Prolonged Activation of cAMP-response Element-binding Protein and ATF-2 Needed for Nicotine-triggered Elevation of Tyrosine Hydroxylase Gene Transcription in PC12 Cells*

Phosphorylation (P-) of cAMP-response element-binding protein (CREB) by protein kinase A or mitogen-activated protein kinases was implicated in mediating the increased tyrosine hydroxylase (TH) gene expression after prolonged exposure to nicotine in vivo and in cell culture. We examined the time course and signaling pathways for phosphorylation of CREB and possible involvement of ATF-2. Treatment of PC12 cells with 200 μm nicotine triggered rapid but transient elevation of P-CREB followed by a second sustained rise after 2-5 h of continuous nicotine. In contrast, ERK1/2 was only phosphorylated with short term nicotine exposure. MEK inhibitor U0126 abolished nicotine-induced rise in P-ERK1/2, but not P-CREB, nor did it inhibit nicotine-evoked elevation in TH promoter activity, indicating that ERK1/2 was not needed for induction of TH gene expression by nicotine. In contrast, protein kinase A inhibitor H-89 or Ca2+/calmodulin-activated protein kinase inhibitor KN-93 reduced the nicotine-triggered rise in P-CREB and TH promoter activity. There was a delayed elevation of P-ATF-2 after 1 h of nicotine treatment, accompanied by increased ATF-2 protein. Upstream kinase JNK, but not p38, was phosphorylated especially after 5 min to 2 h of nicotine exposure. To examine the requirement for CREB and ATF-2, cells were transfected with dominant negative forms of ATF-2 or CREB. Both reduced the basal TH promoter activity and the response to nicotine. Knockdown of ATF-2 or CREB with siRNA did not alter basal TH promoter activity or mRNA but greatly attenuated the response to nicotine. The results suggest that both ATF-2 and CREB mediate activation of TH gene transcription by nicotine.

Administration of nicotine in vivo and in cell culture triggers increased expression of a number of genes related to neurosecretion including the catecholamine biosynthetic enzymes. These changes in gene expression may be involved in neurochemical and cardiovascular effects of smoking. Nicotine is a potent sympathomimetic agent, and its administration, in doses similar to those obtained in smoking, increases heart rate and systolic and diastolic blood pressure in humans and many animal species (for review, see Ref. 1). These cardiovascular effects are largely attributed to the direct stimulation of release of catecholamines, epinephrine and norepinephrine, from the adrenal medulla and peripheral sympathetic nerve endings (2)(3)(4). In this regard a polymorphism in the gene for tyrosine hydroxylase (TH), 2 the rate-limiting enzyme in catecholamine biosynthesis, has been associated with tobacco use (5).
Transcriptional mechanisms are at least partially responsible for the nicotine triggered changes in TH gene expression. The response of TH to nicotine has been proposed to be mediated by CREB. Nicotine treatment both in vivo and in vitro induces phosphorylation of CREB (6 -8). The cAMP/calcium response element (CRE/CaRE) in the TH promoter is required for its transcriptional activation by nicotine in PC12 cells (6). CREB is activated through the phosphorylation at its regulatory site on serine residue 133 (Ser-133) in response to cellular stimulations that result in increased [Ca 2ϩ ] i and/or cAMP levels (9 -11). Phosphorylated CREB (P-CREB) bound to the CRE/CaRE on the promoter of a target gene facilitates its transcription, an event called stimulus-transcription coupling (10,12,13). The phosphorylation of CREB on Ser-133 can be mediated by several kinases including protein kinase A (PKA), protein kinase C, CaM kinase II/IV, or MAP kinases (14).
The involvement of PKA was implicated by the finding that nicotine failed to increase TH mRNA levels in PKA-deficient PC12 cells or in the presence of adenylyl cyclase inhibitors (6,15). In addition, the phosphorylation of CREB in response to nicotine has been proposed to be mediated by MAP kinases (7,16). However, the temporal characteristics of nicotine-triggered CREB phosphorylation are not well established. The effects of nicotine on phosphorylation of CREB were generally examined during short term treatments for up to 1 h (6, 7) but not during the long term exposure. CRE-mediated gene expression was found to correlate more with sustained, rather than transient, P-CREB elevation (10). Although the long term response of P-CREB to nicotine was not yet examined, our previous findings suggest that long-rather than short term elevation of [Ca 2ϩ ] i is more pertinent for the activation of TH gene transcription. The addition of the calcium chelator BAPTA to PC12 cells even 2 h after the addition of nicotine still prevented the elevation in TH mRNA levels.
