A Role for Protein Kinase C-δ in the Regulation of Ornithine Decarboxylase Expression by Oxidative Stress¹

Polyamines are small aliphatic molecules that play a key role in regulating cell proliferation (1). Polyamine biosynthesis is one of the first events to occur in proliferating cells, preceding both nucleic acid and protein synthesis (2). Although most of the specific functions of polyamines have not been identified, they are thought to fulfill structural and regulatory roles in nucleic acid and protein biosynthesis and function (3). ODC³(EC 4.1.1.17) is the initial and rate-limiting enzyme in polyamine biosynthesis and is essential for cell proliferation. However,overexpression of ODC beyond some minimum threshold can induce cell transformation and tumor formation (4); indeed, abnormal levels of the enzyme have been detected in some animal and human cancers (2). ODC is also an important marker for tumor promotion in skin. Studies show that overexpression of ODC in mouse skin induces tumor promotion in the absence of typical tumor promoters,such as TPA (5). ODC activity is transiently induced by various mitogenic and toxic stimuli (6). During ODC induction, the increase in its mRNA is usually much less than the observed activity (7), suggesting that some regulation of ODC activity occurs posttranscriptionally. Isoelectric focusing analyses in various tissues and cells provide evidence for multiple forms of ODC (8, 9, 10), a further indication that the protein may undergo posttranslational modifications. Indeed, it has been demonstrated in several studies that ODC is phosphorylated in situ at serine and threonine residues (11). In RAW264 cells, phosphorylated ODC is more stable and exhibits a 50%higher catalytic efficiency than the unphosphorylated protein(12). Casein kinase II has been identified as one of the kinases that phosphorylates ODC in these cells; however,phosphorylation by this enzyme does not result in increased ODC activity (12).

Induction of ODC in murine keratinocytes treated with the tumor promoter TPA has been associated with PKC activation, and there is data that support transcriptional regulation of ODC by PKC-α(13). The current studies investigate the role of PKC-δin the regulation of ODC by the oxidative agent H₂O₂. Oxidants derived from reactive oxygen species and organic hydroperoxides can contribute to tumor promotion, a process of selection and clonal expansion in which the expression of genes, such as ODC, that regulate cell growth are modulated in initiated cells (14). The pathways controlling the expression of such genes by oxidants in mammalian systems are not well defined. Oxidant tumor promoters also activate PKCs (15). PKC is a multienzyme family of serine/threonine kinases (16) that is classified into three groups:(a) classical PKCs (α, βI, βII, and γ), which are Ca²⁺, phosphatidylserine, and diacylglycerol/TPA dependent; (b) novel PKCs (δ, ε, η, θ, and μ),which are phosphatidylserine and diacylglycerol/TPA dependent; and(c) atypical PKCs (ζ, τ, and λ), which are Ca²⁺ and diacylglycerol/TPA independent. Regulation of ODC and tumor promotion is commonly studied in skin or cultured keratinocytes because skin is a good model for studying two-stage carcinogenesis. The present studies, performed in mouse papilloma cells, focused on the role of PKC-δ in the induction of ODC by H₂O₂, because although various PKCs are expressed in mouse epidermis, mainly the novel types,PKC-δ and PKC-η, are associated with keratinocyte differentiation(17, 18), a course that if deregulated could progress into a neoplastic phenotype in skin. Our choice to study PKC-δ was further influenced by the observations that expression of PKC-η protein in cultured keratinocytes is dependent on high Ca²⁺concentrations, whereas PKC-δ expression is independent of Ca²⁺ levels (17). The mouse PE cells used in these studies were sustained in low Ca²⁺levels to maintain them in a proliferative state. We find that H₂O₂ induces ODC activity severalfold and that this induction appears to be mediated by PKC-δ.

Cell lines were maintained in a 37°C humidified environment containing 5% CO₂ in air. Murine papilloma PE cells (19) were cultured in Eagle’s MEM without CaCl₂ (BioWhittaker, Walkersville, MD)supplemented with Chelex (Bio-Rad, Hercules, CA)-treated fetal bovine serum (8%) and 0.05 mm CaCl₂. Serum starvation was achieved by incubation in Eagle’s MEM supplemented with 0.5% Chelex-treated fetal bovine serum (8%) and 0.05 mmCaCl₂ for at least 16 h prior to treatment with H₂O₂ or TPA in serum-free media.

