Phorbol Ester-induced G1 Phase Arrest Selectively Mediated by Protein Kinase Cδ-dependent Induction of p21*

Although protein kinase C (PKC) has been widely implicated in the positive and negative control of proliferation, the underlying cell cycle mechanisms regulated by individual PKC isozymes are only partially understood. In this report, we show that PKCδ mediates phorbol ester-induced G1 arrest in lung adenocarcinoma cells and establish an essential role for this novel PKC in controlling the expression of the cell cycle inhibitor p21. Activation of PKC with phorbol 12-myristate 13-acetate (PMA) in early G1 phase impaired progression of lung adenocarcinoma cells into S phase, an effect that was completely abolished by specific depletion of PKCδ, but not PKCα. Although the PKC effect was unrelated to the inhibition of cyclin D1 expression, PKC activation significantly up-regulated p21 and down-regulated Rb hyperphosphorylation and cyclin A expression. Elevations in p21 mRNA and protein by PMA were mediated by PKCδ but not PKCα. Studies using luciferase reporters also revealed an essential role for PKCδ in the PMA-induced inhibition of Rb-dependent cyclin A promoter activity. Finally, we showed that the cell cycle inhibitory effect of PKCδ is greatly attenuated by RNA interference-mediated knock-down of p21. Our results identify a novel link between PKCδ and G1 arrest via p21 up-regulation and highlight the complexities in the downstream effectors of PKC isozymes in the context of cell cycle progression and proliferation.

Although protein kinase C (PKC) has been widely implicated in the positive and negative control of proliferation, the underlying cell cycle mechanisms regulated by individual PKC isozymes are only partially understood. In this report, we show that PKC␦ mediates phorbol ester-induced G 1 arrest in lung adenocarcinoma cells and establish an essential role for this novel PKC in controlling the expression of the cell cycle inhibitor p21. Activation of PKC with phorbol 12-myristate 13-acetate (PMA) in early G 1 phase impaired progression of lung adenocarcinoma cells into S phase, an effect that was completely abolished by specific depletion of PKC␦, but not PKC␣. Although the PKC effect was unrelated to the inhibition of cyclin D1 expression, PKC activation significantly up-regulated p21 and down-regulated Rb hyperphosphorylation and cyclin A expression. Elevations in p21 mRNA and protein by PMA were mediated by PKC␦ but not PKC␣. Studies using luciferase reporters also revealed an essential role for PKC␦ in the PMA-induced inhibition of Rb-dependent cyclin A promoter activity. Finally, we showed that the cell cycle inhibitory effect of PKC␦ is greatly attenuated by RNA interference-mediated knock-down of p21. Our results identify a novel link between PKC␦ and G 1 arrest via p21 up-regulation and highlight the complexities in the downstream effectors of PKC isozymes in the context of cell cycle progression and proliferation.
Activation of protein kinase C (PKC) 3 is known to cause major effects on cell proliferation, survival, and differentiation. Phorbol esters, natural compounds that activate PKC and mimic the action of the lipid second messenger diacylglycerol, can impact on cell cycle progression both in positive or negative manners, resulting in either stimulation of mitogenesis or cell cycle arrest, or even in apoptosis, depending on the cell type (1)(2)(3)(4). Such a wide range of biological responses triggered by phorbol esters relates to the multiplicity of intracellular effectors for these compounds, which include the classic PKCs (␣, ␤I, ␤II, and ␥), the novel PKCs (␦, ⑀, , and ), as well as PKC-related kinases and nonkinase phorbol ester receptors (5,6). The observed changes in the expression and function of different PKC isozymes and their downstream effectors in various types of human cancers, including lung, colon, prostate, and breast cancer, suggests that dysregulation of PKC signaling may be a factor in uncontrolled cell proliferation and neoplastic transformation (7)(8)(9)(10)(11)(12).