In addition to CREB, a number of other transcription factors such as CREM, ATF-1, and ATF-2 can also bind the CRE/CaRE element (17,18). Their role in the regulation of TH gene expression in the response to nicotine has not been well studied. Use of various chimeras of transcription factors with Gal4 indicated that in conjunction with CREBbinding protein, they can mediate CRE-dependent transcriptional activation of the TH promoter (19,20). ATF-1 was only marginally effective in increasing TH promoter activity, suggesting that it does not play a major role in the regulation of TH gene expression. ATF-2 has been implicated in transcriptional regulation of TH during neural development (21). Its expression in human brain is neuronal, and cell bodies of catecholamine-synthesizing neurons, such as the substantia nigra and locus coeruleus, are among the regions with especially high levels of ATF-2 expression (22). Therefore, ATF-2 may play a role in the regulation of TH gene expression.
To elucidate the molecular mechanisms and transcription factors mediating the induction of TH transcription by nicotine, we examined the temporal changes in phosphorylation of CREB and ATF-2 and the signaling pathways involved as well as their role in CRE/CaRE-mediated TH promoter. Our results suggest that both CREB and ATF-2 are required for nicotine-triggered elevation of TH transcription.
Cell Culture-PC12 cells were maintained in DMEM supplemented with 10% fetal bovine serum, 5% heat-inactivated donor horse serum, 50 g/ml streptomycin, and 50 IU/ml penicillin in a humidified atmosphere at 37°C and 7% CO 2 as described previously (15). Cells were maintained at a medium density (ϳ3 ϫ 10 5 /cm 2 ). For nicotine treatment, nicotine ditartrate dissolved in sterile water was added to obtain the desired final concentration.
Western Blot Analysis-Cells were collected in 0.5 ml of lysis buffer containing 125 mM Tris-HCl, pH 6.8, 25% glycerol, 2% SDS, 0.01% bromphenol blue, and 2% ␤-mercaptoethanol. Proteins from total cell lysates were resolved on 10% SDS-PAGE and transferred to nitrocellulose membranes (Trans-Blot Transfer Medium, Bio-Rad). Equal loading was ensured by Ponceau S staining. Membranes were blocked in phospho-buffered saline, 0.05% Triton X-100 containing 5% skim milk powder and were then probed overnight with specific primary antibodies (1:1000 phospho-CREB, 1:2000 phospho-ERK. or 1:1000 phospho-ATF-2). Antibodies were detected with the corresponding horseradish peroxidase-linked secondary antibodies. Blots were developed using SuperSignal West Pico chemiluminescent substrate (Pierce) detection reagents. Membranes were stripped with stripping buffer (2% SDS, 100 mM Tris-HCl, 0.1% ␤-mercaptoethanol) for 45 min at 60°C and reprobed with the corresponding antibodies to total CREB, ERK1/2, ATF-2, or ␤-actin for loading control. The membranes were then exposed to x-ray films for various time intervals. The images were cap-tured with a GS-800 calibrated densitometer (Bio-Rad), and the ratio was quantified by densitometric analyses within the linear range of each captured signal.
Plasmids and Transfections-The preparation of the constructs with the first 272 nucleotides of the rat TH promoter driving the expression of a firefly luciferase gene (p5ЈTH-Luc (Ϫ272/ϩ27)) has been previously described (23). The corresponding plasmids were mixed with SuperFect TM in a ratio of 5 g to 10 l according to the manufacturer's instructions (Qiagen), and then the mixtures were added to PC12 cells grown on 6-well plates in 0.6 ml of serum-free DMEM. The cells were incubated with the transfection mixtures for 3 h at 37°C in an incubator with 5% CO 2 . The transfection mixtures were removed, and the cells were washed twice with phospho-buffered saline, which was then replaced with 1 ml of DMEM containing 10% horse serum and 5% fetal bovine serum, and the cells were incubated for an additional 24 h at 37°C in humidified air containing 7% CO 2 and then treated with nicotine for another 16 h. The cells were harvested in 1 ml of phosphobuffered saline and collected by centrifugation.
