INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

AN ADEQUATE SUPPLY of oxygen is essential for the survival of mammalian cells. Chronic hypoxia induces phenotypic remodeling, leading to long-term adaptive responses. The molecular mechanisms underlying the adaptations to chronic hypoxia are poorly understood. Activation of specific genes is considered to be an important mechanism by which hypoxia triggers long-term adaptive responses. In general, the genes that are activated by low oxygen fall into two classes: immediate early genes (IEGs), which are induced within minutes after the onset of hypoxia, and late response genes, which are activated more slowly, over hours (see Ref. 6). c-fos is one of the most extensively investigated members of the IEG family (14). Hypoxia stimulates c-fos expression both in vivo (9, 13, 26) and in vitro (25, 28, 29). The Fos protein forms a heterodimeric complex with Jun, the protein product of another IEG. Fos-Jun heterodimers or Jun-Jun homodimers bind to the DNA sequence TGACTCA, the consensus binding site for the transcription factor activator protein-1 (AP-1) (2). Using reporter gene assays, we have previously shown that hypoxia increases AP-1 activity and antisense c-fos prevents AP-1 activation by low oxygen (17), suggesting that c-fos is essential for the formation of the AP-1 complex during hypoxia. Many of the late response genes that are activated during hypoxia contain AP-1 consensus binding sites in their promoter regions. Consequently, it has been proposed that c-fos regulates the expression of late response genes during hypoxia (7). Consistent with this idea, antisense c-fos (17), as well as mutations in AP-1 binding sites (17, 22), abolishes the activation of the tyrosine hydroxylase gene during low oxygen. These studies suggest that c-fos participates in cellular adaptations during hypoxia by regulating certain late response genes, such as tyrosine hydroxylase, via formation of the AP-1 transcription factor.

Despite its potential role in cellular adaptations during hypoxia, little is known about the mechanisms underlying c-fos activation by low oxygen. Ca²⁺-dependent signaling pathways are critical for c-fos stimulation by a variety of stimuli, including neurotransmitters and growth factors (see Ref. 12). Furthermore, it is well established that hypoxia increases intracellular Ca²⁺ concentration ([Ca²⁺]_i) in many cell types (5). Therefore, in the present study, we determined whether c-fos stimulation by hypoxia requires elevation of [Ca²⁺]_i, and, if so, we aimed to identify the Ca²⁺-dependent signaling pathways associated with the activation of c-fos by low oxygen. Using rat pheochromocytoma-12 (PC-12) cells as a model, we demonstrate that c-fos activation by hypoxia requires Ca²⁺ influx through L-type voltage-gated Ca²⁺ channels and is dependent on the activation of Ca²⁺/calmodulin-dependent protein kinases (CaMKs). Our results further demonstrate that both Ca²⁺/cAMP-responsive and serum-responsive cis-elements (Ca/CRE and SRE, respectively) are essential for transcriptional activation of c-fos by low oxygen.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell culture. PC-12 cells (from Dr. K. Neet, Finch University of Health Sciences/Chicago Medical School; original clone from Dr. L. Greene) were grown in a humidified incubator at 37°C circulated with 10% CO₂ and 21% O₂. The growth medium (DMEM) was supplemented with 10% horse serum and 5% fetal bovine serum (FBS) containing 100 U/ml penicillin and 100 µg /ml streptomycin. Cells were grown to ~80% confluency. Before the experiment, cells were placed in low-serum-containing medium (0.5% FBS) for 18 h. Cells were exposed to normoxia (21% O₂ and 10% CO₂ balanced with N₂) or to hypoxia (1% O₂ and 10% CO₂ balanced with N₂) in an oxygen-regulated incubator (Heraeus) for 3 h unless otherwise indicated. In the experiments involving treatment with drugs, cells were preincubated for 30 min with different concentrations of drugs or vehicle and were then subjected to either hypoxia or normoxia.