Cells expressing dominant-negative PKC-δ(PE_PKC-δDN) were generated by transfection of the plasmid SRD-DK376A (20)with lipofectamine (Life Technologies, Inc., Gaithersburg, MD)according to the supplier’s directions. Control cells were transfected with a blank plasmid (PE_control). G418-resistant clones were isolated by clonal dilution, expanded, and characterized.

After treatment of confluent PE cultures, the cells were rinsed with pyridoxal phosphate (50 μg/ml)-containing PBS and harvested into Eppendorf tubes. The cells were subjected to three cycles of freezing on dry ice and thawing at 37°C. The cleared lysates were used to determine ODC activity as described previously (21).

Treated cells were harvested and lysed in 20 mm Tris-Cl (pH 7.5), 1 mm MgCl₂, 1 mmEDTA, 1 mm EGTA, 4 mmdiisopropylfluorophosphate, 1 mm sodium vanadate, 25μg/ml leupeptin, 25 μg/ml pepstatin, 1% Triton X-100, and 50 mm mercaptoethanol for 30 min on ice. The cleared lysates were used to determine PKC activity by adding 2–4 μg of protein in an assay mixture containing 50 mm Tris-Cl (pH 7.5), 1 mm CaCl₂, 15 mmMgCl₂, 10 μm TPA, 0.25 mg/ml of phosphatidylserine, 50 μm ATP, 1 μCi of[γ-³²P]ATP, 2.5 mm DTT, and 50μ m of a PKC-δ pseudosubstrate region-derived peptide(22). The reactions were incubated at room temperature for 15 min and spotted onto phosphocellulose discs (Life Technologies,Inc.). The discs were washed twice with 1% phosphoric acid and twice with distilled water and analyzed by liquid scintillation. Nonspecific PKC activity was determined as described but in the absence of TPA or phosphatidylserine. The specific PKC activity was obtained by subtracting the nonspecific activity from the total activity.

Cells were lysed in PBS by subjecting the extracts to three cycles of freezing in dry ice and thawing at 37°C. Cleared cell lysates (10μg protein) were analyzed by Western blot analysis as described previously (23) with polyclonal antibodies specific to PKC-δ (Santa Cruz Biotechnologies, Inc., Santa Cruz, CA). ODC protein expression was determined with a polyclonal ODC antibody.

Total RNA was isolated with RNA Stat-60 (Tel-Test, Inc., Friendswood,TX). RNA (20 μg/lane) was fractionated in formaldehyde-agarose gels and transferred onto Nytran membranes (Schleischer-Shuell, Keene, NH)according to the manufacturer’s directions. A mouse ODC cDNA probe was labeled with [α-³²P]dCTP as illustrated in a random primer labeling kit (Boehringer Manneheim, Indianapolis, IN),and hybridization and washes were done as described previously(24). Hybridization to a 24-bp oligonucleotide complementary to 18S RNA (5′-ACGGTATCTGATCGTCTCGAACC-3′) that was end-labeled with [γ³²P]ATP by T4 kinase (Life Technologies, Inc.) was used to control for differences in loading and transfer efficiency.

PE cells were transiently transfected with 3 μg/ml of −4362/+131 ODC-luciferase (25), 0.2 μg/ml of pRLTK (Promega Corp.,Madison, WI), and 12 μl/ml of lipofectamine (Life Technologies, Inc.)according to the manufacturer’s directions. Cells were treated with 200 μm H₂O₂,and luciferase activity was determined 6 h after treatment with the dual-luciferase reporter assay system (Promega). Luciferase activities were determined with a Berthold LB9505 luminometer.