Firefly luciferase activity was determined using luciferase reporter assay system from Promega. The PC12 cells were lysed with 40 l of passive lysis buffer, and 4 l of the lysate was added to 20 l of the luciferase substrate. Luminescence was measured immediately with a Luminometer, model TD-20/20 (Turner, Sunnyvale, CA). Luciferase activity was normalized to a protein concentration of the samples determined with the Bio-Rad protein assay system using the Bradford method. At least three or four cell culture plates were used for each treatment. All experiments were performed at least twice.
siRNA-We tested at least thee different annealing double-stranded siRNAs for CREB or ATF-2 genes to find the siRNA that has the strongest effect in reducing CREB or ATF-2 protein levels. The sense strand for siATF-2 was 5Ј-CCUUCUGUUGUAGAAACAAtt-3Ј, and the antisense strand was 5Ј-UUGUUUCUACAACAGAAGGtt-3Ј. The sense The cells lysates were subjected to Western blot analysis with antibodies to P-ERK1/2 or P-CREB. B, densitometry values for P-CREB/CREB. Nitrocellulose filters were stripped and incubated with ERK1/2 or CREB antibodies, respectively, for loading control. Representative Western blots are shown. Densitometry values are means of three to four gels. strand for siCREB was 5Ј-GGAGUCUGUGGAUAGUGUAtt-3Ј, and antisense strand was 5Ј-UACACUAUCCACAGACUCCtg-3Ј. Downregulation of total CREB or ATF-2 proteins was evaluated by Western blot using specific antibodies as mentioned above. We followed the guidelines (Qiagen) for co-transfection of adherent cells with siRNA and plasmid DNA using TransMessenger TM transfection reagent. First we diluted 2.4 l of Enhancer R in the appropriate volume of buffer EC-R followed by 0.1 g of siRNA and 0.2 g of plasmid DNA. After incubation at room temperature for 5 min, 2.5 l of TransMessenger transfection reagent was added to the siRNA-plasmid DNA-Enhancer R mixture. The mixture was incubated for 10 min and then added to PC12 cells grown on 24-well plates in 0.1 ml of serum-free DMEM. The cells were incubated with the transfection mixtures for 4 h at 37°C in an incubator with 5% CO 2 . The transfection mixtures were removed, the cells were washed twice with phospho-buffered saline, which was then replaced with 0.4 ml of DMEM containing 10% horse serum and 5% fetal bovine serum, and the cells were incubated for an additional 48 h at 37°C and 7% CO 2 and then treated with nicotine for another 16 h.
Isolation of RNA and RT-PCR-The levels of TH mRNA were determined by real-time RT-PCR as described before (24,25). Total RNA was isolated from quadruplicate cell cultures by using RNA-STAT 60 (Tel-Test, Friendswoods, TX), and the concentration was quantified by using Ribo-Green fluorescent dye (Molecular Probes, Eugene, OR). RNA (300 ng) was reverse-transcribed with avian myeloblastosis virus reverse transcriptase (Sigma) and 1 M specific reverse transcription primer for the TH gene, 5Ј-TCAGGCTCCTCTGACAG-3Ј, in 5 l of RT mixture.
RT was carried out at 42°C for 1 h followed by 10 min on 85°C. Quantitative real-time PCR was performed using Light Cycler System with SYBR Green buffer (Roche Applied Science) with the following primers: forward primer, 5Ј-GTGAACCAATTCCCCATG-3Ј; reverse primer, 5Ј-CAGTACACCGTGGAGAG-3Ј. The denaturation program (95°C for 7 min) was followed by a four-segment amplification and quantification program repeated 37 times (95°C for 3s; 59°C for 3 s; 72°C for 18 s; 85°C for 0 s for a single fluorescence measurement) and a melting curve program (70°C to 99°C) and, finally, a cooling program down to 40°C. The threshold cycle was determined using the Fit Points method to provide optimal standard curve values (0.98 -1.0). For quantification assays, a standard curve was used, produced by amplification of several 10-fold dilutions of linearized plasmids constructs containing TH cDNA. The results of real-time PCR were normalized to the amount of total RNA used in the PCR reaction.
