Redox regulation of the calcium/calmodulin-dependent protein kinases.

Reactive oxygen intermediates (ROI) have been viewed traditionally as damaging to the cell. However, a predominance of evidence has shown that ROI can also function as important activators of key cellular processes, and ROI have been shown to play a vital role in cell signaling networks. The calcium/calmodulin-dependent protein kinases (CaM kinases) are a family of related kinases that are activated in response to increased intracellular calcium concentrations. In this report we demonstrate that hydrogen peroxide treatment results in the activation of both CaM kinase II and IV in Jurkat T lymphocytes. Surprisingly, this activation occurs in the absence of any detectable calcium flux, suggesting a novel means for the activation of these kinases. Treatment of Jurkat cells with phorbol 12-myristate 13-acetate (PMA), which does not cause a calcium flux, also activated the CaM kinases. The addition of catalase to the cultures inhibited PMA-induced activation of the CaM kinases, suggesting that similar to hydrogen peroxide, PMA also activates the CaM kinases via the production of ROI. One mechanism by which this likely occurs is through oxidation and consequential inactivation of cellular phosphatases. In support of this concept, okadaic acid and microcystin-LR, which are inhibitors of protein phosphatase 2A (PP2A), induced CaM kinase II and IV activity in these cells. Overall, these results demonstrate a novel mechanism by which ROI can induce CaM kinase activation in T lymphocytes.

The CaM kinases 1 are a family of related kinases activated in response to increased calcium levels (for review, see Refs. [1][2][3]. The primary members of this family that demonstrate broad substrate specificity are CaM kinase I, II, and IV. Other family members are dedicated to the phosphorylation of a specific substrate (i.e. myosin light chain kinase, phosphorylase kinase, and elongation factor 2 kinase/CaM kinase III). All members of the CaM kinase family are activated in response to the binding of calcium/calmodulin, which causes a conformational change revealing the substrate binding domain of the kinase (1). However, there are differences in the way each kinase is regulated.
CaM-KII is a multimeric enzyme composed of 10 -12 catalytic subunits (4 -6). Upon calcium/calmodulin binding, CaM-KII undergoes autophosphorylation in a region of the protein that prevents the autoinhibitory domain from interacting with the kinase domain. Once phosphorylated in this manner, CaM-KII activity becomes independent of calcium (7)(8)(9). To return to an inactive state, dephosphorylation of CaM-KII must occur; both protein phosphatases 1 and 2A (PP1 and PP2A) appear to play important physiological roles in the dephosphorylation of CaM-KII (10).
CaM-KIV is phosphorylated by the kinase CaM-KK (11)(12)(13). CaM-KK is activated by calcium/calmodulin complexes as well, but the potential for other means of regulation has also been suggested (13)(14)(15). Phosphorylation of CaM-KIV by CaM-KK occurs in the pseudosubstrate domain that interacts with the catalytic domain of the kinase. The phosphatase PP2A has been shown to form a complex with CaM-KIV and at least in vitro can cleave the phosphate group off CaM-KIV (16,17). This interaction is also believed to be important in vivo because transfection of cells with the PP2A inhibitor, SV40 small T antigen, is able to potentiate CaM-KIV induction of cAMPresponse element-binding protein (CREB)-dependent transcription (17). Specific cell-permeable inhibitors of the serine phosphatases such as okadaic acid and microcystin-LR can inhibit both PP2A and PP1 through binding to the catalytic domain (18,19). Additionally, hydrogen peroxide has also been shown to inhibit these phosphatases; this is believed to occur via the oxidation of a reactive cysteine residue in the catalytic domain of these phosphatases (20,21). Inhibition of this phosphatase could conceivably result in the activation of the CaM kinases in the absence of an apparent increase in intracellular calcium.
The CaM kinases have been implicated recently in T cell receptor-, hydrogen peroxide-, and PMA-induced IB kinase activation and IB phosphorylation in Jurkat T lymphocytes. Treatment of T lymphocytes with hydrogen peroxide induces calcium fluxes in these cells similar to those caused by antibodies to the T cell receptor (22). PMA treatment of T lymphocytes is not reported to induce a calcium flux. Surprisingly, PMA-induced NF-B activation is reported to be sensitive to calmodulin antagonists and CaM kinase inhibitors (23,24). In * This work was supported by Scientist Development Grant 9930099N and Grant-in-aid 0355834U from the American Heart Association (to R. A. F.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  addition, we have reported previously that hydrogen peroxideinduced IB phosphorylation was not blocked by incubation of the cells with BAPTA and EGTA, although CaM kinase inhibitors did block this response (22). These results suggest that calcium-independent activation of the CaM kinases is occurring. In this report we demonstrate the activation of both CaM-KII and CaM-KIV in the absence of increases in intracellular calcium. This activation is induced by ROI as well as by PMA. Additionally, we report that phosphatase inactivation can also lead to the activation of the CaM kinases. We postulate that both ROI and PMA can induce CaM kinase activation through the inactivation of phosphatases by oxidation.