Emerging evidence points to specialized roles for PKC isozymes due to their differential intracellular localization, substrate specificity, and selective pathway activation (13). Individual PKC isozymes can have either overlapping or opposite functions, which is strictly dependent on the cell context. For example, while PKC␤ has been implicated in the stimulation of proliferation in fibroblasts (14), this PKC isozyme causes growth arrest in colon adenocarcinoma cells (12). Opposite roles in proliferation have been ascribed to discrete members of the novel PKC family, because PKC⑀ is generally regarded as a pro-mitogenic kinase, whereas in most cellular models PKC␦ is either growth inhibitory or pro-apoptotic (15)(16)(17)(18). The mechanisms by which PKCs control proliferation have been a subject of intense investigation. In particular, PKC␦ has become the focus of numerous studies due to its tumor suppressor activity both in cellular and animal models (19 -21). Deciphering isozyme-specific functions in the context of cell cycle regulation should uncover potential signaling targets for the therapy of cancer and other proliferative diseases.
It is well established that normal progression through the cell cycle is a highly coordinated process that depends on cyclins, Cdks (cyclin-dependent kinases) and Cdk inhibitors. Although it is not yet fully understood how individual PKC isozymes control cell cycle progression, it is clear that PKC activation has profound effects on both the G 1 /S transition as well as on progression through G 2 phase. Indeed, substantial evidence supports a role for PKCs in the regulation of the cell cycle machinery, either by translational or post-translational mechanisms. PKCs can influence the G 1 phase by regulating the activity of specific cdk-cyclin complexes and the phosphorylation status of pRB, a critical regulator of G 1 /S transition, ultimately controlling the expression of E2F-regulated genes (2,7). For example, studies in NIH 3T3 cells have shown that the novel PKC inhibits cdk2 activity, leading to G 1 arrest (3). In vascular endothelial cells, PKC effects on cdk2 activity do not involve changes in cdk2 expression but rather reflect changes in the expression levels of cyclins A and E (22,23). Another mechanism by which PKCs influence the G 1 /S transition is through the modulation of cdk inhibitors. Indeed, phorbol 12-myristate 13-acetate (PMA) up-regulates p21 in a p53-independent manner in many cell types (24,25). Evidence linking G 1 arrest to p27 up-regulation in response to phorbol esters has also been presented in U937 leukemia cells (26). A series of elegant studies by Black and coworkers (27)(28)(29) has demonstrated that the G 1 arrest caused by PKC␣ activation in intestinal epithelial cells correlates with the induction of p21 and p27 and the accumulation of the hypophosphorylated form of Rb. PKC␦ was reported to regulate G 1 /S and G 2 /M transitions (2, 7), suggesting a high level of complexity in the regulation of downstream events by this novel PKC. Importantly, the vast majority of these early studies has relied heavily on overexpression strategies or the use of pharmacological tools and dominant-negative mutants of questionable specificity.
In this report, we explore the role of PKC isozymes as mediators of phorbol ester-induced control of proliferation in lung cancer cells. Thus far, little is known about the role of individual PKCs in cell cycle control in lung cancer cells, particularly those isozymes that could be potentially growth inhibitory. Despite the limited mechanistic information available, ongoing efforts aimed at targeting PKC isoforms for lung cancer therapy are being developed (8,30). PKC isozymes have also been linked to a chemo-resistant phenotype in lung cancer and have been implicated in tobacco carcinogenesis (31)(32)(33). By using RNAi to knock-down individual PKCs in a lung cancer cell model, we have determined an absolute requirement of PKC␦ for phorbol ester-induced G 1 arrest. This novel PKC is essential for phorbol ester-induced up-regulation of p21 and inhibition of cyclin A promoter activity.

EXPERIMENTAL PROCEDURES
Materials-Cell culture medium was purchased from Invitrogen. PMA and its inactive isomer 4␣-PMA were obtained from LC Labora-tories (Woburn, MA) and dissolved in ethanol. The pan-PKC inhibitor GF109203X (bisindolilmaleimide I) was from BIOMOL Research Laboratories Inc. (Plymouth Meeting, PA). Gö6976 was purchased from Alexis (San Diego, CA). Propidium iodide was from Sigma.
