Early single cell bifurcation of pro- and antiapoptotic states during oxidative stress.

In a population of cells undergoing oxidative stress, an individual cell either succumbs to apoptotic cell death or maintains homeostasis and survives. Exposure of PC-12-D(2)R cells to 200 microm hydrogen peroxide (H(2)O(2)) induces apoptosis in about half of cells after 24 h. After 1-h exposure to 200 microm H(2)O(2), both antiapoptotic extracellular regulated kinase (ERK) phosphorylation and pro-apoptotic Ser-15-p53 phosphorylation are observed. Microarray and real-time PCR assays of gene expression after H(2)O(2) exposure identified several transcripts, including egr1, that are rapidly induced downstream of ERK. Single cell analysis of egr1 induction and of phospho-ERK and phospho-p53 formation revealed the presence of two distinct cellular programs. Whereas the proportion of cells activating ERK versus p53 at 1 h depended on H(2)O(2) concentration, individual cells showed exclusively either phospho-p53 formation or activation of ERK and egr1 induction. Exposure to H(2)O(2) for 1 h also elicited these two non-overlapping cellular responses in both dopaminergic SN4741 cells and differentiated postmitotic PC-12-D(2)R cells. Repressing p53 with pifithrin-alpha or small interfering RNA increased ERK phosphorylation by H(2)O(2), indicating that p53-dependent suppression of ERK activity may contribute to the bi-stable single cell responses observed. By 24 h, the subset of cells in which ERK activity was suppressed exhibit caspase 3 activation and the nuclear condensation characteristic of apoptosis. These studies suggest that the individual cell rapidly and stochastically processes the oxidative stress stimulus, leading to an all-or-none cytoprotective or pro-apoptotic signaling response.

Reactive oxygen species (ROS) 1 have been implicated in the pathophysiology of several human diseases, including atherosclerosis, pulmonary fibrosis, cancer, neurodegenerative disorders, and aging (1)(2)(3). Although the cytotoxic actions of ROS are well known, ROS are increasingly recognized as compo-nents of cellular signaling that modulate responses in both physiological and pathological conditions (4). For example, ROS are produced in muscle cells upon binding of ligands such as angiotensin II (5). In addition, ROS production has been documented in a number of cells stimulated with cytokines, including tumor necrosis factor-␣, transforming growth factor-␤, and interleukin-1 (6 -8), and with such growth factors as bovine fibroblast growth factor, nerve growth factor, plateletderived growth factor, and epidermal growth factor (9 -11). These observations suggest that ROS can damage various cell components or activate specific physiological signaling pathways, with the relative effects determined by ROS concentration.
Oxidative stress has been reported to activate seemingly contradictory signaling pathways, and the consequences of the response vary widely; the ultimate outcome is dependent on the balance between these stress-activated pathways (12,13). Among the main stress signaling pathways and/or central mediators activated in response to oxidant injury are the extracellular regulated kinase (ERK) (14 -16), c-jun amino-terminal kinase (JNK) (17)(18)(19), p38 mitogen-activated protein kinase (20) signaling cascades, the phosphoinositide 3-kinase/Akt pathway (21), the nuclear factor-B signaling system (22,23), p53 activation (24,25), and the heat shock response (26). In general, the heat shock response, ERK, phosphoinositide 3-kinase/Akt, and nuclear factor-B signaling pathways exert a pro-survival influence during oxidant injury, whereas activation of p53, JNK, and p38 are implicated in apoptosis (see review in Ref. 12).
ROS, including hydrogen peroxide (H 2 O 2 ), are natural byproducts generated by living organisms as a consequence of aerobic metabolism (27). The cellular toxicity of H 2 O 2 is associated with the rapid modification of cellular constituents, including the depletion of intracellular glutathione and ATP, a decrease in NAD ϩ level, an increase in free cytosolic Ca 2ϩ , and lipid peroxidation (28). H 2 O 2 also activates the opening of the mitochondrial permeability transition pore and the release of cytochrome c (29). In the cytoplasm, cytochrome c, in combination with Apaf-1, activates caspase-9 leading to the activation of caspase-3 and subsequent apoptosis (30). The initiating events leading to activation of these different signaling pathways in response to H 2 O 2 are incompletely understood.
