Mouse astrocytes deficient in the mitochondrial form of manganese superoxide dismutase (SOD2) do not survive in culture under atmospheric air with 20% oxygen (O2), which is a common condition for cell cultures. Seeding the cells and maintaining them under mild hypoxic conditions (5% O2) circumvents this problem and allows the cells to grow and become confluent. Previous studies from our laboratory showed that this adaptation of the cells was not attributable to compensation by other enzymes of the antioxidant defense system. We hypothesized that transcriptional activity and upregulation of genes other than those with an antioxidant function are involved. Our present study shows that c-Myc was significantly induced and that it inhibited p21 and induced proteins such as cyclin-dependent kinases, cyclin D, and cyclin E, which are involved in the cell cycle process, along with phosphorylation of the retinoblastoma protein and Cdc2 (cell division cycle 2). These mechanisms contribute to cell proliferation. Small interfering RNA of c-Myc, however, blocked proliferation of SOD2 homozygous (SOD2−/−) astrocytes under mild hypoxia consisting of 5% O2, whereas it did not affect the growth of wild-type astrocytes. Our results indicate that c-Myc plays a critical role in hypoxia-induced proliferation and survival of SOD2−/− astrocytes by overcoming injury caused by oxidative stress.
Superoxide radicals, along with other reactive oxygen species (ROS), are toxic to cells when generated in excess and can cause severe damage to DNA, proteins, and lipids, leading to cell death (Chan, 1996). They also affect the various molecular components of cells, including mitochondria. Mitochondrial dysfunction results in necrotic cell death via loss of ATP production. Several cellular and biochemical signaling pathways in mitochondria also activate the caspases and are thus involved in apoptosis (Chan, 2005). These processes are exacerbated in mice that are deficient in manganese superoxide dismutase (SOD2), suggesting that oxidative stress precedes mitochondrial-derived apoptosis. Previous work from our laboratory showed that mouse astrocytes completely deficient in the mitochondrial form of SOD2 did not grow and subsequently died in 20% atmospheric oxygen (O2). Seeding the cells under 5% O2 circumvented this problem and allowed the cells to grow and become confluent (Copin et al., 2001). The factors induced by mild hypoxia (5% O2) that overcome SOD2 deficiency and contribute to proliferation of homozygous SOD2−/− astrocytes are important and can explain the adaptive response of these astrocytes.
One of the key genes involved in hypoxia is the transcription factor hypoxia-inducible factor-1 (HIF-1) (Semenza, 1999). Carmeliet et al. (1998) showed that induction of HIF-1α activation prevented G1/S transition, suggesting that HIF-1α inhibits cell growth. Current theory on hypoxic signaling relies on ROS as mediators of hypoxia-induced gene transcription (Semenza, 1999; Liu et al., 2005). In addition to HIF-1α, c-Myc is widely known as a crucial regulator of proliferation of normal and neoplastic cells and is induced by ROS (Simon et al., 1998). Most importantly, HIF-1α functionally antagonizes c-Myc (Koshiji et al., 2004). Our study examined the roles that c-Myc and HIF-1α may play in the survival and proliferation of SOD2−/− astrocytes.
Cells that proliferate go through a cycle consisting of the G1, S, G2, and M phases. Cell proliferation is primarily regulated at the G1 phase of the cycle (Liu et al., 1998). The retinoblastoma protein (Rb) is essential for the G1/S checkpoint, and its phosphorylation by the cyclin D and E/cyclin-dependent kinase (CDK) complexes nullifies growth-suppressive activity. Cyclin/CDK complex activity is modulated by binding to CDK inhibitors such as p27 and p21, preventing them from inhibiting the function of cyclin E/CDK2 and allowing the cells to enter S phase.
Growing evidence indicates that hypoxia alters cellular proliferation by programmed cell death and by cell cycle arrest (Shimizu et al., 1995). Most nontransformed hypoxic cells remain viable but are arrested in G1 and do not proliferate (Graeber et al., 1996). Previous studies indicate that hypoxia-induced cell cycle arrest is accompanied by decreased activity of certain CDK complexes and hypophosphorylation of Rb, leading to inhibition of cell cycle progression (Gardner et al., 2001). In contrast, SOD2−/− cells have been shown to proliferate under 5% O2 hypoxia. In this study, we examined the role of cell cycle proteins in the survival of cultured astrocytes derived from SOD2 knock-out mice under 5% O2 hypoxia where, instead of cell cycle arrest, cell proliferation occurs.