Statistical Analysis-Statistical significance was determined by Student's t test for experiments with two groups or by performing an analysis of variance followed by the Fisher least significant difference test for experiments with more than two groups. A level of p Ͻ 0.05 was considered statistically significant.

Temporal Effect of Nicotine on Phosphorylation of ERK1/2 and Long
Term Activation of CREB-Our previous experiments indicated the importance of prolonged elevation of calcium for nicotine-induced changes in TH gene expression (15). Nicotine led to sustained elevation of [Ca 2ϩ ] i for up to 6 h of constant exposure. Therefore, we examined the effect of prolonged exposure to 200 M nicotine, a concentration leading to maximal induction of TH mRNA (6). Results shown in Fig. 1A reveal a sharp transient increase in phosphorylation of ERK1/2 by 5 min followed by a decline to basal levels with no change in the level of phosphorylation afterward. In contrast, although the level of P-CREB was elevated after 5 and 30 min of nicotine treatment and declined to near basal levels by 1 h of treatment, P-CREB increased again at the 2-h time point and rose even further for up to 5 h of exposure to nicotine. Densitometry (Fig. 1B) indicated that there was about a 5-fold increase in P-CREB with transient as well as prolonged treatment with nicotine. The results suggest that long term elevation of P-CREB occurs without concomitant activation of ERK1/2. Exposure to lower concentrations of nicotine (10 and 50 M) for up to 1 h also led to transient but weaker phosphorylation of CREB and ERK1/2 with similar kinetics (data not shown).
Phosphorylation of ERK1/2 Is Not Required for the Nicotine-triggered Regulation of TH Promoter Activity-The phosphorylation of ERK1/2 was found to be very transient, whereas phosphorylation of CREB was sustained with prolonged treatment in PC12 cells. Previous studies implicated the involvement of MAP kinases in the short term nicotinetriggered elevation in P-CREB and TH gene expression (7,8); therefore, we examined whether phosphorylation of ERK1/2 is also necessary for the nicotine-induced phosphorylation of CREB. Pretreatment for 15 min with 10 M MEK inhibitor U0126 prevented the phosphorylation of ERK1/2 triggered by 200 M nicotine as expected ( Fig. 2A). However, under the same conditions this inhibitor did not prevent the phosphorylation of CREB in nicotine-treated cells.
Because the phosphorylation of CREB is implicated in the nicotinetriggered activation of TH transcription, we investigated if nicotine can induce elevation of TH promoter-driven reporter activity in the presence of MEK inhibitor U0126. Cells were transfected with reporter constructs in which luciferase activity was controlled by the first 272 nucleotides of the rat TH promoter (p5ЈTH/Luc(Ϫ272/ϩ27)). This promoter region contains several regulatory elements including an AP-1 like motif and the Egr1/Sp1 site as well as a perfect consensus CRE/CaRE (TGACGTCA, at Ϫ45 to Ϫ38), which is necessary for the response to nicotine (6,26,27). As shown in Fig. 2B, after incubation with 200 M nicotine, there was a 2.5-fold increase in TH promoter-driven luciferase activity. In the presence of 10 M MEK inhibitor U0126, nicotine still was able to elevate TH promoter-driven luciferase activity. These results indicate that the MAP kinase pathway is not involved in the regulation of TH promoter activity in response to nicotine.   The effect of CaMKII inhibitor was even more pronounced. There was no elevation of the level of P-CREB on the 5Ј and no detectable P-CREB after 3 h of nicotine treatment (Fig. 3A). To further elucidate the role of PKA and CaMKII on TH transcription, we transfected PC12 cells with the TH promoter construct. The cells were pretreated with H-89 or KN-93 inhibitors for 30 min before nicotine was added, and respective luciferase activity was measured. Nicotine induced the TH promoter activity more than 2.5-fold compared with the untreated controls. However, both inhibitors prevented the effect of nicotine on TH promoter activity (Fig. 3B). Basal transcription levels were not changed with both inhibitors, suggesting that only the induction is affected.