We have reported recently that H 2 O 2 induces apoptosis in PC-12-D 2 R cells and in the nigral dopaminergic neuronal cell line SN4741 in a concentration-dependent manner (31,32). These observations suggest that when exposed to a level of oxidative stress that can induce apoptosis in a portion of cells, each individual cell must proceed through a decision-making process that ultimately results either in its survival or its death. We report here that early after H 2 O 2 exposure, each cell activates either homeostatic or proapoptotic signaling pathways, but not both. Our results indicate that it may be chal-lenging to develop models of the mechanism of oxidative stressinduced cell death based solely on cell population and biochemical assays.
Cell Culture-PC-12-D 2 R (31, 32) cells were maintained in Dulbecco's modified Eagle's medium supplemented with 500 g/ml G418 (Invitrogen), 10% horse and 5% fetal bovine serum (Invitrogen) in a humidified atmosphere containing 5% CO 2 at 37°C. For differentiation, PC-12-D 2 R cells were plated on to collagen-coated plates in Dulbecco's modified Eagle's medium containing 10% horse serum and 5% fetal bovine serum and allowed to attach overnight. The cells were then induced to differentiate by growing in Dulbecco's modified Eagle's medium supplemented with 0.5% fetal bovine serum and 100 ng/ml nerve growth factor for 7 days. Substantia nigra dopaminergic neuronal cell line SN4741was cultured as described previously (33).
Immunoblotting-Cells were washed twice with ice-cold phosphatebuffered saline and lysed in 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Igepal C630, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovandate, 5 g/ml aprotinin, and mixture of protease inhibitors (Roche Diagnostics GmbH) at 4°C for 20 min. After centrifugation at 14,000 ϫ g for 20 min at 4°C, equal amounts of proteins were resolved by SDS-polyacrylamide gel electrophoresis. The proteins were electrotransferred to nitrocellulose membranes and detected by immunoblotting using the ECL system according to the manufacturer's recommendations. The blots were then stripped in buffer containing 62.5 mM Tris-HCl, pH 6.8, 2% SDS, and 100 mM ␤-mercaptoethanol for 30 min at 50°C and re-probed with respective antibodies.
Gene Expression Analysis-PC-12-D 2 R cells were treated with 200 M H 2 O 2 or vehicle for 1 h, and total RNA was isolated using StrataPrep total RNA miniprep kit according to the manufacturer's protocol. Preparation of cRNA, hybridization, and scanning of the rat genome U34 arrays were performed as described previously (34,35). Affymetrix microarray suite 5.0 was used to analyze the raw data using the criteria of 60% concordance across multiple array comparisons and -fold changes Ն1.6 for outlier detection. For each of the up-regulated targets, quantitative real-time polymerase chain reaction (qPCR) was carried out in an ABI Prism 7900HT (Applied Biosystems, Foster City, CA) using SYBR-Green assay, as described previously (35,36). All the gene-specific primer sets used for qPCR are listed in supplemental table. The authenticity of the PCR products was verified by melting curve analysis and agarose gel electrophoresis. The comparative cycle threshold (C T ) method was used to analyze the data by generating relative values of the amount of target cDNA. Relative quantitation for any given gene, expressed as -fold variation over control (untreated cells), was calculated after determination of the difference between C T of the given gene A and that of the calibrator gene B (␤-actin) in treated cells (⌬C T1 ϭ C T1A Ϫ C TB ) and untreated cells (⌬C T0 ϭ C T0A Ϫ C TB ) using the 2 Ϫ⌬⌬CT (1-0) formula (37). C T values are means of triplicate measurements. Experiments were repeated three to five times. Immunocytochemistry-PC-12-D 2 R cells growing on collagen-coated or SN4741 cells growing on ploy-ornithine-coated cover glass were treated as indicated. The cells were fixed and permeabilized as described previously (31,32), and immunocytochemical staining for phospho-ERK, phospho-p53, or active caspase-3 was carried out. Anti-phospho-ERK (1:400), anti-phospho-p53 (1:500), or active caspase-3 antibody (1:200) was added and incubated overnight at 4°C. For double-immunolabeling, a mouse monoclonal anti-phospho-ERK antibody was used. After washing, cells were incubated with corresponding sec-ondary antibodies for 2 h at room temperature. The cells were washed three times in phosphate-buffered saline, and the nuclei were stained with 1 g/ml (in phosphate-buffered saline) of the fluorescent DNA dye 4Ј,6-diamidino-2-phenylindole dihydrochloride (DAPI) for 10 min and then washed with phosphate-buffered saline. The liquid was drained and the slides were mounted in Vectashield (Vector Laboratories, Burlingame, CA) mounting medium.