Materials and Methods
The animals used in this study were mutated by a method of homologous recombination and subsequent inactivation of the SOD2 gene. These CD-1/SV129 mice were generated from our laboratory and were extensively bred on a CD-1 background for >10 generations (Li et al., 1995). Homozygous SOD2 knock-out animals died at ∼10 d after birth because of a heart abnormality. These SOD2−/− mice were produced by crossing two heterozygous SOD2−/+ mice. Immediately after birth, the animals were typed by PCR amplification of mutant and normal fragments of the SOD2 gene from tail DNA obtained by proteinase K digestion as described previously (Li et al., 1995).
Primary astrocyte culture.
Astrocytes were cultured as described by Chan et al. (1988) with minor modifications. Briefly, cortical tissue of newborn mice was dissociated by 0.25% trypsin and passed sequentially through 80 μm and then 10 μm filtration meshes. Cell suspensions in minimum essential medium (MEM; Invitrogen, Grand Island, NY) supplemented with 20% fetal bovine serum (Gemini Bio-Products, Calabasas, CA) were plated into 24 multiwell plates or 35 or 60 mm dishes (Falcon; BD Biosciences, Franklin Lakes, NJ) at a dilution of one brain per 48 cm2 and incubated at 37°C under 5% CO2–air atmosphere in a water-jacketed incubator (Forma Scientific, Marietta, OH) or under 5% O2–5% CO2–90% N2 in a gas-tight humidified chamber (modular incubator chamber; Billups-Rothenberg, Del Mar, CA). After 4 d, the culture medium was replaced with fresh MEM containing 10% fetal bovine serum, and the astrocytes were fed twice per week. c-Myc small-interfering RNA (siRNA) was administered when cells reached 50% confluence.
Cell death assay.
The 2,3′-azino-bis(ethylbenzothiazoline-6-sulfonic acid) (ABTS)/enzyme-linked immunosorbent assay method used for quantification of cell death is a fast, reliable, and objective procedure (Brooke et al., 1999). Cells were fixed by adding cold methanol. After blocking with 5% dry milk in PBS overnight, glial fibrillary acidic protein antibody was added at a 1:1000 dilution for 30 min. The cells were then washed and incubated in a secondary antibody for 30 min, followed by incubation with the ABC reagent (ABC kit; Vector Laboratories, Burlingame, CA) for 30 min, and then incubated in ABTS for 20–30 min. Colorimetric readings were taken at 405 nm using an ELISA plate reader.
In situ detection of superoxide anion production.
The early production of superoxide anions after hypoxia was investigated using hydroethidine (HEt) as in a previously described method (Bindokas et al., 1996). HEt is selectively oxidized to ethidium by superoxide anions, but not by other ROS such as hydrogen peroxide, hydroxyl radical, or peroxynitrite. HEt solution (1 mg/ml in MEM) was administered directly to the control group (cells that were at 20% O2) and to cells immediately after they were taken out of the 5% O2 hypoxic chamber. The cells were then incubated for 5 min in the solution. They were washed three times in PBS, and after fixation with 4% paraformaldehyde for 10 min, the slides were covered with VECTASHIELD mounting medium with 4′,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories). These sections were observed under a fluorescent microscope at an excitation of 355 nm and an emission of >415 nm for HEt.
Western blot analysis.
To obtain enough cells to represent “normal conditions” for Western blotting, we first cultured SOD2−/− mouse astrocytes in 5% O2 hypoxia for 5 d (when the cells were not confluent) and then continued with normal culture conditions for 3 d. The cells were rinsed with cold PBS, lysed for 15 min on ice in lysis buffer, sonicated, and centrifuged at 4°C. Supernatants were collected, and an equal volume 2× SDS sample buffer (Invitrogen, Carlsbad, CA) was added. Protein samples were boiled for 5 min and separated on 4–12% or 10–20% Tris-glycine Novex gels (Invitrogen) and transferred to polyvinylidene difluoride membranes and then blocked with 5% milk in PBS-Tween 20, followed by incubation with the respective primary antibody overnight at 4°C. The membranes were then washed and incubated in secondary antibody for 1 h and washed with PBS-Tween. The bound antibody was visualized using an ECL-Plus kit (Amersham Biosciences, Piscataway, NJ). The bands were scanned using a densitometer (GS-700; Bio-Rad, Hercules, CA), and quantification was performed using Multi-Analyst 1.0.2 software (Bio-Rad).
