Nuclear mitogen-activated protein kinase activation by protein kinase czeta during reoxygenation after ischemic hypoxia.

We examined the upstream kinases for mitogen-activated protein kinase (MAPK) activation during ischemic hypoxia and reoxygenation using H9c2 cells derived from rat cardiomyocytes. Protein kinase C (PKC)zeta, an atypical PKC isoform mainly expressed in rat heart, has been shown to act as an upstream kinase of MAPK during ischemic hypoxia and reoxygenation by analyses with PKC inhibitors, antisense DNA, a dominant negative kinase defective mutant, and constitutively active mutants of PKCzeta. Immunocytochemical observations show PKCzeta staining in the nucleus during ischemic hypoxia and reoxygenation when phosphorylated MAPK is also detected in the nucleus. This nuclear localization of PKCzeta is inhibited by treatment with wortmannin, a phosphoinositide 3-kinase inhibitor that also inhibits MAPK activation in a dose-dependent manner. This is supported by the inhibition of MAPK phosphorylation by another blocker of phosphoinositide 3-kinase, LY294002. An upstream kinase of MAPK, MEK1/2, is significantly phosphorylated 15 min after reoxygenation and observed mainly in the nucleus, whereas it is present in the cytoplasm in serum stimulation. The phosphorylation of MEK is blocked by PKC inhibitors and phosphoinositide 3-kinase inhibitors, as observed in the case of MAPK phosphorylation. These observations indicate that PKCzeta, which is activated by phosphoinositide 3-kinase, induces MAPK activation through MEK in the nucleus during reoxygenation after ischemic hypoxia.

activated by phosphoinositide 3-kinase (PI3K) participates in the activation of MAPK through MEK in the nucleus.
Cell Culture and Ischemic Hypoxia and Reoxygenation-An embryonic rat heart-derived cell line, H9c2 cells, was plated at a density of 5 ϫ 10 4 cells per dish in 100-mm culture dishes. After incubation in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal calf serum for 72 h, the cells were cultured in serum-free DMEM for 60 -72 h. Simulated ischemia was achieved as described previously (38). Briefly, the cells were incubated in slightly hypotonic Hanks' balanced saline solution (1.3 mM CaCl 2 , 5 mM KCl, 0.3 mM KH 2 PO 4 , 0.5 mM MgCl 2 , 0.4 mM MgSO 4 , 69 mM NaCl, 4 mM NaHCO 3 , and 0.3 mM Na 2 HPO 4 ) without glucose or serum for 2 h at 37°C. Hypoxia was achieved using an air tight incubator from which oxygen was removed by replacement with nitrogen. The oxygen concentration in the incubates was adjusted to 1%. After incubation under the conditions of ischemic hypoxia, the cells were incubated in DMEM without serum under normoxic conditions (20% O 2 , 5% CO 2 ) at 37°C for the indicated times.
Electrophoresis and Immunoblotting-Cells were washed with cold phosphate-buffered saline (PBS) and lysed with lysis buffer (1% Triton X-100, 20 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 2.5 mM sodium pyrophosphate, 1 mM ␤-glycerol phosphate, 1 mM sodium orthovanadate, 1 g/ml leupeptin, 1 mM phenylmethanesulfonyl fluoride). The cellular extracts and molecular mass standards were electrophoresed in 10% (w/v) polyacrylamide gels in the presence of SDS and transferred to polyvinylidene difluroide membranes (0.45 m, Millipore Co., Bedford, MA) in the case of phospho-MAPK and phospho-Elk-1, or nitrocellulose membranes in the case of other proteins. The blots were blocked with 5% non-fat dry milk in Tris-buffered saline containing 0.05% (w/v) Tween 20, and incubated with antibody. The blots were then washed, and the antigens were visualized by enhanced chemiluminescent detection reagents.
Immunofluorescent Staining-H9c2 cells were seeded in a Chamber Slide at a density of 3 ϫ 10 4 cells per well. The cells were subjected to ischemic hypoxia for 2 h, and then fixed with acetone/methanol (50/50) for 3 min at Ϫ20°C. Following fixation, the cells were blocked with 10% fetal bovine serum in PBS for 1 h. The cells were then incubated for 24 h at 4°C with antibody against PKC or/and phospho-MAPK, or phospho-MEK1/2 at 1:100 dilution in PBS containing 3% bovine serum albumin, washed with PBS, and incubated for an additional hour at room temperature with Cy3-conjugated anti-mouse IgG or/and Cy2-conjugated anti-rabbit IgG at 1:800 dilution in PBS containing 3% bovine serum albumin. The cells were viewed with a fluorescent microscope (Axioplan 2, Carl Zeiss Co., Ltd., Heidelberg, Germany).
Transfection of DNAs into H9c2 Cells-Oligonucleotide transfection of PKC (39) and PKC␣ (40) was determined as described previously (41). Briefly, phosphorothioate oligonucleotides with the sequence 5Ј-GTCGGTCCTGCTGGGCAT-3Ј for PKC and 5Ј-CAGCCATGGTTC-CCCCCAAC-3Ј for PKC␣ were synthesized as antisense DNAs. Control phosphorothioate oligonucleotides were also synthesized with the following sequences: 5Ј-ATGCCCAGCAGGACCGAC-3Ј for PKC (sense), 5Ј-GTCGGTACCAAGGGGGGT-3Ј for PKC␣ (sense). The cells (typically 80% confluent in 24-well dishes) were washed three times with PBS. Appropriate dilutions of oligonucleotides in 200 l of serum-free DMEM including liposomes (Tfx-50, Promega Co., Madison, WI) were preincu-bated at room temperature for 15 min. The cells were incubated for 1 h at 37°C in the presence of 5% CO 2 . At the end of the incubation period, 1 ml of medium containing 10% fetal bovine serum was added. The cells were incubated for 12 h, and then incubated in serum-free medium for 60 -72 h. For the transfection of expression vectors, subconfluent cell cultures were transfected with the indicated plasmids using Tfx-50 (Promega) as described above. To obtain H9c2 cells that stably expressed PKC mutants, clones of stable transfectants were isolated by their ability to grow in the presence of G418 (750 g/ml, Promega). Untransfected cells could not proliferate up to 7 days in the presence of G418 (750 g/ml). The expression of PKC was confirmed by immunoblotting using anti-PKC antibody. After incubation under the conditions of ischemic hypoxia and reoxygenation, the cells were used for biochemical assay.
Other Assays-MAPK activity was measured by immunoprecipitation kinase assay using Elk-1 as a substrate as described previously (34,41).

RESULTS
PKC Is Involved in MAPK Activation during Ischemic Hypoxia and Reoxygenation-We used the cells incubated for 60 -72 h in serum-free DMEM since MAPK activity was detected for up to 48 h in untreated cells after the removal of serum. We first examined the time course of MAPK activation during ischemic hypoxia and reoxygenation by immunoblotting using an anti-phospho-MAPK antibody that recognizes the tyrosine phosphorylation of MAPK (Tyr 204 ) necessary for activation. MAPK phosphorylation was increased approximately 10fold after 15 min of reoxygenation following ischemic hypoxia (Fig. 1, A and C), and p42 MAPK was preferentially activated as compared with p44 MAPK . The amount of MAPK remained almost the same during ischemic hypoxia and reoxygenation (Fig. 1B). To confirm the amount of MAPK protein, two MAPK antibodies (K-23, antibody for p42 MAPK and p44 MAPK , and C-16, antibody for p42 MAPK ) were used. These antibodies produced essentially similar results in all experiments (data not shown). MAPK phosphorylation was consistent with MAPK activity measured by immunoprecipitation kinase assay using Elk-1 as a substrate (Fig. 1D). To identify the upstream kinase(s) for MAPK activation, we examined the phosphorylation of MAPK in the presence of various inhibitors after 15 min of reoxygenation following 2 h of ischemic hypoxia. As shown in Fig. 2, A-C, both GF109203X (3 M) and chelerythrine (5 M), inhibitors of the ATP-binding site of PKC (42,43), blocked MAPK phosphorylation to the control level, while calphostin C (1 M), an inhibitor that interacts with the phorbol ester-binding site of PKC (44), suppressed MAPK phosphorylation only weakly. Phorbol 12-myristate 13-acetate (20 nM) significantly increased the phosphorylation of MAPK and MEK1/2 at 15 min after stimulation; this phosphorylation was largely inhibited by 1 M calphostin C (data not shown). The results indicate that calphostin C works as an effective inhibitor of phorbol 12myristate 13-acetate-sensitive PKC. These observations suggest that an atypical PKC isoform lacking a phorbol esterbinding site, may be involved in MAPK activation during reoxygenation. Since PKC, among PKC isoforms expressed mainly in the heart, has no phorbol ester-binding site we transfected an antisense oligonucleotide against PKC into cells using a liposome method. The antisense DNA inhibited PKC expression by 75% compared with the sense DNA (data not shown). These DNAs had no effect on the expression of other PKC isoforms (data not shown). The expression of PKC isoforms was almost the same in three independent experiments. Another member of the atypical PKC isoform, PKC/, was not detected by immunoblotting using the anti-PKC antibody (for the amino acid sequence of human PKC) under the conditions used in this study. Cells in which the PKC protein was depleted by the antisense DNA showed a significant inhibition of MAPK phosphorylation after 15 min of reoxygenation, while sense DNA-treated cells showed significantly more MAPK phosphorylation the same amount as in untreated cells (Fig. 3, A-C). The antisense DNA for PKC␣ caused a significant inhibition of the expression of PKC␣, whereas it had no effect on the expression of PKC. Treatment with an antisense DNA for PKC␣ resulted in the induction of MAPK phosphorylation after 15 min of reoxygenation to almost the same extent as in the presence of the sense DNA for PKC␣. The expression of neither PKC␣ nor PKC was affected by the sense DNA for PKC␣ (data not shown). The results suggest that the decrease in the phosphorylation level of MAPK is not generally caused by treatment with antisense DNA. To confirm the activation of MAPK by PKC, we transfected a kinase-deficient dominant negative (DN) mutant of PKC, in which a critical lysine in the ATPbinding site is replaced by tryptophan, into H9c2 cells. Transfection with DN mutant of PKC led to the overexpression of the PKC mutant (data not shown) as detected by immunoblotting using anti-PKC antibody. In agreement with the data obtained with the PKC antisense DNA, the increased phosphorylation of MAPK observed after 15 min of reoxygenation following ischemic hypoxia was strongly inhibited in cells expressing the DN mutant of PKC K275W (Fig. 4). We then used a constitutively active (CA) mutant of PKC, in which an alanine in the pseudosubstrate domain of PKC is replaced by glutamate. The transfection of the CA PKC mutant into cells increased the expression of the PKC mutant (data not shown). The expression of the CA PKC mutant in cells led to a significant phosphorylation of MAPK after reoxygenation as compared with the mock control (Fig. 5). The biochemical and biological properties of the mutants have been described previously by Ü berall et al. (37). MAPK phosphorylation by 10% fetal calf serum was significantly increased in the mock control, an increase similar to that in cells expressing DN and CA PKC (data not shown). These findings show that PKC mainly participates in the activation of MAPK during reoxygenation after ischemic hypoxia. In unstimulated cells, the expression of CA PKC mutants did not affect MAPK activity, suggesting that the localization of PKC may play a role in MAPK activity.
Nuclear Localization and Activation of PKC/MAPK Pathway during Ischemic Hypoxia and Reoxygenation-We examined the localization of PKC and phosphorylated MAPK during ischemic hypoxia and reoxygenation by immunocytochemical analysis. PKC translocated from the cytoplasm to the nucleus during 2 h of ischemic hypoxia, and was present predominantly in the nucleus after 15 min reoxygenation (Fig. 6, A-C), when phosphorylated MAPK was also detected in the nucleus (Fig. 6F). The nuclear localization of PKC was inhibited by 100 nM wortmannin, an inhibitor of PI3K (Fig. 6, D and E). Consistent with the localization of PKC in the cytoplasm, wortmannin also blocked the phosphorylation of MAPK in a dose-dependent manner (1-100 nM) (Fig. 7, A-C). No PKC staining was observed by a secondary antibody alone (data not shown). Phosphorylated MAPK staining could not be detected in untreated cells or wortmannin-treated cells (data not shown). The expression of PKC as detected by immunoblotting in wortmannin-treated cells was similar to that in untreated cells using anti-PKC antibody (data not shown). Another inhibitor of PI3K, LY294002 (30 M), also largely inhibited MAPK phosphorylation 15 min after reoxygenation (Fig. 7, D-F). These findings suggest that PKC activated by PI3K translocates to the nucleus, which leads to the activation of nuclear MAPK. We then examined the activation and localization of MEKs, upstream kinases of MAPK, using anti-phospho-MEK1/2 antibody that recognizes the phosphorylation site (Ser 217 /Ser 222 ) necessary for the activation. The phosphorylation of MEK1/2 increased significantly with a peak at 15 min after reoxygenation (Fig. 8, A-D), and a time course similar to that for MAPK phosphorylation. The phosphorylation was inhibited by PI3K inhibitors, wortmannin (Fig. 8, E-H) and LY294002 (Fig. 8, I-L), and PKC inhibitors, chelerythrine and GF109203X (data not shown). A MEK inhibitor, PD98059 (50 M), blocked MAPK activation during reoxygenation in a dose-dependent manner, and had no effect on JNK phosphorylation (data not shown). These findings suggest that the PI3K/PKC pathway acts as an upstream kinase of MEK, which participates in MAPK activation during reoxygenation after ischemic hypoxia. Active phosphorylated MEK1/2 was detected mainly in the nucleus during reoxygenation (Fig. 9B), whereas it was retained in the cytoplasm during serum stimulation (Fig. 9C). Phosphorylated MEK1/2 was undetectable in unstimulated cells (Fig. 9A). These findings indicate that nuclear MEK activates MAPK during reoxygenation. DISCUSSION p42 MAPK was preferentially activated during reoxygenation after ischemic hypoxia as compared with p44 MAPK , implying that the activating pathway of MAPK may be different than that induced by serum or PMA stimulation since both p42 MAPK and p44 MAPK are activated by treatment with serum and PMA. The inhibition of PKC expression and the activity significantly suppresses the phosphorylation of MAPK during reoxygenation, demonstrating that PKC is involved in MAPK activation. In cells expressing the CA mutant of PKC, MAPK phosphorylation was undetectable in unstimulated cells. The activation of MAPK may require that PKC translocates from the cytosol to other organelle where the MAPK pathway is present during ischemic hypoxia. This activation pathway might exist in the nucleus based on the following reasons.

FIG. 4. MAPK phosphorylation during reoxygenation after ischemic hypoxia in H9c2 cells expressing a kinase-deficient dominant negative mutant of PKC, K275W.
H9c2 cells were pretreated with 1 g of DNA for 1 h in DMEM containing liposomes (Tfx-50), and cells containing the introduced vector were selected by G418 (750 g/ ml). Cell extracts (40 g of protein) were prepared from cells reoxygenated for 15 min after ischemic hypoxia for 2 h, and subjected to immunoblotting with anti-phospho-MAPK antibody (A) or anti-MAPK antibody (B). The figure shows representative immunoblots obtained from three independent experiments. The levels of MAPK phosphorylation (sum of phosphorylation levels of p42 MAPK and p44 MAPK ) were determined from the immunoblots by densitometric analysis (mean Ϯ S.E., n ϭ 3; *, p Ͻ 0.05 versus control; #, p Ͻ 0.05 versus ischemic hypoxia and reoxygenation) (C).
First, during ischemic hypoxia and reoxygenation, PKC localizes mainly in the nucleus where phosphorylated MAPK is also detected. Second, the nuclear localization of PKC is blocked during ischemic hypoxia and reoxygenation in wortmannintreated cells. Consistent with the localization of PKC in the cytoplasm, the phosphorylation of MAPK during reoxygenation is also suppressed by wortmannin. We also observed the nuclear translocation of MEK, an upstream kinase for MAPK, during ischemic hypoxia, 2 and reoxygenation for 15 min after ischemic hypoxia led to phosphorylation of MEK only in the nucleus although the active form of MEK remained in the cytoplasm for 15 min after serum stimulation. The phosphorylation of MEK during reoxygenation was blocked by PKC inhibitors and PI3K inhibitors as shown in this study. These observations indicate that nuclear PKC activates the MEK/MAPK pathway. It was initially thought that MEK is always localized in the cytoplasm, but Tolwinski et al. recently demonstrated that MEK also translocates to the nucleus in mammalian cells in a regulatable manner during mitosis (45). How does MEK uptake into the nucleus during ischemic hypoxia occur? One possibility is that a decrease in intracellular ATP by depletion of molecular oxygen and glucose during ischemic hypoxia may lead to the nuclear accumulation of MEK by inhibiting nuclear export via a nuclear export signal that is known to be dependent on intracellular ATP (46 -48). Another possibility is that the interaction of MEK to CRM1/exportin in the nuclear membrane (49 -52) may be inhibited by a protein kinase activated during ischemic hypoxia, since protein phosphorylation has been suggested to modulate a nuclear export signal by masking an adjacent protein sequence (53)(54)(55). The nuclear MEK/ MAPK pathway will be able to induce gene expression rapidly under pathophysiological conditions. PKC, working as an upstream kinase of MAPK during reoxygenation after ischemic hypoxia, is activated by PI3K in response to various stimuli including ischemic hypoxia as indicated in this study (20, 30, 56 -58). Joyal et al. (59) show that PI3K is activated by an intracellular Ca 2ϩ -dependent effector protein, calmodulin, and that the increase in Ca 2ϩ induced by ischemic hypoxia may be involved in the activation of PI3K.
Recently, it was reported that phosphatidylinositol 3,4,5triphosphate, a product of PI3K, directly activates PDK, which leads to the activation of PKC (60). The inhibition of PI3K activity can inhibit the increased cell survival induced by growth factors (61)(62)(63). One target of PI3K is c-Akt (also designated protein kinase B), which promotes survival through the phosphorylation of BAD (64). Recently, Burnet et al. (65) reported that c-Akt phosphorylates Forkhead transcription factor, which is involved in the promotion of cell survival in late phase. PI3K can also activate the MAPK pathway (66 -68), and the activated MAPK produces signals capable of improving survival in PC12 cells (69). Another target of PI3K, PKC, may also play a role in cell survival through MAPK during ischemia and reperfusion since PKC can activate the MAPK pathway via Ras-dependent (19) or -independent pathways (20). Consistent with this idea, the inhibitions of the PKC/MAPK pathway induces apoptoic cell death during reoxygenation. 3 Cardiomyocyte-specific proteins and functions were conserved in the H9c2 cells used in this study, and the pathway of apoptotic inhibition might also be observed in a primary cardiac cell system. In cardiac myocytes that have lost their mitogenic activity, the preservation of cell viability by inhibition of apoptotic cell death may be critical for the maintenance of normal cardiac function. Further investigations are required to eluci-  date the role of the PKC/MAPK pathway in pathogenic processes.