Cholinergic Stimulation of Early Growth Response-1 DNA Binding Activity Requires Protein Kinase C and Mitogen-Activated Protein KInase Kinase Activation and Is Inhibited by Sodium Valporate in SH-SY5y Cells

Authors


  • Abbreviations used : Egr, early growth response ; EMSA, electrophoretic mobility shift assay ; ERK1/2, extracellular-regulated kinases 1/2 ; MAP, mitogen-activated protein ; MEK1/2, mitogen-activated protein kinase kinases 1/2 ; PMA, phorbol 12-myristate 13-acetate.

Address correspondence and reprint requests to Dr. R. S. Jope at Department of Psychiatry and Neurobiology, Sparks Center, Room 1057, University of Alabama at Birmingham, Birmingham, AL 35294-0017, U.S.A.

Abstract

Abstract : Activation of muscarinic receptors in human neuroblastoma SH-SY5Y cells with carbachol stimulated a rapid and large increase in early growth response-1 (Egr-1, also called zif268 and NGF1-A) protein levels and DNA binding activity. Egr-1 DNA binding activity was stimulated within 15 min of treatment with carbachol and maintained a maximum 20-fold increase over basal between 1 and 2 h after treatment, and the EC50 was ~ μM carbachol. Carbachol-stimulated Egr-1 DNA binding activity was dependent on protein kinase C, as it was potently inhibited by GF109203× (IC50 ~0.1 mM) and was reduced by 85 ± 5% by down-regulation of protein kinase C. Inhibitors of increases in intracellular calcium levels reduced carbachol-induced Egr-1 DNA binding activity by 25-35%. Carbachol-stimulated activation of Egr-1 was reduced 35% by genistein, a tyrosine kinase inhibitor, and 60% by PD098059, an inhibitor of mitogen-activated protein kinase kinases 1/2 (MEK1/2) that activates extracellular-regulated kinases 1/2 (ERK1/2). A novel inhibitory action was caused by chronic (7-day) administration of sodium valproate but not by two other bipolar disorder therapeutic agents, lithium and carbamazepine. Valproate treatment reduced carbachol-stimulated Egr-1 DNA binding activity by 60% but did not alter carbachol-induced activation of ERK1/2 or p38 or increases in Egr-1 protein levels. These results reveal that muscarinic receptors activate Egr-1 through a signaling cascade primarily encompassing protein kinase C, MEK1/2, and ERK1/2 and that valproate substantially inhibits Egr-1 DNA binding activity stimulated by carbachol or protein kinase C.

Immediate early genes provide a crucial mechanism for rapidly linking signals generated by extracellular molecules to alterations of neuronal function. The early growth response (Egr ; also called zif268, krox 24, NGF1-A, and tis8) proteins are a widely distributed family of products of immediate early genes that include Egr-1, Egr-2, Egr-3, and other related proteins (Beckmann and Wilce, 1997). Various stimuli, such as excitatory amino acids, growth factors, and activators of protein kinase C, can induce neuronal expression of Egr proteins, which function as transcription factors with zinc finger DNA binding domains that bind to the common consensus sequence GCG(T/G)GGGCG (Christy and Nathans, 1989 ; O'Donovan et al., 1999). Several characteristics of Egr-1 suggest that it has a crucial role in neuronal function, as it is regulated by synaptic activity (Worley et al., 1991), has a lower threshold of activation than the prominent immediate early gene product c-fos (Worley et al., 1993), and appears to make an important contribution to neural plasticity, including influencing learning performance (Fordyce et al., 1994).

Cholinergic muscarinic M3 receptors are linked to the phosphoinositide signal transduction system that modulates protein kinase C activity and intracellular calcium levels, and cholinergic function is an integral component of learning and memory (Pacheco and Jope, 1996). These properties suggest that stimulation of muscarinic M3 receptors may influence the activation of Egr, in addition to its well-documented coupling to activation of the AP-1 transcription factor (Jope and Song, 1997). In accordance with this, there have been several reports that activation of muscarinic receptors increases Egr-1 mRNA levels (Arenander et al., 1989 ; Altin et al., 1991 ; Katayama et al., 1993 ; Hughes and Dragunow, 1994 ; Coso et al., 1995 ; Ebihara and Saffen, 1997) and that protein kinase C (Altin et al., 1989 ; Coso et al., 1995 ; Ebihara and Saffen, 1997) and influx of extracellular calcium (Ebihara and Saffen, 1997) contribute to this response. Only recently were increases in Egr-1 protein levels and Egr DNA binding activity demonstrated following muscarinic receptor activation in a study of receptor-transfected HEK293 cells, and these responses were partially inhibited by down-regulation of protein kinase C (von der Kammer et al., 1998). Thus, little is known about signaling cascades or modulatory influences regulating this important response to activation of muscarinic receptors.