Reagents. The following chemicals and reagents were used: lipofectamine, FBS, horse serum, DMEM (Life Technologies, Gaithersburg, MD), primary antibodies against CaMKII and CaMKIV raised against regulatory domain (Transduction Labs, Lexington, KY), and phosphospecific cAMP-responsive element binding protein (CREB) [at serine residue 133 (Ser-133)] and CREB antibodies (New England Biolabs). We obtained affinity-purified horseradish peroxidase-conjugated goat anti-mouse and goat anti-rabbit antibody from Santa Cruz; [-³²P]ATP and [-³²P]dCTP from NEN-DuPont (Boston, MA); nitrendipine from Research Biochemical International (Natick, MA); KN-93, 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA)-AM, and BAY K 8644, from Calbiochem (La Jolla, CA); the luciferase assay kit from Promega; the -galactosidase assay kit from Tropix; and TRIReagent from Molecular Research Center (Cincinnati, OH). All reagents for RT-PCR assay were from Perkin-Elmer Cetus. Other reagents used in this study were of analytical grade and obtained from Sigma Chemical. Stock solutions of KN-93, BAPTA-AM, and nitrendipine were made in DMSO, whereas BAY K 8644 was made in methanol.

Plasmids. The following plasmids were used in the present study: pfos-LUC, pCaMKII, pCaMKII290, pCGN, and pRSV-LacZ. Wild-type and Ca/CRE c-fos promoters with chloramphenicol acetyltransferase reporter were provided by Dr. Gilman (11) and were reconstructed with luciferase reporter gene as described previously (27). A detailed description of the wild-type and active mutant of CaMKII (CaMKII290) was published previously (27). The -galactosidase-expressing plasmid pRSV-LacZ was from American Type Culture Collection (16).

Measurement of c-fos mRNA. c-fos mRNA was analyzed by an RT-PCR assay as described previously (27). Briefly, total RNA was isolated with TRIReagent and then treated with RNase-free DNase I (Boehringer, Indianapolis, IN) for 1 h at room temperature. For reverse transcription, 0.5 µg of total RNA was added to a reaction mixture of 20 µl containing 5 mM MgCl₂, 1 mM of each of deoxynucleotide triphosphates, 2.5 µM of random hexamers, 50 mM KCl, 20 mM Tris · HCl (pH 8.3), 1 unit of RNase inhibitor, and 2.5 units of RT. The reaction mixture was incubated at 25°C for 15 min and then incubated at 42°C for 30 min. The reaction was terminated by heating to 95°C for 5 min and flash cooling on ice. The prepared cDNA was stored at 70°C until further use; 2 µl of the reverse-transcribed material (cDNA) were used for PCR amplification. The reaction mixture (20 µl) contained 50 mM KCl, 10 mM Tris · HCl (pH 8.3), 2 mM MgCl₂, 1.5 µg each of upstream and downstream primers, 1 µCi [-³²P]dCTP (3,000 Ci/mmol), and 0.75 units of Taq polymerase. The reaction mixture was amplified for 30 cycles with the following cycle profile: denaturation for 45 s at 94°C, annealing for 45 s at 56°C, and extension for 1 min at 72°C. In preliminary experiments, we optimized the PCR conditions and product amplifications and found them to be linear up to 30 cycles for both -actin and c-fos. Ten microliters of the PCR products were electrophoresed on 1.5% agarose gels. Gels were stained with ethidium bromide, and radioactive products were visualized after exposure of the dried gels to Kodak film and quantified using PhosphorImager (Molecular Dynamics).

Reporter gene assays. Cells were transfected with one or more of the plasmids, depending on the experimental protocol, as described previously (17). Briefly, cells were plated in 60-mm tissue culture plates at a density of 5 × 10⁵ cells/plate in growth medium containing serum. For transfection, DNA-liposome mixture was prepared using 10 µg of lipofectamine, 1 µg of pfos-LUC, and 0.25 µg of pRSV -galactosidase (internal control for determining the transfection efficiency) in 2 ml of serum-free medium. Cells were incubated in the DNA-liposome mixture (2 ml/plate) for 4 h followed by addition of 2 ml of medium. After 36 h, cells were exposed to hypoxia for the desired durations. Total amount of DNA to be transfected per plate was maintained equal by adding pUC19 DNA. Control experiments were run in parallel in normoxia and also by transfecting cells with vector alone.