Immunoprecipitation was performed as described previously (Saito et al., 2004). Three hundred micrograms of protein from the cytosolic fraction were used for coimmunoprecipitation. The protein sample was incubated with a 50% slurry of protein G-Sepharose (Amersham Biosciences) for 1 h at 4°C, and this mixed sample was centrifuged at 12,000 × g for 1 min. The supernatant was incubated with 2 μg of monoclonal mouse anti-CDK4 antibody (Cell Signaling Technology, Beverly, MA) and 15 μl of protein G-Sepharose (50% slurry) for 1 h at 4°C. The negative control was prepared with protein G-Sepharose without an antibody. The 14,000 × g pellets were washed three times and used as the samples bound to each antibody. After adding the same volume of Tris-glycine SDS sample buffer (Invitrogen) to the samples, they were boiled to remove the Sepharose beads. After centrifugation at 14,000 × g for 1 min, the supernatant was immunoblotted with a 1:600 dilution of anti-cyclin D antibody (Cell Signaling Technology), as described for the Western blot method.
Astrocytes were cultured on glass coverslips. The cells were fixed with 4% paraformaldehyde. They were then incubated with blocking solution followed by incubation overnight with an anti-c-Myc antibody at a dilution of 1:100. Secondary goat anti-rabbit IgG antibody conjugated with Texas Red (Jackson ImmunoResearch, West Grove, PA) was added at a dilution of 1:200. The cells were then mounted with VECTASHIELD medium with DAPI (Vector Laboratories). Fluorescence of Texas Red was observed at excitation of 510 nm and emission of >580 nm. Fluorescence of DAPI was observed at excitation of 360 nm and emission of >460 nm using a laser confocal microscope (LSM 510; Zeiss, Thornwood, NY).
Transfection of c-Myc siRNA in astrocytes.
c-Myc or nontarget mutant (control) siRNA diluted in OPTI-MEM I was added to diluted siPORT Lipid (Ambion, Austin, TX) and incubated at room temperature for 20 min. Cells were washed with OPTI-MEM I, and a transfection agent/siRNA complex with a final concentration of 100 nm was added dropwise and incubated for 4 h. Fresh growth medium was added to each well and incubated for 48 h. Transfection efficiencies were determined by immunofluorescence of targeted cells and were in the range of 80%. Specific silencing of the targeted gene was confirmed by Western blot analysis of the knockdown protein c-Myc.
Treatment of astrocytes with antisense c-Myc.
The antisense oligonucleotide for c-Myc was 5′-CACGTTGAGGGGCATCGTCGC-3′ (Devi et al., 2005). The sense oligomer 5′-GCGACGATGCCCCTCAACGTG-3′ served as a control. A 50 or 100 nm concentration of the oligomers was added directly to the cells and placed under 5% O2. After treatment, cell death was assessed using a lactate dehydrogenase assay. Total RNA was prepared from the astrocytes using a Micro-to-Midi Total RNA Purification System (Invitrogen). For reverse transcriptase (RT)-PCR analysis, SuperScript One-Step RT-PCR kit with Platinum Taq (Invitrogen) was used with a total volume of 50 μl. Specific primer pairs for c-Myc were used as follows (Forte et al., 2005): forward, 5′-CTTCCCCTACCCGCTCAACGAC-3′; reverse, 5′-CACATCAATTTCTTCCTCATCA-3′. As a control, β-actin was amplified with the following primers: forward, 5′-ACCCACACTGTGCCCATCTAA-3′; reverse, 5′-GCCACAGGATTCCATTACCCAA-3′. The mixtures were subjected to RT-PCR on a thermal cycler (Mastercycler Gradient; Eppendorf, Westbury, NY). RT-PCR conditions for c-Myc were 45°C for 30 min, 94°C for 2 min, 30 cycles of 94°C for 15 s, 56°C for 30 s, 72°C for 15 s, and then 72°C for 5 min. RT-PCR products were electrophoresed on 1% agarose gels and stained with ethidium bromide. After electrophoresis, the gels were recorded with a gel documentation system (GS700; Bio-Rad).
Quantification and statistical analysis.
