Tumor Promoter Benzoyl Peroxide Induces Sulfhydryl Oxidation in Protein Kinase C: Its Reversibility Is Related to the Cellular Resistance to Peroxide-Induced Cytotoxicity

Archives of Biochemistry and Biophysics Vol. 363, No. 2, March 15, pp. 246 –258, 1999 Article ID abbi.1999.1100, available online at http://www.idealibrary.com on

Tumor Promoter Benzoyl Peroxide Induces Sulfhydryl Oxidation in Protein Kinase C: Its Reversibility Is Related to the Cellular Resistance to Peroxide-Induced Cytotoxicity 1 Rayudu Gopalakrishna,* ,2 Usha Gundimeda,* Wayne B. Anderson,† Nancy H. Colburn,‡ and Thomas J. Slaga§ *Department of Cell and Neurobiology, School of Medicine, University of Southern California, Los Angeles, California 90033; †Laboratory of Cellular Oncology, National Cancer Institute, Bethesda, Maryland 20892; ‡Laboratory of Viral Carcinogenesis, NCI-Frederick Cancer Research and Development Center, Frederick, Maryland 21702; and §Center for Cancer Causation and Prevention, AMC Cancer Research Center, Denver, Colorado 80214

Received September 16, 1998, and in revised form December 28, 1998

Since tumor promoter benzoyl peroxide (BPO) mimics phorbol esters in some aspects, its effects on protein kinase C (PKC) were previously studied. However, in those studies due to the presence of thiol agents in the PKC preparations, the sensitive reaction of BPO with redox-active cysteine residues in PKC was not observed. In this study, by excluding thiol agents present in the purified PKC preparation, low concentrations of BPO modified PKC, resulting in the loss of both kinase activity and phorbol ester binding (IC 50 5 0.2 to 0.5 mM). This modification, which was not dependent on transition metals, was totally blocked by a variety of thiol agents including GSH, which directly reacted with BPO. Substoichiometric amounts of BPO (0.4 mol/mol of PKC) oxidized two sulfhydryls in PKC and inactivated the enzyme which was readily reversed by dithiothreitol. The regulatory domain having zinc thiolate structures supporting the membrane-inserting region provided the specificity for PKC reaction with BPO, which partitioned into the membrane. Unlike H 2O 2, BPO did not induce the generation of the Ca 21/lipid-independent activated form of PKC. Other redox-sensitive enzymes such as protein kinase A, phosphorylase kinase, and protein phosphatase 2A required nearly 25- to 100-fold higher concentrations of BPO for inactivation. BPO also inactivated PKC in a variety of cell types. In the JB6 (30 P 2) nonpromotable cell line and other normal cell lines, 1

This work was supported by Grant CA62146 from the National Cancer Institute. A preliminary account of this work was presented at the 89th Annual Meeting of the American Association for Cancer Research, New Orleans, March 28 –April 1, 1998. 2 To whom correspondence should be addressed at Department of Cell and Neurobiology, USC School of Medicine, 1333 San Pablo St., MMR-330, Los Angeles, CA 90033. Fax: 323-442-1771. E-mail: [email protected]. 246

where BPO was more cytotoxic, it readily inactivated PKC due to a slow reversibility of this inactivation by the cell. However, in the JB6 (41 P 1) promotable cell line, C3H10T1/2 and B16 melanoma cells, where BPO was less cytotoxic, it did not readily inactivate PKC due to a rapid reversibility of this inactivation by an endogenous mechanism. Nevertheless, BPO inactivated PKC at an equal rate in the homogenates prepared from all these cell types. Inclusion of NADPH reversed this inactivation in the homogenates to a different extent, presumably due to a difference in distribution of a protein disulfide reductase, which reverses this oxidative modification. BPO-induced modification of PKC occurred independent of the cellular status of GSH. However, externally added GSH and cell-impermeable thiol agents prevented the BPO-induced modification of PKC. Since BPO readily partitions into membranes, its reaction with redox-cycling thiols of membrane proteins such as PKC may trigger epigenetic events to prevent cytotoxicity, but favor tumor promotion. © 1999 Academic Press Key Words: benzoyl peroxide; protein kinase C; zinc fingers; oxidation of cysteine residues; tumor promoters.

Benzoyl peroxide (BPO), 3 a widely used free radicalgenerating compound in industry, is a tumor promoter 3

Abbreviations used are: BPO, benzoyl peroxide; PKC, protein kinase C; PMA, phorbol 12-myristate 13-acetate; PDBu, phorbol 12,13-acetate; BSO, D,L-buthionine-S,R-sulfoximine; DTNB, 5,59-dithiobis(2-nitrobenzoic acid); ME, 2-mercaptoethanol; DTT, dithiothreitol; PKA, protein kinase A; PP2A, protein phosphatase 2A; DTPA, diethylenetriaminepentaacetic acid; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; MEM, minimal essential medium. 0003-9861/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

REVERSIBLE SULFHYDRYL OXIDATION OF PKC BY BENZOYL PEROXIDE

and progression agent in mouse skin (1, 2). The tumorpromoting activity of BPO also has been observed in cell transformation assays (3, 4). Similar to the phorbol ester type of tumor promoters, BPO induces regenerative hyperplasia, hyperkeratinization, an increase in dark basal keratinocytes, and ornithine decarboxylase activity (5). Unlike phorbol esters, however, BPO does not induce inflammation and the release of eicosanoids (6). The mechanism by which BPO exerts tumor-promoting activity is not known. BPO is activated by a copper-mediated breakdown into benzoyloxyl radicals, which have been shown to induce DNA-strand scission and the promutagenic DNA base modification such as formation of 8-hydroxydeoxyguanosine (7–9). It was also shown to induce mitochondrial damage under conditions where no formation of free radicals was observed (10, 11). The breakdown of BPO with nucleophilic sulfur compounds without the formation of free radicals was reported several years ago (12). A comparative study carried out with a variety of organic peroxides suggested that their biological activities related to tumor promotion could not be entirely explained based on the nature of the primary free radical formed, but rather suggested that additional properties of these agents may also play a role in this process (6). Furthermore the signal transduction mechanisms that influence the epigenetic events and their relation to genetic changes induced by BPO have yet to be elucidated. Tumor promoters such as phorbol esters, mezerein, and telocidin reversibly bind to protein kinase C (PKC) to activate this kinase, which ultimately leads to its downregulation (13). Furthermore, oxidant tumor promoters such as m-periodate, inflammatory H 2O 2, and cigarette smoke polyphenolic agents can also activate and inactivate PKC, suggesting that this enzyme may be a receptor not only for phorbol esters but also for a variety of tumor promoters (14 –18). Nevertheless, unlike these agents, BPO was shown to have no direct activating effect on purified PKC (19). However, in mouse skin treated with BPO, a decrease in Ca 21dependent PKC activity and a concomitant formation of Ca 21-independent activity was observed (20). In the same study, BPO was shown to have no direct effect on the isolated PKC (20). Two other studies, however, reported that PKC was inhibited by high (0.1 to 10 mM) concentrations of BPO (21, 22). In another study, inhibition of PKC was observed in the test tube at high concentrations of BPO, while there was no effect in intact cells (23). Furthermore, it was suggested that the inhibition of PKC by BPO was not due to an oxidative modification, but rather due to the specific binding of BPO to the regulatory domain by having a structural similarity to the PKC activator, diacylglycerol, and, as a result, PKA was not affected (21). In another study, inhibition of both PKC and PKA suggested a nonspecific binding of BPO to the catalytic domain at