Fluorescent in Situ Hybridization-Egr1 (161 bp) and ␤-actin (150 bp) DNA fragments were amplified from RNA isolated from PC-12-D 2 R cells by reverse transcription PCR using primer sets described in supplemental table. The DNA fragments were purified from agarose gel using QIAEX II gel extraction kit and sub-cloned into pDrive vector. Individual clones were sequenced to determine the orientation of the DNA. To generate cRNA probes, we used T7 promoter for antisense and SP6 promoter for sense probes to minimize the interference of vector sequences in double fluorescent in situ hybridization (FISH). Aminoallyl-UTP incorporated cRNA probes were generated using Maxiscript in vitro transcription kit. The yield and integrity of riboprobes was confirmed by gel electrophoresis. The egr1 and ␤-actin cRNA probes were labeled with Alexa Fluor 488 and CY3, respectively, according to manufacturer's protocols. The riboprobes were purified on atlas nucleospin columns. The cells grown on cover slips were fixed and permeabilized as described previously (31,32). After prehybridization in 5ϫ SSC, 50% formamide, and 1 mg/ml tRNA at room temperature for 30 min, the denatured probe was added to the prehybridization buffer. Hybridization was carried out for 2 h at 52°C. After two 5-min washes in 5ϫ SSC, 50% formamide, 0.1% SDS, and twice in 2ϫ SSC, the nuclei were stained with DAPI, and slides were mounted in Vectashield mounting medium. When immunostaining was carried out after FISH, the slides were incubated with respective antibodies as described above.
RNA Interference-Custom SMARTpool plus small interfering RNA (siRNA) to target rat p53 (GenBank TM accession number NM_030989) was designed and synthesized by Dharmacon (Lafayette, CO). siRNA (50 pmol) was transfected into PC-12-D 2 R cells using transit-TKO transfection reagent (Mirus, Madison, WI) as described previously (32). After 48 h of transfection, cells were treated with H 2 O 2 (200 M) or vehicle for 1 h, and total RNA or cell extract was prepared. A nonspecific RNA duplex (Dharmacon) was used in control experiments.

ERK and p53
Pathways Are Activated in Response to Oxidative Stress-PC-12-D 2 R cells and substantia nigra dopaminergic SN4741 cells undergo apoptosis when exposed to H 2 O 2 in a concentration-and time-dependent manner (31,38). Oxidative stress is known to activate multiple signal transduction pathways in many experimental systems (12). To identify the signaling mechanisms activated by PC-12-D 2 R cells in response to H 2 O 2 , we have assessed the activation of ERK, JNK, p38kinase, and p53, using Western blot analysis with antisera against phospho-ERK, phospho-JNK, phospho-p38 kinase, and phospho-p53. We found that 200 M H 2 O 2 rapidly induced the phosphorylation of ERK but not of JNK or p38 kinases in PC-12-D 2 R cells (Fig. 1A, B, and C). The activation of ERK by H 2 O 2 was rapid and sustained ( Fig. 1A, top). Anti-phospho-Ser15 antibody was used to detect oxidative stress-induced phosphorylation of p53, presumably caused by DNA damage (39 -41) after exposure to H 2 O 2 for periods up to 6 h. In PC-12-D 2 R cells, p53 phosphorylation was significantly enhanced within 30 min after H 2 O 2 exposure and continued to increase for up to 2 h (Fig. 1D). The level of total p53 protein on Western blot analysis was unchanged after 6 h of incubation with H 2 O 2 (Fig. 1D). These results demonstrate the early activation of ERK and p53 signaling pathways in response to oxidative stress in PC-12-D 2 R cells.
Characterization of ERK-activated Gene Program in Response to Oxidative Stress-ROS, through its effects on cell signaling, alters the expression of specific genes (42). To identify the genomic response during oxidative stress, the gene expression profile associated with H 2 O 2 exposure was studied using oligonucleotide microarrays and regulated transcripts were confirmed by qPCR. As shown in Table I, genes that encode transcription factors including egr1, c-fos, c-jun, pc3, and a zinc finger protein (copeb) were up-regulated after 200 M H 2 O 2 for 1 h. Other up-regulated genes include inner mitochondrial membrane component ATP synthase subunit c, stress response gene 70-kDa heat shock protein (hsp70), and the immediate-early inducible small GTP binding protein rhoB. We also found that the mitogen-activated protein kinase phosphatase-1 (mkp1) was increased by Ͼ3-fold in cells treated with H 2 O 2 . Several immediate early genes, such as egr1, c-fos, and c-jun, have been reported to be transcriptionally activated by increased cellular oxidation (16,43,44). Oxidative stress has also been shown to induce rhoB (45) and hsp70 (26).