Modulation of signaling cascades leading to regulation of transcription factor activities and gene expression appears to be important for the therapeutic responses to lithium, carbamazepine, and sodium valproate, agents used in the treatment of bipolar disorder (Jope, 1999). Recent reports have shown that lithium and sodium valproate influence the activation of the immediate early gene transcription factor AP-1 (Williams and Jope, 1995 ; Chen et al., 1997 ; Jope and Song, 1997 ; Ozaki and Chuang, 1997 ; Unlap and Jope, 1997 ; Asghari et al., 1998 ; Yuan et al., 1998), as well as the DNA binding activity of the cyclic AMP response element and NFκB transcription factors (Jope and Song, 1997 ; Ozaki and Chuang, 1997 ; Wang et al., 1999). For example, acute treatments of SH-SY5Y cells with lithium and sodium valproate have been reported to increase basal AP-1 DNA binding activity, and lithium inhibited AP-1 activation in response to stimulation of cholinergic muscarinic receptors (Jope and Song, 1997 ; Asghari et al., 1998 ; Yuan et al., 1998). Notably lacking are studies of the effects of any of these agents on the activation of Egr or of the effects of chronic administration of these agents, which is required for therapeutic responses, on the receptor-coupled stimulation of any transcription factor, except for a study of AP-1 in rat brain after pilocarpine administration (Williams and Jope, 1995).

The present investigation was undertaken to address the issues of muscarinic receptor signaling to Egr activation and modulation by agents therapeutic for bipolar disorder. Specifically, we tested if stimulation of endogenous M3 muscarinic receptors in human neuroblastoma SH-SY5Y cells was coupled to activation of Egr DNA binding activity, identified intermediates in the signaling pathway linking muscarinic receptors to activation of Egr-1, and determined if chronic administration of therapeutically relevant concentrations of lithium, carbamazepine, or sodium valproate modulated basal or stimulated Egr DNA binding activity.

MATERIALS AND METHODS

Materials

Reagents were obtained from the following sources : RPMI 1640 medium from Cellgro (Herndon, VA, U.S.A.) ; horse serum, L-glutamine, and penicillin/streptomycin from Life Technologies (Gaithersburg, MD, U.S.A.) ; fetal clone II from Hyclone (Logan, UT, U.S.A.) ; antibodies to Egr-1, Egr-2, and Egr-3 from Santa Cruz Biotechnology (Santa Cruz, CA, U.S.A.) ; carbachol, phorbol 12-myristate 13-acetate (PMA), diaminobenzidine, genistein, dantrolene, nickel chloride, lithium chloride, carbamazepine, and sodium valproate from Sigma (St. Louis, MO, U.S.A.) ; GF109203X, Gö6976, and Ro 31-8220 from Alexis Biochemicals (San Diego, CA, U.S.A.) ; BAPTA-AM, SB203580, and PD098059 from CalBiochem (La Jolla, CA, U.S.A.) ; antibodies to phospho-extracellular-regulated kinase 1/2 (phospho-ERK1/2) and phospho-p38 from New England Biolabs (Beverly, MA, U.S.A.) ; and Lumiglo chemiluminescent substrate from Kirkegaard and Perry Laboratories (Gaithersburg).

Cell culture

Human neuroblastoma SH-SY5Y cells were grown on Corning 100-mm-diameter tissue culture dishes (Corning, NY, U.S.A.) in RPMI 1640 medium containing 10% horse serum, 5% fetal clone II, 2 mM L-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. Cells were maintained in humidified, 37°C chambers with 5% CO2. Cells were plated at a density of 105 cells per 100-mm-diameter dish and were harvested ~48 h later, following treatments described in Results.

Isolation of nuclear extracts

SH-SY5Y cells were washed two times with 4 ml of phosphate-buffered saline and lysed with 4 ml of Nonidet P-40 lysis buffer [10 mM Tris-Cl (pH 7.4), 3 mM MgCl2, 10 mM NaCl, and 0.5% Nonidet P-40]. Cell lysates were centrifuged at 4,000 g for 5 min at 4°C. The supernatant was discarded, and the pellet containing nuclear material was resuspended in 50 μl of buffer [20 mM HEPES (pH 7.9), 20% glycerol, 0.3 M NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 1 mM dithiothreitol, 0.1 mMβ-glycerophosphate, 0.5 mM vanadate, 1 mM phenylmethane-sulfonyl fluoride, and 1 μg/ml each pepstatin A, leupeptin, and aprotinin]. After a 30-min extraction on ice, samples were centrifuged at 16,000 g for 15 min at 4°C. The supernatant containing nuclear extracts was transferred to a sterile microfuge tube, and protein concentrations were determined by the method of Bradford (1976).