All data were obtained from three to five independent experiments and are expressed as mean ± SD, and for each experiment multiple measurements were made on three to six different dishes or wells. Comparisons among multiple groups were performed using a one-way ANOVA (Fisher’s protected least significant difference test) with appropriate post hoc tests (StatView, version 5.01; SAS Institute, Cary, NC), whereas comparisons between two groups were achieved using Student’s t test. p ≤ 0.05 was considered statistically significant.
Five percent oxygen greatly improved the survival of SOD2−/− astrocytes under normal culture conditions
Cell death was measured using a modification of an immunohistochemical staining method (ABTS) with an astrocyte-specific antibody against glial fibrillary acidic protein, which allows for quantification of astrocytes using an enzyme-linked immunosorbent assay plate reader. Under standard culture conditions, heterozygous SOD2−/+ and wild-type astrocytes had the same survivability, whereas homozygous SOD2−/− cells did not grow. Under 5% O2 hypoxic conditions, SOD2−/− astrocytes dramatically proliferated and reached confluence similar to normal astrocytes (Fig. 1A). The level of SOD2 was examined in the wild-type, SOD2−/+, and SOD2−/− astrocytes. There was no SOD2 expression in the SOD2−/− astrocytes, whereas 50% expression was observed in the SOD2−/+ compared with the wild-type astrocytes (Fig. 1A). Under 5% O2 hypoxia, ROS production as shown by HEt staining was significantly lower in the SOD2−/− astrocytes compared with normal culture conditions in atmospheric 20% O2 (Fig. 1B), whereas there was no significant difference in the wild-type astrocytes under such conditions (supplemental Fig. 1, available at www.jneurosci.org as supplemental material). HEt staining was punctate, which supports our previous report (Sugawara et al., 2002) describing double staining of ethidium and the mitochondrial marker MitoTracker, which demonstrated that many of the punctate superoxide signals were colocalized with mitochondria, confirming mitochondria as a major source of superoxide production.
Improvement in survival with 5% O2 hypoxia is not through expression of Bcl-2-family proteins
Bcl-2 is known to protect against enhanced oxidative stress caused by various ROS-inducing agents. To investigate whether they had a role in rescuing the SOD2−/− astrocytes, we examined several members of the Bcl-2 family of proteins. Western blot analysis demonstrated a basal level in Bcl-2 expression in the wild-type astrocytes. Five percent O2 hypoxia did not have any influence on this expression. However, there was an increase in Bcl-2 in the SOD2−/+ astrocytes compared with the wild-type astrocytes under both normal conditions and 5% O2 hypoxia (data not shown). In the SOD2−/− astrocytes, expression of the Bcl-2 protein significantly increased compared with both the wild-type and SOD2−/+ astrocytes, whereas under 5% O2 hypoxia, a decrease in the expression of Bcl-2 in the SOD2−/− astrocytes was observed (Fig. 2). Bcl-XL showed a similar pattern of expression, with a decrease in SOD2−/− astrocytes under 5% O2 hypoxia (Fig. 2). Expression of Bad, a proapoptotic protein, also decreased in the SOD2−/− astrocytes under 5% O2 hypoxia. There was no difference in Bax protein expression between normal conditions and 5% O2 hypoxia (Fig. 2). In summary, the Bcl-2 family of proteins did not play a significant role in promoting survival and growth of the SOD2−/− astrocytes under 5% O2 hypoxia (Fig. 2).
c-Myc rather than HIF-1α was induced in SOD2 knock-out astrocytes under 5% O2 hypoxia
HIF-1α is known to be induced by hypoxia. HIF-1α protein expression was not induced in either the wild-type or the SOD2 knock-out astrocytes under normal conditions or 5% O2 hypoxia (Fig. 3A). In cells subjected to anoxic conditions (0% O2), HIF-1α induction was observed (positive control in Fig. 3A). Along with HIF-1α, c-Myc is widely known as a crucial regulator of proliferation of normal and neoplastic cells and is induced by ROS (Simon et al., 1998). There was no difference in c-Myc expression in the wild-type astrocytes under normal conditions or 5% O2 hypoxia. In the SOD2−/+ astrocytes, c-Myc expression was lower than in the wild-type astrocytes under normal conditions but was induced under 5% O2 hypoxia. There was no c-Myc expression in the SOD2−/− astrocytes under normal conditions, but it was significantly induced with 5% O2 hypoxia (Fig. 3B,C), and c-Myc staining was strong in cells that underwent mitosis (Fig. 3D). Thus, a critical difference in c-Myc expression was seen between wild-type astrocytes and SOD2-deficient astrocytes.