Nicotine Triggered Long Term Phosphorylation of ATF-2 as Well as JNK-Next, we examined the effect of nicotine on phosphorylation of ATF-2. For these experiments PC12 cells were treated for different time periods with 200 M nicotine, and the levels of phosphorylation of ATF-2 were examined by Western blots using antibodies to P-ATF-2 (Th-71) and to total ATF-2 (Fig. 4). Short term incubation with nicotine (5-30 min) did not change the levels of P-ATF-2. However, both phosphorylation of ATF-2 and the protein levels were induced after 1 h of continual incubation with nicotine. They continued to increase and remained significantly elevated for up to 5 h (longest time examined) (Fig. 4A). Densitometry was used to estimate the level of induction, and the results are shown in Fig. 4B. The elevation of P-ATF-2 accompanied the rise in ATF-2 protein levels.
Stress-activated kinases, such as p38 (28) and JNK (29), are involved in the stimulation of ATF-2. Therefore, we used Western blotting to determine the level of phosphorylation of JNK with antibody against the P-JNK. We found an increase in the level of P-JNK, reaching maximal levels at 5 min and remaining high for 2 h, then slowly decreasing to the basal level at 5 h (Fig. 5A). The level of total JNK was not changed at any of the time points. Our experiments showed no change in the level of P-p38 (data not shown). Data from the densitometry of the changes in the levels of P-JNK and total JNK protein are shown (Fig. 5B).

Attenuation of the Nicotine-triggered Activation of TH Promoter by Dominant Negative CREB or ATF-2-To analyze whether ATF-2 and
CREB are directly involved in the regulation of TH gene expression in PC12 cells, the cells were co-transfected with p5ЈTH-Luc(Ϫ272/ϩ27) and either expression plasmid for d/n ATF-2 or d/n CREB. Untreated controls and nicotine-treated cells were co-transfected with p5ЈTH-Luc(Ϫ272/ϩ27) reporter construct and empty expression vector (pcDNA3). The results showed a marked decrease in TH promoter activity in cell cultures transfected with either of these dominant negative constructs (Fig. 6). The addition of nicotine increased TH promoter driven luciferase activity in control cells but did not elicit any significant change in the cells expressing d/n-ATF-2 or d/n CREB. Thus, both d/n-ATF-2 and d/n-CREB reduced the basal expression and prevented the response to nicotine. These results suggest that ATF-2 and CREB transcription factors are likely required for the nicotine-triggered regulation of TH gene promoter in PC12 cells.
Knockdown of CREB or ATF-2 Inhibits the Basal and Nicotine-stimulated TH Promoter Activity and the Level of TH mRNA-To further study the requirement for CREB and ATF-2, we used siRNAs to selectively knock down these genes. First we used fluorescent dye-labeled siRNA to determine the efficiency of siRNA uptake by transfection. PC12 cells were transfected with Cy TM 3-labeled glyceraldehyde-3phosphate dehydrogenase siRNA. One day after transfection, 90% of the cells were labeled, demonstrating successful uptake of the siRNA (Fig.  7A). Additional preliminary experiments were performed to evaluate the ability of CREB or ATF-2 siRNAs to inhibit the levels of targeted proteins and to select the appropriate siRNA concentration and incubation time. PC12 cells were transfected with CREB or ATF-2 siRNAs or scrambled RNA, and levels of CREB or ATF-2 proteins were determined by Western blots. After 2 days of transfection, both CREB and ATF-2 protein levels were greatly reduced compared with the controls (Fig. 7B). Therefore, cells were co-transfected with either CREB or ATF-2 siRNAs and p5ЈTH-Luc(Ϫ272/ϩ27). After 2 days, nicotine was added, and cells were further incubated for 16 h. Control cells were transfected with equal concentrations of scrambled RNA instead of the siRNA. Relative luciferase activity was significantly reduced, indicating that both CREB and ATF-2 transcription factors are required for the nicotine-triggered rise in TH promoter activity (Fig. 7C).  Because both CREB and ATF-2 siRNAs were able to attenuate TH promoter activity, it was important to verify the levels of endogenously expressed TH mRNA. PC12 cells were transfected with the respective siRNAs, and after 48 h the cells were treated with nicotine overnight.