Electrophoretic mobility shift assay (EMSA)

A 22-bp double-standed oligonucleotide containing the consensus sequence for Egr 5′-(ATGCCCGGCGCGGGGCGAGGG)-3′ was used for EMSAs (GIBCO, Grand Island, NY, U.S.A.). Double-stranded oligonucleotide (200 pmol) was radiolabeled by incubating for 1 h at 37°C in 20 μl containing 10 × GIBCO React 2 buffer, 0.5 mM deoxynucleotide triphosphate, DNA polymerase I (Klenow enzyme ; GIBCO), and 100 μCi of [α-32P]dCTP (Amersham, Arlington Heights, IL, U.S.A.). Following incubation, samples were diluted to 100 μl with sterile TE buffer [10 mM Tris-HCl (pH 8.0) and 1 mM EDTA], and free probe was removed by centrifugation at 1,000 rpm for 45 s on a Sephadex G-50 column.

DNA binding was measured by incubating nuclear extracts (10 μg of protein) in 20 μl of binding buffer containing 20 mM HEPES (pH 7.0), 4% glycerol, 500 mM KCl, 1 mM MgCl2, 0.5 mM dithiothreitol, 1 mg of poly(dI-dC), and ~10,000 cpm of radiolabeled Egr oligonucleotide for 30 min at 4°C. For supershift experiments, nuclear extracts were incubated with antibody (0.5 μg) to Egr-1, Egr-2, or Egr-3 for 30 min before incubation with binding buffer. Reaction mixtures were electrophoresed on 6% nondenaturing polyacrylamide gels in 0.25× TBE (22.3 mM Tris, 22.3 mM boric acid, and 0.5 mM EDTA) for 1.5 h at 150 V. The gels were then vacuum-dried, exposed to a phosphorscreen overnight, and quantitated using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA, U.S.A.).

Immunoblotting

Nuclear extracts (25 μg) were mixed with Laemmli sample buffer (2% sodium dodecyl sulfate) and placed in a boiling water bath for 10 min. Proteins were resolved in 7.5% sodium dodecyl sulfate-polyacrylamide gels, transferred to nitrocellulose, and incubated with polyclonal Egr-1, Egr-2, or Egr-3 antibody (1 : 1,000). Immunoblots were developed using horseradish peroxidase-conjugated secondary antibody followed by incubation in diaminobenzidine solution. Phosphotyrosine immunoblots were incubated with polyclonal phospho-ERK1/2 or phospho-p38 antibody and were developed using horseradish peroxidase-conjugated secondary antibody followed by enhanced chemiluminescence.

Statistical analysis

All treatments were tested at least three times with separate batches of cells, and statistical significance was determined by ANOVA using the INSTANT program.

RESULTS

Stimulation of muscarinic receptors increases Egr-1 DNA binding activity

To investigate the effects of cholinergic stimulation on Egr activation, DNA binding activity was measured in nuclear extracts prepared from human neuroblastoma SH-SY5Y cells that had been treated with the cholinergic agonist carbachol (1 mM for 1 h). Carbachol treatment significantly increased Egr DNA binding activity (Fig. 1). An antibody specific for Egr-1 completely supershifted the carbachol-stimulated band to one of slower mobility, whereas antibodies specific for Egr-2 and Egr-3 only slightly diminished the band intensity. Therefore, carbachol stimulated Egr DNA binding activity in SH-SY5Y cells, and Egr-1 was the primary protein in the complex.

Figure 1.

EMSA and supershift analysis of Egr DNA binding activity in SH-SY5Y cells. SH-SY5Y cells were treated without (Ctl) or with 1 mM carbachol for 1 h, nuclear extracts from carbacholtreated cells were preincubated with antibodies for Egr-1, Egr-2, or Egr-3, and Egr DNA binding activity was measured using EMSA analysis as described in Materials and Methods. Egr-1 DNA binding activity, the supershifted (SS) band, and a nonspecific (ns) band are indicated with arrows.

FIG. 1.

Figure 2 shows the time course (15 min-6 h) of carbachol-stimulated Egr-1 DNA binding activity. Stimulation of Egr-1 DNA binding activity was evident within 15 min of treatment with carbachol, a maximal increase that was ~2,000% of control values was maintained between 1 and 2 h of treatment, and this was followed by a rapid decline to ~500% of control following 3 h of treatment with carbachol. Preincubation with the muscarinic antagonist atropine (10 μM) completely blocked carbachol-stimulated Egr-1 DNA binding activity, confirming that Egr-1 activated in response to carbachol was due to stimulation of muscarinic receptors.

Figure 2.