Constitutively active PKC-δ (20) was subcloned into the modified adenoviral shuttle vector pAdEG1 (26), and virus was purified essentially as described (26, 27, 28). PE_PKC-δDN cells were infected with virus (100 plaque-forming units/cell) for 2–4 h in serum-free media. Infection media was then replaced with normal growth medium, and cells were allowed to grow for 24 h, after which they were treated with H₂O₂ and analyzed for ODC activity. The fluorescence of green fluorescent protein was monitored with a confocal fluorescence microscope (Carl Zeiss, Jena, Germany) at 488 nm argon excitation, fitted with a 515-nm filter.

Statistical significance of the differences in the means was assessed by one-way ANOVA, followed by Bonferroni’s multiple comparison test.

ODC activity can be induced in response to a variety of cellular stresses. Induction of ODC activity is a hallmark for the action of tumor promotion, in vivo and in vitro. PE cells,a murine keratinocyte cell line derived from mouse papillomas produced in a classical two-stage protocol, were selected for the current studies because of their sensitivity for induction of ODC by TPA(19) and organic hydroperoxide tumor promoters(29). As shown in Fig. 1 A, treatment of PE cells with graded concentrations of H₂O₂ for 5 h resulted in a dose-dependent induction of ODC activity with maximal activity observed at 200 μmH₂O₂. Cells treated with 200 μm of H₂O₂ for 5 h typically exhibited a 10–12-fold increase in ODC activity compared with a 25-fold increase with the tumor promoter TPA (100 ng/ml), a well-established inducer of ODC activity (data not shown).

The following set of studies attempted to identify the molecular pathway by which oxidants activate ODC. The fact that H₂O₂ activates PKC(30), coupled with our observation that H₂O₂ induces ODC activity,led us to question whether PKC plays a role in this process. Pretreatment of cells with 1 μm calphostin, a nonspecific PKC inhibitor, prior to treatment with 200 μmH₂O₂ resulted in complete inhibition of ODC activity compared with cells treated with H₂O₂ only (Fig. 1 B). Calphostin alone had no effect on constitutive levels of ODC activity.

To further probe the role of PKC types in the regulation of ODC by oxidants, cells were stably transfected with a dominant-negative PKC-δ construct and characterized by immunoblot analysis and PKC activity measurements. PKC-δ appeared as a doublet of phosphorylated and unphosphorylated proteins as determined by immunoblot analysis of control and transfected cells (Fig. 2,A). As demonstrated in Fig. 2,A, Lanes 2and 3, there was a 3-fold increase in mutant protein in PE_PKC-δDN cells compared with native protein in PE_control cells. The dominant-negative PKC-δ is mutated in its ATP-binding site and therefore lacks the ability to autophosphorylate and activate itself. However, a slower migrating phosphorylated protein was still obtained in PE_PKC-δDN cells. We speculate that this outcome is likely attributable to the fact that wild-type PKC-δ and its kinase-inactive mutant are both phosphorylated on tyrosine residues by tyrosine kinases(30). PKC activity, as determined with a peptide designed against the PKC-δ pseudosubstrate region, was increased 3-fold in cell extracts isolated from H₂O₂-treated PE_control cells. However, this activity was completely inhibited in extracts from H₂O₂-treated PE_PKC-δDN cells (Fig. 2 B), consistent with a functional loss of PKC-δ. PKC activity in extracts isolated from TPA-treated PE_control cells was increased 22-fold, but this activity was only decreased by 50% in PE_PKC-δDN cells after treatment with TPA. Presumably, PKC activity was only partially inhibited in extracts from TPA-treated PE_PKC-δDN cells because,although the PKC-δ pseudosubstrate peptide has high specificity for PKC-δ, it is not absolute and can be phosphorylated by other PKC types. Also, in the presence of TPA, the concentration of the dominant-negative PKC-δ in PE_PKC-δDN cells may not be high enough to overcome the particularly robust activation of multiple types of PKCs.