Real time PCR revealed that both CREB or ATF-2 siRNAs effectively prevented nicotine-induced elevation of TH mRNA (Fig. 8). The basal level of TH mRNA was not affected by CREB and ATF-2 siRNAs.

DISCUSSION
This paper suggests that both CREB and ATF-2 transcription factors can play significant roles in the regulation of TH gene expression in response to nicotine. CREB as well as ATF-2 were found to be phosphorylated with nicotine treatment. Examination of the temporal changes revealed that CREB is rapidly and transiently phosphorylated with short term exposure to nicotine. Prolonged nicotine treatment led to a subsequent second sustained increase in CREB phosphorylation. ATF-2 was also phosphorylated in nicotine-treated cells, but the response was delayed and only evident with prolonged treatment (longer than 1 h). The elevation in P-ATF-2 was accompanied by increased ATF-2 protein levels. Although ERK1/2 was also phosphorylated in response to short term nicotine treatment, it was not required for the activation of CREB or for the induction of TH promoter activity. PKA inhibitor H-89 and CaMKII inhibitor KN-93 attenuated the nicotine-evoked phosphorylation of CREB as well as TH promoter activity in response to nicotine. Nicotine-stimulated induction of TH promoter activity was inhibited by dominant negative forms of either CREB or ATF-2 and by their knock down with selective siRNA.
Although CREB was previously shown to be phosphorylated in nicotine-treated PC12 cells (7,15), the effect of prolonged exposure to nicotine was not examined. This could be an important issue since the  ). B, Western blots were used to estimate the levels of CREB or ATF-2 proteins after transfection with siCREB or siATF-2 RNAs for 48 h. C, PC12 cells were co-transfected with p5ЈTH-Luc (Ϫ272/ϩ27) and siCREB or siATF-2 RNAs or scrambled RNA as a control. After treatment with nicotine (24 h), luciferase activity was measured and standardized to the respective protein levels. *, p Ͻ 0.05; **, p Ͻ 0.01, compared with the respective controls. #, p Ͻ 0.05; ##, p Ͻ 0.01 between different groups of nicotine-treated cells. duration of P-CREB maintenance can determine the range of genes affected (30). The induction of CRE-mediated gene expression is reported to correlate with sustained, rather than transient P-CREB in the nucleus (10).
In this study we found that the level of P-CREB was elevated about 5-fold after 5 and 30 min exposure to nicotine and declined to near basal levels by 1 h of treatment. There was a second sustained elevation of P-CREB at the 2-h time point that rose even further for up to 5 h of exposure to nicotine. In this regard, stimulation of CREB phosphorylation by retinoic acid in the 123.7 PC12 cell line as well as in A126 -1B2 cells was sustained with P-CREB levels and remained elevated for at least 8 h (31). CREB phosphorylation by prolonged nicotine treatment was observed in other cell types as well. Continual exposure to nicotine for 1 h induced sustained phosphorylation of CREB in the nuclei of about 50% of rat hippocampal cultured cells (32). Nicotine also elicited phosphorylation of CREB in ciliary ganglion neurons in culture in a dose-dependent fashion (33). Incubation with 10 -100 M nicotine for 5 min led to nuclear staining of P-CREB in 80% of the cells. When nicotine treatment was combined with blockage of voltage-gated calcium channels by cadmium, elevation of P-CREB was observed for as long as 3 h, which subsided to basal levels by 4 h.
Studies in vivo have given some indication that the phosphorylation of CREB in response to nicotine administration might be sustained. CREB phosphorylation in rat adrenal medulla was evident 6 h after a single injection of nicotine (34) or after several repeated injections 30 min apart (35). Phosphorylation of CREB was also increased in the prefrontal cortex by chronic exposure to nicotine, although it was decreased in the nucleus accumbens (8). Pandy et al. (36) examined P-CREB in different rat brain areas after 10 days of nicotine injections and found that the levels of P-CREB were mostly unchanged. However, significant reductions in CREB protein levels and in P-CREB were observed in several specific structures of the cortex and the amygdala after nicotine withdrawal (36).