Western blot analysis. Proteins were separated on nonreducing 10% SDS-PAGE gels. Resolved proteins were transferred onto Immobilon membranes (Millipore, Bedford, MA) at 40 V/mm² for 2 h at 20°C using a Bio-Rad transblot apparatus. After transfer, the membranes were blocked overnight in TBS-T (0.1% Tween 20 and 20 mM Tris-buffered saline, pH 7.6) containing 5% BSA at 4°C. Membranes were incubated with the appropriate primary antibodies for 1 h at 25°C and then washed three times in TBS-T every 5 min. Monoclonal antibodies raised against the regulatory domain of CaMKII or CaMKIV (Transduction Laboratories) were used. Membranes were then incubated for 1 h with the appropriate secondary antibodies conjugated with horseradish peroxidase in TBS-T containing 1% BSA. Membranes were washed three times for 5 min each in TBS-T. Protein bands were detected by an enhanced chemiluminescence detection system (Amersham). The membranes were exposed to Kodak XAR films.

Measurements of CaMKII activity by in vitro kinase assay. Cells were plated at a density of 4 × 10⁶/100-mm dish. Eighteen hours before experiments, cells were placed in medium containing low serum (0.5% FBS). After the hypoxic exposure, cells were lysed in buffer containing 25 mM HEPES (pH 7.4), 150 mM NaCl, 2.5 mM MgCl₂, 0.1 mM EDTA, 0.5 mM dithiothreitol, 20 mM glycerophosphate, 1 mM sodium orthovanadate, 2 µg/ml leupeptin, 100 µg/ml phenylmethylsulfonyl fluoride, and 1% NP-40 on ice for 20 min. Cell debris was removed by centrifugation at 10,000 g for 10 min. Protein was assayed by a protein assay kit (Bio-Rad), using BSA as a standard. The standard curve for the kinase assay was constructed according to the protocols suggested by the manufacturer (CaMKII assay kit; Upstate Biotechnology). The reaction mixture was 30 µl of solution containing 50 mM HEPES (pH 7.5), 10 mM MgCl₂, 0.1 mM [-³²P]ATP, 125 µM autocamtide, and 400 ng of calmodulin. Activities of other serine/threonine kinases such as protein kinase A and protein kinase C were inhibited by 0.2 µM of inhibitor peptide. The reaction was initiated by adding 10 µl of cell supernatant (equal amount of protein). After incubation at 30°C for 15 min, 10-µl aliquots were spotted on p81 phosphocellulose paper. After 3 min, the paper was immersed in 0.75% phosphoric acid and washed 10-15 times and then washed in acetone for 5 min with five changes. The radioactivity was counted by a scintillation counter.

Measurements of Ca²⁺ currents. Ca²⁺ current was monitored using the whole cell configuration of the patch-clamp technique as described previously (23). Briefly, currents were recorded using an Axopatch 200A voltage-clamp amplifier, filtered at 5 kHz, and sampled at a frequency of 10 kHz using an IBM-compatible computer with a Digidata 1200 interface and pCLAMP software (Axon Instruments). Currents were not leak subtracted. Current-voltage relations were elicited from a holding potential of 80 mV using 50-ms steps (5 s between steps) to test potentials over a range from 50 to +50 mV in 10-mV increments. Current at each potential was measured as the average over a 2.5-ms span at the end of the step. Ca²⁺ current was isolated using K⁺- and Na⁺-free intracellular and extracellular solutions. The intracellular solution had the following composition (in mM): 130 CsCl, 20 tetraethylammonium chloride, 5 MgATP, 0.1 Tris-GTP, 5 EGTA, and 5 HEPES; pH was adjusted to 7.2 with CsOH. The extracellular solution contained (in mM) 140 N-methylglucamine chloride, 5.4 CsCl, 10 BaCl₂, 10 HEPES, and 11 glucose; pH was adjusted to 7.4 with CsOH. Extracellular solutions were made hypoxic by degassing under vacuum for 1 h and then continuously bubbling with hypoxic gas mixture before and during experiments. The extracellular solution was changed using a fast-flow apparatus consisting of a linear array of borosilicate glass tubes (23). In these experiments, Ba²⁺ was the charge carrier. For simplicity, Ba²⁺ current conducted by Ca²⁺ channels will be referred to as Ca²⁺ current.