247

the nucleotide-binding site (22). None of these studies, however, took into account the reactivity of thiols with BPO. Therefore, the mercapto agents (10 to 50 mM) used in the preparation of PKC were not removed during the incubation of PKC with BPO (20 –23). Moreover, these studies did not consider a high susceptibility of PKC to oxidative modification and its reversibility in the cell. Both the regulatory and catalytic domains of PKC have unique cysteine-rich regions, which are very susceptible to oxidative modification (17). Depending on the type of oxidant, the site, and the extent of modification, PKC can be either activated or inactivated (14 – 18). The regulatory domain has four zinc fingers coordinated by 12 cysteine residues present within the vicinity of a hydrophobic region, which inserts this kinase into the membrane (24 –26). Given the fact that zinc-finger thiolates are highly redox sensitive (27, 28), membrane-associated PKC may be a logical target for lipophilic peroxides such as benzoyl peroxide, which partition into the membrane. Furthermore, BPO is not broken down by glutathione peroxidase (10). Therefore, it is important to determine whether the redox modification sulfhydryls in PKC by BPO act as a sensor to couple peroxide breakdown, to signaling for tumor promotion, and/or to prevent the cytotoxicity induced by this peroxide. While BPO is a strong cytotoxic agent, prior initiation with a carcinogen was shown to provide a transformed phenotype exhibiting a resistance to this toxicity. Furthermore, it was noted that sublethal oxidative stress elicited by BPO might give a selective growth advantage to the initiated cell for promotability (29). In addition, studies carried out with promotable (41 P 1) JB6 skin epidermal cell lines with oxidants suggested that antioxidant enzymes may contribute to the relative resistance of the transformed cell to cytotoxic oxidants (30, 31). However, it is not known whether the breakdown of BPO and/or the reversal of BPO-induced protein sulfhydryl modifications are high in transformed cells. In this communication, we show BPO-induced oxidation of cysteine residues in PKC and its reversal by a cellular reductive mechanism utilizing NADPH/glucose metabolism and discuss the significance of this process in overcoming the cytotoxicity which is required for promotion of the transformed cell phenotype by a strong oxidant. MATERIALS AND METHODS Materials. BPO and diethylenetriaminepentaacetic acid (DTPA) were from Aldrich. 5,5-Dithiobis(2-nitrobenzoic acid) (DTNB), BSO, protein kinase A, and phosphorylase kinase were from Sigma. Dioleoylphosphatidylcholine and dioleoylphosphatidylserine were from Avanti Polar Lipids. [20- 3H]Phorbol 12,13-dibutyrate (sp act 20 Ci/ mmol) was from Dupont NEN, and [g- 32P]ATP (sp act 20 Ci/mmol) was from ICN. A mixture of a, b, and g isoenzymes of PKC was purified from rat brain (17). In some cases, individual PKC isoen-

248

GOPALAKRISHNA ET AL.

zymes, which were separated by hydroxylapatite (32), were used. The catalytic and regulatory domains of PKC were separated by trypsin treatment (24). Fresh stock solutions of BPO were prepared in ethanol and sonicated prior to use. The final concentration of ethanol was maintained at less than 0.5% in all samples including the control. There was a solubility problem with BPO at higher (250 mM) concentrations. Similar results were obtained using acetone as the solvent. PKC assay and phorbol ester binding. Both protein kinase activity using histone H1 as the substrate and phorbol ester binding using 3 [ H]phorbol 12, 13-dibutyrate (PDBu) as the ligand with the conditions standardized for purified enzyme (method 1) were determined using multiwell filtration assays as described before (33). PKC activity was expressed as units, where one unit of enzyme transfers 1 nmol of phosphate to histone H1 per min at 30°C. Quantitation of cysteine sulfhydryls in BPO-modified PKC. Highly purified PKC (0.5 nmol), free from thiol agents, was incubated in triplicates with various amounts of BPO (0.25 to 10 mol/mol of PKC) in 100 ml of 20 mM Tris–HCl, pH 7.5/1 mM CaCl 2 for 10 min at 30°C. Excess BPO was removed by a centrifuge gel filtration using Bio-Spin 6 (Bio-Rad) columns, and then the samples were transferred to a 96-well plate. Sulfhydryls were quantitated by a modified DTNB method (34, 35). To each well, 50 ml of 1.2 mM DTNB was added and the absorbance was read at 405 nm. To measure the total sulfhydryls present in PKC, 50 ml of 10% SDS was added to all samples, and the absorbance at 405 nm was read again after 5 min. N-Acetylcysteine was used as the standard for sulfhydryl quantitation. Insertion of BPO into liposomes. Liposomes were prepared by using nonperoxidizable phospholipids, which maintain membrane fluidity as well as support PKC binding and activation. Dioleoylphosphatidylcholine (0.4 mmol), dioleoylphosphatidylserine (0.1 mmol), and cholesterol (0.24 mmol) along with BPO (1 to 100 nmol) were mixed in chloroform and the organic solvent was evaporated. To the dried lipid layer, 5 ml of 20 mM Tris–HCl, pH 7.5, was added and multilammellar vesicles were prepared by sonication for 3 min. Initially, these liposomes, having increasing amounts of BPO, were incubated with PKC along with 1 mM CaCl 2 in a total volume of 250 ml for 5 min at 30°C. Then, without delay, aliquots of these modified proteins were used to determine the residual kinase activity and PDBu binding. Isolation of PKC from cells treated with BPO. Cells were grown in 100-mm petri dishes in minimal essential medium (MEM) supplemented with heat-treated 5% fetal calf serum. When indicated, cells were serum starved at confluency by incubation with serum-free MEM (0.1% serum) for 24 h. In some cases, serum-starved cells were treated with 2-deoxyglucose (30 mM) for 4 h. Cells in the indicated medium were treated with BPO, and then were homogenized in buffer A (20 mM Tris–HCl, pH 7.5/1 mM EDTA/ 0.5 mM phenylmethylsulfonyl fluoride/150 nM pepstatin A) containing 1% Nonidet P-40. Unless otherwise indicated, mercapto compounds were omitted from all the buffers. The cell extracts were subjected to DEAEcellulose chromatography as described previously (33). Evaluation of cytotoxicity of BPO. The cytotoxicity of BPO with various cell types was assessed by following MTT reduction and cell detachment. Initially, cells were grown in 96-well plates and treated with BPO (1 to 100 mM) for 2 to 18 h in a serum-containing medium or serum-free medium. Then, MTT solution was added and incubated for 4 h. The formazan formed was dissolved in dimethylsulfoxide and the absorbance was read at 550 nm (36). For the cell detachment assay, cells were washed and the cells that remained attached were stained with sulforhodamine B, which was then measured (37). Other methods. In a manner similar to PKC, protein kinase A and phosphorylase kinase were preincubated with BPO and the activities were determined. Instead of Ca 21/lipids and histone H1, 5 mM cAMP and unfractionated histone were used to determine pro-

tein kinase A activity; 0. 1 mM CaCl 2 and phosphorylase B were used in the phosphorylase kinase assay. Protein phosphatase 2A (PP2A) activity was determined by multiwell filtration method using PKCphosphorylated histone H1 as the substrate (38). GSH was measured using an enzymatic recycling assay (39). In all cases, the enzyme activities were determined in duplicates or in triplicates and were expressed in units, where one unit of enzyme produces 1 nmol of product per minute at 30°C. Although the data presented in the figures came from one experiment, similar results were obtained in at least two more experiments.