Five percent O2 hypoxia induced SOD2−/− astrocytes to pass through a cell cycle checkpoint
Because c-Myc expression was upregulated in SOD2−/− astrocytes under mild hypoxia, the downstream proteins under c-Myc that are involved in the cell cycle pathway were examined. Under 5% O2 hypoxia, there was significant phosphorylation of Rb at Ser807/811 in the SOD2−/− cells compared with the cells grown under normal culture conditions. This upregulation in phosphorylation of Rb was not seen in the wild-type astrocytes. Expression of cyclin D, CDK4, CDK6, cyclin E, and CDK2 increased in the SOD2−/− astrocytes with 5% O2 hypoxia as demonstrated by Western blot analysis. In the wild-type astrocytes, there was no significant difference in phosphorylated cell division cycle 2 (Cdc2) expression between normal conditions and 5% O2 hypoxia, whereas, in the SOD2−/− astrocytes, Cdc2 phosphorylation was significantly upregulated with 5% O2 hypoxia compared with normal culture conditions. p21 was undetectable in the wild-type astrocytes under both normal and 5% O2 hypoxia conditions. The p21 protein level was much higher in the SOD2−/− astrocytes under normal conditions compared with the wild-type astrocytes. However, there was a significant decrease in the SOD2−/− astrocytes under 5% O2 hypoxia (Fig. 4).
p53 activation in SOD2−/− astrocytes under 5% O2 hypoxia
There was no consistent activation of p53 in the wild-type astrocytes under 5% O2 hypoxia compared with normal culture conditions, but both expression and activation of p53 increased significantly in the SOD2−/− astrocytes after 5% O2 hypoxia. Expression of the downstream gene, p53-upregulated modulator of apoptosis (PUMA), however, was minimal in the SOD2−/− astrocytes under both normal and 5% O2 hypoxia conditions (Fig. 5).
c-Myc plays a critical role in the survival of SOD2−/− astrocytes
To examine the effects of c-Myc on astrocyte survival and proliferation, c-Myc siRNA was used to knock down c-Myc expression. Fluorescent-labeled rhodamine-conjugated siRNA demonstrated transfection efficiency to be ∼80%. A representative of the transfected astrocytes can be seen in Figure 6A. c-Myc protein expression decreased by 60% within 48 h of c-Myc siRNA administration as demonstrated by Western blot analysis and quantitation (Fig. 6B). Survival of the SOD2−/− astrocytes transfected by c-Myc siRNA decreased significantly to 60% under 5% O2 hypoxia compared with the nontarget transfected SOD2−/− astrocytes (Fig. 6C). The wild-type astrocytes transfected by c-Myc siRNA showed no significant difference in survival (Fig. 6C).
Antisense c-Myc was also used to knock down c-Myc expression. Treatment of SOD2−/− astrocytes with 100 nm of antisense c-Myc reduced both RNA and protein expression of c-Myc (Fig. 7A) and reduced the viability of SOD2−/− astrocytes by ∼25% (Fig. 7B). This confirms the results obtained from c-Myc siRNA transfection.
Moreover, expression of phosphorylated Rb, phosphorylated CDK2, Cdc2, cyclin D, cyclin E, and p53 decreased, whereas p21 expression increased in the transfected SOD2−/− astrocytes (Fig. 8A). Quantitative expression of the proteins is represented in Figure 8B. Binding of CDK4 and cyclin D was also inhibited by the transfection of c-Myc SiRNA (Fig. 8C).
Our study demonstrates that, under 5% O2 hypoxia, c-Myc was upregulated in the SOD2−/− astrocytes. Without the induction of HIF-1α, c-Myc was dominant in the inhibition of p21 and the induction of proteins involved in the cell cycle process. Through direct or indirect stimulation of CDKs and cyclin, Rb is phosphorylated, and cells that would not otherwise grow pass through the G1/S checkpoint. Phosphorylated Cdc2, one of the Rb target genes, was also induced, causing cells to go through the G2/M phase. Our results suggest that this cascade of events plays a role in SOD2−/− astrocyte proliferation under 5% O2 hypoxia (supplemental Fig. 3, available at www.jneurosci.org as supplemental material).