Data analysis. The data are expressed as means ± SE from three to five individual experiments, each run in duplicate. Statistical analysis was performed by ANOVA or by paired t-test where appropriate. P values <0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of BAPTA on c-fos mRNA induction and transcriptional activation by hypoxia. To determine whether activation of c-fos mRNA requires increases in [Ca²⁺]_i, cells were exposed for 30 min to varying concentrations of BAPTA, a Ca²⁺ chelator; cells were then challenged with hypoxia (1% O₂) for 3 h. Experiments without BAPTA in the medium served as controls. As shown in the example in Fig. 1A, top, c-fos mRNA increased during hypoxia and BAPTA abolished the effects of low oxygen in a concentration-dependent manner. Average results are summarized in Fig. 1A, bottom. BAPTA at submicromolar concentrations (e.g., 0.1 µM) reduced and BAPTA at micromolar concentrations (3 µM) abolished c-fos activation by hypoxia. However, under normoxia, BAPTA (3 µM) by itself had no effect on c-fos expression. Furthermore, neither hypoxia nor BAPTA-AM had any significant effect on -actin mRNA, which served as controls.

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Effects of 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA) on c-fos activation by hypoxia. A, top: effect of BAPTA-AM on c-fos mRNA activation by hypoxia. N, normoxia; H, hypoxia (1% O2 for 3 h). -Actin mRNA was monitored to normalize sample-sample variations in PCR analysis. A, bottom: average data from 3 individual experiments run in duplicate. Data are means ± SE. BAPTA inhibited c-fos mRNA induction by hypoxia in a concentration-dependent manner. -Actin mRNA was unaffected by either hypoxia or BAPTA-AM. B, top: schematic presentation of c-fos promoter. SIE, Sis-inducible element FAP: Fos-AP-1-like element. B, bottom: effect of BAPTA on c-fos promoter activation by hypoxia; c-fos promoter activity was measured by a luciferase (Luc) expression vector. -Galactosidase (-gal) activity was monitored as an internal control for assessing the transfection efficiency. Data are means ± SE from 5 individual experiments run in duplicate. * P < 0.05 (ANOVA). BAPTA significantly attenuated c-fos promoter activation by hypoxia. SRE, serum-responsive cis-element; Ca/CRE, Ca2+/cAMP-responsive cis-element.

Effects of L-type voltage-gated Ca²⁺ channel agonist and antagonist on c-fos activation by hypoxia. The results described thus far indicate that an increase in [Ca²⁺]_i is required for c-fos activation by hypoxia. An increase in [Ca²⁺]_i could occur as a result of mobilization of intracellular Ca²⁺ stores and/or influx through membrane voltage-gated Ca²⁺ channels. In PC-12 cells, activation of voltage-gated Ca²⁺ channels appears to be necessary for the Ca²⁺ influx during hypoxia (15). PC-12 cells express predominantly L-type voltage-activated Ca²⁺ channels (24). Therefore, we investigated whether hypoxia-induced c-fos activation is linked to the activation of L-type Ca²⁺ channels. As shown in Fig. 2A, nitrendipine, a specific L-type Ca²⁺ channel blocker, abolished c-fos mRNA induction by hypoxia in a concentration-dependent manner. Average data showed that 1 µM nitrindipine reduced c-fos mRNA activation by 60%, whereas 3 µM completely prevented the response to hypoxia. At higher concentrations (10 µM), nitrendipine tended to affect basal c-fos expression under normoxia, whereas 3 µM nitrendipine had no significant effect on basal c-fos expression. Therefore, in all further experiments, we chose to study 3 µM nitrendipine. To further establish the involvement of L-type Ca²⁺ channels, the effect of ()BAY K 8644, an L-type Ca²⁺ channel agonist, was tested on c-fos mRNA activation by hypoxia. As illustrated in Fig. 2B, BAY K 8644 (1 µM) significantly enhanced c-fos activation by hypoxia (controls vs. BAY K 8644, P < 0.01). Figure 2C summarizes the effects of nitrendipine and BAY K 8644 on c-fos promoter activation by hypoxia. As can be seen, 3 µM nitrindipine prevented and 1 µM BAY K 8644 enhanced c-fos promoter activation by hypoxia (P < 0.01; n = 5).