Previous studies implicated a role for the MAP kinase pathway in CREB-mediated regulation of gene transcription by nicotine in PC12 cells (7,16) and chick cilliary ganglion cultures (33). Our results indicate that phosphorylation of ERK1/2 is most probably not a necessary event for the nicotine-induced phosphorylation of CREB since the MEK inhibitor, U0126, did not prevent the phosphorylation of CREB in nicotine-treated cells. There is no elevation of P-ERK1/2 with long term treatment, which is in contrast with the intensive elevation of P-CREB. Our experiments revealed that U0126 did not reduce the TH promoter response to nicotine, indicating that ERK1/2 kinases are not needed for the regulation of TH gene expression in response to nicotine. The apparent discrepancies between our results and a previous report (7) on the effect of U0126 on CREB phosphorylation may reflect differences in expression of specific nAChRs subtypes or signaling molecules necessary for "cross-talk" between MAP kinase pathway and CREB. In this regard, Nakayama et al. (39) found that the ␣3␤4 nAChR subtype, but not ␣7 nAChR, appears to be important for phosphorylation of ERK1/2 in PC12h cell subtype. Our earlier studies indicated that activation of the ␣7 nAChR is involved in the nicotine-triggered activation of TH gene expression in PC12 cells (37). Levels of ␣7 nAChR differ markedly among various populations of PC12 cells (38) and are reported to be low in the PC12h cells, the subtype used by Nakayama et al. (39). It is possible that ␣7 nAChR plays a critical role in the activation of CREB in an ERK1/2-independent manner in our cell model.
In contrast to ERK1/2 inhibitors, PKA inhibitor H-89 and CaMK inhibitor KN-93 greatly reduced the nicotine-triggered phosphorylation of CREB as well as TH promoter activity. The crucial importance of the PKA kinase in the regulation of TH gene expression by nicotine was previously noted. Pretreatment with the adenylyl cyclase inhibitor 2Ј,5Јdideoxyadenosine (DDA) prevented both the nicotine-elicited phosphorylation of CREB and the subsequent induction of TH mRNA (15). PKA-deficient PC12 cells treated with nicotine were unable to support many of the alterations in gene expression observed in normal cells, including the elevation of TH mRNA levels (6,40). The possible mechanism whereby activation of CaMKs regulates the expression of target genes is not well understood. We obtained results demonstrating that CaMK inhibitor KN-93 antagonized CREB activation by nicotine. It is known that CREB is one of the effectors for CaMK; for instance, it has been shown to phosphorylate CREB at the Ser-133 residue (18). Our results suggest that both CaMKII and PKA are involved in the activation of CREB with nicotine treatment.
It is well known that activation of CaMK is directly dependent on elevated intracellular calcium [Ca 2ϩ ] i , which also activates several different types of adenylyl cyclases, thus activating cAMP-dependent pathways, including PKA kinases. Many studies have demonstrated the ability of nicotine to elicit rapid elevations in [Ca 2ϩ ] i and activation of different signaling pathways (41,42). Our previous data demonstrated that several minutes after the initial transient rise in [Ca 2ϩ ] i , there was a second smaller elevation that was sustained for as long as 6 h with continual exposure to nicotine (15). The second elevation of [Ca 2ϩ ] i appears to be a necessary event, since the elevation in TH mRNA levels was inhibited by Ca 2ϩ chelator BAPTA-AM even when added 2 h after nicotine. Although a prolonged rise of [Ca 2ϩ ] i is needed for induction of TH mRNA by nicotine, such a sustained increase was not required for induction by elevated K ϩ or bradykinin (43). We also have shown before that ␣7 nAChR agonists, 3-(2,4-dimethoxybenzilidene)anabeseine (DMXB) and 3-CA, increased and maintained relatively sustained levels of [Ca 2ϩ ] i similar to the second elevation with nicotine, and they were at least as effective as nicotine in increasing TH mRNA levels (37).