Cell Proliferation Assays-Cells (1 ϫ 10 5 ) cells were seeded 6-well plates. After incubation for 24 h in normal medium, cells were treated with different concentration of PMA for 1 h. Cells were trypsinized at different times and counted in a hemocytometer. For the MTS assay, H358 cells were seeded in 96-well plates. After incubation for 24 h in normal medium, cells were treated with different concentrations of PMA for 1 h. Cells were then washed with phosphate-buffered saline and incubated for an additional 24-h period. Absorbance at 490 nm was determined with the CellTiter 96 Aqueous One Solution Reagent (Promega, Madison, WI). When PKC inhibitors were used, they were added to the culture 30 min before and during PMA treatment. Cell Cycle Assays-H358 cells were seeded in 60-mm dishes (ϳ50% confluency) and serum-starved for 24 h. Approximately 80% of the cells were synchronized in G 0 after serum starvation. Synchronized cells were incubated in RPMI 1640 medium with or without PMA for 1 h and then collected at selected times for propidium iodide staining (0.1 mg/ml) followed by flow cytometry analysis, as reported previously (4).
Western Blot Analysis-Cells were lysed in a buffer containing 50 nM Tris-HCl, pH 6.8, 10% glycerol, 2% SDS, and 5% ␤-mercaptoethanol. Cell extracts (20 g of protein/lane) were subjected to SDS-PAGE and transferred to nitrocellulose membranes (Millipore, Bedford, MD). After blocking with 5% milk in 0.05% Tween 20/phosphate-buffered saline, membranes were incubated with the primary antibody for 1 h. Either anti-mouse or anti-rabbit horseradish peroxidase (1:3000, Bio-Rad) were used as secondary antibodies. Bands were visualized with a chemiluminescence detection kit (ECL, Amersham Biosciences). Densitometric analysis was performed under conditions that yielded a linear response using ImageJ software.
Quantitative Real-time RT-PCR-Cells were lysed in 500 l of TRIzol (Invitrogen), and total RNA was extracted according to the manufacturer's protocol. Total RNA (20 ng) from each sample was used for cDNA synthesis using reverse transcription reagents from Applied Biosystems Inc. Each sample was analyzed in duplicate, and mRNA levels were quantified by reference to a standard curve using the Prism 7000 sequence detection software. p21 mRNA expression was normalized to cdk4 mRNA expression, which did not vary during serum stimulation or PMA treatment.
Cell Transfections and Promoter Analyses-H358 cells (ϳ5 ϫ 10 5 , ϳ80% confluency) in 35-mm dishes were transiently co-transfected with ϩCRE and ϪCRE cyclin A promoter-luciferase vectors (34) using 5 l of Lipofectamine Plus reagent, 1 g of cyclin A promoter-luciferase plasmid, and either 1 g of a human papilloma virus type-18 E7 expression vector or 1 g of empty vector. A Renilla luciferase expression plasmid (0.01 g, pRL-CMV, Promega) was co-transfected for normalization of transfection efficiency. After an overnight recovery, cells were serum-starved for 24 h, treated with either 10 nM PMA or vehicle for 1 h, and stimulated with 10% FBS for different times. Cyclin A promoterdriven luciferase activity was determined and normalized to Renilla luciferase, as previously described (34).

PMA Induces G 1 Arrest in Bronchoalveolar Cells via PKC Activation-
To determine whether phorbol esters affect proliferation of lung adenocarcinoma cells, asynchronous H358 and H441 cells were treated with different concentrations of PMA (1 nM to 1 M, 1 h). PMA was removed by extensive washing, and cell number was determined at different times. Fig. 1A shows that PMA treatment caused a dose-dependent reduction in cell number, both in H358 and H441 cells. No evidence of apoptosis was detected upon PMA treatment, as judged by the lack of nuclear fragmentation and the absence of a sub-G 0 /G 1 population upon flow cytometry analysis (data not shown). Assessment of cell proliferation in H358 cells with an MTS assay also revealed a significant dose-dependent inhibition by PMA. The inactive analog 4␣-PMA (100 nM), on the other hand, was ineffective. The "pan" PKC inhibitor GF109203X (bisindolylmaleimide I) completely blocked the effect of PMA (Fig. 1B). Thus, the effect was mediated by PMA-responsive PKC isozymes and not by other endogenous phorbol ester receptors present in these cells (13).