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Effect of L-type Ca2+ channel antagonist and agonist on c-fos activation by hypoxia. Cells were treated with the L-type Ca2+ channel antagonist or agonist, nitrendipine (nitren; A) or BAY K 8644 (BayK; B), respectively, for 30 min and then were exposed to hypoxia (1% O2) for 3 h. In A and B, effects of nitrendipine and BAY K 8644 (1 µM) on c-fos mRNA during hypoxia are shown at top, and average results are summarized at bottom. Average data represent means ± SE from 3 individual experiments run in duplicate. C: effects of nitrendipine and BAY K 8644 on c-fos promoter activity during hypoxia. Data represent means ± SE from 5 individual experiments. * P < 0.05. Nitrendipine inhibited and BAY K 8644 potentiated c-fos promoter activation by hypoxia.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Hypoxia augments Ca2+ currents in pheochromocytoma-12 (PC-12) cells. A: example of current (I) traces showing high-threshold Ca2+ current before (control, Con) and during exposure to extracellular solution equilibrated with hypoxic gas. Currents were elicited by a 50-ms step to 0 mV from a holding potential of 80 mV. B: comparison of the magnitude of current augmentation [(hypoxia/control  1) × 100] at different membrane potentials (Vm) by hypoxia. * Current values were significantly different under control and hypoxia (P < 0.05, paired t-test; n = 7).

Effects of CaMK inhibitor and ectopic expression of the active mutant of CaMKII on c-fos promoter activation by hypoxia. One of several mechanisms by which Ca²⁺ regulates c-fos gene expression is through the activation of CaMKs, especially CaMKII and CaMKIV (see Ref. 12). As shown in Fig. 4A, a protein band corresponding to CaMKII could readily be detected in PC-12 cells, whereas protein corresponding to CaMKIV was not evident. However, protein bands corresponding to both CaMKII and CaMKIV could readily be seen in Hep 3B cells, which served as controls. To determine whether hypoxia increases CamKII activity, cells were exposed to hypoxia and CaMKII activity was determined by in vitro kinase assay. Hypoxia resulted in a significant activation of CaMKII (Fig. 4B). This activation was transient in that it returned to basal levels after 1 h of hypoxia (Fig. 4B). Furthermore, KN-93 (10 µM), a selective pharmacological inhibitor of CaMKs, prevented hypoxia-induced CaMKII activation (P < 0.01; Fig. 4C).

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Effect of hypoxia on Ca2+/calmodulin-dependent protein kinase (CaMK) activity. A: immunoblot analysis of CaMKII and CaMKIV in PC-12 cells. Hep 3B cells were used as controls. PC-12 expressed CaMKII but not CaMKIV. Hep 3B cells expressed both CaMKII and CaMKIV. B: PC-12 cells were exposed to normoxia [21% O2, control (C)] or to varying durations of hypoxia (1% O2), and CaMKII activity was analyzed by in vitro kinase assay. CaMK activity was expressed as percentage of normoxic controls. CaMKII activity increased within 15 min of hypoxia and returned to basal level by extending the duration of hypoxic exposure to 1 h. C: activation of CaMKII by hypoxia was inhibited by 10 µM KN-93, a CaMK inhibitor. Data in B and C are means ± SE from 4 individual experiments run in duplicate. * P < 0.05 (ANOVA).

View larger version (19K): [in this window] [in a new window]   Fig. 5.   CaMK inhibitor prevents c-fos promoter activation by hypoxia. A: PC-12 cells transfected with c-fos-luciferase construct were exposed to 10 µM KN-93, a CaMK inhibitor, for 30 min and then exposed to hypoxia (1% O2) for 3 h. Data are means ± SE from 4 experiments run in duplicate. Hypoxia increased c-fos promoter activity, and this effect was prevented by KN-93. B: ectopic expression of a constitutively active mutant of CaMKII mimics the effects of hypoxia on c-fos promoter activity. PC-12 cells were transiently transfected with fos-luciferase construct along with either pCMV (expression vector) or pCaMKII (wild type; wt) or pCaMKII290 (active mutant). CaMK constructs used in this study are schematically presented at top (Reg, regulatory; Assoc, associate). c-fos promoter activity was significantly greater in cells expressing CaMKII290 (active mutant) than cells with wild-type CaMKII or the basic plasmid pCMV. Data represent means ± SE from 4 experiments run in duplicate. * P < 0.05 (ANOVA).