Time course of carbachol (Cb)-stimulated Egr-1 DNA binding activity. Egr-1 DNA binding activity was measured by EMSA analysis in nuclear extracts from SH-SY5Y cells incubated with 1 mM Cb for 15, 30, 60, 120, 180, or 360 min. A : A representative EMSA of the time course of Cb-stimulated Egr-1 DNA binding activity. Ctl, untreated control. Also shown is a representative EMSA demonstrating that 10 μM atropine (Atr) blocked Cb-stimulated Egr-1 DNA binding activity (measured 1 h after treatment). B : Quantitation of Egr-1 DNA binding activity. Values are given as percentages of Egr-1 DNA binding activity in Ctl cells. Data are mean ± SEM (bars) values (n = 3). *p < 0.05 compared with Ctl cells.

FIG. 2.

The relationship between Egr-1 DNA binding and Egr-1 protein levels was examined by using quantitative immunoblots to determine the effects of carbachol on Egr-1 protein levels. Carbachol treatment increased Egr-1 protein levels in whole-cell lysates (data not shown) and in nuclear extracts (Fig. 3A and B), and the time course of this increase paralleled that of Egr-1 DNA binding after carbachol treatment. Although a labeled band corresponding to Egr-3 DNA binding activity was not observed in the EMSAs, carbachol stimulated a large increase in the level of Egr-3 in nuclear extracts (Fig. 3C). Egr-2 was not detected in immunoblots of nuclear extracts with or without carbachol treatment (data not shown).

Figure 3.

Egr-1 protein levels increase after carbachol treatment. Nuclear extracts were prepared from SH-SY5Y cells treated with 1 mM carbachol for 0, 15, 30, 60, 120, or 180 min. A : A representative immunoblot of carbachol-induced nuclear Egr-1 protein levels compared with untreated control (Ctl) cells. B : Quantitation of carbachol-stimulated Egr-1 protein levels in nuclear extracts. Data are mean ± SEM (bars) values (n = 4). *p < 0.05 compared with Ctl cells. C : An immunoblot of carbachol-induced nuclear Egr-3 protein levels.

FIG. 3.

The concentration dependence of carbachol-stimulated Egr-1 DNA binding activity was investigated in SH-SY5Y cells. Preliminary experiments indicated that maximal Egr-1 DNA binding activity was induced by carbachol concentrations between 10 and 100 μM ; therefore, carbachol concentrations of 0.1, 1, 3, 10, and 100 μM were used. These experiments revealed that the EC50 for carbachol-stimulated Egr-1 DNA binding activity was ~1 μM, and maximal stimulation was achieved following treatment with ~10 μM carbachol (Fig. 4).

Figure 4.

Carbachol concentration-dependent stimulation of Egr-1 DNA binding activity. Egr-1 DNA binding activity was measured by EMSA analysis in nuclear extracts from SH-SY5Y cells incubated with 0.1, 1, 3, 10, or 100 μM carbachol for 1 h. Values are given as percentages of Egr-1 DNA binding activity in untreated cells (controls). Data are mean ± SEM (bars) values (n = 5).

FIG. 4.

Signaling intermediates mediating carbachol-stimulated Egr-1 DNA binding activity

The roles of the second messengers produced by the muscarinic receptor-coupled phosphoinositide signaling system were examined in the mediation of carbachol-stimulated Egr-1 DNA binding activity and protein levels in SH-SY5Y cells. The involvement of protein kinase C, which is activated by stimulation of the phosphoinositide signaling system, was investigated initially using the protein kinase C activator PMA and the inhibitor GF109203X. Exposure of cells to 0.5 μM PMA for 1 h resulted in a substantial incresae in Egr-1 DNA binding activity, and this was completely blocked by pretreatment (5 min) with 10 μM GF109203X (Fig. 5A and B). GF109203X also concentration-dependently decreased carbachol-induced Egr-1 DNA binding activity, with the low concentration of 0.1 μM GF109203X causing a 45% inhibition of the response to carbachol and 10 μM GF109203X causing complete blockade. Immunoblots revealed that GF109203X also inhibited carbachol- and PMA-stimulated increases in nuclear Egr-1 protein levels (Fig. 5C). Carbachol-induced increases in Egr-1 DNA binding activity also were reduced by other treatments that inhibit protein kinase C. These include the selective inhibitors Ro 31-8220 and Gö6976, as well as down-regulation of protein kinase C caused by a 24-h pretreatment with 1 μM PMA (Fig. 5D). The effects of PMA-induced down-regulation of protein kinase C are especially notable because this treatment may be the most reliable for specifically reducing protein kinase C activity, and it inhibited carbachol-stimulated Egr-1 DNA binding activity by 85 ± 5% (n = 5). These findings demonstrate that protein kinase C activity is necessary for carbachol-stimulated Egr-1 DNA binding activity.

Figure 5.