HEt staining of SOD2−/− astrocytes under normal atmospheric 20% O2 showed punctate staining, possibly representing mitochondrial localization, which was also observed by Sugawara et al. (2002). Thus, these SOD2-deficient astrocytes are susceptible to oxidative stress; however, when placed under 5% O2 (mild hypoxia), they survived and proliferated (Fig. 1 and supplemental Fig. 2, available at www.jneurosci.org as supplemental material), suggesting the onset of an alternate trigger to compensate for the oxidative stress. In the present study, HIF-1α did not appear to be upregulated and thus may not be involved in the survival of the SOD2−/− astrocytes. HIF-1α has been reported to be indirectly involved with p21 and c-Myc (Koshiji et al., 2004); however, in the present study, the upregulation of c-Myc suggests that it has a role in overcoming oxidative stress and assisting the SOD2−/− astrocytes in survival and proliferation.
The results of our analysis of Bcl-2 expression were similar to those of a study that showed that complex 1-deficient cells that do not induce SOD2 did induce the Bcl-2 protein (Robinson, 1998). Bcl-2 is known to protect against enhanced oxidative stress (Hochman et al., 1998; Papadopoulos et al., 1998; Xu et al., 1999). Overexpression of the Bcl-2 or Bcl-XL genes can protect against hydroxyl radical production by allowing cells to adapt to ROS overproduction (Esposti et al., 1999; Ouyang et al., 2002). In our experiments, there was an increased expression of Bcl-2 and Bcl-XL in the SOD2−/− astrocytes under normal oxygen conditions, which protected the cells from ROS. However, under mild hypoxia, the expression of both proteins decreased, thus indicating that they are not critical factors in cell proliferation.
SOD2 activity in vivo is often downregulated in cancer cells compared with adjacent normal cells and the manganese SOD immunoreactive protein and its activity are barely detectable in contrast to other antioxidant enzymes (copper/zinc SOD, catalase, and glutathione peroxidase) (Oberley and Oberley, 1995). An inverse correlation between manganese SOD activity and cellular growth potential has been well documented in mitochondria of both cancer and nonmalignant cells (Oberley, 2001). This may explain the incapability of SOD2−/− cells to survive in atmospheric 20% O2.
Previous reports showed that hypoxia promotes cell cycle arrest in diverse primary cell types (Åmellem et al., 1996). Genetic studies from mouse HIF-1α null cells indicate that HIF-1α is required for hypoxia-induced cell cycle arrest. Koshiji et al. (2004) showed that HIF-1α upregulates p21, a key CDK inhibitor that controls the G1 checkpoint. In their study, the PAS (Per-ARNT-Sim) domain of a stable HIF-1α mutant displaced c-Myc binding from the p21 promoter and functionally counteracted c-Myc repression of p21, even in the absence of a hypoxic signal. c-Myc has also been shown to inhibit Bcl-2 and Bcl-XL expression (Greider et al., 2002), supporting our observation of decreased expression of these proteins in the astrocytes under 5% O2 hypoxia.
In our study, stable HIF-1α protein expression was not detected under mild hypoxia (5% O2), whereas c-Myc was strongly induced. In agreement with our findings, Jiang et al. (1996) reported that HIF-1α is induced at much lower O2 tensions with half-maximal expression at 2% and maximal expression at 0.5%. Moreover, c-Myc, is induced by ROS (Simon et al., 1998). In our study, we observed different levels of c-Myc expression in all three groups of astrocytes (wild-type, SOD2−/+, and SOD2−/−). The results of c-Myc siRNA transfection and antisense c-Myc treatment of SOD2−/− astrocytes suggest that c-Myc is important for cell survival in the SOD2−/− astrocytes. Survivability decreased significantly, by ∼60%, in c-Myc siRNA-transfected SOD2−/− astrocytes under 5% O2 hypoxia, whereas the wild-type astrocytes transfected by c-Myc siRNA were not affected. Furthermore, in the transfected SOD2−/− astrocytes, expression of the downstream cell cycle proteins, Cdc2, cyclin D, cyclin E, and phosphorylated CDK2, decreased, whereas p21 increased. Binding of CDK4 and cyclin D was also inhibited. The schematic diagram in supplemental Figure 3 (available at www.jneurosci.org as supplemental material) suggests the pathway that SOD2 astrocytes might follow to proliferate with 5% O2 hypoxia.