To determine whether phorbol ester treatment influences the G 1 /S transition, we synchronized H358 cells in G 0 by serum starvation and examined the effect of PMA upon serum release. The maximum percentage of cells in G 0 (ϳ80%) was achieved upon 24-h serum starvation (data not shown). H358 cells entered S phase ϳ12 h after serum release, with the maximum percentage of cells in S phase observed at 18 h (Fig.  1C). At 24 h, the majority of cells were in G 2 (data not shown). When H358 cells were treated with PMA (100 nM, 1 h) at t ϭ 0 h, cells were unable to progress into the S phase (Fig. 1, C and D). The arrest caused by PMA was fully blocked with the GF109203X but not by Gö6976, an inhibitor of the classic PKCs. Representative flow cytometry experiments are depicted in Fig. 1D. Analysis of the expression of cyclins in untreated cells revealed that cyclin D1 is significantly elevated 6 h after serum release, which corresponded with an increase in Rb phosphorylation (Fig. 1E). As expected, cyclins A and B become elevated at later time points. Treatment of H358 cells with PMA significantly impaired cyclin A and cyclin B up-regulation. Analysis of cdk inhibitors revealed that PMA caused a strong and sustained up-regulation of p21, without any significant up-regulation in the levels of p27. Because up-regulation of p21 prevents Rb dephosphorylation, we determined phospho-Rb levels using a phospho-specific antibody (phospho T821 Rb). Consistent with the p21 up-regulation, a marked reduction in phosphorylated Rb was detected in PMA-treated cells.
Studies in various cell models have suggested differential roles for PKC in early-and late-G 1 phase. Thus, we decided to examine the consequence of adding PMA (100 nM) for 1 h at different times after serum release in H358 cells (Fig. 2). The inhibitory effect of PMA was only observed when added at early times in G 1 but not at later time points. Indeed, addition of PMA at t ϭ 0 -8 h impaired cell progression into the S phase, whereas addition of PMA at t Ն12 h was clearly ineffective, suggesting that only PKC activation in early-mid G 1 phase caused cell cycle arrest.
The Inhibitory Effect of PMA on G 1 /S Progression Is Mediated by PKC␦-H358 cells express three phorbol ester-responsive PKCs: one classic PKC (PKC␣) and two novel PKCs (PKC␦ and PKC⑀). In addition, they express the phorbol ester unresponsive PKC (Fig. 3A). Growth inhibitory roles have been ascribed to PKC␣ and/or PKC␦ in various cell models. The lack of effect of Gö6976, the inhibitor of classic PKCs (see Fig. 1D), led us to speculate that PKC␦ may be a key mediator of the growth inhibitory effect of PMA in H358 cells. To address this issue we depleted individual PKCs in H358 cells using RNAi. Upon transfection of specific dsRNAs for either PKC␣ or PKC␦, reductions of 69 Ϯ 19% and 81 Ϯ 9% in PKC␣ and PKC␦ levels were achieved, respectively (Fig.  3, B and C). Analysis of cell proliferation in knock-down cells revealed remarkable differences in each case. Cell cycle analysis revealed that normal progression into the S phase upon serum stimulation (in the absence of PMA treatment) was not affected when either PKC was knocked down. However, in PKC␦-depleted cells, PMA was unable to cause G 1 arrest, because the cells progress normally into the S phase. On the other hand, PMA fully arrested H358 cells in which PKC␣ has been depleted (Fig. 3D), suggesting that this isozyme was dispensable for the PMA effect. Taken together, these results indicate that PKC␦ mediates the anti-proliferative effect of PMA in H358 cells.
p21 Is Required for PMA-and PKC␦-induced G 1 Arrest in H358 Cells-Up-regulation of p21 in response to phorbol esters has been reported in several cellular models. H358 cells are p53-null (35), thus p21 up-regulation in response to PMA in these cells (see Fig. 2E) is p53-independent. To determine the requirement of p21 in PMA-induced G 1 phase arrest, a specific dsRNA was designed, which depleted this cdk inhibitor by 66% upon delivery into H358 cells (Fig. 4A). No effect was observed using an unrelated dsRNA designed to deplete GFP (data not shown). Next, we examined the effect of PMA on G 1 /S progression in p21-depleted cells. Indeed, upon p21 RNAi the G 1 arrest induced by PMA was significantly impaired (Fig. 4, B and C). Moreover, the loss of Rb phosphorylation in response to PMA was also significantly inhibited upon depletion of p21 (Fig. 4D). Thus, p21 is required for PMA-induced G 1 arrest in H358 cells.
Overexpression of PKC␦ Inhibits H358 Cell Proliferation-To further establish a role for PKC␦ as inhibitor of proliferation in H358 cells, we assessed the effect of PKC␦ overexpression using an adenoviral delivery approach. After infection of H358 cells with increasing multiplicities of infection (m.o.i. values) of a PKC␦ AdV, elevated levels of this PKC were readily detected (Fig. 5A). PKC␦ overexpression caused a significant reduction in cell number. No evidence of apoptosis was observed (data not shown). On the other hand, a control LacZ AdV, which has the same backbone as the PKC␦ AdV, did not reduce H358 cell number at an m.o.i. ϭ 300 pfu/cell (Fig. 5B). Notably, PKC␦ overexpression in H358 cells led to a significant elevation in p21 levels. Similar results were observed in H441 cells (Fig. 5A).
We next assessed whether p21 depletion could affect the inhibitory effect of PKC␦. H358 cells were subjected to p21 RNAi and then infected with the PKC␦ AdV (m.o.i. ϭ 100 pfu/cell). The anti-proliferative effect observed by PKC␦ overexpression (assessed 48 h after infection with the PKC␦ AdV) was significantly impaired in p21-depleted cells compared with control cells (Fig. 5C). The rescue effect was partial (ϳ50%), probably due to the strong and persistent nature of the stimulation caused by PKC␦ overexpression and to the fact that p21 knock-down in H358 cells was not complete (see Fig. 4A). Nevertheless, this effect was not observed when an unrelated control dsRNA (GFP) was delivered into H358 cells. Taken together, these experiments strongly support an antiproliferative role for PKC␦ in human bronchoalveolar cells and indicate that p21 is a major effector of PKC␦ in this context. PKC␦ Mediates PMA-induced Elevations in p21-Based on the information presented above, we hypothesized that PKC␦ may be required for p21 up-regulation in response to PMA. To prove this concept we determined p21 levels in response to PMA in PKC knock-down cells. Notable differences in p21 expression were found between PKC␣and PKC␦-depleted cells. Indeed, although in PKC␣ knock-down H358 cells PMA-induced elevations in p21 protein levels were minimally affected, the PMA effect was blunted after PKC␦ RNAi. PKC␦ RNAi also impaired PMA-induced up-regulation of p21 in H441 cells (Fig. 6A). To further strengthen these results, we next assessed p21 mRNA levels by Q-PCR. A time-course analysis showed that p21 mRNA levels in H358 cells are elevated as a consequence of PMA treatment (Fig. 6B). Maximum levels were observed 12 h after incubation with PMA. In agreement with the results observed at the protein level, p21 mRNA up-regulation in response to PMA was essentially blunted in PKC␦ knockdown H358 cells but not affected upon depletion of PKC␣ (Fig. 6C). These results unambiguously define the requirement of PKC␦ for p21 up-regulation in response to PMA treatment in H358 cells.
PKC␦ Controls the Expression of the Cyclin A Promoter-The inhibitory effect of PMA on the expression of cyclin A in H358 cells (see Fig.  1E) prompted us to explore whether PKC, and more specifically PKC␦, controls the activity of the cyclin A promoter. The cyclin A promoter contains several regulatory elements, including CRE (cAMP response element) and E2F sites clustered near the transcription start sites. We used a cyclin A promoter construct, p284/cyclinA-Luc (or ϩCRE), which comprises 284 bp spanning from Ϫ120 to ϩ164 of the cyclin A promoter upstream of the luciferase reporter (34). This vector was transfected into H358 cells, and promoter activity was subsequently determined at different times after serum stimulation, either in the absence or presence of PMA (100 nM, 1 h, added at t ϭ 0). A time-dependent increase in luciferase activity was observed in control cells.
However, this activity was essentially blunted by PMA treatment (Fig.  7A). As expected, luciferase activity of the CRE-mutated promoter (p225/cyclin A, or ϪCRE, which comprises bp Ϫ61 to ϩ164 of the cyclin A promoter (34)) was less than that containing the wild-type CRE (Fig. 7, compare scales in A and B). However, its activity was also impaired by PMA (Fig. 7B). This result shows that the CRE site is not required for the inhibitory effect of PMA and suggests that the E2F site may be the target of PMA action. Indeed, we found that transfection of human papillomavirus-18 E7, which inactivates Rb and allows for constitutive E2F activity, largely eliminated the inhibitory effect of PMA on the cyclin A promoter. Moreover, the rescue of promoter activity by E7 was independent of the CRE site (Fig. 7C). Together, these data indicate that the inhibitory effect of PMA on cyclin A promoter activity and gene expression is mediated through the E2F site, a result that agrees well with the stimulatory effect of PMA on p21 levels.