RESULTS

BPO-induced inhibition of PKC activity and PDBu binding in the absence of mercapto agents. Initially, a direct effect of BPO on the phosphotransferase activity of purified PKC was studied by including BPO in the PKC assay. In the absence of 2-mercaptoethanol (ME), BPO effectively inhibited PKC activity with an IC 50 of 0.2 mM (Fig. 1A). There was no appreciable increase in Ca 21/lipid-independent activity. However, in the presence of as low as 100 mM ME, IC 50 for BPO-induced inhibition was increased by as high as 150-fold. With a further increase in concentration of ME, the concentration of BPO required to inhibit PKC correspondingly increased. A similar type of protection was observed with a variety of thiol agents such as DTT, GSH, and cysteine. However, oxidized DTT, GSSG, and cystine had no protective effect, indicating that a reduced thiol agent is required to protect the enzyme from BPO. Similarly, PDBu binding was inhibited by BPO with an IC 50 of 0.5 mM, which was also blocked by thiol agents (Fig. 1B). Since the PDBu binding assay is less sensitive than the kinase assay, approximately two- to threefold higher amounts of PKC were employed in the PDBu binding studies. As a result, the observed IC 50 value of BPO was correspondingly higher for the inhibition of PDBu binding than that for the inhibition of kinase activity. The kinase activity and PDBu binding of the proteolytically separated catalytic and regulatory domains, respectively, were also inhibited by BPO with the same sensitivity as the holoenzyme. Furthermore, the Ca 21/lipid-independent protamine phosphotransferase activity of the holoenzyme was also inhibited to the same extent by BPO (data not shown). It is considered that the phosphorylation of protamine sulfate represents the activity directly contributed by the catalytic domain of PKC without the need for the binding of lipids to the regulatory domain. Conceivably, the oxidation sensitive sites for BPO reaction are present within both regulatory and catalytic domains. The stable breakdown product of BPO, benzoic acid, had no effect on either PKC activity or PDBu binding even at a high (1 mM) concentration. Reaction of various sulfhydryl compounds with BPO. Whether thiol agents can directly react with BPO to protect PKC was determined. Various thiol agents at a concentration of 1 mM were incubated with increasing concentrations of BPO for 15 min at 30°C. BPO inter-

REVERSIBLE SULFHYDRYL OXIDATION OF PKC BY BENZOYL PEROXIDE

FIG. 1. BPO-induced inhibition of (A) PKC activity and (B) PDBu binding in the absence and presence of variable concentrations of ME. In the wells of a multiwell filtration plate, desalted PKC (0.025 units) was preincubated with various concentrations of BPO for 5 min at 30°C in the absence (open circles) or presence of indicated concentrations (filled symbols) of ME. Then lipids (2.5 mg phosphatidylserine/0.1 mg diolein) along with prewarmed reaction mixture were added. The complete reaction mixture was incubated for 5 min at 30°C to determine PKC histone phosphotransferase activity. For determining BPO-induced inhibition of PDBu binding, PKC (0.07 units) was similarly preincubated with BPO. Phosphatidylserine (2.5 mg), followed by [ 3H]PDBu binding mixture, was added, and then [ 3H]PDBu bound to PKC was determined as described previously (33).

feres with the sulfhydryl quantitation method by reacting with chromogenic thionitrobenzoic acid formed by the reaction of sulfhydryls with DTNB. Therefore, the excess BPO was extracted with organic solvent and the thiols present in aqueous medium were then quantitated using DTNB. As shown in Fig. 2, incubation with BPO decreased thiols by approximately 4 to 5 mol/mol of BPO. Similarly, other agents such as DTT were also oxidized to a similar extent based on their thiol content. These thiol modifications occurred even in the presence of the metal chelator, DTPA, suggesting that there is no need for a transition metal in this reaction. Thiol reversible and irreversible oxidation of PKC. To determine whether BPO inhibited PKC activity either by reversibly binding to PKC or by inducing a

249

modification of the enzyme, PKC was initially treated with BPO in the absence of mercapto agents. Detergent was then added and PKC was isolated free from BPO by DEAE-cellulose chromatography. The treatment with BPO in the presence of the chelator, DTPA, decreased PKC activity with an IC 50 of 0.8 mM (Fig. 3). This loss in enzyme activity was not recovered by extensive dialysis, suggesting that BPO, indeed, induced a modification of PKC. Inclusion of 5 mM CuCl 2 or FeCl 2 did not further increase the rate of inactivation. However, the inclusion of 1 mM CaCl 2 enhanced PKC inactivation induced with low concentrations of BPO. Since Ca 21 induces a hydrophobic site on PKC (17), it may act by facilitating the interaction of BPO with cysteine residues present within the hydrophobic region. When BPO-modified PKC was incubated with DTT, PKC activity was substantially regenerated, especially when lower (,2 mM) concentrations of BPO were used for PKC modification. Nevertheless, the ability of DTT to regenerate PKC activity was gradually diminished with an increase in concentration of BPO used for PKC modification. BPO-induced oxidation of cysteine residues in PKC. The quantitation of PKC sulfhydryls, remaining after BPO treatment was carried out in two steps. First, purified PKC (0.5 nmol) was treated with various amounts of BPO and the excess, unreacted BPO was removed by gel filtration. Then the accessible sulfhydryls present in the modified PKC were determined in the absence and presence of SDS. Since the Ca 21-dependent PKC isoenzymes a, b, and g exhibited identi-

FIG. 2. Direct reaction of BPO with various thiol agents. N-Acetylcysteine, ME, and GSH (1 mM) in 2 ml of 20 mM Tris–HCl, pH 7.5/1 mM DTPA were incubated with the indicated concentrations of BPO for 15 min at 30°C. Then BPO was extracted with 4 ml chloroform:methanol (2:1, v/v). The unreacted thiols were measured by taking 50 ml of the aqueous layers into wells of a 96-well microtiter plate and then adding 150 ml of 0.5 mM DTNB. The absorbance of thionitrobenzoate formed was read at 405 nm.

250

GOPALAKRISHNA ET AL.

FIG. 3. The phosphotransferase activity of PKC modified with BPO in the presence and absence of Ca 21. PKC was incubated with the indicated concentrations of BPO in the presence of either 1 mM DTPA or 1 mM CaCl 2 for 10 min at 30°C. The detergent NP-40 (1% final) was added to all samples. Where indicated, DTT (10 mM final) was added, and the samples were left on ice for 10 min. Excess unreacted BPO was removed by applying these samples to a DEAE-cellulose column (0.5 ml), and the enzyme was eluted with 0.2 M NaCl.

cal susceptibility to BPO and have 16 of 20 (average) conserved cysteine residues in their sequences, the mixture of all three isoenzymes was employed for quantitation of sulfhydryls. Only 18 sulfhydryls were

titrated with DTNB after SDS denaturation of the control untreated protein (Fig. 4). With a low amount of BPO (0.4 mol/mol of PKC), DTNB-titratable sulfhydryls were decreased by two residues and PKC activity was decreased by approximately 80% of the control untreated enzyme. This activity was substantially recovered by the incubation of BPO-treated PKC with DTT. However, with an increase in BPO concentration, there was further modification of sulfhydryls in PKC, but the enzyme activity was not readily reactivated by DTT treatment. Even with a high amount of BPO (10 mol/mol of PKC), nearly 10 cysteine sulfhydryls were not modified. These cysteine residues may be deeply buried within the protein and thus not accessible for the BPO-induced modification. Reaction of liposomal membrane-inserted BPO with the regulatory domain of PKC. Since BPO is lipophilic and is not soluble in aqueous medium, it is expected to partition into the membrane. Therefore, we tested whether BPO inserted into the lipid bilayer can react with PKC or not. Since BPO induces lipid peroxidation (40) and the lipid peroxidation products can damage proteins, liposomes were prepared with nonperoxidizable phospholipids that also maintain certain degree of unsaturation needed for membrane fluidity. Furthermore, the phospholipid composition used (phosphatidylcholine and phosphatidylserine, 4:1 mol/mol) was previously shown to be optimal to support the Ca 21dependent binding and insertion of PKC into lipid

FIG. 4. Quantitation of cysteine sulfhydryls in PKC modified with BPO and reversal of this modification by DTT. Highly purified PKC (0.5 nmol) was incubated in triplicate with the indicated amounts of BPO. After removing the excess BPO by centrifuge gel filtration the sulfhydryls present in PKC were determined using DTNB (solid lines). The total sulfhydryls in PKC were measured after the addition of SDS. Simultaneously, small aliquots of treated PKC, after removal of BPO, were diluted in a buffer with or without 10 mM DTT and PKC activity (dotted lines) was determined. The values represent means of triplicate estimations.

REVERSIBLE SULFHYDRYL OXIDATION OF PKC BY BENZOYL PEROXIDE

251

FIG. 5. Inactivation of PKC and the regulatory and catalytic domains by BPO inserted into liposomal membrane. BPO was inserted into liposomal membrane during the preparation of multilamellar vesicles. A suspension of lipid vesicles (50 ml) having the indicated amounts of BPO was incubated with PKC holoenzyme or its proteolytically separated regulatory and catalytic domains along with 1 mM CaCl 2 for 5 min. Then aliquots of these modified proteins were used to determine the residual kinase activity (solid line) or PDBu binding (dotted line). Closed circles, data obtained with PKC holoenzyme; open circle with solid line, kinase activity of the catalytic domain; open circle with dotted line, PDBu binding by the regulatory domain.

membrane (41, 42). As shown in Fig. 5, BPO inserted into liposomes inactivated both kinase activity and PDBu binding of PKC holoenzyme and PDBu binding of the proteolytically separated regulatory domain. However, it was not effective in inactivating the kinase activity of the proteolytically separated catalytic domain lacking the membrane-inserting region. Thus, the proteolytically separated catalytic domain, although reacted with BPO when added as a suspension, weakly reacted with BPO inserted into the membrane. Whereas the holoenzyme or the proteolytically separated regulatory domain having the membrane-inserting region can readily react with BPO inserted into the membrane. Specificity of BPO-induced modification of PKC. In the absence of thiol agents, protein kinase A was inactivated only at higher concentrations of BPO with an IC 50 of 5 mM, which was 25-fold higher than that required to inactivate PKC. Furthermore, this inactivation required the presence of cAMP; in the absence of cAMP, still higher concentrations of BPO (IC 50 5 125 mM) were required to inactivate PKA. Both phosphorylase kinase and PP2A activities were inactivated only at higher concentrations of BPO with an IC 50 of 18 and 7 mM, respectively. Unlike PKC, these enzymes not only required higher concentrations of BPO for inactivation, but they also were not readily regenerated by incubation with DTT following BPO-induced inactiva-

tion. Furthermore, under the conditions described in the legend to Fig. 5, BPO inserted into liposomes did not inactivate these enzymes. This suggests that the partitioning of BPO into the membrane may further add specificity to BPO reaction with certain proteins. BPO-induced inactivation of PKC in intact JB6 (41 P 1) cells: Relationship to glucose metabolism and cell population density. When confluent JB6 (41 P 1) cells were treated with BPO in a serum-containing medium, the extent of PKC inactivation was low even with a high concentration of BPO (Fig. 6). When serum was removed to prevent BPO from reacting with serum proteins having sulfhydryls such as albumin, the rate of PKC inactivation was increased. When the cells were serum starved for 24 h to decrease metabolic activity, PKC inactivation by BPO was further increased. To specifically suppress glucose metabolism, the analog 2-deoxyglucose was used as a metabolic antagonist of glucose. 2-Deoxyglucose blocks the cellular uptake of glucose and also prevents the metabolism of glucose 6-phosphate (43). Overall, these conditions may decrease glycolysis and the steady-state generation of cellular reducing equivalents and ATP. With 2-deoxyglucose pretreatment, BPO-induced inactivation of PKC was dramatically increased. Under these conditions, no increase in membrane association (translocation) of PKC was observed as measured not only by the kinase activity assay but also by immuno-

252

GOPALAKRISHNA ET AL.

FIG. 6. The differential rate of inactivation of PKC in JB6 (41 P 1) cells treated with BPO under various conditions. Four sets of confluent JB6 cells (approx 5 3 10 6) were treated with the indicated concentrations of BPO for 1 h in a serum-containing medium or serum-free medium (serum removed prior to treatment), or after serum starvation for 24 h or serum starvation followed by pretreatment with 30 mM 2-deoxy-D-glucose for 4 h. Total PKC was extracted from these treated cells and its activity was determined.

blotting using antibodies specific to a isoenzyme of PKC. Although an increase in immunoreactivity of PKC was observed by Western blotting, this was due to the enhanced immunostaining of PKC after oxidative modification. Previous elegant studies revealed that PKC isoenzymes after oxidative modification exhibit enhanced cross-reactivity with the antibodies raised against other isoenzymes of PKC (44). At a low cell population density (10% confluency), even in the presence of serum, treatment with BPO for 1 h inactivated PKC with an IC 50 of 10 mM. However, in serum starved conditions, 10 mM concentrations of BPO inactivated nearly 90% of the activity. Nevertheless, when the IC 50 values for BPO were normalized to cell number, these were 7.1 and 8.5 mM per million cells for subconfluent and confluent cell population, respectively. These values were not appreciably different. Therefore, the effect of BPO depends on the amount of BPO in the medium and the number of cells in the petri dish, i.e., the amount of BPO that can partition into one cell. However, with low population density cells, there is a need to use larger numbers of petri dishes per experiment to determine PKC activity. Therefore, experiments were carried out with confluent cells using the required high concentrations of BPO. Addition of either CuCl 2 (10 mM) or its chelator, bathocuproine disulfonic acid (100 mM), to the medium did not influence the rate of inactivation of PKC by BPO (data not shown), which suggests that copper has no influence on the reaction of BPO with PKC sulfhydryls.

Differential rate of inactivation of PKC in various cell types and its relationship to BPO-induced cytotoxicity. Various cell types were seeded and were growth arrested by serum starvation to attain comparable cell density (4 to 5 3 10 6 cells per petri dish). These cells were then treated with BPO for 1 h. Although PKC was inactivated by BPO in all these cell types, the IC 50 for BPO in JB6 (30 P 2), C3H10T1/2, JB6 (41 P 1), and B16 (F10) was 10, 20, 60, and 150 mM, respectively (data not shown). When the time course of PKC inactivation by BPO was followed at higher concentrations well above the IC 50 for BPO, PKC activity was completely inactivated. However, with BPO concentrations lower than the IC 50, especially at two-thirds of the IC 50, there was an initial loss of PKC activity, which returned to the control level at a later time period. In B16 melanoma cells, treatment with 100 mM BPO caused a rapid decrease in PKC activity within 15 min, but this activity returned to the untreated control level within 60 min (Fig. 7). However, in JB6 (41 P 1) cells, 100 mM BPO induced a decrease in PKC activity that did not return to the untreated control level. Yet, when the concentration of BPO was decreased to 40 mM, a reversible inactivation of PKC was observed. Similarly, 10T1/2 cells exhibited a reversible change in PKC activity when these cells were exposed to a low (12 mM) concentration of BPO. Unlike the other cell lines presented in Fig. 7, treatment of JB6 (30 P 2) cells with 7 mM BPO (two-thirds of IC 50), did decrease PKC activity, but the PKC activity did not return to the untreated control level. By further decreasing BPO con-

FIG. 7. Transient inactivation of PKC in various cell types treated with BPO. Confluent B16 (F10), JB6 (41 P 1), JB6 (30 P 2), and 10T 1/ 2 cells, which were maintained in serum-free medium for 24 h, were treated with BPO at 100, 40, 7, and 12 mM concentrations, respectively. These concentrations selected were approximately two-thirds of the IC 50 values of BPO to inactivate PKC in a 1-h time period in these cell types. Then at various intervals of time, total PKC activity present in the treated cells was determined.