The time course for the induction of ODC activity by H₂O₂ was investigated in the PE_control and PE_PKC-δDN cells after treatment with 200 μmH₂O₂. A transient induction in ODC activity was observed in PE_control cells,with maximal ODC activity observed between 5 and 7 h of treatment(Fig. 3,A), followed by a rapid decline in enzymatic activity. This transient induction of ODC by various stimuli is attributed to a rapid degradation of ODC through an antizyme-dependent mechanism(31). ODC activity was not increased at any time point in PE_PKC-δDN cells upon treatment with H₂O₂,suggesting that PKC-δ may be required for induction of ODC expression by H₂O₂ (Fig. 3,A). By contrast, a partial inhibition of ODC induction in response to TPA was seen in PE_PKC-δDN cells compared with PE_control cells at a low dose (20 ng/ml;Fig. 3 B). No inhibition was seen at a higher dose (100 ng/ml) of the phorbol ester (data not shown). This result suggests that for TPA, unlike H₂O₂,inhibition of PKC-δ in PE_PKC-δDN cells has limited impact upon ODC induction and that elevation of other PKC enzymes is important in signaling for enzyme induction.

To further confirm that PKC-δ is a mediator in the induction of ODC by H₂O₂,PE_PKC-δDN cells were infected with virus expressing a constitutively active PKC-δ attached to green fluorescent protein. This approach was taken to achieve a high level of constitutively active PKC- δ expression that could overcome the dominant-negative effects observed in PE_PKC-δDN cells. About 80–90% of the cells expressed constitutively active PKC-δ, as determined by the presence of green fluorescent protein (Fig. 4,B). The distribution of PKC-δ was concentrated in the perinuclear region, although some protein was also observed in the cytoplasm. In cells infected with virus containing green fluorescent protein only (positive control), the protein was expressed diffusely throughout the cells (Fig. 4,C). Overexpression of constitutively active PKC-δ in PE_PKC-δDN cells increased basal ODC activity 4-fold in untreated cells. Moreover, this maneuver restored ODC inducibility as activity was further enhanced 8-fold after H₂O₂ treatment (Fig. 4 D).

Having established that PKC-δ is important in the regulation of ODC enzyme activity by H₂O₂,the next step was to determine whether the effect on induction was transcriptional and/or posttranscriptional. The kinetics of appearance of ODC message and protein levels in cells treated with H₂O₂ was analyzed by Northern and Western analyses. Significant increases in ODC mRNA levels in PE_control cells were initially observed after 4 h of treatment, and peak message levels (3-fold increase) were maintained up to 7 h of treatment (Fig. 5). No induction of ODC mRNA occurred in H₂O₂-treated PE_PKC-δDN cells (Fig. 5).

Treatment of PE_control cells with TPA (positive control) resulted in a time-dependent increase in ODC protein with maximal levels observed at 6 h (14-fold), as determined by immunoblot analysis (Fig. 6,A). A similar time-dependent, albeit smaller increase in maximal ODC protein (7-fold at 6 h) was observed in PE_control cells after H₂O₂ treatment (Fig. 6,B). However, no increase in ODC protein was observed in PE_PKC-δDN cells (Fig. 6 C), except for trace elevation of ODC protein after 7 h of H₂O₂ treatment, which did not correspond to any increase in ODC activity at the same time point.

To confirm the observations from Northern analysis that PKC-δ is involved in the transcriptional regulation of ODC, activation of an ODC promoter-luciferase reporter construct by H₂O₂ was investigated in PE_control and PE_PKC-δDN cells. Normalized luciferase activity was increased 2.7-fold in PE_control cells after H₂O₂ treatment; however,similar treatment of PE_PKC-δDN cells did not elicit any increase in luciferase activity (Fig. 7).

ODC regulation has been the subject of numerous investigations in different models of multistage carcinogenesis but most commonly in skin. For instance, transgenic animals that express both elevated ODC and activated ras develop spontaneous skin tumors in the absence of carcinogens or tumor promoters, suggesting a cooperation between these genes during tumor development (32). The mechanism for such a cooperation, however, is not clear. Nonetheless,this and other observations point to the importance of deregulation of ODC in the neoplastic process.