In other mammalian cells, as in PC-3 cells, SOD2 not only regulates cell survival by scavenging superoxide radicals but also affects cell proliferation by retarding the G1-to-S transition (Venkataraman et al., 2005). Because mitochondrial proteins are encoded in both mitochondria and the nuclear genome, biogenesis is a coordinated effort in which mitochondria transmit signals to the nucleus and vice versa (Felty and Roy, 2005). How mitochondria transmit these signals in the process of cell proliferation is still unknown. Recently, a model was proposed in which mitochondria are able to translate changes in mitochondrial ROS production into a signal that is relayed to the cytoplasm (Chen et al., 2004). This signal was shown to modulate nuclear factor-κ B mediation in tumor necrosis factor-induced apoptosis (Hughes et al., 2005). Transcription factors like nuclear factor-κ B and others may then mediate expression of other genes such as c-fos, c-jun, and c-Myc (Kamata and Hirata, 1999). Most transcription factors are known to be redox-sensitive (Kamata and Hirata, 1999; Harris, 2002). In the SOD2-deficient astrocytes, c-Myc induction by this mechanism may be responsible for cell rescue.
c-Myc is directly recruited to the p21 (Cip1) promoter by the DNA-binding protein Miz-1 (c-Myc-interacting zinc-finger protein) and inhibits p21 activity, allowing the cells to proceed to the G1/S phase. c-Myc also directly induces cyclin and CDKs, as shown in our study. Induction of c-Myc is sufficient to drive quiescent cells into the cell cycle (Eilers et al., 1989). Overexpression of c-Myc genes alone can also cause REF52 cells to divide (Smith et al., 1995). This deregulated c-Myc induction could make cells tolerant of the resulting genomic change (Smith et al., 1995), and this could explain SOD2 cell growth and proliferation through activation of cell cycle proteins, although this could also be detrimental to cancer cells. In our study, the expression of c-Myc and cell cycle proteins and the results of c-Myc transfection support the involvement of c-Myc induction in the survival and proliferation of the SOD2−/− astrocytes. With our experimental conditions, we also observed an elevated p53 level caused by c-Myc. p53 is a tumor suppressor protein and functions by inducing cell cycle arrest and apoptosis in response to DNA damage.
A growing number of p53 target genes has been implicated in mediating the apoptotic effects of p53. These include Bax, a proapoptotic Bcl-2 family member, and PUMA. The present study shows no significant induction of PUMA, a target gene of p53. There was no increase in Bad or Bax. PUMA deficiency blocks cell death throughout the developing nervous system after gamma-irradiation treatment, and endogenous wild-type p53 is not sufficient, in the absence of PUMA, for inducing efficient cell death (Jeffers et al., 2003). Hence, our present study indicates that the cells under 5% O2 hypoxia were allowed to pass through the cell cycle checkpoint and that, in the absence of PUMA upregulation, the cells proliferated. In this study, p53 activation may serve as a warning signal of DNA damage that occurs as a result of cell proliferation. It appears that a threshold for the p53 protein level may exist, below which neither growth arrest nor cell death can occur; however, in our study, with the elevation of p53, the threshold was reached more quickly once DNA damage occurred.
The rapid disappearance of p21 after inhibition of transcription may contribute to p53 upregulation (Blagosklonny et al., 2002). SOD2 has also been suggested to have a reciprocal effect on p53 by downregulating its transcription (Drane et al., 2001). Despite the predominance of its proapoptotic effects during oxidative stress, p53 activation can activate other pathways and genes that may be involved in a compensatory mechanism to alleviate adverse effects of oxidative stress. Here, we report that the response of cell cycle proteins to SOD2−/− astrocytes subjected to hypoxia was to rescue the cells and to allow them to proliferate. It is not clear, however, whether the same hypoxic conditions can rescue newborn SOD2−/− mice from death by activating the c-Myc signaling pathway. Nevertheless, our study presents a new molecular strategy that uses hypoxia to compensate for oxidative stress-induced mitochondrial dysfunction that occurs in many neurological disorders.
This work was supported by National Institutes of Health Grants P50 NS14543, RO1 NS25372, RO1 NS36147, and RO1 NS38653 and an American Heart Association Bugher Foundation Award.
- Correspondence should be addressed to Dr. Pak H. Chan, Neurosurgical Laboratories, Stanford University, 1201 Welch Road, MSLS #P304, Stanford, CA 94305-5487. Email: