Protein Kinase C (cid:1) Stimulates Apoptosis by Initiating G 1 Phase Cell Cycle Progression and S Phase Arrest *

Overexpression of protein kinase C (cid:1) (PKC (cid:1) ) stimulates apopto-sisinawidevarietyofcelltypesthroughamechanismthatisincom- pletely understood. PKC (cid:1) -deficient cells are impaired in their response to DNA damage-induced apoptosis, suggesting that PKC (cid:1) is required to mount an appropriate apoptotic response under conditions of stress. The mechanism through which it does so remains elusive. In addition to effects on cell survival, PKC (cid:1) elicits pleiotro-pic effects on cellular proliferation. We now provide the first evidence that the ability of PKC (cid:1) to stimulate apoptosis is intimately linked to its ability to stimulate G 1 phase cell cycle progression. Using an adenoviral-based expression system to express PKC (cid:2) , - (cid:1) , and - (cid:3) in epithelial cells, we demonstrate that a modest increase in PKC (cid:1) activity selectively stimulates quiescent cells to initiate G 1 phase cell cycle progression. Rather than completing the cell cycle, PKC (cid:1) -infectedcellsarrestinSphase,aneventthattriggerscaspase-dependentapoptoticcelldeath.Apoptosiswasprecededbytheacti- vationofcellcyclecheckpoints,culminatinginthephosphorylation activation, and apoptosis. S phase entry was inhibitedbyLY294002,suggestingthatPI3KisrequiredfortheeffectsofPKC (cid:4) onG 1 toS phase cell cycle progression. Blockade of S phase entry using LY294002 prevented checkpointactivationandapoptosis.Thepan-caspaseinhibitorQVDblockedapoptosisandincreasedtheproportionofSphasecells.Althoughnotshownhere,itispossiblethatadditionalpathwaysactivatedbyPKC (cid:4) contribute to checkpoint activation and apoptosis.

responsive to diacylglycerol, whereas the atypical isoforms are diacylglycerol-insensitive (reviewed in Ref. 1). Most cells express multiple PKC isozymes that require distinct cofactors and exhibit unique intracellular localizations. Pharmacological inhibitors have been used to assess the roles of individual PKC isozymes. Much of what is known regarding the biological roles of PKC␦ has been derived from studies using the PKC␦-selective inhibitor rottlerin, the specificity of which is in question. The isolation of cell lines overexpressing or lacking individual isozymes has been used to decipher the physiological roles of select PKC isozymes. A major limitation of this approach is that it does not readily distinguish between the effects of PKC activation versus the subsequent down-regulation of PKC expression. We opted to use adenoviruses to drive selective, modest increases in the activities of PKC␣, -␦, and -⑀, a powerful approach that has provided substantial insight into the regulation of cellular survival by PKC (2)(3)(4). We selected thyroid epithelial cells for this analysis because alterations in PKC activity, expression, and structure (5-9) have been documented in human thyroid tumors. The results of this analysis revealed the existence of a novel mechanism through which PKC␦ stimulates apoptosis and demonstrate a tight linkage between the effects of PKC␦ on cell proliferation and cell death.
Cell Culture-Wistar rat thyroid cells were propagated in three hormone-containing growth medium (3H, containing thyroid-stimulating hormone, insulin, and serum) as described previously (10). Cells were rendered quiescent by starvation in Coon's modified Ham's F-12 medium devoid of thyroid-stimulating hormone and growth factors (basal medium) for 48 -72 h.
Adenoviral Infection-Adenoviruses for PKC␣, -␦ (2, 3), and -⑀ and kinase-defective PKC␦(K376R) (11) were used. Quiescent cells were infected with adenoviruses for PKC␣ (2,000 or 10,000 particles (p)/cell), PKC␦ (5,000 p/cell), or PKC⑀ (7,000 p/cell) in basal medium for 16 h (day 1 post-infection). Virus was removed, and the cells washed twice and incubated in basal medium for the times indicated. In experiments where TPA was used, TPA (25 nM) was added at day 2 post-infection for 60 min, and the cells were washed and then incubated in basal medium for the times indicated. Inhibitors were added at the time of infection and again at day 1 post-infection.
Flow Cytometry-Floating and trypsinized cells were collected and pelleted at 1,000 rpm for 5 min at 4°C. Cells were fixed in ice-cold MeOH overnight and resuspended in 250 l each RNase (200 units/ml) and propidium iodide (0.1 mg/ml in 0.1% Triton X-100, 0.037 mg/ml EDTA) in phosphate-buffered saline. Cells were analyzed on a BD Biosciences FACScan and analyzed using CellQuest TM Pro software.
DNA Laddering-Floating and adherent cells were collected and resuspended in TE/Triton buffer (0.2% Triton X-100, 10 mM Tris, pH 8.0, 1 mM EDTA). Following incubation on ice for 10 min, an aliquot of the lysate containing total DNA was removed, and the remaining lysate was centrifuged at 14,000 ϫ g for 15 min at 4°C. Supernatant containing low molecular weight DNA was transferred to a fresh tube and treated with DNase-free RNase A (60 g/ml) for 1 h at 37°C. SDS (0.5%) and proteinase K (150 g/ml) were added, and the samples were incubated for 1 h at 50°C. DNA was precipitated by addition of 0.1 volume of 0.5 M NaCl and 1.0 volume isopropanol, followed by incubation on ice for 10 min. Following centrifugation, DNA was resuspended in TE buffer (10 mM Tris, pH 8.0, 1 mM EDTA) and fractionated on 2% agarose gels.
Immunostaining-DNA synthesis was assessed by BrdUrd incorporation as described in Kupperman et al. (10). For active caspase-3 and ␥-H2AX staining, cells were fixed in 3.7% formaldehyde/phosphatebuffered saline, permeabilized in 0.2% Triton X-100 for 2 min, and stained with primary and secondary antibodies for 1 h at 37°C. Nuclei were stained with 4Ј,6-diamidino-2-phenylindole (DAPI). Images were captured on a Zeiss Axiophot microscope fitted with a Hamamatsu ORCA-ER digital camera using Axiovision 4.2 software.
Western Blotting-Cells were washed in ice-cold phosphate-buffered saline and lysed in radioimmune precipitation assay buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 0.5% sodium deoxycholate, 1% Nonidet P-40, 0.1% SDS) plus protease inhibitors. Protein content in clarified lysates was assayed using the Bio-Rad DC protein assay. Incubations with primary antibodies were performed overnight at 4°C, followed by incubation with horseradish peroxidase-conjugated secondary antibodies and detection via chemiluminescence.
Kinase Assays-For PKC assays, infected cells were scraped in isotonic buffer (50 mM Tris, pH 7.4) containing protease inhibitors, sonicated, and clarified by centrifugation. Protein concentration was determined, and samples were dispensed into 20-l aliquots in triplicate. 30 Two experiments were performed with similar results. B, cells were infected as described above and harvested at day 2 post-infection, and kinase activity was assessed using PKC␣-pseudosubstrate peptide as substrate (12). The results of a representative experiment performed in triplicate are shown. Similar results were observed in three independent experiments. C, cells infected as described above were harvested at day 2, and PKC expression was assessed by Western blotting with PKC␣, -␦, and -⑀-specific antibodies.

RESULTS
PKC␦ Induces Aberrant Cell Cycle Progression-Phorbol esters stimulate proliferation and impair differentiated function in thyroid epithelial cells. Because differentiated function is an important prognostic indicator for patients with thyroid tumors (14), the identity of the PKC isozymes responsible for these effects is important to ascertain. To elucidate the contributions of PKC␣, -␦, and -⑀, the phorbol ester-responsive isozymes expressed in Wistar rat thyroid cells (15), to the regulation of thyroid cell proliferation and function, we utilized adenoviruses to overexpress individual PKC isozymes. Quiescent Wistar rat thyroid cells were infected with adenoviruses expressing PKC␣, -␦, or -⑀ at varying multiplicities of infection and kinase activity measured in cell lysates prepared at day 2 post-infection. Conditions were derived under which each isozyme exhibited a modest 2-fold increase over endogenous PKC activity, and could be selectively activated with a dose of TPA to which uninfected cells did not respond (2) (Fig. 1B and data not shown). Western blotting indicated that the PKC isozymes were substantially overexpressed under these conditions (Fig. 1C). Immunostaining experiments conducted with PKC-selective antibodies revealed that Ͼ95% of the cells were infected under these conditions (data not shown).
Flow cytometry analysis of propidium iodide-stained cells revealed that PKC␦ selectively induced cell cycle entry. Expression of PKC␦ decreased the proportion of G 1 phase cells and significantly increased the number of S phase cells (Fig. 1A). The effects of PKC␦ on S phase entry were similar in the presence and absence of TPA activation. Strikingly, the effects of PKC␦ on cell cycle progression were not reproduced by PKC␣ or -⑀ even following expression at higher levels than PKC␦ and in the presence or absence of phorbol ester treatment. At the concentration used, TPA failed to stimulate cell cycle progression in mock infected cells (Fig. 1A, lower panel). To corroborate these data, the effects of PKC␦ on molecular markers of cell cycle progression were analyzed. PKC␦ stimulated the expression of cyclins A and E, as well as of Cdk-2, the catalytic partner for the G 1 phase cyclins ( Fig. 2A). Expression was increased beginning at day 2 post-infection and sustained over   Intriguingly, PKC␦-infected cells did not progress into G 2 /M ( Fig.  1A), suggesting that these cells fail to complete the cell cycle. This is most clearly seen in Fig. 3A where PKC␦ stimulated a time-dependent increase in S phase cells (days 2-3) without a coordinate increase in G 2 phase cells (days 2-5). This pattern was strikingly different from that observed following mitogenic (3H) treatment of quiescent cells (Fig.  3C). PKC␦-infected cells arrested with a DNA content that was greater than that of G 1 phase (3H, 30 h) cells and less than that of G 2 /M phase (3H plus nocodazole) cells. In agreement with these data, PKC␦ failed to stimulate cyclin B1 expression, a marker of G 2 phase cells ( Fig. 2A). Together, these results indicate that a modest increase in PKC␦ activity is sufficient to induce exit from quiescence, but not to stimulate cell proliferation. Rather, PKC␦-expressing cells proceed through G 1 phase and subsequently arrest in S phase.
PKC␦-stimulated S Phase Arrest Is Followed by Apoptosis-The proportion of PKC␦-infected S phase cells declined over time, commensurate with an increase in the proportion of cells containing hypodiploid DNA (Figs. 3A and 4). Despite their similar activities, PKC␣ and -⑀ did not stimulate hypodiploid DNA content in the presence or absence of TPA activa-tion (Fig. 4). Similar to its effects on cell cycle progression, hypodiploid DNA content induced by PKC␦ did not require TPA activation.
To document that the effects of PKC␦ on hypodiploid DNA content reflected apoptosis, caspase-3 cleavage was examined. Western blotting using an antibody that detects pro-caspase-3 and activated (cleaved) forms of caspase-3 revealed the presence of cleaved caspase-3 in lysates prepared from detached (floating) PKC␦-infected cells (Fig. 5A). The abundance of intact pro-caspase-3 was markedly reduced in these cells. Longer exposures of the Western blot shown in Fig. 5A revealed the presence of cleaved caspase-3 in adherent cells (data not shown), indicating that cleavage was not secondary to cell detachment. Nonetheless, to confirm that apoptosis was initiated in adherent cells, we analyzed the expression of active caspase-3 using an antibody that selectively recognizes the cleaved form of caspase-3. Staining for active caspase-3 was first detected at day 3 post-infection and increased over time (Fig. 5B). The caspase-3 positive cells invariably exhibited an apoptotic nuclear morphology assessed by DAPI staining. Active caspase-3 staining was abolished following treatment with the pancaspase inhibitor Q-VD-OPh (QVD) (Fig. 5B). Finally, PKC␦ stimulated QVD-sensitive DNA laddering, confirming that cell death was apoptotic (Fig. 5C).  PKC␦ is subject to caspase-dependent cleavage, generating a C-terminal fragment containing an intact kinase domain (16). DeVries et al. (17) reported that overexpression of the PKC␦ fragment resulted in its accumulation in the nucleus and the induction of apoptosis. Western blotting with a C-terminal-directed antibody failed to reveal the presence of significant amounts of the PKC␦ fragment in apoptotic thyroid cells (data not shown), hence we do not believe PKC␦ cleavage initiates apoptosis in these cells.
Given that apoptosis temporally follows the accumulation of cells in S phase, we reasoned that PKC␦-infected cells die from S phase. Indeed, inhibition of apoptosis with QVD prevented the accumulation of cells with hypodiploid DNA and markedly increased the proportion of S phase cells (Fig. 3B). This result, together with the absence of cyclin B1 expression as well as of G 2 /M phase cells (Figs. 1 and 2), strongly suggests that expression of PKC␦ stimulates S phase arrest that culminates in apoptosis.
PKC␦ Activates the DNA Damage Response Pathway-We set out to confirm the effects of PKC␦ on S phase entry by monitoring effects on DNA synthesis. DNA synthesis was examined by incorporation of the thymidine analog bromodeoxyuridine (BrdUrd) into replicating DNA (10). Unlike mitogen-treated thyroid cells where BrdUrd labeling is uniform (13,18,19), BrdUrd localized to punctate foci in PKC␦-infected cells (Fig. 6A). The significance of these foci remains to be determined; however, they are reminiscent of the foci observed following stalled replication and/or DNA damage (20,21). This prompted us to investigate whether PKC␦ activated the DNA damage response pathway.
ATR-, ATM-, and DNA-dependent protein kinase (DNA-PK) are key mediators of the cellular response to DNA damage. These PI3K-like kinases are activated by replication stress and DNA double strand breaks. Activation of ATR/ATM results in the phosphorylation and activation of downstream substrates, notably Chk-1 and Chk-2. Stalled replication is a robust stimulator of ATR activity and culminates in the activating phosphorylation of Chk-1 on serine 345 (22). Overexpression of PKC␦, but not PKC␣, stimulated Chk-1 phosphorylation (Fig. 6B). Phosphorylation was maximal at day 2 post-infection, similar to the time course over which PKC␦ stimulated DNA synthesis (data not shown). To confirm the activation of cell cycle checkpoints, the effects of PKC␦ on p53 phosphorylation were examined. ATR and ATM phosphorylate p53 on serine 15 (23). PKC␦ stimulated p53 phosphorylation at this site, whereas PKC␣ failed to do so (Fig. 6C). Phosphorylation of p53 in PKC␦-expressing cells was maximal at day 2 post-infection with significant phosphorylation also observed at day 3. To further document checkpoint activation, we assessed the effects of PKC␦ on the phosphorylation of histone H2AX (designated ␥-H2AX), an indicator of DNA double strand breaks (24). Immunostaining experiments revealed an increase in ␥-H2AX staining as early as day 2 post-infection (Fig. 6D). Thus, ␥-H2AX staining temporally coincided with BrdUrd incorporation and Chk-1 and p53 phosphorylation. In contrast, caspase-3 activation was first detected at day 3 post-infection (Fig. 5B), suggesting that it followed checkpoint activation. To eliminate the possibility that DNA cleavage itself induced checkpoint activation, the effects of PKC␦ on checkpoint activity were examined in the presence of the caspase inhibitor QVD. QVD had no effect on the ability of PKC␦ to stimulate Chk-1, p53, or H2AX phosphorylation (Fig. 6, B and C, and data not shown). These data support a model wherein PKC␦ stimulates replication stress, leading to checkpoint activation and ultimately, apoptosis.
PKC␦-stimulated Apoptosis Requires Cell Cycle Entry-PKC␦ has been reported to activate JNK (25,26) and to require p38 or ERK for induction of apoptosis (4,27). Pharmacological inhibitors of JNK (SP600125), p38 (SB203580), and MEK1 (UO126) were used to determine whether PKC␦stimulated apoptosis requires activity of these MAPK family members. None of these inhibitors impaired DNA laddering induced by PKC␦ (Fig.   7A). Intriguingly, treatment with the pan-PKC inhibitor GF109203X also failed to impair DNA laddering. To further examine whether kinase activity was required for PKC␦-stimulated apoptosis, the effects of kinase-deficient PKC␦ were examined. Following expression at levels similar to wild type PKC␦ (data not shown), kinase-deficient PKC␦ stimulated apoptosis as assessed by hypodiploid DNA content (Fig. 7B). Therefore, kinase activity is not strictly required for the effects of PKC␦ on apoptosis in thyroid cells. Indeed, kinase-dependent and -independent effects of PKC␦ on apoptosis have been previously reported (28,29).
Given that apoptosis temporally follows a delay in S phase, we set out to determine whether inhibition of cell cycle entry would prevent apoptosis. Phosphatidylinositol 3-kinase (PI3K) activity is required for thy- The same blot was probed with actin to demonstrate equal loading. Two time-course experiments were performed with similar results. PKC␦ was observed to phosphorylate Chk-1 at day 2 in more than five experiments. Lower panels, lysates prepared from PKC␦and PKC␣-(10,000 p/cell) infected cells harvested at day 2 were subjected to Western blotting with the p-Chk1(Ser-345) antibody. Pretreatment with QVD (20 M) had no effect on Chk-1 phosphorylation in response to PKC␦. The same blot was probed with an anti-actin antibody to document equal protein loading. Two experiments were performed with similar results for PKC␣ and four experiments for QVD on PKC␦ effects. In C: Upper panels, total cell lysates prepared from PKC␦-infected cells on days 2-3 post-infection were subjected to Western blotting with antibodies to p-p53 (Ser-15) and actin. Three experiments were performed with similar results. Lower panels, Western blot of total cell lysates prepared from PKC␦-and PKC␣ (10K p/cell)-infected cells harvested at day 2 post-infection in the presence or absence of QVD (20 M) and probed for p-p53 (Ser-15) and actin. Two experiments were performed with similar results for PKC␣, and four experiments for QVD on PKC␦ effects. D, PKC␦-infected cells were fixed at day 2 post-infection and stained with an antibody detecting histone H2AX phosphorylated at Ser-139 (␥-H2AX) and with DAPI. Images were exposed for equal times. Three experiments were performed with similar results. roid cell proliferation (19). Therefore, the PI3K inhibitor LY294002 was used to investigate the consequences of cell cycle inhibition on PKC␦stimulated apoptosis. Unlike the other inhibitors tested, LY294002 abolished DNA laddering (Fig. 7A). FACS analysis confirmed that LY294002 blocked the accumulation of cells containing hypodiploid DNA and that this inhibitor prevented S phase entry (Fig. 7C). Consistent with the latter effect, LY294002 abolished DNA synthesis in PKC␦expressing cells (Fig. 7D). PDK-1 phosphorylates PKC␦ on T505, a modification that increases kinase activity. Because PDK-1 is downstream of PI3K, we considered the possibility that LY294002 inhibited PKC␦ kinase activity. Pretreatment of PKC␦-infected cells with LY294002 (data not shown), or addition of LY294002 directly to in vitro kinase assays, did not impair PKC␦ kinase activity (Fig. 7E). Additionally, LY294002 had no effect on PKC␦ expression (Fig. 8A). Therefore, we speculate that the inhibitory effects of LY294002 on apoptosis are mediated predominantly through its ability to block S phase entry. In agreement with this notion, the JNK, MEK1, and p38 inhibitors were less effective than LY294002 in preventing S phase entry (Fig. 7C) and DNA synthesis (Fig. 7D). The PKC inhibitor GF109203X only partially reduced PKC␦-stimulated DNA synthesis.
To further establish that checkpoint activation arose as a consequence of delayed progress through S phase, the effects of LY294002 on PKC␦-stimulated Chk-1 and p53 phosphorylation were investigated. Interestingly, treatment with PI3K (LY294002), MEK1 (UO126), p38 (SB203580), or JNK (SP600125) inhibitors blocked or significantly impaired Chk-1 phosphorylation (Fig. 8A). However, these inhibitors differed in their ability to block PKC␦-induced p53 phosphorylation. Only the PI3K inhibitor (LY294002) completely inhibited p53 phosphorylation. Because LY294002 is known to inhibit ATR, ATM, and DNA-PK activity (30 -33) its effects on Chk-1 phosphorylation were assessed. In contrast to its effects on PKC␦-stimulated Chk-1 phosphorylation, LY294002 failed to inhibit Chk-1 phosphorylation in response to aphidicolon (an inhibitor of replication, and positive control for ATR activation/Chk-1 phosphorylation) (Fig. 8B). Therefore, at the concentration used in these studies, LY294002 does not directly inhibit ATR activity. Cumulatively, these data suggest that replication stress in response to PKC␦ expression initiates a checkpoint response that results in apoptosis (Fig. 9).

DISCUSSION
Using adenoviruses to elicit a 2-fold increase in PKC activity, we compared the cellular consequences associated with selective overexpression of PKC␣, -␦, and -⑀. Remarkably, overexpression of PKC␦ in quiescent thyroid epithelial cells was sufficient to stimulate G 1 /S phase cell cycle progression. Rather than completing the cell cycle, PKC␦expressing cells arrested in S phase, and ultimately perished by caspasedependent apoptosis. Blockade of apoptosis restored the accumulation of S phase cells, indicating that PKC␦-expressing cells perish from S phase. These highly unusual effects of PKC␦ could not be reproduced by similar or higher levels of PKC␣ or PKC⑀ activity. These data reveal that increased PKC␦ activity is sufficient for exit from the quiescent state, but insufficient for completion of the cell cycle.
Experiments to examine the effects of PKC␦ on DNA synthesis revealed a highly unusual pattern of BrdUrd incorporation in PKC␦infected cells. Rather than the uniform labeling of nuclei observed in mitogen-treated cells, BrdUrd localized to heterogeneous nuclear foci in PKC␦-infected cells. The distinct pattern of BrdUrd incorporation, together with the protracted S phase observed in PKC␦-infected cells, prompted us to investigate whether PKC␦ activated cell cycle checkpoints. ATR, ATM, and DNA-PK are activated by DNA damage and replication stress and play an essential role in the maintenance of genomic stability through their ability to halt cell cycle progression and facilitate DNA repair. Under conditions of replication stress or irreparable DNA damage, checkpoint kinases facilitate cell death. Overexpression of PKC␦, but not PKC␣ stimulated Chk-1 and p53 phosphorylation, indicators of checkpoint activation. PKC␦-infected cells also exhibited phosphorylation of histone H2AX, a marker of DNA double strand breaks.
If apoptosis was a direct consequence of replication stress induced by S phase arrest, then inhibition of S phase entry would be expected to rescue PKC␦-expressing cells from apoptosis. Indeed, treatment with the PI3K inhibitor LY294002 blocked S phase entry, checkpoint activation, and apoptosis. In contrast, inhibitors of MAPK cascades did not prevent apoptosis. However, MEK1, JNK, and p38 inhibitors were much less effective than the PI3K inhibitor in blocking S phase entry in PKC␦infected thyroid cells. Hence, our data strongly suggest that it is the accumulation of PKC␦-expressing cells in S phase that induces apoptosis and that apoptosis is a consequence of replication stress.
Whether checkpoint activation is essential for PKC␦-mediated apoptosis remains to be determined. As members of the phosphatidylinositol kinase family, ATM, ATR, and DNA-PK are inhibited by LY294002 (30 -33). To ascertain whether LY294002 rescued apoptosis through the direct inhibition of checkpoint kinase activity, its effects on aphidicolin-stimulated Chk-1 phosphorylation were examined. At the concentration used in these studies, LY294002 blocked Chk-1 phosphorylation in response to PKC␦, but not that stimulated by aphidicolin. Nonetheless, we cannot exclude a role for other phosphatidylinositol kinases (i.e. ATM or DNA-PK) in the inhibition of apoptosis by LY294002. PKC␦ associates with and inhibits the activity of DNA-PK (34), raising the possibility that PKC␦ stimulates S phase entry and coordinately inhibits DNA repair, resulting in the accumulation of singlestranded DNA and checkpoint activation. Full-length PKC␦ has been found in the nucleus (17), and we observed nuclear staining following overexpression of PKC␦ in thyroid cells (data not shown). Therefore, PKC␦ is in the appropriate cellular compartment to elicit effects on DNA synthesis and repair. Complexes containing PKC␦ and c-abl, a proteintyrosine kinase activated by oxidative stress and DNA damage, have also been reported (35,36). PKC␦ associates with many cellular proteins, including p73␤ (37), members of the phospholipid scramblase family PSL-1 (38) and PSL-3 (39), lamin B (40), and src family kinases (reviewed in Ref. 41). Complexes between PKC␦ and mammalian target of rapamycin (mTOR), as well as PI3K, have also been described (42)(43)(44). It is possible that PKC␦containing protein complexes elicit functional consequences on DNA repair and apoptosis, either through alterations in protein phosphorylation or through a scaffolding function. If PKC␦ acts via the latter mechanism, this would explain why PKC␦-stimulated apoptosis is independent of PKC␦ kinase activity in thyroid and other (29) cells.