A direct role for ODC in tumor promotion is evident in studies where targeted expression of high levels of ODC to mouse skin produced tumors after initiation with a carcinogen, in the absence of administration of a tumor promoter (5). The present studies examined the regulation of ODC by H₂O₂,primarily because oxidants are known tumor promoters in vivo, and secondly, the precise mechanisms by which oxidants cause promotion are not clear. The current studies demonstrate that ODC activity is induced in murine keratinocytes by H₂O₂, an action common to most, if not all, tumor promoters in skin. There appear to be several mechanisms by which oxidants can trigger signal transduction pathways leading to enhanced gene expression. For instance, butylated hydroperoxide and H₂O₂activate the mitogen-activated protein kinase cascade (33, 34), a critical signaling pathway for cellular proliferation. This well-characterized pathway is stimulated by growth factors,mitogens, and stress responses and is initiated by tyrosine kinases that sequentially activate ras, c-raf, and other downstream kinases. Deregulation of this pathway, as might occur in ras-transformed cells, is associated with ODC overexpression(35).

Another important signaling pathway in the cell, the multienzyme PKC family, which mediates many of the actions of the potent tumor promoter TPA, can also be activated by H₂O₂. However,H₂O₂ activates PKC-δ by a mechanism that differs from TPA activation of this kinase. Unlike TPA,stimulation of cells with H₂O₂ does not require PKC-δ translocation to the membrane for activation, and the activity of PKC-δ isolated from H₂O₂-treated cells can be measured independently of lipid cofactors (36). Investigations into the mechanisms of activation of PKC indicate that oxidants may modify cysteine residues on several PKCs, leading to their activation (37). Other studies show tyrosine phosphorylation of major PKC isoforms, including PKC-δ, in the presence of H₂O₂, and suggest that this may be a mechanism of activation by this oxidant(30).

Cellular expression of the dominant-negative PKC-δ completely prevented the inducibility of ODC by H₂O₂, suggesting that this PKC isoform specifically regulates ODC in response to oxidants and perhaps other actions of oxidants in mouse skin. However, there are contrary views as to the role of PKC-δ in tumor promotion and cell transformation. For instance, whereas a dominant-negative PKC-δmutant inhibited sis-induced transformation and platelet-derived growth factor-BB-mediated anchorage-independent colony formation in NIH3T3 cells (38), expression of this mutant in 3Y1 fibroblasts induced c-src-mediated transformation of these cells and TPA-induced anchorage-independent colony formation(39). These data reflect how little is known on the physiological role of PKC-δ. The complexity of PKC-δ regulation and function is further demonstrated by recent studies that find PKC-δ to be a substrate for c-src, which phosphorylates this kinase at tyrosine residues and promotes its degradation (40).

Overexpression of PKC-δ in vivo and in vitrohas also been shown to inhibit tumor promotion and induce apoptosis,respectively (41, 42), supporting a role for PKC-δ as a tumor suppressor. However, the physiological relevance of overexpressing this protein remains to be determined because overexpression may overwhelm endogenous protein levels, leading to nonspecific PKC-δ localization and substrate specificity. This consideration is especially important in light of very recent data showing differential localization of PKC-δ to cellular, nuclear, or Golgi membranes in response to different agonists (43). Such a differential localization may influence substrate specificity of PKC-δ, and hence, its biological response, especially if there is colocalization with a substrate. Unfortunately, knowledge of the physiological substrates of PKC-δ that may aid in understanding its diverse physiological response is scant.

The fact that we demonstrate an increase in ODC mRNA (3-fold) in response to H₂O₂ that does not correspond to an increase in ODC protein (7-fold) or enzyme activity (12-fold) suggests the additional involvement of posttranscriptional regulation of ODC by oxidants. It is well documented that ODC undergoes rapid degradation upon stimulation; it has also been shown that ODC can be phosphorylated in situand in vitro and that phosphorylated ODC is more stable and has a higher catalytic activity than the unphosporylated protein(12). Although we report that PKC-δ is required for the induction of ODC activity by oxidants and although our data support a role for transcriptional regulation, we cannot rule out the possibility of posttranscriptional effects, such as phosphorylation of ODC by PKC-δ, to explain the full increase in ODC activity.

We thank Dr. Syu-ichi Hirai for providing the PKC-δconstructs, Dr. David Johns for the viral vectors, Dr. Ajit Verma for the ODC promoter-luciferase vector, and Dr. Thomas O’Brien for the ODC antibody.