To identify the component of the oxidative stress-induced gene program downstream of ERK activation, we inhibited ERK with PD98059. Addition of PD98059 (100 M) 1 h before H 2 O 2 treatment decreased the induction of egr1, pc3, and mkp1 (Fig. 2), genes that have been identified as downstream of ERK in other experimental systems (46 -48). PD98059 did not prevent the induction of c-fos, copeb, c-jun, hsp70, rhoB, ATP synthase subunit c, and expressed sequence tag (GenBank TM accession number AI639167).
Characterization of egr1 Induction in Response to Oxidative Stress-To investigate the cellular segregation of the diverse responses to oxidative stress, we next studied the induction of egr1 mRNA using FISH. We exposed cells to 200 M H 2 O 2 , a concentration that induces apoptosis in approximately half of the cells (31). When cells were exposed to 200 M H 2 O 2 , we observed that approximately half of the cells had strong fluorescent signals (Fig. 3). The signal in the subset of cells not showing egr1 induction at this concentration of H 2 O 2 was in-distinguishable from that of control cells. Cells hybridized with sense-oriented probe for egr1 showed no fluorescent signal in any cells (data not shown). Simultaneous double-FISH for egr1 and ␤-actin mRNA showed no change in ␤-actin mRNA expression during H 2 O 2 treatment (Fig. 3). These data demonstrate that H 2 O 2 induces a high level of egr1 induction in a subset of cells and no detectable change in egr1 levels in others.
ERK-mediated egr1 Induction and p53 Activation Were Present in Discrete Subsets of Cellular Populations-In many experimental systems, signaling through ERK is known to be prosurvival (16,30,49,50) and p53 activation is known to be proapoptotic (25,(51)(52)(53). The p53 protein plays a central role in the cellular response to DNA damage that leads to phosphorylation and activation of p53 (53)(54)(55). To characterize the cellular segregation of signaling pathways simultaneously activated by ROS, we monitored the phosphorylation of ERK and p53 signaling pathways after H 2 O 2 exposure using immunocytochemical staining of PC-12-D 2 R cells. In response to H 2 O 2 , enhanced ERK phosphorylation was detected both in the cytoplasm and in the nucleus (Fig. 4A, top). However, phospho-p53 was mainly localized in the nucleus (Fig. 4A, bottom). Approximately half of the cells showed ERK or p53 phosphorylation in response to 200 M H 2 O 2 (Fig. 4B).
We next used double labeling to study whether there was overlap of p53 and pERK/egr1 induction within the same cells. As shown in Fig. 5A, egr1 mRNA co-localized to cells showing ERK activation. However, the induction of egr1 mRNA by H 2 O 2 was absent in cells showing p53 activation (Fig. 5B). These results indicate that oxidative stress activates ERK or p53 signaling pathways in separate cell subpopulations.
To explore whether the ROS-mediated segregation of signaling pathways in separate cell subpopulations observed in PC-12-D 2 R cells was present in a different cellular context, we studied this signaling pathways in a mouse immortalized nigral dopaminergic cell line SN4741 (33). Incubation of these cells with H 2 O 2 was found to induce cell death in a concentrationdependent manner (31,38). SN4741 cells undergo cell death at low concentrations of H 2 O 2 (50 -100 M) compared with PC-12-D 2 R cells. We found that H 2 O 2 induced phosphorylation of both ERK and p53 (data not shown). At the single cell level, 50 M H 2 O 2 phosphorylated ERK and p53 in separate populations of cells (Fig. 7B). However, 100 M H 2 O 2 activated p53 in almost all the cells (Fig. 7C). These results suggest that ROS activates opposing signaling pathways in discrete cell sub-populations of dopaminergic neuronal cell line.   (Fig. 9A). In an attempt to elucidate the fate of cells showing either ERK or p53 activation, we examined activation of caspase-3, a mediator of cell death (31). When examined after 24 h of H 2 O 2 treatment, condensed nuclei and active caspase-3, which are hallmarks of apoptosis, were found exclusively in phospho-ERK-negative cells (Fig. 9B). These data indicate that by 1 h, the cells have segregated into two populations and suggest that those that activate p53, a signaling molecule upstream of caspase-3, will proceed to apoptosis and those that activate ERK/egr1 will maintain homeostasis.