At first glance, the ability of PKC␦ to stimulate G 1 phase cell cycle progression appears to conflict with numerous reports in the literature that PKC␦ mediates growth suppression. However, there is increasing evidence that PKC␦ positively regulates some aspects of cell cycle progression. Kitamura et al. (45) reported that serum stimulated the biphasic activation of PKC␦ in fibroblasts. The second wave of PKC␦ activity was temporally correlated with entry into S phase. Overexpression of PKC␦ enhanced serum-stimulated DNA synthesis, whereas kinase-defective mutants impaired it, effects that were observed in both 3Y1 and NIH3T3 fibroblasts. In a separate study, overexpression of PKC␦ led to increased E2F promoter activ-  . Model for PKC␦-stimulated apoptosis. PKC␦ stimulates entry into S phase, followed by S phase arrest, checkpoint activation, and apoptosis. S phase entry was inhibited by LY294002, suggesting that PI3K is required for the effects of PKC␦ on G 1 to S phase cell cycle progression. Blockade of S phase entry using LY294002 prevented checkpoint activation and apoptosis. The pan-caspase inhibitor QVD blocked apoptosis and increased the proportion of S phase cells. Although not shown here, it is possible that additional pathways activated by PKC␦ contribute to checkpoint activation and apoptosis.
ity in 3Y1 fibroblasts (46). Therefore, accumulating evidence supports a conserved role for PKC␦ in the positive regulation of early events in the cell cycle. Despite its requirement for serum-stimulated DNA synthesis, chronic overexpression of PKC␦ mediated growth inhibition in 3Y1 cells (45). Similarly, transient expression of PKC␦ decreased the proportion of M phase cells. In capillary endothelial cells, overexpression of PKC␦ caused delayed exit from S phase (47), whereas in PKC␦-overexpressing CHO cells, treatment with phorbol esters stimulated G 2 /M arrest (48). Although apoptosis was not reported in these studies, apoptosis is a common end point of stalled cell cycle progression. PKC␦-expressing thyroid cells exhibited sustained increases in Cdk-2 expression and activity. This would be expected to result in the stabilization of E2F, an event known to trigger apoptosis. Moreover, replication stress stabilizes E2F via ATM-dependent phosphorylation, and DNA damage leads to the acetylation of E2F1 and its recruitment to apoptotic promoters (49). Together, these data indicate that PKC␦ is capable of promoting G 1 to S phase transition and either inhibits cell cycle progression at a step between S and G 2 /M phases, or requires additional signals to efficiently complete the cell cycle. The failure of PKC␦-overexpressing cells to complete the cell cycle could be due to an imbalance in the level of PKC␦ activity compared with that of other PKC enzymes or cell cycle regulators.
In conclusion, our findings reveal a novel mechanism through which PKC␦ stimulates apoptosis. Apoptosis arises as a consequence of inefficient cell cycle progression, culminating in checkpoint activation and apoptosis (Fig. 9). This is likely to be a well conserved mechanism through which PKC␦ regulates cell survival, given the reports that PKC␦ regulates not only apoptosis, but also cell cycle progression. The signaling pathways through which PKC␦ stimulates apoptosis may vary by cell type; however, we suggest that the mediators of PKC␦ effects on apoptosis are those that convey effects on cell proliferation. In thyroid epithelial cells, it is clear that the targets of PKC␦ include members of the phosphatidylinositol kinase family, thereby identifying phosphatidylinositol kinases as downstream mediators of PKC␦ effects. It will be important to investigate whether PKC␦ plays a role in thyroid-stimulating hormone-driven proliferation, and/or downstream from thyroid oncogenes such as Ras and Ret/PTC given their effects on both cell proliferation and survival (50 -54).