Inhibition of protein kinase C blocks carbachol- and PMA-induced Egr-1 DNA binding activity and increased nuclear Egr-1 protein levels. A-C : Egr-1 DNA binding activity and nuclear Egr-1 levels were examined by EMSA and immunoblot analysis, respectively, in nuclear extracts prepared from SH-SY5Y cells pretreated with 0, 1, 3, 10, or 30 μM GF109203× (GF109) for 5 min followed by incubation with 1 mM carbachol or 0.5 μM PMA for 1 h. A : A representative EMSA of the inhibition by GF109 of carbachol- and PMA-induced Egr-1 DNA binding activity. Ctl, untreated control. B : Quantitation of the effects of GF109 on carbachol-stimulated Egr-1 DNA binding activity. Values are given as percentages of Egr-1 DNA binding activity in cells treated with carbachol alone. Data are mean ± SEM (bars) values (n = 3-5). *p < 0.05 compared with carbachol-stimulated Egr-1 DNA binding activity. C : A representative immunoblot demonstrates the inhibition by GF109 of carbachol- and PMA-induced nuclear Egr-1 protein levels. D : Egr-1 DNA binding activity was measured in cells treated with the protein kinase C inhibitors Ro 31-8220 (Ro31 ; 15 min) or Gö6976 (15 min) at 3 and 10 μM or after protein kinase C down-regulation induced by a 24-h treatment with 1 μM PMA, followed by incubation with 1 mM carbachol (for 1 h).

Figure 6.

Effects of calcium modulators on carbachol-stimulated Egr-1 DNA binding activity. Egr-1 DNA binding activity was measured in nuclear extracts from SH-SY5Y cells preincubated with 20 μM BAPTA-AM (BAPTA ; for 30 min), 1 μM dantrolene (Dant ; for 10 min), 2 mM nickel (NiCl ; for 10 min), or 30 μM KN-62 (for 10 min), followed by incubation with carbachol (1 mM for 1 h). Values are given as percentages of Egr-1 DNA binding activity in carbachol-treated cells. Data are mean ± SEM (bars) values (n = 3). *p < 0.05 compared with carbachol-stimulated Egr-1 DNA binding activity.

FIG. 5.

FIG. 6.

The role of calcium was examined in the Egr-1 response to carbachol treatment because intracellular calcium concentrations are increased following stimulation of the phosphoinositide signaling system. As shown in Fig. 6, preincubation of SH-SY5Y cells with BAPTA-AM (20 μM) caused a 35% reduction in carbachol-induced Egr-1 DNA binding activity. Also, pretreatment with dantrolene (1 μM), an inhibitor of calcium release from the endoplasmic reticulum, had no effect on the Egr-1 response. Pretreatment with nickel chloride (2 mM), an inhibitor of agonist-induced calcium influx, or KN-62, an inhibitor of calcium/calmodulin-dependent kinase II, caused 25-32% reductions in carbachol-induced Egr-1 DNA binding activity. These results indicate that an increased intracellular calcium level is necessary for optimal activation of Egr-1 by carbachol, although this response demonstrated a greater dependence on protein kinase C activity.

The involvement of protein tyrosine kinase activity and mitogen-activated protein (MAP) kinases in the Egr-1 response to carbachol was examined using selective protein kinase inhibitors. The tyrosine kinase inhibitor genistein (100 μM for 10 min) significantly attenuated carbachol-stimulated Egr-1 DNA binding activity (Fig. 7A). PD098059 is an inhibitor of MAP kinase kinases 1/2 (MEK1/2), which activates ERK1/2. Examination of the inhibition of ERK1/2 activation by 10, 25, and 50 μM PD098059 revealed that 50 μM PD098059 was necessary to obtain substantial inhibition (Fig. 7B) and 50 μM PD098059 caused a 60% inhibition of carbachol-stimulated Egr-1 DNA binding activity. However, pretreatment with SB203580 (20 μM for 10 min), an inhibitor of p38, had no effect on carbachol-stimulated Egr-1 DNA binding activity. These findings indicate that ERK1/2 activation is necessary for full muscarinic-receptor stimulated activation of Egr-1.

Figure 7.

Signaling intermediates mediating carbachol-stimulated Egr-1 DNA binding activity. A : Egr-1 DNA binding activity was measured by EMSA analysis in nuclear extracts from SH-SY5Y cells pretreated for 10 min with 100 μM genistein (Gen), 50 μM PD098059 (PD), or 20 μM SB203580 (SB) followed by incubation with 1 mM carbachol for 1 h. Values are given as percentages of Egr-1 DNA binding activity in carbachol-treated cells. Data are mean ± SEM (bars) values (n = 3-5). *p < 0.05 compared with carbachol-stimulated Egr-1 DNA binding activity. B : A representative immunoblot (n = 3-4) of ERK1/2 phosphotyrosine immunoreactivity measured in cell lystates from cells pretreated with 0, 10, 25, or 50 μM PD (for 30 min) followed by incubation with 1 mM carbachol for 15 min. Ctl, untreated control.

FIG. 7.

Chronic sodium valproate treatment inhibits stimulated Egr-1 DNA binding activity

The effects of chronic (7-day) pretreatment with therapeutically relevant concentrations of lithium, carbamazepine, or sodium valproate on carbachol-stimulated Egr-1 DNA binding activity were examined. Chronic treatment with lithium (1 mM) or carbamazepine (0.05 mM) did not alter carbachol-stimulated Egr-1 DNA binding activity (Fig. 8). However, chronic treatment with sodium valproate (0.5 mM) caused a large inhibition (60 ± 5%) of carbachol-stimulated Egr-1 DNA binding activity. This was not due to a direct effect of sodium valproate on DNA binding by Egr-1 because inclusion of 0.5 mM valproate in the binding assay with nuclear extracts from carbachol-treated cells did not alter Egr-1 DNA binding (data not shown). As expected from the EMSA results, pretreatment with neither lithium nor carbamazepine altered carbachol-stimulated increases in Egr-1 protein levels in nuclear extracts (data not shown). Unexpectedly, however, sodium valproate treatment also did not affect increases in the nuclear Egr-1 protein levels induced by carbachol in the same samples in which valproate drastically reduced Egr-1 DNA binding activity (Fig. 9A).

Figure 8.

Effects of chronic treatment with lithium (Li), carbamazepine (CBZ), or sodium valproate (VPA) on carbachol-stimulated Egr-1 DNA binding activity. Egr-1 DNA binding activity was measured in nuclear extracts from SH-SY5Y cells treated with 100 μM carbachol with or without 7-day pretreatments with Li, CBZ, or VPA. Values are given as percentages of Egr-1 DNA binding activity stimulated by carbachol in cells not treated with antibipolar drugs. Data are mean ± SEM (bars) values (n = 5). *p < 0.05 compared with carbachol-stimulated Egr-1 DNA binding activity.

Figure 9.

Effects of chronic sodium valproate (VPA) treatment on carbachol (Cb)-induced nuclear Egr-1 protein levels and PMA-induced Egr-1 DNA binding activity and nuclear protein levels. Egr-1 DNA binding activity and nuclear Egr-1 levels were measured by EMSA and immunoblot analysis, respectively, in SH-SY5Y cells treated chronically with VPA. A : A representative immunoblot (n = 3) of nuclear Egr-1 levels from untreated control (Ctl) cells or cells pretreated with 0.5 mM VPA for 7 days followed by incubation with 1 mM Cb for 1 h. B and C : A representative EMSA (n = 4) and immunoblot (n = 3), respectively, of Ctl cells compared with cells pretreated with 0.5 mM VPA for 7 days followed by incubation with 0.5 μM PMA for 1 h.

FIG. 8.

FIG. 9.

Because protein kinase C was found to play a predominant role in mediating carbachol-stimulated activation of Egr-1 DNA binding activity, the effects of sodium valproate were tested on protein kinase C-activated Egr-1 to determine if its inhibitory effect on carbachol-stimulated Egr-1 DNA binding activity was targeted upstream or downstream of protein kinase C. Chronic valproate treatment (0.5 mM for 7 days) reduced PMA-induced Egr-1 DNA binding activity by 50 ± 6% (Fig. 9B). However, similar to results obtained for carbachol-stimulated Egr-1, valproate did not alter PMA-induced nuclear Egr-1 protein levels (Fig. 9C).

To determine if valproate inhibited carbachol-stimulated Egr-1 DNA binding activity upstream or downstream of MAP kinases, the phosphotyrosine immunoreactivities of ERK1/2 and p38 were examined. Treatment with carbachol stimulated large increases in the phosphotyrosine immunoreactivities of ERK1/2, a response that was unaltered by chronic valproate treatment, and p38, which was significantly increased in cells treated chronically with valproate (Fig. 10). Thus, the inhibition by sodium valproate of carbachol-stimulated Egr-1 DNA binding activity cannot be attributed to an inhibitory effect of valproate on the activation of ERK1/2 or p38.

Figure 10.

Effect of chronic valproate (VPA) treatment on ERK1/2 and p38 phosphotyrosine immunoreactivities. ERK1/2 and p38 phosphotyrosine immunoreactivities were measured by immunoblot analysis of cell lysates prepared from SH-SY5Y cells pretreated with 0.5 mM sodium VPA for 7 days followed by incubation with 1 mM carbachol (Cb) for 15 min. Chronic VPA treatment did not significantly alter activation of ERK1/2, but the carbachol-stimulated increase in p38 phosphotyrosine immuno-reactivity was 112% greater (p < 0.05 ; n = 5) after treatment with chronic sodium VPA. Ctl, untreated control.