REVERSIBLE SULFHYDRYL OXIDATION OF PKC BY BENZOYL PEROXIDE

FIG. 8. Differential reversibility of BPO-induced inactivation of PKC by NADPH in crude homogenates prepared from various cell types. Crude cell homogenates (2 ml) prepared from JB6 30 P 2 (circle), JB6 41 P 1 (triangle), and B16(F10) (squares) cells were adjusted to 0.5 mg protein/ml and then treated for 5 min with the indicated concentrations of BPO either in the presence (closed symbols) or in the absence (open symbols) of 1 mM NADPH. The detergent Nonidet P-40 (1% final) then was added to the homogenates, and PKC was isolated by DEAE-cellulose chromatography.

centrations below this (7 mM) level, the extent of apparent inactivation of PKC was also decreased. Therefore, it was difficult to evaluate the reversible nature of this modification in JB6 (30 P 2) cells. These results suggest that each cell type has its own threshold limit at which BPO-mediated inactivation of PKC can be reversed by an endogenous mechanism. The BPO threshold required for irreversible inactivation of PKC was in agreement with the cytotoxicity elicited by BPO in these cell types: JB6 (30 P 2) . 10T1/2 . JB6 (41 P 1) . B16 (F10). A relatively rapid rate of BPOinduced inactivation of PKC along with higher cytotoxicity was also observed with certain normal cell types such as bovine aortal endothelial cells and rat lung epithelial cells compared to a slower rate of PKC inactivation and lower cytotoxicity noted with certain cancer cell types such as C6 glioma and LL/2 lung carcinoma cells (data not shown). Reversal of BPO-induced PKC oxidation by NADPH in cell homogenates. To determine possible differences in protein disulfide reductase activity in crude homogenates which can reverse this PKC oxidative modification, homogenates prepared from three different cell types were treated with BPO in the presence and absence of NADPH. Unlike the results observed with intact cells, PKC activity was inactivated by BPO to a similar degree in the homogenates prepared with JB6 (41 P 1), 30 (P 2), and B16 (F10) cell types (Fig. 8). However, the ability of NADPH to reverse this modification varied among the different homogenates. The

253

reversal of PKC modification by NADPH was highest in homogenates of B16 melanoma cells, followed by JB6 (41 P 1) cells, and was minimal in JB6 (30 P 2) cells. In another approach, crude homogenates were initially treated with BPO; the excess BPO was removed by PD-10 gel filtration columns and then the cell extracts were treated with NADPH. This approach also resulted in the regeneration of PKC activity. However, when PKC was isolated from the cell extract by a DEAE-cellulose column, NADPH failed to reverse the BPO-induced modification of PKC, suggesting that the factor which reverses PKC modification was separated from PKC during the DEAE-cellulose chromatography. This factor was eluted with 0.25 M NaCl. Its activity was found to be heat stable at 55°C. Apparent increase in Ca 21/lipid-independent PKC activity with BPO-induced inactivation of PP2A. Previous studies showed that oxidants such as H 2O 2, mperiodate, and polyphenolic agents can modify Ca 21/ lipid-dependent PKC (peak A) to the Ca 21/lipid-independent form (peak B) (14 –18, 45). Oxidatively activated PKC (peak B) is eluted with higher concentrations of NaCl than required for the elution of the Ca 21/lipid-dependent form of PKC (peak A) (14 –18, 45). Since this fraction (peak B) eluted from DEAEcellulose has a high amount of PP2A activity, it is important to use a PP2A inhibitor to achieve a reliable determination of PKC activity in this fraction. Although microcystin-LR (10 nM) inhibits both PP2A and PP1 (46), histone H1, phosphorylated by PKC, is preferentially dephosphorylated by PP2A but not by PP1 (47). As shown in Fig. 9, by inclusion of microcystin-LR in the PKC assay, the observed activity of Ca 21/lipidindependent form of PKC (peak B) was higher in the untreated control cells, which decreased by BPO treatment. On the contrary, by omission of microcystin-LR in the PKC assay, the observed activity of peak B was apparently lower in the untreated control, which transiently elevated by BPO treatment. Furthermore, a concomitant decrease in PP2A activity was observed in the BPO-treated cells. Conceivably, an apparent increase in the activity of Ca 21/lipid-independent form (peak B) of PKC was due to the inactivation of PP2A present within this fraction, which interfered with the PKC assay carried out in the absence of PP2A inhibitors. Thus, BPO treatment did not convert PKC to the Ca 21/lipid-independent activated form (peak B). Compartmentalization of PKC oxidation. Since GSH directly reacts with BPO to prevent oxidative modification of PKC in the test tube, we determined whether a depletion of cellular GSH increases PKC inactivation by BPO. When JB6 (41 P 1) cells were pretreated with BSO (100 mM) for 24 h in a serumcontaining medium, there was a substantial (.90%) decrease in GSH from 22.1 nmol/mg protein in the control to 1.8 nmol/mg protein in BSO-treated cells.

254

GOPALAKRISHNA ET AL.

FIG. 9. Increase in Ca 21/lipid-independent PKC (peak B) activity in BPO-treated cells and its relationship to the inactivation of PP2A eluting in this peak B fraction. Serum-starved confluent JB6 (41 P 1) cells were treated with indicated concentrations of BPO for 60 min. The detergent-solubilized cell extract was applied to a DEAE-cellulose column (0.5 ml), the Ca 21/lipid-dependent PKC activity (peak A) was eluted with 0.1 M NaCl, and then Ca 21/lipid-independent PKC activity (peak B) was eluted with 0.25 M NaCl. The Ca 21/lipid-independent activity of PKC (solid line) in the presence and absence of the PP2A inhibitor (10 nM microcystin-LR), as well as PP2A activity (dotted line) present in this fraction, were determined.

Nevertheless, BSO-pretreated cells did not show any increased inactivation of PKC by the BPO treatment. Furthermore, there was no clear difference between JB6 (41 P 1) and JB6 (30 P 2) cell lines in the amount of GSH per milligram of protein. Although the endogenous cellular levels of GSH had no effect on PKC oxidation, the thiols N-acetylcysteine, GSH, and cell-impermeable mercaptoethanesulfonic acid (2 mM), when added to the medium just prior to the addition of BPO (100 mM), completely blocked the BPO-induced inactivation of PKC occurring within the cells. DISCUSSION

These studies strongly suggest that PKC can serve as a sensitive redox target for tumor-promoting agent benzoyl peroxide. Therefore, PKC may serve as a receptor for structurally diverse tumor promoters including phorbol ester type of agents and various oxidants. Although the effect of BPO on isolated PKC had previously been studied, mercapto agents were not removed from the PKC preparation in all these studies (20 –23). Therefore, the high degree of sensitivity of PKC to BPO, which was observed in the current study, was not observed in the previous studies (20 –23). The protective effect of the reagent thiols to minimize PKC inactivation is due to their direct reaction with BPO as well as to the reversal of PKC oxidative modification.