Effect of hypoxia on phosphorylation of CREB at Ser-133. The following series of experiments were performed to identify signaling pathways in c-fos expression by hypoxia downstream to CaMKs. CREB is a transcription factor that is involved with transcriptional regulation of several genes, including c-fos (12). CaMKs are one of the CREB kinases that stimulate phosphorylation at Ser-133, and phosphorylation of CREB at Ser-133 is essential for stimulation of c-fos transcription by other stimuli (12). To determine whether hypoxia increases CREB phosphorylation at Ser-133 and whether CaMKs are involved in this response, cells were exposed to hypoxia and phosphorylation of CREB at Ser-133 was analyzed by immunoblot assay using antibodies specific for the phosphorylated form of CREB at residue Ser-133. As shown in Fig. 6, hypoxia increased CREB phosphorylation at Ser-133 but had no effect on CREB protein levels (Fig. 6). Furthermore, the CaMK inhibitor, KN-93 (10 µM), significantly attenuated hypoxia-induced CREB phosphorylation at Ser-133 (P < 0.01; n = 4).

View larger version (26K): [in this window] [in a new window]   Fig. 6.   Attenuation of hypoxia-induced CREB phosphorylation at Ser-133 by KN-93. Phosphorylated (at Ser-133) and total CREB protein were analyzed by immunoblot assay using specific antibodies. Top: example illustrating effect of hypoxia on CREB phosphorylation at Ser-133 with and without KN-93 (10 µM). Bottom: average results from 3 individual experiments run in duplicate. Data represent means ± SE. * P < 0.05 (ANOVA). Hypoxia increases CREB phosphorylation (phospho) at Ser-133, whereas total CREB protein was unaffected. KN-93 attenuated hypoxia-induced increases in CREB phosphorylation at Ser-133.

Effect of point mutations in Ca/CRE and SRE on c-fos promoter activation by hypoxia. CREB constitutively binds to Ca/CRE in the c-fos promoter. Phosphorylation of CREB at Ser-133 promotes c-fos transcription by transactivation of Ca/CRE (12). To assess whether CREB phosphorylation induced by hypoxia is linked to c-fos activation, we transfected cells with c-fos-luciferase plasmid with point mutations in Ca/CRE (C-to-G conversions, point mutation 3), which prevents constitutive binding of CREB to Ca/CRE. Transfected cells were exposed to either hypoxia or 20% FBS, another potent stimulator of c-fos promoter activity (27); the latter served as controls. The results are summarized in Fig. 7. Hypoxia increased c-fos promoter activity in cells transfected with wild-type construct. However, point mutation in Ca/CRE significantly attenuated but did not completely block c-fos promoter activation by hypoxia (Fig. 7). As expected, FBS-induced stimulation of the c-fos promoter activity, which can activate c-fos independent of Ca/CRE, was unaffected by point mutation in Ca/CRE.

View larger version (18K): [in this window] [in a new window]   Fig. 7.   Effects of point mutations in Ca/CRE and SRE on c-fos promoter activation by hypoxia. Cells were transfected with the fos-luciferase promoter constructs, indicated schematically at left (wt, wild type; pm3, point mutation at Ca/CRE; pm12, point mutation at SRE). Transfected cells were exposed to either normoxia (open bars) or hypoxia (solid bars) (1% O2; 3 h) or to 20% fetal bovine serum (shaded bars). Inactivating mutations in Ca/CRE markedly attenuated and mutation in SRE abolished c-fos promoter activation by hypoxia. Data are means ± SE from 3 individual experiments run in duplicate.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The objective of the present study was to identify the signaling pathways that link the hypoxic stimulus to c-fos gene expression. Our results demonstrate that Ca²⁺ influx through L-type voltage-activated Ca²⁺ channels is essential for c-fos activation by hypoxia and depends on CaMK and CREB phosphorylation at Ser-133. In addition, transcriptional activation of c-fos by hypoxia requires Ca/CRE and SRE.