Activation of ERK and p53 in Postmitotic PC-12-D 2 R Cells in Response to H 2 O 2 -We
Activation of p53 in Response to Oxidative Stress Downregulates ERK-Our results demonstrated that H 2 O 2 activates both ERK and p53 in PC-12-D 2 R cells. To elucidate the mechanism underlying these signaling pathways, we studied the cross-talk between ERK and p53 signaling mechanisms. The phosphorylation of ERK in response to H 2 O 2 was blocked by using the selective inhibitor of ERK, PD98059 (56). In Western immunoblots using phospho-ERK and phospho-p53 antibodies, it was determined that addition of PD98059 (100 M) 1 h before H 2 O 2 treatment prevented the phosphorylation of ERK (Fig.  10A). However, PD98059 did not affect the phosphorylation of p53 in response to H 2 O 2 (Fig. 10A).
To examine the effect of p53 on the activation of ERK, we  Double-label FISH study for egr1 mRNA (green) and ␤-actin mRNA (red) of PC-12-D 2 R cells exposed to vehicle or 200 M H 2 O 2 for 1 h. The nuclei were counterstained with DAPI (blue). Note that egr1 is strongly induced in a subpopulation of cells. The field shown is representative, and these results were replicated in four independent experiments. used the p53 inhibitor pifithrin-␣, which blocks p53 transcriptional activation and subsequent apoptosis (57). In Western immunoblots using phospho-ERK and phospho-p53 antibodies, we have found that pifithrin-␣ augmented the activation of ERK in presence of H 2 O 2 (Fig. 10B). These results suggest that H 2 O 2 -induced ERK phosphorylation is negatively regulated by activation of p53. However, we found that 40 M pifithrin-␣ did not inhibit the phosphorylation of p53 in response to H 2 O 2 .
To confirm the regulation of ERK by p53, we reduced the levels of p53 expression in PC-12-D 2 R cells using RNA interference. After transfection with p53-specific or control siRNA, cultures were assessed for p53 mRNA expression by qPCR. As shown in Fig. 10C, p53 expression was substantially repressed (ϳ4-fold) by 48 h after transfection. The involvement of p53 in oxidative stress-induced ERK activation was studied in PC-12-D 2 R cells transfected with p53 or control siRNA. After 48 h of transfection, the cells were incubated with H 2 O 2 and assessed for the phosphorylation of ERK. In control siRNA-transfected cells, H 2 O 2 activated ERK similar to that observed in cells not transfected with siRNA (Fig. 10D). However, p53 repression by siRNA augmented the ERK phosphorylation (Fig. 10D). These results suggest that cells that activate p53 in response to oxidative stress suppress the activation of ERK.

DISCUSSION
In this study, we demonstrate that within the first hour of cells' exposure to oxidative injury, they activate specific signaling pathways that indicate whether the cells will ultimately succumb to or tolerate the insult. ERK activation by oxidative stress marks cells that have chosen to maintain homeostasis. In contrast, cells that activate p53 proceed to cell death. Our data indicate that ROS-mediated anti-and pro-apoptotic signaling events are triggered in each cell early after exposure to oxidative stress. These responses are sustained and mutually exclusive. These early, non-overlapping single cell responses are observed in both proliferating and differentiated PC-12-D 2 R cells and in immortalized dopaminergic SN4741 neurons.
We find that activation of ERK and the induction of egr1 within 1 h mark cells destined to survive after the initial oxidative insult. In PC-12 cells, ERK is mainly activated by growth factors and has been shown to be associated with cell proliferation, differentiation, and promotion of cell survival (13,16,58,59). H 2 O 2 increased the mRNA expression of the ERK-dependent genes egr1, c-jun, c-fos, and mkp1. These genes are known to be transcriptionally activated by increased cellular oxidation (16,43,44) and by nerve growth factor (46,60). Evidence for an antiapoptotic role for ERK has been reported in PC-12 cells after growth factor withdrawal (13,59) and exposure to oxidative stress (16). ERK has also been reported to function as a suppressor of ROS in superior cervical ganglion neurons (61). Activation of the ERK via the Ras/Raf/MEK pathway has further been shown to support survival of neurons in the nervous system (49,50).