EXPERIMENTAL PROCEDURES
Cells and Reagents-The human Jurkat T cell line was obtained from ATCC (Manassas, VA) and cultured in RPMI 1640 with 5% fetal calf serum, 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin. Ionomycin, PMA, and microcystin-LR were purchased from Calbiochem (San Diego, CA) and dissolved in Me 2 SO. Protein kinase C inhibitor peptide (A.A. 19 -31), protein A/G beads, and okadaic acid were also purchased from Calbiochem. Mouse immunoglobulin (IgG1), syntide-2, and protein kinase inhibitor peptides were purchased from Sigma. Anti-CaM kinase II, anti-CaM kinase IV, anti-Akt, antiphospho-Thr-308 Akt, and anti-PP2A antibodies were purchased from Cell Signaling Technology (Beverly, MA). Fluo-3 was purchased from Molecular Probes (Eugene, OR) and dissolved in Me 2 SO. N ␣ -(3-Maleimidylpropionyl)biocytin was also purchased from Sigma.
Sample Preparation for Kinase Assay-Cells were washed in phosphate-buffered saline and resuspended in serum-free RPMI 1640. Fifteen ml containing 1.25 ϫ 10 6 cells/ml were added to 15-ml conical tubes and placed at 37°C for at least 1 h prior to stimulation. EGTA (2 mM) was added to the cells 30 min prior to stimulation to produce calciumfree conditions. Cells were stimulated with hydrogen peroxide (10 mM), ionomycin (500 nM), or PMA (100 nM) for the time period indicated. Cells were treated with microcystin-LR (500 nM) or okadaic acid (1 M) for 15 min. In the experiments employing catalase (10,000 units), it was added for 10 min prior to stimulation. Following treatments, cells were centrifuged, and the supernatant was removed. The cell pellets were resuspended in 300 l of cold lysis buffer (50 mM HEPES, pH 7.5, 2 mM EDTA, 2 mM EGTA, 2 mM benzamidine, 10 mM NaPO 4 , 50 mM NaFl, 5 mM dithiothreitol, 0.5% IGEPAL, 1 mM Na 3 VO 4 , 0.5 mM phenylmethylsulfonyl fluoride, and 10 g/ml each of the following: soybean trypsin inhibitor, aprotinin, antipain, leupeptin, pepstatin A, and N ␣ -p-tosyl-Llysine chloromethyl ketone), transferred to microcentrifuge tubes, and placed on ice for 20 min. Lysates were centrifuged for 15 min at 14,000 rpm in a refrigerated Eppendorf microcentrifuge. The supernatant was transferred to a new microcentrifuge tube and frozen until needed.
Immunoprecipitation-To preclear the lysates, 25 l of protein A/G bead suspension was added to 300 l of the supernatant from above, incubated with rocking for 2 h at 4°C, and spun, and the supernatant was placed in a fresh tube for further analysis. Either the anti-CaM kinase IV or II antibody (2.5 g) was added to the precleared lysates, which were then placed on a rocker for 2 h at 4°C. Protein A/G beads (25 l) were then added to the lysate antibody mixture and rocked for an additional 1 h at 4°C. Samples were centrifuged at 14,000 rpm, and the beads were washed twice in the lysis buffer. The beads were then resuspended in 40 l of CaM kinase dilution buffer (50 mM HEPES, pH 7.5, 1 mg/ml bovine serum albumin, 10% ethylene glycol). A portion of the immunoprecipitate was used to ensure equal amounts of the kinase were immunoprecipitated. To do this, 5 l were combined with 2 l of 3.3ϫ sample buffer (200 mM Tris-HCl, pH 6.8, 33% glycerol, 6.6% SDS, 16.6% 2-mercaptoethanol, 0.04% bromphenol blue), boiled for 5 min, CaM kinase IV (B). Immunoprecipitates were used in a kinase assay in the absence of calcium to measure the phosphorylation of the synthetic peptide, syntide-2 (A and B). In C and D calcium and calmodulin were added to the kinase assays to measure total kinase activity obtainable in the samples. Counts were measured and are shown as the average of two separate assays. Data are expressed as cpm (CPM) Ϯ S.D. cpm values shown are the average of two separate assays. In E, CaM-KII or CaM-KIV was immunoprecipitated following stimulation as indicated above. These immunoprecipitates were then immunoblotted with the immunoprecipitating antibody. Binding of the primary antibody was detected using an AP-labeled goat anti-mouse immunoglobulin and an AP-BCIP/NBT developing system. and frozen until utilized in immunoblot analysis. The remaining immunoprecipitate was used in a kinase assay immediately. In some cases immunoprecipitates were performed using 2.5 g of mouse immunoglobulin to ensure the specificity of immunoprecipitations.
Kinase Assay-Kinase assays were performed as described previously by Park et al. (16). Five l of the immunoprecipitates were assayed for either CaM kinase II or IV activity in 25 l of a reaction mixture containing 50 mM HEPES, pH 7.5, 10 mM magnesium acetate, 2 mM dithiothreitol, 5 M protein kinase inhibitor, 2 M protein kinase C inhibitor peptide, 1 M microcystin-LR, 200 M ATP (4 Ci of [␥-32 P]ATP), and 40 M peptide substrate syntide-2. Total kinase activity was determined using 800 M Ca 2ϩ and 1 M calmodulin. Autonomous activity was determined in the presence of 5 mM EGTA. Duplicates were performed for each reaction, as well as kinase reactions lacking the syntide-2 peptide to account for autophosphorylation (background activity). The background activity was subtracted from the samples containing syntide-2 peptide to determine the kinase activity. The reactions were carried out for 12 min at 30°C and terminated by the addition of 5 l of 200 mM EDTA. Twenty-five l of the kinase reaction mix were spotted on Whatman P81 filter paper. The filter papers were washed three times in 0.5% phosphoric acid, air-dried, and counted on a liquid scintillation counter. In one series of experiments hydrogen peroxide (10 mM) was added directly to the CaM-KII or CaM-KIV immunoprecipitates to determine whether the effects of hydrogen peroxide were mediated directly on the CaM kinases.
Measurement of Intracellular Calcium-Intracellular calcium was measured by loading the cells with Fluo-3 as described previously (22). Briefly, cells were washed in phosphate-buffered saline and resuspended at 5 ϫ 10 6 /ml in RPMI containing 5% serum and 10 M Fluo-3. Cells were incubated at room temperature for 30 min. Cells were then washed and resuspended at 1 ϫ 10 6 cells/ml in RPMI 1640 (serum-and phenol red-free) in a black opaque 96-well microtiter plate. Cells were stimulated as indicated in the figures, and the fluorescence was monitored over time on a Bio-Tek FL600 fluorescence/absorbance spectrophotometer (Winooski, VT). The wavelength for excitation was 485 nm, and emission was measured at 530 nm. Sample Preparation for Immunoblot Analysis-Cells were washed and resuspended in serum-free RPMI 1640. One ml containing 1.25 ϫ 10 6 cells was added to microcentrifuge tubes and warmed at 37°C for at least 1 h prior to the start of the experiment. Cells were stimulated with indicated treatments. The tubes were then centrifuged in a microcentrifuge for 30 s, and the supernatants were removed. Cell pellets were then resuspended in 110 l of cold lysis buffer (25 mM Tris-HCl, pH 7.4, 50 mM NaCl, 0.5% sodium deoxycholate, 2% IGEPAL, 0.2% SDS, 1 mM phenylmethylsulfonyl fluoride, 50 g/ml aprotinin, 50 M leupeptin, 0.5 mM Na 3 VO 4 ) and placed on ice for 15 min with intermittent shaking. Lysates were centrifuged for 10 min at 14,000 rpm in a refrigerated Eppendorf microcentrifuge, and the supernatants (98 l) were removed and mixed with 42 l of 3.3ϫ sample buffer. Samples were boiled (5 min) and frozen (Ϫ20°C) until used in the immunoblot analysis.
Phosphatase Assay-Cells were washed in phosphate-buffered saline and resuspended in serum-free RPMI 1640. Fifteen ml containing 1.25 ϫ 10 6 cells/ml were added to 15-ml conical tubes and placed at 37°C for at least 1 h prior to stimulation. Cells were treated with hydrogen peroxide (1 mM) or microcystin-LR (4 nM) for 15 min. Cells were lysed in 300 l of cold lysis buffer (25 mM Tris-HCl, pH 7.4, 50 mM NaCl, 0.5% sodium deoxycholate, 1% IGEPAL, 1 mM phenylmethylsulfonyl fluoride, 50 g/ml aprotinin, 50 M leupeptin, 0.5 mM Na 3 VO 4 ). Phosphatase activity was determined using the serine/threonine phosphatase assay system from Promega (Madison, WI). This system measures the amount of free phosphate generated in a reaction by measuring the absorbance of a molybdate-malachite green-phosphate complex (25,26). Briefly, free phosphate was removed from the samples (250 l) using a Sephadex® G-25 spin column (600 ϫ g for 5 min at 4°C) after the columns were equalized with 10 ml of cold lysis buffer. The phosphatase reaction was performed in a 96-well plate containing 10 l of 5ϫ PP2A reaction buffer (250 mM imidazole, pH 7.2, 1 mM EGTA, 0.1% ␤-mercaptoethanol, 0.5 mg/ml bovine serum albumin), 5 l of phosphopeptide (sequence of RRA(pT)VA), 5 l of the sample, and 30 l of double distilled H 2 O. The plate was incubated at 30°C for 30 min, and the reaction was stopped by adding 50 l of the Molybdate dye solution. The plate was left at room temperature for 15 min to allow for color development, and optical density of the samples was read on a plate reader with a 600-nm filter. Values were compared with a standard curve constructed using free phosphate, and values are expressed as pmol of phosphate/min/g of protein.
Thiol Labeling-Reduced thiol groups were labeled using N ␣ -(3-maleimidylpropionyl)biocytin (MPB) (27,28). Cells were washed in phosphate-buffered saline and resuspended in serum-free RPMI 1640. Fifteen ml containing 1.25 ϫ 10 6 cells/ml were added to 15-ml conical tubes and placed at 37°C for at least 1 h prior to stimulation. Cells were treated with hydrogen peroxide (1 mM) or PMA (100 nM) for 15 min. Cells were lysed in 300 l of cold lysis buffer (25 mM Tris-HCl, pH 7.4, 50 mM NaCl, 0.5% sodium deoxycholate, 2% IGEPAL, 1 mM phenylmethylsulfonyl fluoride, 50 g/ml aprotinin, 50 M leupeptin, 0.5 mM Na 3 VO 4 ). Protein concentration of the lysates was determined by using the Bio-Rad DC protein assay kit following the manufacturer's protocols. MPB was added at a concentration of 0.1 mg/mg of protein, and lysates were incubated overnight at 4°C in the dark. Lysates were immunoprecipitated as described above using 2.5 g of the anti-CaM KIV antibody. Immunoprecipitates were washed twice in cold lysis buffer and resuspended in 25 l of cold lysis buffer. Immunoprecipitates were mixed with 10 l of 3.3ϫ sample buffer. Samples were boiled (5 min) and frozen. SDS-PAGE was performed, and proteins were electrophoretically transferred to polyvinylidene fluoride membranes. The membranes were blocked overnight at 4°C in blocking buffer (25 mM Tris-HCl, pH 8.0, 125 mM NaCl, 0.1% Tween 20, 1% bovine serum albumin, 0.1% sodium azide). Membranes were then incubated with AP-conjugated streptavidin diluted in blocking buffer for 1 h. The blots were washed twice in TBST and developed with the colorogenic substrates BCIP and NBT (Promega protoblot AP system).

RESULTS
We and others have suggested the possibility for the activation of the CaM kinases in the absence of a calcium flux (22,23). To test this possibility, we stimulated Jurkat cells in the presence and absence of 2 mM EGTA. We have shown previously that this amount of EGTA is sufficient to block any calcium flux into the cells and also strips calcium from the cells, preventing the release of internal calcium stores (22). We then measured the ability of immunoprecipitates of CaM kinase II and IV from these cells to phosphorylate the CaM kinasespecific substrate syntide-2. This kinase assay was performed in the presence of calcium and calmodulin to measure total activity but also in the presence of EGTA to measure autonomous activity. The presence of EGTA in the cultures prior to stimulation elucidates the role of calcium in the activation process of the kinases. The presence of EGTA in the kinase reaction determines the autonomous activity of the kinase; the kinase has already activated and is now free of any calcium requirement. We found that CaM-KII and CaM-KIV activity was increased following stimulation with hydrogen peroxide both in the presence and absence of EGTA during stimulation (Fig. 1, A and B). These increases in CaM kinase activity were similar to the levels induced by ionomycin, a calcium ionophore. However, unlike hydrogen peroxide-induced activity, ionomycin-induced CaM kinase activity was blocked by the addition of EGTA to the cultures. Thus, hydrogen peroxide could induce CaM kinase activity both in the presence or absence of a calcium flux, whereas ionomycin only induced CaM kinase activity in the presence of a calcium flux (Fig. 1, A and  B). We also examined the CaM kinase activity in these samples (Fig. 1, C and D) following the addition of calcium and calmodulin to the kinase reactions. The levels of activity of both CaM kinase II and CaM kinase IV were increased by the addition of exogenous Ca 2ϩ and calmodulin confirming that the kinase activity found in the immunoprecipitates can be attributed to the CaM kinases. The addition of Ca 2ϩ /CaM leads to the full activation of CaM kinase II because of autophosphorylation (8,29). As can be seen in Fig. 1C the level of activity does not change with treatment. These data indicate two things; 1) equal amounts of CaM-KII are present in the immunoprecipitates, and 2) the activity is because of CaM-KII. If the activity were because of a calcium-independent protein following stimulation with hydrogen peroxide then the differences between the treatment groups would still exist on top of the total CaM-KII activity. These results are similar to the results that many other investigators report when calcium and calmodulin are added to CaM-KII kinase assays (29,30). In Fig. 1D the results are interpreted a little differently; CaM-KIV does not autophosphorylate but is instead phosphorylated by CaM-KK, which is not present in the immunoprecipitates. When calcium and calmodulin are added to these immunoprecipitates, you would expect to see increases in activity, but the highest levels of activation cannot be reached without the phosphorylation by CaM-KK. Thus, differences between the treatment groups will still exist because the CaM-KIV is differentially phosphorylated in the treatment groups. These results are also in agreement with what others report for total CaM-KIV activity (14,16,31). Immunoblots of CaM-KII and CaM-KIV immunoprecipitates indicate that the treatments did not influence the amount of kinase present in the immunoprecipitates (Fig. 1E). When an irrelevant antibody (mouse immunoglobulin) was used to perform similar immunoprecipitates, no changes in kinase activity could be noted following hydrogen peroxide treatment (control, 14,144 Ϯ 5577, versus hydrogen peroxide, 15,803 Ϯ 8269). Additionally, no changes in kinase activity could be noted when assayed in the presence of calcium and calmodulin. It should be noted that the same antibody serves as an irrelevant control for both CaM-KII and CaM-KIV.
Hughes et al. (23) demonstrated that PMA was capable of inducing NF-B activation in these cells and that this response was inhibited by the addition of CaM kinase inhibitors. We have reported previously that similar to their findings, PMA can induce IB degradation and that this degradation can be blocked by the addition of a CaM kinase inhibitor (22). Additionally, it has been suggested that PMA can cause the production of oxygen radicals in Jurkat cells and that binding of NF-B to a probe is inhibited by both catalase and antioxidants was then immunoprecipitated from cell lysates and used in a kinase assay containing EGTA to measure the autonomous phosphorylation of the synthetic peptide, syntide-2. The cpm of 32 P incorporated into the peptide was measured on a scintillation counter. Data are expressed as cpm (CPM) Ϯ S.D. Iono, ionomycin. C, Jurkat cells were loaded with Fluo-3 (10 M) for 30 min at room temperature. Cells were washed and resuspended in phenol red-free RPMI 1640 in a black opaque 96-well plate. Where indicated, cells were pretreated with EGTA (2 mM) before stimulation with okadaic acid (1 M), microcystin-LR (500 nM), or cross-linked OKT3 (5 g/ml). Fluo-3 fluorescence was measured using a Bio-Tek fluorescent microtiter plate reader (excitation, 495 nm, and emission, 520 nm) over the indicated time period (in s). (32)(33)(34). PMA does not cause a calcium flux in these cells as measured using Fluo-3 (Fig. 2). Stimulation of the cells with PMA results in the activation of both CaM kinase II and IV (Fig. 3, A and B). Because PMA does not cause a calcium flux, this activation must also be occurring in a manner independent of increases in intracellular calcium. Additionally, PMA was also able to activate the CaM kinases in the presence of EGTA, and as before, ionomycin-induced CaM kinase activity was blocked by the addition of EGTA to the cultures (Fig. 3, A and  B) suggesting that PMA-induced activation of the CaM kinases also occurs in the absence of increases in intracellular calcium. To determine whether this activity was dependent on the induction of oxygen radicals we stimulated Jurkat cells with PMA in the presence of catalase. Catalase was shown to inhibit the activation of both CaM kinase II and IV by PMA (Fig. 3, C  and D). Catalase has no effect on ionomycin-stimulated cells demonstrating that catalase is specific in its actions. These results suggest that PMA is acting via the production of oxygen radicals. In support of this, glucose oxidase, another compound that causes the production of peroxide, is also able to activate both CaM kinase II and IV (Fig. 4). These results, taken together, suggest that both CaM kinase II and IV can be activated in a calcium-independent manner in response to ROI.
Because of their sensitivity to oxidation, one potential mechanism for the increase in CaM kinase activity is the inactivation of regulatory phosphatases. When the PP1/PP2A inhibitors okadaic acid or microcystin-LR were added to the cultures, activation of both CaM kinase II and IV occurred (Fig. 5, A and  B). Both of these inhibitors induce a transient calcium flux in these cells that can be blocked by EGTA pretreatment (Fig.  5C). OKT3 is an antibody directed toward the T cell receptor and serves as a positive control for inducing an intracellular calcium flux in these cells. Similar to the results obtained with hydrogen peroxide, EGTA has no effect on the ability of these inhibitors to induce kinase activity suggesting that activation of the CaM kinases by phosphatase inhibitors also occurs in a calcium-independent manner (Fig. 5, A and B).
We have shown previously that hydrogen peroxide can cause the phosphorylation of the antiapoptotic kinase Akt (22). This phosphorylation was inhibited by CaM kinase inhibitors but was insensitive to the calcium chelators EGTA and BAPTA-AM. We can also show that the phosphatase inhibitors okadaic acid and microcystin-LR induce the phosphorylation of Akt on Thr-308 (Fig. 6). These results show that inhibition of phosphatase activity is sufficient in itself to induce Akt phosphorylation, mimicking the effects of hydrogen peroxide and PMA. Taken with the above results, we have shown that phosphatase inhibition can activate the CaM kinases in a manner similar to treatment with hydrogen peroxide or PMA. Further, downstream CaM kinase targets are also phosphorylated after phosphatase inhibition.
Although it has been shown that hydrogen peroxide can inhibit phosphatase activity in other cell types, we determined whether this occurs in the Jurkat cell line. After treatment with hydrogen peroxide, we tested the ability of whole lysates to cleave the phosphate from a synthetic peptide in PP2Aspecific phosphatase buffer. Free phosphate could be detected by the use of a molybdate dye and measured in a spectrophotometric assay. We found that following treatment with hydrogen peroxide, the phosphatase activity in these lysates decreased (Fig. 7). Furthermore, this activity was similar to the activity seen after treatment with the phosphatase inhibitor microcystin-LR at a level consistent with inhibition of PP2A (Fig. 7). Others have shown that PP2A will form a complex with CaM-KIV and can be co-immunoprecipitated with CaM-KIV (17). It is believed that in vivo, PP2A is integral in the inactivation of CaM-KIV (16,17).
We attempted to determine the phosphatase activity in immunoprecipitates of CaM-KIV, but we found that PP2A is inactivated because of oxidation during the immunoprecipitation process. As an alternative method, we used N ␣ -(3-maleimidylpropionyl)biocytin, which will only react with reduced cysteine residues (27,28). Oxidation of the active cysteine on PP2A results in its inactivation. We immunoprecipitated CaM-KIV and performed an immunoblot to demonstrate that PP2A

FIG. 6. Phosphatase inhibition causes phosphorylation of Akt.
Jurkat cells were washed and resuspended at 1.25 ϫ 10 6 cells/ml in RPMI 1640 containing 5% fetal calf serum. Cells were treated with either (A) okadaic acid (1 M) or (B) microcystin-LR (500 nM) for the indicated time. Cells were lysed and subjected to immunoblot analysis using antibodies against phospho-Akt (Thr-308) or total Akt. Bands were visualized using an AP-BCIP/NBT developing system.

FIG. 7. PP2A phosphatase activity is inhibited by peroxides.
Jurkat cells were washed and resuspended in RPMI 1640 at a concentration of 1.25 ϫ 10 6 cells/ml. Cells were treated with either hydrogen peroxide (1 mM) for 10 min or microcystin-LR (M-LR) (500 nM) for 30 min. Cells were lysed in SDS-free radioimmune precipitation assay buffer and run on a column to remove free phosphate. Ten l of whole cell lysate were then incubated in a phosphate buffer specific for the measurement of PP2A activity. Dephosphorylation of a synthetic peptide was measured using a molybdate dye, and absorbance was measured on a plate reader at 630 nm. Values were compared with a standard curve of free phosphate.
does indeed co-immunoprecipitate with CaM-KIV (Fig. 8A). These data also indicate that an equal amount of PP2A immunoprecipitates with CaM-KIV regardless of the treatment group. It should be noted that the PP2A seen in the immunoblots in Fig. 8, A and B, represents only the PP2A associated with CaM-KIV and likely represents only a small portion of the total PP2A in the cell. Control immunoprecipitates using an irrelevant antibody (mouse immunoglobulin) failed to immunoprecipitate PP2A (Fig. 8E). Cysteine labeling shows that following hydrogen peroxide treatment, PP2A is indeed oxidized (Fig. 8B). Additionally, PMA treatment also results in the oxidation of PP2A (Fig. 8B), supporting our earlier results showing catalase inhibition of PMA-induced CaM kinase activity. This oxidation of PP2A can be noted both in the subset of PP2A that associates with CaM-KIV (Fig. 8B) as well as the total cellular PP2A (Fig. 8D). Cysteine labeling demonstrates that CaM-KIV is also oxidized under these conditions (Fig. 8C).
To determine whether this change in cysteine oxidation could directly activate CaM-KIV we added hydrogen peroxide directly to CaM-KIV or CaM-KII immunoprecipitates. When kinase assays were performed on these samples, a reduction in kinase activity could be noted indicating that the direct oxidation of CaM-KIV or CaM-KII was not mediating the increase in activity (Fig. 9). Considering both the phosphatase assay as well as cysteine labeling, we can demonstrate that hydrogen peroxide causes the oxidation and subsequent inactivation of the phosphatase PP2A. Inactivation of this phosphatase is a potential mechanism for CaM kinase activation by ROI. DISCUSSION We and others have suggested the possibility of a mechanism that activates the CaM kinases in the absence of a calcium flux (22,23). In this paper we elucidate this novel form of regulation of the CaM kinases. Using in vitro kinase assays, we were able to demonstrate that hydrogen peroxide can activate both CaM kinase II and IV within Jurkat T cells (Fig. 1). To the best of our knowledge, no one has demonstrated that the CaM kinases are sensitive to redox changes in the cell. Although H 2 O 2 can cause a calcium flux in these cells, we found that blocking this calcium flux with EGTA has no effect on the activation of the CaM kinases (Fig. 1). These results are novel as an increase in intracellular calcium has always been thought to be necessary for the activation of these kinases.
Hughes et al. (23) demonstrated that PMA was capable of inducing NF-B activation in Jurkat cells and that this response was inhibited by the addition of CaM kinase inhibitors. In this work we demonstrate that PMA treatment activates both CaM kinase II and IV (Fig. 3). It has been shown previously that PMA treatment of human T cells resulted in an increase in intracellular hydrogen peroxide as measured using a fluorescent indicator (35). In addition, treatment of cells with either N-acetylcysteine, an antioxidant (33), or catalase (34) could block PMA-induced NF-B activity. The addition of catalase to the cells before PMA stimulation was able to prevent the activation of both CaM kinase II and IV (Fig. 3, C and D). Because catalase is specific for hydrogen peroxide, we have shown that the PMA-induced activation of the CaM kinases is directly related to hydrogen peroxide production in the cell. Because catalase can block PMA-induced activation of the CaM kinases, we postulate that both PMA and hydrogen peroxide are activating the kinases in a similar manner.
Although calcium is not required for activation of the CaM kinases, it was shown previously that calmodulin antagonists FIG. 8. The cysteine residues of CaM kinase IV-associated PP2A are oxidized. Jurkat cells were washed and resuspended in RPMI 1640 at a concentration of 1.25 ϫ 10 6 cells/ml. Cells were treated with either hydrogen peroxide (1 mM) or PMA (10 nM) for 10 min. Cells were lysed using a SDS-free radioimmune precipitation assay lysis buffer and incubated with N ␣ -(3-maleimidylpropionyl)biocytin (0.1 g/ml) at 4°C overnight. Immunoprecipitation was performed using an antibody against CaM kinase IV. A, immunoblot analysis was carried out on the immunoprecipitates using antibodies against CaM kinase IV and PP2A. B and C, N ␣ -(3-maleimidylpropionyl)biocytin-labeled proteins were detected using APconjugated streptavidin. B represents N ␣ -(3-maleimidylpropionyl)biocytin-labeled labeling of CaM-KIV-associated PP2A, and C represents labeling of total CaM-KIV. D, PP2A was immunoprecipitated from cells as treated above. This blot represents N ␣ -(3-maleimidylpropionyl)biocytinlabeled total PP2A. Bands were visualized using an AP-BCIP/NBT developing system. In E, either mouse immunoglobulin IgG1 or CaM-KIV was used to immunoprecipitate, and the immunoprecipitates were immunoblotted with either PP2A or CaM-KIV. The lane labeled WCL contains whole cell lysates from Jurkat cells.
could block the CaM kinase-dependent phosphorylation of IB in response to both hydrogen peroxide (22) and PMA (24). Additionally, although EGTA blocks a calcium flux into these cells, basal levels of calcium within the cell may still be required for this process. It is believed that calmodulin weakly associates with the calmodulin binding domain in the kinases even before binding to calcium (36). When calcium levels increase, the binding of calmodulin to calcium causes a conformational change that opens a hydrophobic pocket that enables it to bind to target proteins with a greater affinity (36). There are several possible roles that calmodulin may play in this system even in the absence of a calcium flux. Other cellular processes are regulated by calcium-free calmodulin (apo-CaM) (37)(38)(39)(40); potentially calmodulin is able to affect the CaM kinases when not bound to calcium as well. Additionally, calmodulin may be phosphorylated posttranslationally, and this modification may alter the affinity calmodulin has for either Ca 2ϩ or other proteins (41,42). In fact, the tyrosine phosphorylation of calmodulin caused a decrease in the concentration of calmodulin required for half-maximal activation of CaM kinase II (41). The affinity of CaM kinase II has also been shown to be increased for calmodulin once autophosphorylation of the protein has occurred. The affinity of CaM-KII for calmodulin increases over 1000-fold, converting it from a relatively poor CaM-binding enzyme into one with an extremely high affinity, an effect termed "CaM trapping" (43). This process has not yet been shown for CaM kinase IV. Finally, it may be that once CaM binds to the kinase that it may have a higher affinity for cellular Ca 2ϩ . This has been demonstrated for another CaM kinase family member, myosin light chain kinase, where the affinity of CaM for Ca 2ϩ is increased upon binding to the kinase, then further increased once the kinase binds to its substrate (44,45). Any of these changes in binding affinity may suggest how the kinases are activated in the absence of cal-cium, as well as the role that calmodulin plays in this activation.
It should be pointed out that once phosphorylated, both CaM kinase II and IV become independent of Ca 2ϩ changes within the cell. In many cases, phosphatase and kinase activity balance in a resting cell. It is well established that nonspecific inactivation of tyrosine phosphatases results in an overall increase in phosphorylation within a cell. Increases in kinase activity result in increased protein phosphorylation in the absence of any change in phosphatase activity. It would also be expected that a decrease in phosphatase activity in the absence of any change in kinase activity could also result in the increased phosphorylation of a protein.
Regulation of phosphatases by reactive oxygen species is an attractive way for cells to lower the threshold for the activation of many different signaling pathways simultaneously. In many cells, reactive oxygen species are produced after receptor ligation (46 -49). Although receptor ligation results in the phosphorylation of upstream kinases, if inhibitory phosphatases are concurrently inactivated by reactive oxygen species, then the threshold for the activation of a particular pathway is decreased. The initial lowering of the activation threshold may be required for the initiation of certain pathways and then can be dispensable once a certain number of kinases have been activated. Cross-linking of the T cell receptor leads to the production of both H 2 O 2 and O 2 . , and both may play a role in the subsequent activation of the T cell (48). Because PP2A is involved in the regulation of both CaM kinases and based on the used concentrations of the phosphatase inhibitors, we focused on the effects of hydrogen peroxide on PP2A activity within the Jurkat cell. Based on the phosphatase assay, we found that PP2A activity is inhibited by hydrogen peroxide in Jurkat cells (Fig. 7). The effects of oxidants on the activity of PP2A in other cell types have been varied. PP2A activity has been shown to be decreased (20,21), unaffected (50,51), or even enhanced (52) after treatment with oxidants. These divergent results may be because of several different factors. There are several different isoforms of PP2A, of which one or more can be expressed in a particular cell type (53,54). In addition, the relative redox status of one cell as compared with another can be vastly different.
In summary, we have demonstrated that hydrogen peroxide can cause the activation of both CaM kinase II and IV. Most significantly, this activation occurred in the absence of any detectable calcium flux into the cell. In addition, PMA activation of the CaM kinases is also dependent on peroxide production and also occurs in the absence of a calcium flux. This activation of the CaM kinases is likely because of the oxidation and inactivation of phosphatases within the cell. These results demonstrate a novel mechanism by which the CaM kinases can be activated. These results may be significant for the activation of T lymphocytes in the inflammatory environment. Additionally, any cellular process resulting in the production of oxygen radicals can potentially activate the CaM kinases even in the absence of any calcium flux. labeled Control represent untreated cells. Cells were immediately centrifuged, and cells were lysed. Lysates were used to immunoprecipitate CaM kinase II or CaM kinase IV. Immunoprecipitates of either CaM kinase II or IV from untreated cells were incubated with 10 mM hydrogen peroxide in cold lysis buffer for 7 min (in vitro). Immunoprecipitates were used in a kinase assay in the absence of calcium to measure the phosphorylation of the synthetic peptide, syntide-2.