To determine whether promoter activity was controlled in a PKC isozyme-dependent manner, we performed similar experiments in PKC␣and PKC␦-depleted cells. Remarkably, the inhibitory effect of PMA on cyclin A promoter activity was completely abolished upon PKC␦ RNAi. PKC␣ knock-down, on the other hand, was unable to rescue the inhibitory effect of PMA on the promoter activity (Fig. 7D).

DISCUSSION
Because phorbol esters exert profound effects on cell proliferation, it is essential to understand the mechanisms by which their cellular effec- tors control the different constituents of the cell cycle machinery. In that regard, our studies have established that PKC␦, a member of the novel PKC family, mediates the anti-proliferative effect of PMA in lung adenocarcinoma cells. PKC␦ is responsible for the elevations in the cdk inhibitor p21 in response to PMA, which consequently leads to inhibition of G 1 /S progression. Striking differences exist with regards to PKC specificity in this process, as PKC␣, the only classic PKC expressed in H358 cells, does not mediate p21 up-regulation and inhibition of cell cycle progression in response to PMA.
Evidence accumulated in the last years points to distinct roles for individual PKC isozymes in the control of cell cycle progression, both at the levels of G 1 /S and G 2 /M transitions (2,3,28). Early studies in vascular endothelial cells have shown that phorbol esters potentiate the mitogenic effect of growth factors when added in early G 1 phase, but they inhibit DNA synthesis in late G 1 phase (22). These bi-directional effects closely associate to the control of G 1 cdks and result from either activation or inhibition of Rb protein phosphorylation. Hypophosphorylation of Rb and other pocket proteins (p107 and p130) upon PKC activation has been detected in several models, including intestinal epithelial cells, vascular smooth muscle cells, and fibroblasts (2,22,28,36), although the underlying mechanisms are not fully understood and may vary from one cell type to another. In H358 cells, we found that the G 1 arrest caused by PMA after serum release only occurs when the phorbol ester was added at the initial time points, suggesting that PKC was indeed targeting early regulatory events. Interestingly, PMA does not inhibit cyclin D1 expression in H358 cells, as we detected a substantial induction of cyclin D1protein (Fig. 1E) and mRNA by the phorbol ester (data not shown). In other cell types, such as in microvascular endothelial cells, PKC␦ delays the induction of cyclin D1 following serum stimulation (37). Other studies have reported no changes in cyclin D1 in response to PKC activation (36). Despite these major cell type differences in the control of cyclin D1 expression by PKC, our data are suggestive of alternative mechanism for PMA-induced G 1 phase arrest that involves p21 in lung cancer cells.
PMA causes a fast and robust increase in the levels of p21 (mRNA and protein) in H358 bronchoalveolar cells. In this model the induction of p21 represents a key event in the G 1 arrest by phorbol ester activation. Indeed, the effect of PMA was significantly impaired when p21 was depleted with RNAi. Our results also revealed a remarkable selectivity for PKC isozymes in the control of p21 induction, both at the level of protein and mRNA. p21 up-regulation was basically abolished in PKC␦depleted cells, whereas PKC␣ RNAi did not cause any effect. These results contrast with those reported recently in intestinal epithelial cells, where p21 up-regulation correlates with PKC␣ activation and was not affected by PKC␦ RNAi (27). This is not surprising, because major differences in PKC isozyme-specific functions exist among different cell types. Unlike PKC␦, which is generally regarded as a growth inhibitory protein in the vast majority of cell lines, PKC␣ promotes mitogenic signaling in various cell types (38 -41), which suggests that the inhibitory role for PKC␣ is probably restricted to a few cell types. These cell type differences are normally related to differential intracellular localization and access to targets and may well explain the differences observed among the different studies. It has been reported that, in cells in which PKC␦ activation is apoptogenic, such as in prostate cancer cells (4,17), cell death is preceded by p21 up-regulation and Rb dephosphorylation (42). Thus, p21 is probably a key component of anti-proliferative and apoptotic responses mediated by PKCs, particularly those involving PKC␦. Although the generalization of this concept still needs to be determined, our experiments using p21 knock-down cells strongly support that this is the case in lung cancer cells.