c-fos activation by hypoxia requires Ca²⁺ influx through L-type voltage-gated Ca²⁺ channels. Several observations in the present study suggest that activation of Ca²⁺ influx through L-type voltage-gated Ca²⁺ channels is an initial step in the signaling cascade that regulates the c-fos expression during hypoxia. First, nitrendipine, an antagonist of L-type voltage-gated Ca²⁺ channels, blocked and BAY K 8644, an agonist of these channels, enhanced c-fos induction by hypoxia. Second, the Ca²⁺ chelator BAPTA, which is expected to prevent the rise in [Ca²⁺]_i, inhibited c-fos mRNA induction by hypoxia. However, whether hypoxia activates voltage-gated Ca²⁺ channels is controversial. In pulmonary arterial smooth muscle cells (cells from resistant vessels), hypoxia augments Ca²⁺ currents (10). However, in carotid body glomus cells, hypoxia inhibits Ca²⁺ currents (19). In the present experiments on PC-12 cells, hypoxia augmented Ca²⁺ currents in a voltage-dependent manner. These observations are similar to those reported in pulmonary artery smooth muscle cells (10). The currents were sensitive to nitrendipine, suggesting that hypoxia-sensitive Ca²⁺ currents are conducted by L-type voltage-gated Ca²⁺ channels in PC-12 cells. However, from these experiments, we cannot rule out the involvement of other types of high-voltage-activated Ca²⁺ channels. None the less, these observations with patch-clamp experiments further support the idea that c-fos activation by hypoxia in PC-12 cells requires Ca²⁺ influx through activation of L-type Ca²⁺ channels. Hypoxia also stimulates c-fos mRNA in nonexcitable cells such as Hep 3B (25), which may not express voltage-gated Ca²⁺ channels. Therefore, further experiments are necessary to elucidate what role, if any, Ca²⁺ plays in c-fos activation by hypoxia in nonexcitable cells.

CaMKs and CREB phosphorylation are the downstream signaling events in transcriptional activation of c-fos by hypoxia. Having established a role for Ca²⁺, we identified several downstream effectors, including CaMKs, involved in c-fos promoter activation by hypoxia. Of the several CaMKs, CaMKII and CaMKIV are of particular interest because of their association with the regulation of the c-fos gene by other stimuli (12). We found that PC-12 cells expressed CaMKII but not CaMKIV. The absence of CaMKIV in PC-12 cells is not due to the inability of the antibody to detect this protein because control experiments with Hep 3B cells revealed the presence of both CaMKII and CaMKIV. Our results are consistent with those reported by Enslen et al. (8), who also found no evidence for CaMKIV in PC-12 cells. More importantly, our results demonstrate that hypoxia stimulates CaMKII activity, as evidenced by in vitro kinase assay (Fig. 4). Hypoxia-induced CaMKII activation is rapid and occurs within minutes after the onset of the stimulus. However, it is transient in that enzyme activity returned to basal levels with longer exposure to hypoxia. This pattern of CaMKII activation by hypoxia is similar to that reported with other stimuli (18). A recent study reported that CaMKII activity was unaltered or downregulated in PC-12 cells (4). However, these investigators examined the CaMKII activity in cells exposed to 1 h or longer durations of hypoxia. Although the long-term effects of hypoxia (i.e., more than 3 h) were not investigated in this study, we did observe that CaMK activity returns to baseline after 1 h of hypoxia. Most importantly, a number of findings in the present study argue for a role of CaMKII in transcriptional regulation of c-fos by hypoxia. First, CaMKII activation preceded c-fos transcriptional activation, which required hours. Second, a pharmacological antagonist of CaMK, KN-93, not only prevented CaMKII activation but also, more importantly, blocked c-fos promoter activation by hypoxia and had no effect on basal fos-luciferase activity. Third, ectopic expression of an active mutant (pCaMKII290), but not the wild-type CaMKII, mimicked the effects of hypoxia on c-fos transcriptional activation. Together, these observations support a role for CaMKII in the activation of the c-fos promoter by low oxygen. Future experiments with a dominant negative mutant of CaMKII may further support the role of CaMKII. It should, however, be pointed out that Ca²⁺ influx also activates CaMKI in PC-12 cells (1). It remains to be established whether CaMKI also contributes to transcriptional activation of c-fos by hypoxia.