When PKC was treated with BPO as an aqueous suspension, both catalytic and regulatory domains were equally modified as measured by loss of kinase activity and PDBu binding, respectively. However, the membrane-inserted BPO selectively reacted with the regulatory domain that has four membrane-inserting zinc thiolate sites. This modification, occurring at lower concentrations of BPO, was readily reversible by DTT. Other enzymes such as PKA, phosphorylase kinase, and PP2A required nearly 25- to 100-fold higher concentrations of BPO for inactivation, and the modification occurring at higher concentrations of BPO was not readily reversible. Furthermore, membrane-inserted BPO did not readily react with these protein kinases. Thiol reversible inactivation of Na,K-ATPase was previously observed with an IC 50 of 2 mM (48), which is 10-fold higher than that required for PKC inactivation. Therefore, while several protein thiols and low-molecular-weight thiols are oxidized by increasing the concentration of BPO, the membrane-inserting zinc thiolates present within the regulatory domain of PKC (25, 26) may be providing certain specificity for the reaction of PKC sulfhydryls with a limited amount of BPO partitioned into the membrane. The mechanism of reaction of BPO with various organic sulfides and disulfides has been previously reported (12). This mechanism involves a nucleophilic attack of

REVERSIBLE SULFHYDRYL OXIDATION OF PKC BY BENZOYL PEROXIDE

the sulfur compound on the O-O bond of BPO, resulting in formation of an intermediate which subsequently breaks down into benzoic acid and benzoic anhydride (12). This BPO breakdown reaction does not require a transition metal and there is no production of free radicals. A similar type of mechanism may be involved in the reaction of BPO with biothiols such as protein sulfhydryls and GSH. The BPO-induced oxidation of GSH to GSSG was previously reported (10). However, the stoichiometry of GSH molecules oxidized per molecule of BPO is not known. Furthermore, in the previous study it was noted that the disulfide, GSSG, did not block the BPO effects (10). In the current study, GSSG also failed to block the BPO-induced modification. It is not known if there is any difference in the mechanism of reaction of BPO with the biothiols compared to the previously studied organic sulfides and disulfides (12). Oxidation of four or more sulfhydryls of PKC or low-molecular-weight compounds by one molecule of benzoyl peroxide may be due to the formation of reactive intermediates such as sulfenic acid, which in turn can react with additional sulfhydryls to form disulfides. Besides the metal ion-independent breakdown of BPO by sulfur compounds, BPO is known to be decomposed by another mechanism involving electron-transfer reaction mediated by transition metals (such as copper), resulting in the formation of free radicals (7–9, 12). Previous studies showed enhanced formation of free radicals as well as 8-hydroxy-29-deoxyguanosine in DNA in BPO-treated cells, which were inhibited by a pretreatment of cells with a copper chelator, bathocuproine disulfonic acid (9). In the current study, absence of any inhibitory effect of the copper chelator on the BPO-induced inactivation of PKC was due to the lack of a requirement for copper or other transition metals for this reaction. Previous studies have shown an increase in BPO-induced DNA base modification upon depletion of cellular GSH (9). Given the fact that GSH reacts with BPO, it can decrease the amount of BPO available for the reaction with chromatin-associated copper. However, in the present study, a lack of enhanced inactivation of PKC by GSH depletion may be due to the reaction of BPO with PKC sulfhydryls occurring compartmentally separated from water-soluble GSH. Although PKC initially was considered to be localized to the cytosol, it is now considered to be loosely associated with the membrane and cytoskeletal components within the cell and is dissociated during cell homogenization (49, 50). Water-soluble cytosolic GSH may not interrupt this “cage” type shielded redox modification occurring between the membrane-inserted BPO and PKC which is loosely associated with the membrane. However, GSH and other thiol agents, including cell-impermeable mercaptoethanesulfonic acid, when added to the culture medium just prior to the addition of BPO, prevented BPO-induced inactivation of PKC occurring within the cell. This suggested

255

that the breakdown of BPO with thiols present within the culture medium could minimize the amount of BPO that partitions into the cell membrane. The two different mechanisms of BPO breakdown (formation of radical versus ions) are stimulated by different agents (copper versus thiols) and may be compartmentally separated (nuclear versus membrane) to produce different cellular effects (genetic versus epigenetic signaling). Nevertheless, the two mechanisms may influence each other. For example, extensive breakdown of BPO by redox cycling protein thiols or GSH may deprive BPO of the copper-catalyzed DNA base oxidation. Alternatively, BPO oxidation may inactivate thiol-dependent DNA-repairing enzymes and thereby increase promutagenic DNA base oxidations. The observed changes in PKC activity in response to BPO treatment of cells represents a balance between the rate of PKC modification and the rate of its reversal by a cellular reductive mechanism. That treatment of confluent cells with low (,10 mM) concentrations of BPO does not result in a decrease in PKC activity, which may be due to a rapid reversal of this modification. However, it is difficult to distinguish a no net change in PKC modification from a true lack of modification. By decreasing glucose metabolism using 2-deoxyglucose, it is possible to demonstrate such modifications. The observed differences in PKC inactivation by BPO in JB6 promotable and nonpromotable cell lines were unlikely due to differences in the profile of PKC isoenzymes, as these cell types have been reported to have a similar PKC isoenzyme profile (51). Previous studies have shown that glutathione peroxide cannot catalyze the breakdown of BPO (10). Moreover, recent elegant studies have shown that redox-cycling cysteine residues present within some proteins can catalyze a “thiol peroxidase” activity coupled to the thioredoxin system (52–54). An initial inactivation of PKC followed by a regeneration of its activity with a lag period suggests an induction of adaptive mechanism to oxidative stress in the BPO-treated cells. Given the fact that the rate of generation of reducing equivalents is low in the resting cell, initially protein disulfide reductases may be functioning at a slow rate to reduce the protein disulfides formed by BPO-induced oxidation of sulfhydryls in PKC. The protein disulfides that remained in PKC or other sensitive membrane proteins may serve as sensors for the peroxide toxicity. At this stage, although PKC is inactivated with respect to its kinase activity, it may still transduce BPO-induced modification to other thiol proteins by a thiol/disulfide exchange reaction, a key regulatory mechanism as previously suggested for a variety of proteins (55, 56). Such mechanisms may stimulate glucose metabolism and an increase in steady-state generation of reducing equivalents. This results in enhanced activities of protein disulfide reductases, and the regenerated protein sulfhydryls react with additional BPO molecules. These sulfhydryl-disulfide redox cycles ulti-

256

GOPALAKRISHNA ET AL.

mately lead to increased breakdown of BPO and prevention of cell death. Certainly, further studies are needed to determine whether BPO-induced redox modification of PKC sulfhydryls contributes for a significant degradation of BPO in the cell or acts as a sensing mechanism to trigger signaling for the metabolic activation. Current studies show that cell types which can tolerate higher concentrations of BPO have a greater ability to reverse the BPO-induced modification of PKC. Although the nature of the factor was not identified in this study, it appears to resemble thioredoxin reductase, based on its heat stability, elution from a DEAE-cellulose column with 0.25 M NaCl, and a requirement for NADPH. Thioredoxin reductase was reported to be very low in normal cells but to be expressed at much higher levels in some tumor cells (57, 58). Overexpression of thioredoxin has also been reported to give a selective growth advantage to certain tumor cells as well as to overcome apoptosis (59, 60). Moreover, overexpression of thiol peroxidase also has been shown to suppress apoptosis (61). Further studies are certainly needed to determine whether the thioredoxin reductase system or thiol peroxidase is higher in the promotable cell type than in the nonpromotable cell type. Previous studies suggested that the initiation process may generate a transformed phenotype having resistance to the toxicity induced by BPO and that this would give a selective growth advantage over the normal cell population (29). The JB6 family of mouse epidermal clonal genetic variants that are transformation promotable (P 1) or nonpromotable (P 2) provides a suitable in vitro model for studying molecular events that are involved in tumor promotion induced by a variety of tumor promoters including oxidants (3, 62). Oxidants generated by a xanthine oxidase system were reported to induce a higher cytotoxicity in the nonpromotable JB6 (30 P 2) cell line relative to the promotable JB6 (41 P 1) cell line (30, 31). Furthermore, the nonpromotable JB6 (30 P 2) cell line was shown to have low superoxide dismutase and catalase activities compared to the promotable cell line (30, 31). In the current study, we observed that the oxidative inactivation of PKC along with cell death occurred to a greater extent in the JB6 (30 P 2) nonpromotable cell line compared to the promotable cell type (41 P 1). Although oxidative stress is important for tumor promotion, weaker cellular defenses may lead to cell death, while stronger antioxidant defenses may lead to cancer prevention. Therefore, tumor promotion/progression may occur under the conditions of sublethal oxidative stress, which can induce cell signaling and mutations. In this study, PP2A activity was determined to show this protein phosphatase as a contaminant which interferes with the assay of Ca 21/lipid-independent PKC activity in the peak B fraction. As shown in the current study as well as in the study previously reported (20), there was an increase in activity of Ca 21/lipid-indepen-

dent form of PKC in BPO-treated cells. However, as shown in the present study, the inactivation of the contaminant PP2A in peak B fraction was responsible for the apparent increase in Ca 21/lipid-independent activity of PKC. Previous studies showed that certain substrate proteins (such as histone H1 phosphorylated by PKC) are exclusively dephosphorylated by PP2A, suggesting that there is an intimate relationship between the activity of PP2A and PKC (47). Furthermore, inhibitors of PP2A such as okadaic acid, which inhibits PP2A with high affinity, can elicit certain cellular actions of activators of PKC (63). An oxidative inactivation of protein tyrosine phosphatases has been shown to increase protein tyrosine phosphorylation (64, 65). Therefore, it is possible that the BPO-induced inactivation of PP2A may initially enhance the phosphorylation state of certain PKC target proteins, despite the decrease in Ca 21/lipid-dependent PKC activity by BPO. Unlike phorbol esters and certain oxidants such as H 2O 2 which induce initial activation and subsequent inactivation of PKC, BPO induced only the inactivation of this enzyme, but not its direct activation. The anthralin type of tumor promoters, such as chrysarobin, has also been shown to induce only the inactivation of PKC (66). The role of PKC in tumor promotion is not clear. There are suggestions that its so-called “downregulation” may be important in tumor promotion (13, 67). DNA repair was found to be slow in both phorbol ester- and BPO-treated cells (68, 69). Moreover, PKC has been suggested to play an important role in influencing the activity of some DNA-repair enzymes (70). It is possible that the downregulation of PKC may prevent the DNA repair process. Thus, an increase in DNA base oxidation and a decrease in DNA repair will result in an increase in promutagenic DNA base oxidations in BPO-treated cells. Unlike the normal cells, initiated cells or tumor cells can overcome BPO-induced cell death, leading to clonal expansion of transformed cells accumulating BPO-induced mutations. To better understand the role of PKC in tumor promotion, it is important to clarify the functional role of the highly oxidation susceptible zinc thiolate structures present within the regulatory domain of this enzyme. There have been studies that indicate that PKC may have other protein phosphorylation-independent functions mediated by its regulatory domain, in addition to the known protein kinase function mediated by its catalytic domain (71–75). However, the details of such mechanisms are not known. Given the fact that oxidants are tumor promoters and that antioxidants are chemopreventive agents, the redox modulation of PKC may have mediator as well as preventive roles in tumor promotion. ACKNOWLEDGMENTS We thank Zhen-Hai Chen, Vivian Bernardo, and Michelle Tse for their technical assistance.

REVERSIBLE SULFHYDRYL OXIDATION OF PKC BY BENZOYL PEROXIDE

REFERENCES 1. Slaga, T. J., Klein-Szanto, A. J. P., Triplett, L. L. Yotti, L. P., and Trosko, J. E. (1981) Science 213, 1023–1025. 2. O’Connell, J. F., Klein-Szanto, A. J. P., DiGiovanni, D. M., Fries, J. W., and Slaga, T. J. (1986) Cancer Res. 46, 2863–2865. 3. Gindhart, T. D., Srinivas, L., and Colburn, N. H. (1985) Carcinogenesis 6, 309 –311. 4. Kennedy, S. H., Manevich, Y., and Biaglow, J. (1995) Biochem. Biophys. Res. Commun. 212, 118 –125. 5. Klein-Szanto, A. J. P., and Slaga, T. J. (1982) J. Invest. Dermatol. 79, 30 –34. 6. Gimenez-Conti, I., Viaje, A., Chesner, J., Conti, C., and Slaga, T. J. (1991) Carcinogenesis 12, 563–569. 7. Akman, S. A., Kensler, T. W., Doroshow, J. H., and Dizdaroglu, M. (1993) Carcinogenesis. 14, 1971–1974. 8. Hazlewood, C., and Davies, M. J. (1996) Arch. Biochem. Biophys. 332, 79 –91. 9. King, J. K., Egner, P. A., and Kensler, T. W. (1996) Carcinogenesis 17, 317–320. 10. Kennedy, C. H., Winston, G. W., Church, D. F., and Pryor, W. A. (1989) Arch. Biochem. Biophys. 271, 456 – 470. 11. Kennedy, C. H., Dyer, J. M., Church, D. F., Winston, G. W., and Pryor W. A. (1989) Biochem. Biophys. Res. Commun. 160, 1067– 1072. 12. Pryor, W. A., and Bickley, H. T. (1972) J. Org. Chem. 37, 2885– 2893. 13. Nishizuka, Y. (1984) Nature 308, 693– 698. 14. Gopalakrishna, R., and Anderson W. B. (1989) Proc. Natl. Acad. Sci. USA 86, 6758 – 6762. 15. Kass, G. E. N., Duddy, S. K., and Orrenius, S. (1989) Biochem. J. 260, 499 –507. 16. Larsson, R., and Cerutti, P. (1989) Cancer Res. 49, 5627–5633. 17. Gopalakrishna, R., Chen, Z. H., and Gundimeda, U. (1995) Methods Enzymol. 252, 134 –148. 18. Gopalakrishna, R., Chen, Z. H., and Gundimeda, U. (1994) Proc. Natl. Acad. Sci. USA 91, 12233–12237. 19. Leach, K. L., and Blumberg, P. M. (1985) Cancer Res. 45, 1958 – 1963. 20. Donnelly, T. E., Jr., Pelling, J. C., Anderson, C. L., and Dalby, D. (1987) Carcinogenesis 8, 1871–1874. 21. Kumar, R., and Holian, O. (1991) Invest. Dermatol. 96, 490 – 494. 22. Hagemann, I., Toso, S. M., Kitay, K., and Webster, G. F. (1994) Br. J. Dermatol. 130, 569 –575. 23. Matsui, M. S., Mintz, E., and Deleo, V. A. (1995) Skin Pharmacol. 8, 130 –138. 24. Quest, A. F. G., Bloomenthal, J., Bardes, E. S. G., and Bell, R. M. (1992) J. Biol. Chem. 267, 10193–10197. 25. Zhang, G., Kazanietz, M. G., Blumberg, P. M., and Hurley, J. H. (1995) Cell 81, 917–924. 26. Kazanietz, M. G., Wang, S., Milne, G. W., Lewin, N. E., Liu, H. L., and Blumberg, P. M. (1995) J. Biol. Chem. 270, 21852– 21859. 27. Maret, W., and Vallee, B. L. (1998) Proc. Natl. Acad. Sci. USA 95, 3478 –3482. 28. Crow, J. P., Beckman, J. S., and McCord, J. M. (1995) Biochemistry 34, 3544 –3522. 29. Hartley, J. A., Gibson, N. W., Kilkenny, A., and Yuspa, S. H. (1987) Carcinogenesis 8, 1827–1830. 30. Cerutti, P., Krupitza, G., Larsson, R., Muehlematter, D., Crawford, D., and Amstad, P. (1988) Ann. N. Y. Acad. Sci. 551, 75– 82.

257

31. Amstad, P., Peskin, A., Shah, G., Mirault, M. E., Moret, R., Zbinden, I., and Cerutti, P. (1991) Biochemistry 30, 9305–9313. 32. Huang, K.-P., and Huang, F. L. (1991) Methods Enzymol. 200, 241–251. 33. Gopalakrishna, R., Chen, Z. H., Gundimeda, U., Wilson, J. C., and Anderson, W. B. (1992) Anal. Biochem. 206, 24 –35. 34. Ellman, G. L. (1959) Arch. Biochem. Biophys. 82, 70 –77. 35. Gopalakrishna, R., Gundimeda, U., and Chen, Z. (1997) Arch. Biochem. Biophys. 348, 25–36. 36. Camichael, J., DeGraff, W. G., Gazdar, A. F., Minna, J. D., and Mitchell, J. B. (1987) Cancer Res. 47, 936 –942. 37. Skehan, P., Stereng, R., Scudiero, D., Monks, A., McMahaon, J., Vistica, D., Warren, J. T., Bokesch, H., Kerney, S., and Boyd, M. R. (1990) J. Natl. Cancer. Inst. 82, 1107–1112. 38. Gopalakrishna, R., Gundimeda, U., Wilson, J. C., and Chen, Z. (1993) Anal. Biochem. 212, 296 –299. 39. Baker, M. A., Cerniglia, G. J., and Zaman, A. (1990) Anal. Biochem. 190, 360 –365. 40. Tsujimoto, Y., Nagashima, K., Yamazaki, M., and Furuyama, S. (1988) Int. J. Biochem. 20, 591–594. 41. Boni, L. T., and Rando, R. R. (1985) J. Biol. Chem. 260, 10819 – 10825. 42. Bazzi, M. D., and Nelsestuen (1991) Biochemistry 30, 7961– 7969. 43. Wick, A. N., Drury, D. R., Nakada, H. I., and Wolfe, J. B. (1957) J. Biol. Chem. 224, 963–969. 44. Allen, B. G., Andrea, J. E., and Walsh, M. P. (1994) J. Biol. Chem. 269, 29288 –29298. 45. Konishi, H., Tanaka, M., Takemura, Y., Matsuzaki, H., Ono, Y., Kikkawa, U., Nishizuka, Y. (1997) Proc. Natl. Acad. Sci. USA 94, 11233–11237. 46. Honkanen, R. E., Zwiller, J., Moore, R. E., Daily, S. L., Khatra, B. S., Dukelow, M., and Boynton, A. L. (1990) J. Biol. Chem. 265, 19401–19404. 47. Jakes, S., and Schlender, K. K. (1988) Biochim. Biophys. Acta 967, 11–16. 48. Kono, Y., Abiko, Y., Hayakawa, M., and Yamazaki, M. (1981) Comp. Biochem. Physiol. 70C, 35–39. 49. Liao, L., Ramsay, K., and Jaken, S. (1994) Cell Growth. Differ. 5, 1185–1194. 50. Gopalakrishna, R., Chen, Z., and Gundimeda, U. (1997) Arch. Biochem. Biophys. 348, 37– 48. 51. Smith, B. M., and Colburn, N. H. (1988) J. Biol. Chem. 263, 6424 – 6431. 52. Yim, M. B., Chae, H. Z., Rhee, S. G., Chock, P. B., and Stadtman, E. R. (1994) J. Biol. Chem. 269, 1621–1626. 53. Netto, L. E. S., Chae, H. Z., Kang, S-W., Rhee, S. G., and Stadtman, E. R. (1996) J. Biol. Chem. 271, 15315–15321. 54. Chae, H. Z., Uhm, T. B., and Rhee, S. G. (1994) Proc. Natl. Acad. Sci. USA 91, 7022–7026. 55. Gilbert, H. F. (1990) Adv. Enzymol. Rel. Areas Mol. Biol. 63, 69 –172. 56. Ziegler, D. M. (1985) Annu. Rev. Biochem. 54, 305–329. 57. Gladyshev, V. N., Jeang, K. T., and Stadtman, T. C. (1996) Proc. Natl. Acad. Sci. USA 93, 6146 – 6151. 58. Tamura, T., and Stadtman, T. C. (1995) Proc. Natl. Acad. Sci. USA 93, 1006 –1011. 59. Powis, G., Gasdaska, J. R., and Baker, A. (1997) Adv. Pharmacol. 38, 329 –359. 60. Baker, A., Payne, C. M., Briehl, M. M., and Powis, G. (1997) Cancer Res. 57, 5162–5167.

258

GOPALAKRISHNA ET AL.

61. Zhang, P., Liu, B., Kang, S. W., Seo, M. S., Rhee, S. G., and Obeid, L. M. (1997) J. Biol. Chem. 272, 30615–30618. 62. Colburn, N. H., Former, B. F., Nelson, K. A., and Yuspa, S. H. (1979) Nature 281, 589 –591. 63. Gopalakrishna, R., Chen, Z. H., and Gundimeda, U. (1992) Biochem. Biophys. Res. Commun. 189, 950 –957. 64. Hefletz, D., Bushkin, I., Dror, R., and Zick, Y. (1990) J. Biol. Chem. 265, 2896 –2902. 65. Grinstein, S., Furuya, W., Lu, D. J., and Mills, G. B. (1990) J. Biol. Chem. 265, 318 –327. 66. Reiners, J. J. Jr., Cantu, A. R., and Schoeller, A. (1992) Biochem. Biophys. Res. Commun. 186, 970 –976. 67. Kischel, T., Harbers, M., Stabel, S., Borowski, P., Muller, K., and Hilz, H. (1989) Biochem. Biophys. Res. Commun. 165, 981–987. 68. Trosko, J. E., Chang, C. C., Yotti, L. P., and Chu, E. H. Y. (1977) Cancer Res. 37, 188 –192.

69. Birnboim, H. C. (1986) Carcinogenesis 7, 1511–1517. 70. Zhang, W., Hara, A., and Sakai, N. (1993) Neurosurgery 32, 432– 437. 71. Weinstein, I. B. (1988) Adv. Exptl. Med. Biol. 234, 127–140. 72. Conricode, K. M., Brewer, K. A., and Exton, J. H. (1992) J. Biol. Chem. 267, 7199 –7202. 73. Lehel, C., Olah, Z., Jakab, G., and Anderson, W. B. (1995) Proc. Natl. Acad. Sci. USA 92, 1406 –1410. 74. Singer, W. D., Brown, H. A., Jiang, X., and Sternweis, P. C. (1996) J. Biol. Chem. 271, 4504 – 4510. 75. Hammond, S. M., Jenco, J. M., Nakashima, S., Cadawallader, K., Gu, Q., Cook, S., Nozawa, Y., Prestwich, G. D., Frohman, M. A., and Morris, A. J. (1997) J. Biol. Chem. 272, 3860 – 3868.