The involvement of the ATF-2 transcription factor in TH gene expression in response to nicotine exposure has not been previously considered. We found that nicotine triggered a delayed but sustained elevation of activated ATF-2. ATF-2 contains a phosphorylation-dependent transcriptional activation domain on its N terminus (44) that has been shown to be activated via phosphorylation of threonine residues 69 and 71. Both phosphorylation and protein levels of ATF-2 were found to be elevated after 1 h of incubation with nicotine and increased even further for up to 5 h, reaching levels 6-fold higher than in the untreated cells. JNK and p38 were shown to phosphorylate ATF-2 on these threonine residues in the trans-activation domain (45). Nicotine did not alter activation of p38 at any of the times examined. However, JNK was found to be rapidly activated in nicotine-treated cells, and this activation persisted for at least 2 h and preceded the maximal phosphorylation of ATF-2. Thus, we can speculate that the JNK may mediate the nicotine-elicited increase in ATF-2 phosphorylation.
The present study showed that d/n ATF-2 or d/n CREB as well as knock-down of either CREB or ATF-2 with siRNA reduced the nicotine-triggered elevation of TH promoter-driven reporter activity, supporting a role for both transcription factors in the regulation of TH gene expression. The d/n ATF-2 or d/n CREB reduced both basal TH promoter activity and the induced response to nicotine. Previously, the CRE/CaRE motif of the TH promoter was shown to be crucial for basal as well as induced TH expression by several stimuli including cAMP, membrane depolarization, hypoxia, and nicotine (6,20,46). The results with d/n CREB on basal TH transcription are consistent with previous studies showing that basal transcription depends on transcription factor activity at the CRE promoter element (20). However, in contrast, siRNA for CREB or ATF-2 did not affect basal TH promoter activity or mRNA levels, suggesting that other CRE-binding proteins may replace them or that lower concentrations of these transcription factors are sufficient for basal activity. This is the first study to reveal that d/n ATF-2 can reduce basal TH promoter activity as well as the response to nicotine. ATF-2, like CREB, binds to the canonical CRE 5Ј-TGACGTCA-3Ј (47), which is identical to the CRE/CaRE at position Ϫ45 to Ϫ38 in the TH promoter (48,49). The d/n forms of CREB and ATF-2 used in this study contain mutations in the DNA binding domain. Therefore, both act as dominant repressors by forming inactive dimers, which are unable to bind CRE promoter elements. However, the dominant negative forms of CREB and ATF-2 will also inhibit the action of other transcription factors, which are able to heterodimerize with CREB and ATF-2, respectively. CREB can bind CREs as a homodimer or as a heterodimer with several other bZIP transcription factors including ATF-1 and CREM (for review, see Ref. 18). ATF-2 can also bind CREs as a homodimer or as a heterodimer with c-Jun or ATF-3 (50,51). If d/n ATF-2, by way of heterodimerization with CREB, prevented the binding of CREB to the CRE, its effect on TH promoter activity could be indirect. The interactions between ATF-2 and CREB are somewhat controversial. Dimerization was observed between CREB and intact ATF-2, but not ATF-2, with truncations at the N or C terminus. This led to a model of binding requiring the dimerization domains of CREB and ATF-2 as well a zinc finger domain near the C terminus of ATF-2 (52). Some recent experiments indicate that ATF-2 and CREB do not heterodimerize (53). However, these studies used mutant forms of ATF-2 containing the C-terminal portion of ATF-2 an N-terminal FLAG-tagged activation domain from CREB2. In addition, the b-ZIP region of ATF-2 was shown to bind the co-activator CREB-binding protein. This may dilute out the available CREB-binding protein for trans-activation of CREB during overexpression of d/n ATF-2. This could also play a role in the reduction of the response to nicotine. The experiments with siRNA point to a more direct requirement for both CREB and ATF-2 and do not suffer from the above-mentioned possible artifacts due to overexpression of dominant negative forms of the transcription factors. The results suggest that expression of either CREB or ATF-2 alone is not sufficient for the response of TH promoter to nicotine. Future studies remain to further elucidate the mechanism for interactions between ATF-2 and CREB and whether heterodimers between them are required for the regulation of TH gene expression by nicotine.