FIG. 10.

DISCUSSION

Transcription factors are pivotal in cell signaling processes, relaying signals generated by extracellular agents to the nucleus to regulate gene expression. Egr-1 is one of a limited number of transcription factors encoded by immediate early genes that can be activated very rapidly (Beckmann and Wilce, 1997 ; O'Donovan et al., 1999), and it has been shown to be linked to synaptic activity and neural plasticity (Worley et al., 1991 ; Fordyce et al., 1994). The present study demonstrated that muscarinic receptor stimulation induces a rapid and robust increase in Egr-1 protein levels and Egr-1 DNA binding activity and that this stimulation is dependent on the activation of protein kinase C, increased intracellular calcium, tyrosine kinase activity, and the MAP kinase signaling cascade. Furthermore, a therapeutically relevant pretreatment with sodium valproate was found to inhibit stimulated Egr-1 DNA binding activity.

Muscarinic receptor stimulation has been reported to induce Egr-1 mRNA levels in rat astrocytes (Arenander et al., 1989), PC12 cells (Altin et al., 1991), NG108-15 cells (Katayama et al., 1993), receptor-transfected NIH 3T3 cells (Coso et al., 1995), PC12D cells (Ebihara and Saffen, 1997), and receptor-transfected HEK293 cells (von der Kammer et al., 1998). The present study demonstrated that the time course of muscarinic receptor-stimulated increases in Egr-1 protein levels and DNA binding activity was more prolonged than that of previously reported increases in Egr-1 mRNA levels. For example, in astrocytes, PC12D cells, and NIH 3T3 cells, muscarinic receptor stimulation increased the Egr-1 mRNA level maximally after 1 h, but within 2 h it had returned to the basal level (Arenander et al., 1989 ; Coso et al., 1995 ; Ebihara and Saffen, 1997). In contrast, it is evident that a much longer functional response was attained by muscarinic receptor stimulation, as Egr-1 DNA binding activity remained significantly elevated for at least 6 h in carbachol-stimulated SH-SY5Y cells. Thus, muscarinic receptor stimulation causes a rapid and longlasting activation of Egr-1.

The signaling intermediates involved in muscarinic receptor stimulation of Egr-1 have not been definitively established. Muscarinic receptor stimulation of Egr-1 mRNA levels was reported to require protein kinase C activation (Altin et al., 1989 ; Coso et al., 1995 ; Ebihara and Saffen, 1997) and the influx of extracellular calcium (Ebihara and Saffen, 1997). von der Kammer et al. (1998) found that carbachol-stimulated Egr-1 mRNA and protein levels both are at least partially dependent on protein kinase C activation. In agreement with this, the protein kinase C inhibitor GF109203X potently inhibited carbachol-induced increases in Egr-1 protein levels, and this paralleled the inhibition of carbachol-stimulated Egr-1 DNA binding activity. Other inhibitors of protein kinase C also blocked activation of Egr-1 stimulated by carbachol, with PMA-induced down-regulation of protein kinase C causing almost complete inhibition of this response. Increases in intracellular calcium levels also were necessary for optimal activation of Egr-1, but this response was less dependent on calcium than on protein kinase C, as maximal reductions of 25-35% were obtained with calcium inhibitors. Although inhibition of calcium/calmodulin-dependent kinase II by KN-62 was reported to block calcium-induced Egr-1 mRNA expression in PC12 cells (Enslen and Soderling, 1994) and in ELM-I-1 cells (Schaefer et al., 1998), KN-62 reduced carbachol-stimulated Egr-1 DNA binding activity by 25%. These results indicate that muscarinic receptor stimulation of Egr-1 DNA binding activity and nuclear protein levels is highly dependent on protein kinase C activation, but less dependent on calcium mobilization, in human neuroblastoma SH-SY5Y cells.

Muscarinic receptor stimulation activates multiple intermediary kinases, such as MEK1/2, ERK1/2, and p38 (Gutkind, 1998). The tyrosine kinase inhibitor genistein reduced carbachol-stimulated Egr-1 DNA binding activity, demonstrating the necessity for tyrosine kinase activation in muscarinic receptor-linked activation of Egr-1. This is in accordance with a study by Humblot et al. (1997) demonstrating that increased Egr-1 mRNA levels and DNA binding activity in response to stimulation of serotonin 5-HT2 receptors, which are coupled to the phosphoinositide signal transduction system, require protein tyrosine kinase activity in PC12 cells. PD098059, an inhibitor of MEK1/2, recently was reported to attenuate growth hormone-induced Egr-1 gene expression in 3T3-F442A cells (Hodge et al., 1998) and inhibited lysophosphatidic acid-induced Egr-1 mRNA expression in rat mesangial cells (Reiser et al., 1998). In SH-SY5Y cells, PD098059 significantly reduced carbachol-stimulated Egr-1 DNA binding activity and nuclear protein levels. SB203580, an inhibitor of p38, reduced anisomycin-induced Egr-1 mRNA expression in NIH3T3 cells (Lim et al., 1998) but had no effect on lysophosphatidic acidinduced Egr-1 mRNA levels in PC12 cells (Hodge et al., 1998) or on carbachol-stimulated Egr-1 DNA binding activity or nuclear protein levels in this study. Taken together, these results indicate that activation of MEK1/2, which is coupled to activation of ERK1/2, but not of p38 kinase, is required for stimulation of Egr-1 DNA binding activity in response to carbachol in human neuroblastoma SH-SY5Y cells.

Evidence obtained during the last few years indicates that modulation of signaling cascades, transcription factor activities, and gene expression may contribute to the therapeutic effects of drugs used to treat bipolar disorder, which include lithium, sodium valproate, and carbamazepine (Jope, 1999). Although inhibitory actions of these drugs on transcription factors (predominantly AP-1) have been reported, as described in the introductory section, no reports have appeared concerning activation of Egr-1 or direct receptor-coupled activation of any transcription factor after chronic administration of any of these agents, although chronic administration is necessary for therapeutic responses. This study found that chronic treatment with a therapeutically relevant concentration of sodium valproate greatly reduced carbacholinduced Egr-1 DNA binding activity, which is in marked contrast to the lack of effects of lithium or carbamazepine. Furthermore, a novel inhibitory mechanism accounted for this effect of valproate, as it did not inhibit the signaling cascade originating at the muscarinic receptor and encompassing activation of protein kinase C and ERK1/2, which led to increased levels of Egr-1 protein, but the functional DNA binding activity of Egr-1 following stimulation of muscarinic receptors or of protein kinase C was markedly inhibited. Although the mechanisms that regulate the competency of Egr-1 binding to DNA remain to be identified, this effect of valproate provides a precise mechanism whereby Egr-1 activity can be regulated by valproate in isolation from other downstream effectors of the muscarinic receptor-coupled signaling cascades.

Although activation of muscarinic receptors linked to the phosphoinositide signal transduction system stimulates both AP-1 and Egr-1 DNA binding activities in SH-SY5Y cells, some intriguing differences in the signaling cascades and regulation of these responses have been revealed. Identical treatments with nickel, which blocks the plateau phase of carbachol-induced increases in intracellular calcium, or KN-62, an inhibitor of calcium/calmodulin-dependent kinase II, cause at least twice as great an inhibition of carbachol-induced AP-1 than of Egr-1 DNA binding activity (Jope and Song, 1997 ; Pacheco and Jope, 1998). Thus, activation of Egr-1 appears less dependent than that of AP-1 on the calcium arm of the phosphoinositide signaling system in SH-SY5Y cells. These findings are in accordance with a previous report that muscarinic receptor-coupled activation of c-Jun kinase displayed a much greater dependence on modulation by an increased intracellular calcium than did activation of ERK1/2 (Mitchell et al., 1995). Also, whereas 7 days of treatment with sodium valproate greatly diminished carbachol-induced Egr-1 DNA binding activity, but lithium and carbamazepine were ineffective, carbachol-induced AP-1 DNA binding activity was unaffected by 7 days of treatment with sodium valproate but was inhibited by ≥50% by 7 days of treatment with lithium or carbamazepine (Pacheco and Jope, 1998). Thus, the three bipolar disorder therapeutic agents each inhibit muscarinic receptor-mediated transcription factor activation but with varying selectivities for Egr-1 and AP-1.

In conclusion, this study presents two novel findings concerning Egr-1 activation. First, the results define a signaling pathway of muscarinic receptor stimulation leading to Egr-1 activation that primarily comprises protein kinase C, MEK1/2, and ERK1/2. Second, the data provide the first evidence that the antibipolar agent sodium valproate substantially attenuates carbachol-stimulated Egr-1 DNA binding activity and that this inhibition occurs through a novel mechanism other than reducing nuclear Egr-1 protein levels. Clearly, the robust Egr-1 activation in response to muscarinic receptor stimulation points to Egr-1 as a significant downstream target of this pathway. Also, the substantial inhibition of muscarinic receptor-stimulated Egr-1 DNA binding activity by therapeutic concentrations of sodium valproate provides an effect of valproate that may be relevant to its therapeutic mechanism of action in bipolar disorder.

Acknowledgements

This work was supported by grant MH 38752 from the National Institutes of Health.

Ancillary