In contrast to the ERK response, we find that the early activation of p53 by H 2 O 2 predicts the later induction of capsase-3 and apoptosis. The inability of a sublethal dose of H 2 O 2 to activate p53 supports the involvement of p53 in apoptosis in these cells. The p53 tumor suppressor protein has been proposed as a key mediator of stress responses because it plays an essential role in the death of many cell types, including neurons (for review, see Ref. 62). Exposure to ROS can cause nuclear DNA double-stranded breaks that are detected by enzymes from the phosphoinositide 3-kinase family (63), resulting in phosphorylation of serine 15 of p53 and its consequent stabilization and accumulation (40,41,54). It has been suggested that modification of this serine regulates p53 stability by altering Mdm2-p53 interactions (40). Activation of p53 results in the up-regulation of proteins implicated in apoptosis (such as proapoptotic BAX and caspase-3) in many experimental systems (25,51,52,64).
We demonstrate that activation of the ERK signaling pathway in response to ROS has no effect on the phosphorylation of p53, whereas p53 inhibition leads to ERK activation. It has been reported previously that ERK activation in response to cisplatin in ovarian cancer cells can phosphorylate p53 in vitro (65). In our experimental system, the pharmacological inhibition of ERK did not affect phosphorylation p53 in response to H 2 O 2 . In contrast, we found that inhibition of p53 by pifithrin-␣ or repression of p53 by siRNA augmented the activation of ERK by H 2 O 2 . These results suggest the existence of a negative signaling cross-talk pathway from p53 to ERK. This ERK suppression pathway most likely contributes to the structure of a signaling network switch that forces the cell to rapidly select among these two mutually exclusively patterns of response to oxidative stress.
We find that the recruitment of cells to the response state marked by phospho-p53 increases as the concentration of H 2 O 2 increases and that the early response bifurcation is observed equally in dividing and NGF-differentiated cells. These findings suggest that the initial conditions of the cells that present opposite responses to oxidative stress are likely to be similar. For the individual cell, the choice between these two competing and mutually exclusive response states is stochastic. This model, we suggest (Fig. 11), is analogous to the random and exclusive divergence of initially pluripotential cells during development. We propose that the divergent outcome results not from initial differences in the state of the cells but from random selection forced by the design of the oxidative stress signaling circuits. This distinction is important in developing therapeutic strategies to intervene in this process. Our results and interpretation suggest that attempts to identify differences between the subset of cells that survive and those that succumb may be fruitless. On the other hand, further elucidating the structure of the signaling network switch responsible for this bi-stability and determining when the response state becomes unalterable are likely to help in devising rationale strategies to improve the odds for survival of an individual cell.
It has been hypothesized that ROS activates contradictory signaling pathways and that the dynamic balance between these pathways may be important in determining whether a cell survives or undergoes apoptosis (for review, see Ref. 12). Many studies using biochemical assays of cell homogenates have identified the concomitant activation of both pro-and antiapoptotic responses that have served as the basis for models of the underlying mechanisms (e.g. see Refs. 13,[66][67][68][69]. We find that in PC-12-D 2 R and SN4741 cells, even at early time points, that divergent responses do not coexist within the same cells. Activation of effector caspases is considered the final step in many apoptosis pathways. Consistent with our findings, it has been reported that caspase activation during apoptosis occurs in an all-or-none fashion (70). Studies using biochemical assays that monitor the average cellular response may obscure the actual decision-making signaling mechanisms by detecting simultaneous responses that are in fact segregated within different cell subpopulations. Our data suggest that "to die or not to die" is a question resolved quickly by each individual cell. Each individual cell rapidly responds to stress and achieves a coherent physiological state directed toward either apoptosis or survival.  A and B, the blots were stripped and reprobed using the antibodies recognizing total ERK or p53 proteins. p-ERK, phospho-ERK; p-p53; phospho-p53. C, repression of p53 mRNA by p53-siRNA. The cells were transfected with control or p53-siRNA, RNA was isolated after 48 h, and p53 mRNA levels were determined by qPCR (graph) or by reverse transcription PCR for 30 cycles and analyzed on agarose gel electrophoresis (bottom). ␤-Actin from the same samples was amplified as control. Lane 1, control; lane 2, control siRNA; lane 3, p53 siRNA. D, repression of p53 augmented the phosphorylation of ERK in response to H 2 O 2 . 48 h after transfection with control or p53-siRNA, cells were untreated or treated with H 2 O 2 (200 M) for 1 h and cell extracts were prepared. Equal amounts of protein were subjected to immunoblot analysis using phospho-ERK or p53 antibodies. The experiments were repeated three times with similar results.