specific in their cellular targets (12,17).
The most studied NO target is soluble guanylate cyclase (20). A number of recent studies have also demonstrated S-nitrosylation of protein Cys as a critical redox-driven mechanism for modulation of protein function (12,15,16,(21)(22)(23). The Snitrosylation of Cys residues can modify protein activity or simply serve as a NO reservoir through the proteins they are attached to (24 -26). Examination of S-nitrosylated proteins has revealed a degenerate consensus sequence X(K/R/H)C(D/E) (16,27). This consensus sequence is found in hundreds of proteins including caspases, cyclin D1, Rb, and BRAC1/2. This motif does not have to be linear; it can be derived from the tertiary structure of the protein (28). Not all proteins containing the consensus sequence become S-nitrosylated, since other factors exist that control S-nitrosylation such as the location of Cys in hydrophobic compartments. Hydrophobic environments enhance the rate of S-nitrosylation by concentrating the lipophilic O 2 and NO molecules necessary for the generation of the nitrosylating species N 2 O 3 (12,16,29). Of specific interest to the present study are the findings that exogenous RNS donors inhibit PTP activity (e.g. see . In two previous studies, we demonstrated that a mild oxidative stress produced at low clinically relevant doses of ionizing radiation generated a transient signal involving activation of the Ca 2ϩ -dependent NOS-1 in CHO and other epithelial cells (11,12,33). Using a combination of chemical inhibitors and genetic manipulation of NOS-1 activity, we were able to show that the radiation-induced transient activation of ERK1/2 signaling in these cells was dependent on NOS activity and RNS generation.
One mechanism postulates that RNS transiently inhibit PTP by reversible oxidation of PTP active site Cys, thereby shifting the relative balance of Tyr kinase/PTP activities in favor of enhanced Tyr phosphorylation. In this scenario, enhanced ERK1/2 activity is the result of inhibiting PTPs that block the activities of upstream kinases (e.g. epidermal growth factor receptor, Raf kinase) and/or inhibit PTPs downstream that dephosphorylate and inactivate ERK1/2 (e.g. MKP-1 and MKP-3). Since the relative activities of PTPs are 100 -1000 times greater than the activities of Tyr kinases, only modest changes in PTP activity can have significant effects on the net Tyr phosphorylation state of a protein (1,2). Inhibiting PTP activity provides a mechanism for the findings that ionizing radiation enhances total cellular protein Tyr phosphorylation (34,35) and apparently activates diverse specific Tyr kinases (e.g. ERK1/2, epidermal growth factor receptor, c-Abl, and c-Lyn) (36 -40).
In the present study, we show that mild oxidative stresses such as ionizing radiation or low concentrations of H 2 O 2 reversibly inhibit bulk cellular PTP activity and the activities of specific PTPs (SHP-1 and SHP-2). The mechanism of inhibition involves the activation of Ca 2ϩ -dependent NOS and the Snitrosylation of the PTPs. This RNS-dependent mechanism of PTP inhibition represents one example of how an oxidative signaling event is converted into a nitrosative signal (12).

MATERIALS AND METHODS
Cells and Plasmids-CHO-K1, MCF-7, and MRC-5SV cell lines purchased from American Type Culture Collection (Manassas, VA) were cultured as previously described in RPMI 1640 supplemented with 5% fetal calf serum (33). CHO and MRC-5SV cells were transfected using the Lipofectamine PLUS kit according to the manufacturer's protocols (Invitrogen). Transfection efficiency exceeded 80% as attested by transfection with a green fluorescent protein-encoding plasmid and fluorescence microscopy. Plasmids encoding wild type SHP-1 and a dominant negative mutant with the active site Cys 453 3 Ser mutation were obtained from Dr. C. Susini (41). Dr. J. Pessin (42) provided wild type SHP-2. A NOS-1 dominant negative mutant (HemeRedF) has been described (11,43). Expression of HemeRedF in CHO cells inhibits radiation-induced endogenous NOS-1 activity (11).
Pharmacological Inhibitors and Radiation Treatments-The NOS inhibitors, L-NAME at 1 mM and L-NNA at 100 nM, obtained from Sigma, were added to the cells 1 h prior to radiation exposure or treatment with H 2 O 2 . H 2 O 2 was added to 10 -100 M; ionomycin (Sigma) was added to a final 1 M concentration. Prior to adding H 2 O 2 or ionomycin, cells were washed thrice and incubated for 1 h with serum-free culture medium. This minimized serum catalase activity and serum protein binding of ionomycin. Cells were irradiated at room temperature with a Picker 60 Co source at a dose rate of 1.8 -2.0 Gy/min and transferred to a slide-warming plate to maintain temperature at 37°C. All treatments and experimental manipulations were performed under low light conditions.
S-Nitrosylated Protein Purification-S-Nitrosylated proteins were purified using a modification of the "biotin switch" method (44). At the indicated times of treatment, cells were rinsed three times in ice-cold phosphate-buffered saline and lysed in ice-cold lysis buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholic acid, and 0.1% SDS. Cells were scraped and collected into amber polypropylene microcentrifuge tubes to further reduce light exposure. Cell lysates were spun at 13,200 rpm for 5 min. The resulting supernatant (150 l) was incubated for 40 min at 50°C with 450 l of free sulfhydryl blocking buffer consisting of 250 mM Hepes (pH 7.7), 1 mM EDTA, 0.1 mM neocuproine, 2.5% SDS, and 20 mM methylmethanethiosulfonate as blocking agent (Pierce). Proteins were precipitated with 1.0 ml of acetone for 30 min at Ϫ20°C and collected by microcentrifugation. The pellets were resuspended in 1.0 ml of Ϫ20°C acetone, centrifuged, and solubilized in 180 l of HENS buffer, containing 25 mM Hepes (pH 7.7), 0.1 mM EDTA, 10 M neocuproine, 1% SDS. To each sample, 60 l of 4 mM N- [6-(biotinamido)hexyl]-3Ј-(2Јpyidyldithio)propionamide and 2.4 l of 100 mM ascorbate were added. Ascorbate reduces the S-NO bond (but not disulfide or sulfenic) to the free sulfhydryl, and the N- [6-(biotinamido)hexyl]-3Ј-(2Јpyidyldithio)propionamide reacts specifically with and biotinylates the free sulfhydryl. After a brief vortexing, the samples were incubated for 1 h at room temperature. Proteins were precipitated with acetone and subsequently solubilized in 200 l of HENS buffer. A small aliquot was used for protein determination and loading controls (actin or PTP Western blots) and the remaining supernatant was mixed with 400 l of neutralization buffer (20 mM Hepes, 100 mM NaCl, 1 mM EDTA, 0.5% Triton X-100, pH 7.7). Streptavidin-agarose (20 l, 50% suspension) (Sigma) were added to each sample and rotated for 1 h at 4°C. The streptavidin-agarose beads were washed four times with washing buffer (20 mM Hepes, 600 mM NaCl, 1 mM EDTA, 0.5% Triton X-100, pH 7.7) and once with 20 mM Hepes, 100 mM NaCl, 1 mM EDTA (pH 7.7). Proteins were eluted by incubating the beads with 60 l of the same buffer containing 120 mM ␤-mercaptoethanol for 20 min to release the biotin group, regenerating the protein to its original unmodified form. After centrifugation, the supernatants were mixed with 15 l of 5ϫ Laemmli sample buffer, and proteins were fractionated by electrophoresis on 8% SDS-polyacrylamide gels.
Immunoblotting-After electrophoresis, proteins were transferred to nitrocellulose membranes and analyzed using standard Western blotting procedures (e.g. see Ref. 11). All primary antibody incubations were performed in 5% bovine serum albumin in Tris-buffered saline with 0.1% Tween at 4°C overnight. Primary rabbit antibodies were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA): SHP-1 (sc-287) and SHP-2 (sc-424). A goat anti-actin polyclonal antibody was also obtained from Santa Cruz Biotechnology. After incubation with primary antibodies, blots were washed five times for 5 min each before secondary antibodies were added. Incubation with secondary antibody was performed in 5% dry milk solution in Tris-buffered saline with 0.1% Tween at room temperature for 1 h. The secondary antibody, conjugated to alkaline phosphatase, was diluted according to the manufacturer's protocols (Promega, Madison, WI). After extensive washing, the blots were developed with CDP-Star chemiluminescence reagent (PerkinElmer Life Sciences), diluted 1:5 with deionized water, and mixed with a 1:40 dilution of Nitro-Block II (Applied Biosystems, Foster City, CA).
Assay of Total and Immunopurified PTP Activities-Redox modulation of total cellular PTP activity was measured by modifications of an assay developed by Li and Whorton (1,45). This assay, like the protocol for purification of S-nitrosylated proteins, is based on protection of the active site Cys from NEM alkylation by reversible oxidation of the Cys by either S-nitrosylation, S-glutathiolation, or sulfenic acid formation. To distinguish between S-nitrosylation and either S-glutathiolation or sulfenic acid formation, PTP activity is measured after treating NEMtreated lysates or purified PTP with either dithiothreitol to reduce all three oxidative modifications or ascorbate to selectively reduce the S-NO bond (44). Protection against NEM alkylation measured by an increase in dithiothreitol or ascorbate-recoverable activity is equated to the amount of PTP inhibition obtained after exposing cells to either low doses of radiation or H 2 O 2 .
For total cellular PTP activity measurements, lysates from cells cultured in 3.5-cm dishes were prepared by washing cells once in ice-cold PBS followed by lysis in 200 l of lysis buffer containing 1% Triton X-100, 100 mM NaCl, 25 mM Hepes (pH 7.4), 1 mM EDTA, with or without 20 mM NEM. After 10 min on ice, lysates were centrifuged and assayed for PTP activity using a commercially available PTP 96well activity kit (Molecular Probes, Inc., Eugene, OR). This assay procedure uses the PTP substrate, DiFMUP, and a Ser/Thr phosphatase inhibitor mixture to ensure that only PTP activity is measured. To meet the requirements of the present studies, the substrate-phosphatase inhibitor mix was solublized in 0.1% Tween 20 and 25 mM MOPS (pH 7.0) plus either 25 mM dithiothreitol or 50 mM ascorbate. Cell lysates initially diluted between 1:2 and 1:10 depending on the cell preparation were further diluted by adding 5-200 l of the complete assay buffer to reduce the NEM concentration and initiate catalytic activity. Activity was monitored over the next 10 -20 min with a Packard fluorescence plate reader. A linear regression analysis and normalization to protein measured by the Bradford assay was used to calculate PTP specific activities. Fig. 1A shows typical raw data and linear regression analysis used for these measurements.
SHP-1 and SHP-2 PTP activities were measured with immunoprecipitates isolated from cells transfected with plasmids expressing these PTPs. Cell lysates prepared as above with NEM were incubated with 2 g of either anti-SHP-1 or anti-SHP-2 and protein A-agarose (Oncogene Sciences, Cambridge, MA) for 90 min followed by four washes in lysis buffer and two in the same buffer without Triton but containing Ser/Thr phosphatase inhibitors (10 mM pyrophosphate, 10 mM glycerophosphate, 20 mM NaF). Enyzme activity was initiated by adding 20 M 6,8-difluoro-4-methylumbelliferyl phosphate to the beads. Beads were continuously rotated at 37°C for 30 min before microcentrifugation to remove the beads. Fluorescence of the supernatants was measured. Control experiments using vanadate as a PTP inhibitor verified the selectivity of the assay and verified that the measured activities were linear for at least 30 min of incubation. Activities were normalized with respect to cell lysate protein concentrations.
Cell Ca 2ϩ Measurements-Cytosolic free [Ca 2ϩ ] was measured with the fluorescent Ca 2ϩ -sensitive dye, fura-2, as previously described (46), but with ratiometric imaging software provided by Universal Imaging and detection with a Photometrics Sensys CCD camera. Cells on coverslips were loaded with dye by incubating cells at room temperature in serum-free medium containing a 2.5 M concentration of the membrane-permeant acetoxymethyl ester derivative of fura-2 for 30 -60 min. A subsequent incubation for 60 min without dye was used to maximize ester cleavage and dye retention in the cells. Coverslips were mounted in a perfusion chamber prior to microscopic analysis.
NOS Activity-Cellular NOS activity was measured by the arginine-citrulline conversion assay using [ 3 H]arginine as previously described (11).
Statistical Analysis-Each experimental figure shown is one experiment representative of at least two experiments. The data points in each figure represent the mean plus or minus the S.D. of at least two measurements. For the analysis of data from several experiments, results were normalized with respect to a control value (e.g. t ϭ 0, or 0 Gy) to facilitate comparisons. Two-sided t tests were used to compare the considered groups. Since the sample numbers are small, it is assumed that the data are sampled from a normal distribution. The t value is calculated as the absolute value of (mean 1 Ϫ mean 2)/square root of (s1 2 /n1 ϩ s2 2 /n2), where n represents sample size and s is the S.D. value. p Ͻ 0.05 was determined to be of statistical significance.

Radiation Inhibits Cellular PTP Activity by a NO-sensitive
Mechanism-Our previous investigations showed that a mild oxidative stress induced by exposing cells to low doses of ionizing radiation activated a constitutive, Ca 2ϩ -dependent NOS activity in diverse cell types (11,33). This activity was inhibited by a dominant negative mutant of NOS-1 and enhanced by expression of a wild type NOS-1, suggesting that radiation activates a NOS-1 isoform in CHO cells. Furthermore, we showed that radiation-induced activation of the ERK1/2 signaling pathway in CHO cells as revealed by measurements of enzyme activity and Tyr phosphorylation of ERK1/2 could be blocked by NOS inhibitors such as L-NAME or expression of a dominant negative mutant of NOS-1. We proposed that one mechanism for this activation might be inhibition of a counteracting PTP acting at one step of the ERK1/2 activation pathway by S-nitrosylation of the PTP active site Cys. In support of this proposal, preliminary studies showed that radiation stimulated the S-nitrosylation of the active site Cys of two prominent cellular PTPs, SHP-1 and SHP-2 (12).
More direct evidence for this mechanism has come from measuring total cellular PTPase activity before and after irradiating cells. As described under "Materials and Methods," protection from NEM alkylation was used to assess active site Cys modification and thus indirectly provide a measure of PTP inhibition. The protection by either S-nitrosylation, oxidation to sulfenic acid, or S-glutathiolation was revealed by treating cell lysates with the SH reducing agent, dithiothreitol. The sensitivity of S-NO to reduction by ascorbate was used to distinguish S-nitrosylation from the ascorbate-insensitive sulfenic acid formation or S-glutathiolation. Fig. 1A shows a typical time course used in the PTP assay. Specific activities were determined from the linear regression analyses and normalized with respect to protein concentration. The results in Fig. 1B compare the relative total PTP specific activities before and after treatment with NEM followed by recovery with either dithiothreitol or ascorbate. Maximal cellular PTP specific activity was defined as that measured in cell lysates not treated with NEM but in the presence of 25 mM dithiothreitol. Based on this definition, 60 -80% of total cellular PTP activity was irreversibly inhibited by treatment with NEM (n ϭ 8 separate experiments). The remaining 20 -40% of basal PTP activity was protected from NEM alkylation and inactivation by a mechanism reversible by dithiothreitol treatment. Approximately half as much was also protected by a mechanism reversed by ascorbate treatment (17 Ϯ 8% of total basal PTP activity; n ϭ 6). Two-sided t tests revealed statistical differences between the dithiothreitol and ascorbate treatments in six independent experiments with p values of Ͻ0.01.
Cells were irradiated (5 Gy), and 5 min postirradiation, cell lysates were prepared with NEM. Maximal radiation-induced NOS activity and ERK1/2 activation in CHO cells are observed at 5 min postirradiation (11). As shown in Fig. 1B, an additional modest but significant protection from alkylation was observed following radiation and reversal by treatment with dithiothreitol or ascorbate. The increase in protection produced by a 5-Gy radiation exposure ranged between 5 and 20% of total cellular PTP activity (n ϭ 4 separate experiments, p Ͻ 0.02). Dithiothreitol was consistently more effective than ascorbate in unmasking this radiation-induced PTP inhibition. However, the p values for these comparisons ranged between 0.03 and 0.26, and thus the difference between the two reductants in their effectiveness in revealing radiation-induced PTP activity was not statistically significant.
The effects of NOS inhibitors on basal and postradiation total cellular PTP activities were also assessed. Initial experiments established that preincubating cells with 100 nM L-NNA for 1 h prior to irradiation inhibited basal and radiation-stimulated NOS activity by 71 Ϯ 12% (n ϭ 3, data not shown). This compares with 50% inhibition observed previously with another NOS inhibitor, L-NAME, at 1.0 mM for 1 h (11). Under these treatment conditions, at least 25% of basal cellular PTP activity insensitive to NEM is sensitive to L-NNA (25-55%, n ϭ 3, p Ͻ 0.05; e.g. see Fig. 2B). The radiation-stimulated protection was completely inhibited under these conditions (n ϭ 3; p Ͻ 0.001). Similar results were obtained with L-NAME (e.g. see Fig. 5). These results, combined with the partial ascorbate sensitivity in recovery of PTP activity and our previous findings demonstrating active site S-nitrosylation of the PTP, SHP-1, following radiation treatment of cells (12), support the conclusion that radiation-induced inhibition of PTP activity is a consequence of S-nitrosylation. Pharmacological inhibition of NOS also completely inhibited radiation-induced protection of PTP activities in MCF-7 breast carcinoma cells and MRC-5SV fibroblasts (data not shown).
The findings with the pharmacological inhibitors were verified in CHO cells by genetic manipulation of NOS-1 activity using the dominant negative mutant, HemeRedF (11). At 48 h post-transfection with the mutant, Ͼ80% of the radiation-stimulated NOS activity is inhibited (11). Results in Fig. 3 from a single experiment representative of two performed show that expression of HemeRedF under these conditions completely blocks the transient PTP inhibition observed following radiation. In contrast to what was found with the chemical inhibitors, a prolonged 48-h genetic inhibition of NOS activity had no significant effect on basal NEM-insensitive PTP activity with respect to vector control cells (Fig. 3 (inset); n ϭ 3, p Ͻ 0.6). This suggests possible compensatory mechanisms for redox protection of PTP function under conditions of sustained NOS inhibition.

SHP-1 and SHP-2 Are Inhibited by a Radiation-induced NO-sensitive Mechanism-The activities of immunopurified
PTPs, SHP-1 and SHP-2, were also measured. The PTPs were immunopurified as described under "Materials and Methods" from cell lysates prepared at times directly before radiation and 5 min postradiation. Catalytic activities were measured with the same fluorescent substrate used in the bulk PTP measurements and with the assays initiated by the addition of substrate and either dithiothreitol or ascorbate to the immune complex beads. The results for both PTPs in Fig. 4 are of a single representative experiment repeated three times with p values shown for that experiment. The results in Fig. 4A with purified SHP-1 parallel those obtained with total cellular PTP measurements. After irradiation, there is an approximate 40% increase in NEM-insensitive PTP activity measured with dithiothreitol present in the assay (42 Ϯ 11%, n ϭ 3 independent experiments, p Ͻ 0.008). With ascorbate, there is a 20% increase in NEM-insensitive PTP activity following a radiation exposure (21 Ϯ 6%, n ϭ 3 independent experiments, p Ͻ 0.02). That only 50% is recovered with ascorbate relative to dithiothreitol may in part be explained by the prolonged incubation times (30 min). The absence of a sufficiently strong reducing agent such as dithiothreitol can result in increased oxidation of active site Cys and PTP inhibition (e.g. see Ref. 45). The results and statistical analysis in Fig. 4, B and C, demonstrate that   FIG. 2. The NOS inhibitor, NNA, enhances the sensitivity of basal PTP activity to NEM alkylation and blocks the radiationstimulated protection of bulk PTP activity to NEM alkylation. A, the rate curves for the 3-min postirradiation curves are shown and demonstrate that the NOS inhibitor, NNA, blocks the radiation-induced protection of PTP activity from NEM alkylation. The radiation dose was 5 Gy. B, the complete time course for radiation-induced protection of bulk PTP activity. Cells were treated with NNA as described under "Materials and Methods" and irradiated, and cell lysates were prepared and processed as described in the legend to Fig. 1

FIG. 4. Radiation (5 Gy) inhibits SHP-1 and SHP-2 activities in CHO cells by a mechanism sensitive to inhibition by the NOS
inhibitor, L-NAME, and revealed by either dithiothreitol or ascorbate. Cells (6-cm dishes) were transfected with 2 g of plasmid DNA encoding SHP-1 or SHP-2, and 48 h post-transfection, cells were treated with or without 1.0 mM L-NAME for 1 h and irradiated at 5 Gy. Cells were harvested at 5 min postirradiation. Cell lysates were normalized according to protein levels, and SHP-1 or SHP-2 was immunoprecipitated and assayed as described under "Materials and Methods." A, the relative effects of dithiothreitol (DTT) and ascorbate in the assay buffer on SHP-1 activity. A representative experiment of three performed is shown. B, the effect of L-NAME inhibition of NOS on SHP-1 activity with dithiothreitol in the assay buffer. C, the effect of L-NAME inhibition on SHP-2 activity with dithiothreitol in the assay buffer. The results in B and C are from a single representative experiment performed in duplicate. The data points in all three panels are the averages Ϯ S.D. of triplicate samples with the p values calculated for the individual experiment shown. was observed. As was found in the radiation studies, the H 2 O 2induced PTP protection was mostly blocked by prior incubation with the NOS inhibitor, L-NAME (Fig. 5).
Similar results were obtained when the activity of SHP-1 immunopurified from H 2 O 2 -treated cells was measured (Fig.  6). SHP-1 purified from H 2 O 2 -treated cells was protected from NEM alkylation by a mechanism blocked by L-NAME and reversed by treatment with either dithiothreitol or ascorbate. The H 2 O 2 -induced protection was more effectively reversed by dithiothreitol than ascorbate, similar to what was found in measurements of total NEM-insensitive PTP activity after radiation. In contrast to the results obtained for total cellular PTP activity, the H 2 O 2 -induced protection of SHP-1 was transient. The transient nature of the protection, however, fits with the findings shown in Fig. 8 [Ca 2ϩ ] and Stimulate NOS Activity-In our previous investigation, we demonstrated that ionizing radiation activated cellular NOS-1 activity in CHO cells as assayed by an arginine to citrulline conversion assay (11). These experiments were repeated with H 2 O 2 over the concentration range of 10 -40 M. At these low concentrations, H 2 O 2 stimulates NOS activity about 2-3-fold, comparable with what is achieved with the Ca 2ϩ ionophore, ionomycin, or a radiation dose of 2 Gy (Fig. 7A) (11).
Several investigations have demonstrated that low H 2 O 2 concentrations and other mild oxidative stresses including ionizing radiation reversibly stimulate increases in cytoplasmic free [Ca 2ϩ ] (e.g. see Refs. 46 and 47). We verified this in the cell lines used in the present study at the single cell level using microscopic fluorescence measurements of the Ca 2ϩ -sensitive dye, fura-2 (Fig. 7B) (Figs. 2B and 4). To establish the linkage between S-nitrosylation of PTPs and their inhibition by H 2 O 2 , S-nitrosylated SHP-1 and SHP-2 were purified by the biotin switch method following treatment of cells with either H 2 O 2 or ionomycin (Fig. 8). Maximal S-nitrosylation of SHP-1 and SHP-2 was observed between 3 and 6 min after adding H 2 O 2 , corresponding to the time period during which these PTPs were maximally protected from NEM alkylation. Basal and H 2 O 2 -induced S-nitrosylation were mostly inhibited by a prior 1-h incubation with L-NAME. Following from its effects on NOS activity, it was not surprising to find that ionomycin also stimulated the S-nitrosylation of SHP-2 (Fig. 8B) and SHP-1 (data not shown). We conclude from the experiments in Figs. 7 and 8 that low concentrations of H 2 O 2 , like low doses of ionizing radiation, induce intracellular Ca 2ϩ transients, stimulating Ca 2ϩ -dependent NOS activity and the S-nitrosylation of PTPs, and by so doing transiently inhibiting PTP activities.
Previously, we demonstrated using cells transfected with wild type and a mutant SHP-1 with the active site Cys 453 mutated to Ser that radiation-induced S-nitrosylation was exclusively targeted to Cys 453 (12). These experiments were repeated with 100 M H 2 O 2 as the stimulant (Fig. 8C). As was found with radiation, H 2 O 2 -induced S-nitrosylation of SHP-1 was undetectable in cells expressing the Cys 453 3 Ser mutant except for the minor probably endogenous wild type component. Similar unambiguous experiments cannot be performed with the SHP-2 active site Cys mutant, since overexpression of the dominant negative SHP-2 mutant inhibits endogenous NOS-1 activity (11). Overexpression of wild type SHP-1 or the Cys 453 3 Ser mutant is without effect on NOS activity in these FIG. 5. H 2 O 2 protects bulk PTP activity in CHO cells by a mechanism inhibited by the NOS inhibitor L-NAME. As discussed under "Materials and Methods," cells were washed three times with medium without serum to remove catalase activities in the culture medium. Other procedures are the same as described in the legend to  6. H 2 O 2 transiently protects SHP-1 activity from NEM alkylation by a mechanism sensitive to the NOS inhibitor L-NAME. Cells were transfected with a plasmid expressing SHP-1 and treated with L-NAME as described in the legend to Fig. 4. After treatment with or without 90 M H 2 O 2 , cell lysates were prepared at the designated times, and SHP-1 was immunoaffinity-purified and assayed with ascorbate (A) or dithiothreitol (B) as described in the legend to cells (11). These results with SHP-1 are compatible with the effects of radiation and H 2 O 2 on enzyme activity and provide a mechanism for how ionizing radiation or H 2 O 2 transiently inhibit SHP-1 by S-nitrosylation. DISCUSSION It has been previously argued that a mild oxidative stress such as clinically relevant doses of ionizing radiation may activate cellular redox response pathways by mechanisms involving RNS such as NO rather than ROS (11). In contrast to most ROS, NO is very specific and readily reversible in its reactions, and intracellular NO levels are highly regulated by both anabolic and catabolic mechanisms. By these criteria, NO represents the "prototypic redox signaling molecule" (15,16,48). We provide evidence in the present report with H 2 O 2 and past publications with ionizing radiation (11,33,46) that mild oxidative events, by stimulating transient increases in intracellular [Ca 2ϩ ], activate a Ca 2ϩ -dependent NOS. Herein we show that one consequence of oxidative stimulation of NO generation is the transient S-nitrosylation and inhibition of PTPs critical in cellular signal transduction pathways.
Evidence supporting our hypothesis falls into three categories. First, as already mentioned, mild oxidative events stimulate Ca 2ϩ -dependent NOS activity. Second, we demonstrate that low doses of ionizing radiation or H 2 O 2 protect both total cellular and selected immunopurified PTPs from NEM alkylation. Operationally, this indicates that mild oxidative stresses such as ionizing radiation and H 2 O 2 transiently inhibit PTP activities. The underlying mechanism of S-nitrosylation is revealed in part by the protection moiety's relative sensitivity to reduction by ascorbate. The S-NO bond is selectively reduced by ascorbate, whereas other modifications, including oxidation to sulfenic acid and S-glutathiolation, are impervious to ascorbate treatment and are only reduced with potent reducing agents such as dithiothreitol (30,44,45). We demonstrate that a significant proportion of the protected PTP activity observed after radiation or H 2 O 2 treatments is sensitive to ascorbate reduction ( Figs. 1 and 4A). Importantly, pharmacological and genetic inhibition of cellular NOS activity completely blocked either H 2 O 2 -or radiation-induced protection of PTP activity, whether measured in cell lysates or after immunopurification of selected PTPs (Figs. 2-6). The third line of evidence follows from the demonstrations that mild oxidative stresses, whether induced by ionizing radiation, low [H 2 O 2 ], or a Ca 2ϩ ionophore, stimulate the S-nitrosylation of specific PTPs (Fig. 8) (12). The assay for S-nitrosylation depends on ascorbate reduction certifying the S-NO formation. We conclude from these three lines of evidence that mild oxidative stress induces S-nitrosylation of PTPs and as a consequence transiently inhibits PTP activity.
In analyzing the data, three quantitative issues need addressing. One issue concerns the relatively small but statistically significant amount of PTP activity that appears sensitive to inhibition by S-nitrosylation and induced by the mild oxidative stresses employed herein. The modest and transient inhibition (5-20% of total cellular PTP activity) must be considered in light of the relative enzymatic rates of PTPs and protein tyrosine kinases. It is the balance of the two apposing enzymatic activities that determines net phosphorylation of the target proteins. The catalytic rate constants for PTPs are, in general, 100 -1000 times that of kinases, and thus a small change in PTP activity can have a significant impact on the net phosphorylation of a protein (1,2). This consideration is especially important for signal transduction pathways that are autocrine-regulated (e.g. epidermal growth factor receptor in some tumor cells). With these cells, the kinase signal is always "on" due to autocrine-generated ligand binding or receptor mutations that make the kinase constitutively active. This predicts that autocrine-stimulated Tyr phosphorylation is in large part regulated by the apposing PTP activities.
The relative intracellular juxtaposition of the activated NOS, the target PTPs, and their phosphorylated targets is an additional important consideration given the limited cellular diffusion distances of most RNS (12). A localized increase in NO may S-nitrosylate all PTPs within a small intracellular volume but with considerably lesser effect further from the RNS source. Experimentally, this would be measured as a small overall change in PTP activity when measured in bulk as performed here but may have important signaling consequences when localized near target PTPs. NOS-1 is localized in the endoplasmic reticulum of some cells (49,50); there is evidence for a NOS-1 isoform in mitochondria of diverse cell types (51,52), and isoforms of NOS-1 have a PDZ domain that presumably is essential for targeting (19). The other constitutive Ca 2ϩ -activated, NOS, NOS-3, or endothelial NOS, is palmitoylated, and this appears important for its plasma membrane tethering (53). PTPs are also geographically targeted within cells. PTP1B, for example, is predominantly found in the endoplasmic reticulum (54). SHP-1 has a nuclear or perinuclear localization in epithelial cells but is cytoplasmic in hematopoietic cells (55). SHP-2, on the other hand, interacts via its SH2 domains either directly or indirectly through adaptor proteins to plasma membrane tyrosine kinase receptors (56,57). SHP-2 also localizes in the nuclei of cells (58). Thus, the relative effect of S-nitrosylation on a specific PTP activity and the resultant effects on the phosphorylation state of target proteins will depend in part on the proximity of the PTP to the NOS that is activated (12). These same considerations apply when attempting to compare the relative effects of exogenous and endogenous ROS/RNS. The action of extracellular ROS/RNS on cells, whether added directly or with donors, is unlikely to mimic the results obtained by the localized ROS/RNS generated by cellular mechanisms.
The relative effects of dithiothreitol, ascorbate, and NOS inhibition with chemical inhibitors or genetically with a dominant negative NOS mutant on PTP activities were compared. Dithiothreitol and pharmacological inhibition of NOS with L-NAME were equivalently effective in revealing PTP activity protected from NEM alkylation following either radiation or H 2 O 2 treatments. Compared with dithiothreitol, ascorbate reduction recovered about half as much PTP activity. One explanation is that after the initial reduction of the S-nitrosylated PTP active site Cys by ascorbate, reoxidation not reversible by ascorbate occurs during the assay. An alternative mechanism has the reversal of S-nitrosylation by the cell involving an intermediate step, S-glutathiolation (5,59). The disulfide bond formed from S-glutathiolation is insensitive to ascorbate but is readily reduced by dithiothreitol. Thus, at any specific time, the PTP pool would consist of unmodified, S-glutathiolated and S-nitrosylated PTP with relative proportions of each dependent on the redox state of the cell. It is also worth noting that peroxynitrite, the reaction product of NO and superoxide anion, is produced at low levels after oxidative treatments (e.g. radiation) (11). Peroxynitrite can oxidize protein sulfhydryls to sulfenic acid (60).
There was some interexperimental variability in the absolute PTP specific activities measured. Previous studies have shown that total cellular activity and the activities of individual PTPs can vary as much as 15-fold, depending on cell culture density (61,62). Since we did not apply stringent precautions in regulating cell density, this is a likely explanation. In the assays of specific PTPs, variations in transfection efficiency and expression levels may also be factors. Similar considerations may also apply to why L-NAME significantly inhibits total basal PTP activity but not the basal activities of purified SHP-1 or SHP-2 ( Figs. 4 and 6). Other explanations are possible, including the relative location of the PTPs and NOS and other cellular NO donors less sensitive to short term NOS inhibition (e.g. GSNO (52)). Importantly, these variations do not alter the overall conclusions that mild oxidative treatments stimulate PTP S-nitrosylation and as a result inhibit both bulk PTP activity and the activities of purified PTPs.
The focus of this report has been on ROS/RNS signaling and how this intersects with key components of tyrosine phosphorylation-dependent signal transduction pathways. The analysis has been limited to measurements of bulk cellular PTP activity and the activities of two purified PTPs. Since PTPs are characterized by highly homologous active sites, it would not be surprising to find that other PTPs are sensitive to inhibition by S-nitrosylation following a mild oxidative stress. Of particular interest may be the CDC25 family of mixed function PTPs, since they are critical in regulating cell cycle traversal and cell cycle arrest following oxidative stress (63). Besides PTPs, several recent studies have demonstrated S-nitrosylation and functional modulation of a number of other proteins including Ras, Raf, some Ser/Thr phosphatases, transcription factors, and caspases (e.g. 21, 48, 64 -68). Based on our findings, these proteins may also be regulated during oxidative signaling by RNS-dependent pathways.