Our studies also reveal that in H358 cells, PKC activation impairs the expression of cyclin A, which is normally required for the activation of cdk2 in S phase. Cell cycle-dependent expression of the cyclin A promoter is regulated by E2F-pocket protein complexes. The cyclin A promoter also contains a cAMP-responsive element (CRE) element that is probably constitutively occupied in G 1 -phase cells and regulated by mitogenic signaling via phospho-CREB. We found that PKC activation markedly inhibits the activity of a cyclin A-luciferase reporter and that this effect was independent of the CRE site. Interestingly, the PMA effect was overcome by the expression of human papillomavirus E7, which inactivates Rb and allows for constitutive E2F signaling. Our results therefore suggest that the inhibitory effect of PKC on cyclin A promoter activity is probably secondary to the pocket protein inactivation and E2F release. This effect is mediated by PKC␦ and not by PKC␣, as determined by RNAi. All of these results agree well with our data showing that PMA and PKC␦ up-regulate p21, a cdk inhibitor that blocks pocket protein inactivation.
In summary, our studies established PKC␦ as a key component of the phorbol ester response leading to p21 induction, prevention of Rb dephosphorylation, and G 1 arrest in lung cancer cells. In addition to uncovering a relevant mechanism underlying PKC control of cell cycle, our studies may have important implications in carcinogenesis. For example, polycyclic aromatic hydrocarbon metabolites, which are constituents of tobacco smoke, have been recently found to inhibit the activity of PKC isozymes, including that of PKC␦ (43). One may speculate that impairing the activation of growth inhibitory PKCs could greatly impact on cell cycle regulatory mechanisms and therefore represent a causative or added factor to lung carcinogenesis. Various PKC FIGURE 7. PKC␦ inhibits the activity of the cyclin A promoter. A and B, H358 cells were transiently co-transfected with either ϩCRE (A) or ϪCRE (B) cyclin A promoter-luciferase reporter plasmids, together with a pRL-CMV Renilla luciferase vector. After 14 h, cells were serum-starved for 24 h and then stimulated with 10% FBS. Cells were treated with either 10 nM PMA (ϩPMA) or vehicle (ϪPMA) for 1 h at t ϭ 0 h. Luciferase activity was determined and normalized to Renilla luciferase activity. Promoter activity was plotted as -fold stimulation relative to the FBS-treated cells at t ϭ 6 h in the presence or absence of PMA. Data are presented as mean Ϯ S.E. of triplicate measurements. A second experiment gave similar results. C, activity of the ϩCRE or ϪCRE cyclin A promoters was determined either in the absence (ϪE7) or presence (ϩE7) of E7, and upon treatment with either 10 nM PMA (ϩPMA) or vehicle (ϪPMA) for 1 h at t ϭ 0 h. Luciferase activity was determined and normalized to Renilla luciferase activity. Promoter activity was plotted as -fold stimulation relative to the FBS-treated cells in the absence of E7. Data are presented as mean Ϯ S.E. of triplicate measurements. Two additional experiments gave similar results. D, H358 cells were co-transfected with the ϩCRE cyclin A luciferase reporter plasmid and the corresponding dsRNAs. Cells were serum-starved and treated with either 10 nM PMA (ϩPMA) or vehicle (ϪPMA) at t ϭ 0 h. Luciferase activity was determined 24 h after stimulation with FBS 10% and normalized to Renilla luciferase activity. Data were plotted as percent maximal activity relative to control cells in the absence of PMA. The results are representative of two independent experiments. analogs (including phorbol esters) are currently undergoing clinical trials for the treatment of various hyperproliferative and neoplastic diseases (1, 44 -47); our studies may also have important implications for understanding the molecular basis of their therapeutic effects. The characterization of the novel PKC␦-p21 link highlights the complexities in the signaling events downstream of PKC activation in the context of cell proliferation and provides mechanistic rationale for the development of isozyme specific strategies for therapeutic interventions.