Induction of acetylcholine receptor gene expression by ARIA requires activation of mitogen-activated protein kinase.

Transcription of genes encoding nicotinic acetylcholine receptor (AChR) subunits (α, β, γ or ϵ, and δ) is highest in nuclei localized to the synaptic region of the muscle, which contributes to maintain a high density of AChRs at the postjunctional membrane. ARIA (AChR inducing activity) is believed to be the trophic factor utilized by motor neurons to stimulate AChR synthesis in the subsynaptic area. To elucidate the signaling mechanism initiated by ARIA, we established stable C2C12 cell lines carrying the nuclear lacZ gene under the control of the mouse ϵ subunit promoter or chicken α subunit promoter. ARIA stimulated tyrosine phosphorylation of erbB proteins in these C2C12 cells within 15 s with a peak at 5 min. Immediately following tyrosine phosphorylation of erbB proteins, mitogen-activated protein (MAP) kinase was activated which occurred within 30 s and peaked at 8 min after ARIA stimulation. Concomitantly, expression of AChR genes was induced by ARIA. ARIA-induced AChR subunit transgene expression was observed only in differentiated myotubes and not in myoblasts, suggesting that downstream signaling component(s) are regulated in a manner dependent on the myogenic program. Inhibition of the MAP kinase activity by using a specific MAP kinase kinase inhibitor or by overexpressing dominant negative mutants of Raf or MAP kinase kinase attenuated or abolished the ARIA-induced activation of AChR α and ϵ subunit gene expression. These results indicate that regulation of AChR gene expression by ARIA in C2C12 cells requires activation of the MAP kinase signaling pathway.

Transcription of genes encoding nicotinic acetylcholine receptor (AChR) subunits (␣, ␤, ␥ or ⑀, and ␦) is highest in nuclei localized to the synaptic region of the muscle, which contributes to maintain a high density of AChRs at the postjunctional membrane. ARIA (AChR inducing activity) is believed to be the trophic factor utilized by motor neurons to stimulate AChR synthesis in the subsynaptic area. To elucidate the signaling mechanism initiated by ARIA, we established stable C2C12 cell lines carrying the nuclear lacZ gene under the control of the mouse ⑀ subunit promoter or chicken ␣ subunit promoter. ARIA stimulated tyrosine phosphorylation of erbB proteins in these C2C12 cells within 15 s with a peak at 5 min. Immediately following tyrosine phosphorylation of erbB proteins, mitogen-activated protein (MAP) kinase was activated which occurred within 30 s and peaked at 8 min after ARIA stimulation. Concomitantly, expression of AChR genes was induced by ARIA. ARIA-induced AChR subunit transgene expression was observed only in differentiated myotubes and not in myoblasts, suggesting that downstream signaling component(s) are regulated in a manner dependent on the myogenic program. Inhibition of the MAP kinase activity by using a specific MAP kinase kinase inhibitor or by overexpressing dominant negative mutants of Raf or MAP kinase kinase attenuated or abolished the ARIAinduced activation of AChR ␣ and ⑀ subunit gene expression. These results indicate that regulation of AChR gene expression by ARIA in C2C12 cells requires activation of the MAP kinase signaling pathway.
The AChRs 1 mediate postsynaptic depolarization at the neuromuscular junction. This ligand-gated ion channel is a 250-kDa pentameric complex of four different subunits in a stoichiometry of ␣(2), ␤, ␥ or ⑀, and ␦ (1, 2). The AChRs are highly enriched at the crests of the postjunctional folds at a concentration of 10 4 receptors/m 2 , which is at least 1,000-fold higher than in the extrasynaptic membrane (3,4). During early devel-opment, however, AChRs appear throughout the membrane in myotubes. Innervation of myotubes leads to an increase in AChR synthesis, clustering of AChRs at the synapse, and subsequently to loss of extrasynaptic AChRs (4). The adult-type AChR contains an ⑀ subunit in place of ␥ and displays different electrophysiological properties.
A description of the molecular mechanisms underlying synapse formation at the neuromuscular junction is far from complete. Synthesis of AChRs is stimulated by ARIA, a factor that was identified initially in chick brain on the basis of its ability to promote AChR synthesis in cultured skeletal muscle cells (5). This factor is a 42-kDa protein that has sequence homology with the Neu differentiation factor (NDF), heregulin, and glial growth factors (GGF), all of which are recently discovered ligands for erbB receptor tyrosine kinases (6 -9). ARIA, NDF, heregulin, and GGF are encoded by alternatively spliced transcripts of the same gene (9). ARIA is synthesized in motor neurons (6) and is thought to be released from motor neurons at the neuromuscular junction. In adult skeletal muscle, ARIA is localized in synaptic basal lamina (10 -12) and remains at original synaptic sites following damage to nerve and muscle (10). The sequence homology between ARIA and heregulins immediately suggested a novel pathway for the regulation of muscle AChR synthesis by motor neurons, i.e. through tyrosine phosphorylation (for review, see Ref. 13). In fact, ARIA was found to induce tyrosine phosphorylation of 185-kDa proteins in the muscle (14). Currently three erbB proteins (erbB2/HER2 (15,16), erbB3/HER3 (17, 18), and erbB4/HER4 (19)), of approximately the same molecular mass (ϳ185 kDa), have been identified in adult skeletal muscle and shown to be tyrosine phosphorylated in response to ARIA (10,20). Among them, erbB3 (20,21) and erbB4 (20) may be localized at the neuromuscular junction. The mouse C2 muscle cells express erbB2 and erbB3, and little, if any, erbB4 (10). This suggests that one or a combination of these erbB receptor tyrosine kinases may represent the ARIA receptor.
The signal transduction pathway initiated by ARIA after activation of the erbB receptor tyrosine kinases has not been elucidated. We established C2C12 cell lines, which stably possessed the nlacZ gene under control of the promoter of the AChR ␣ or ⑀ subunit to study the signal transduction mechanism of the ARIA-induced activation of AChR gene expression. In this report, we provide evidence that ARIA-mediated induction of AChR gene expression requires the activation of MAP kinase.

EXPERIMENTAL PROCEDURES
Materials-The recombinant ARIA (rHRG ␤1 177-244, a peptide of HRG ␤1 residues 177-244) was generously provided by Dr. Mark Sliwkowski (22). This peptide stimulates AChR gene expression in myotubes in primary culture (11,21). PD98059 was a gift from Dr. A. R. Saltiel (23). Epidermal growth factor (EGF) was from Boehringer Mannheim. Rapamycin was purchased from Research Biochemicals International. Cell culture medium was purchased from Life Technol-ogies, Inc. All other chemicals were from Sigma.
Cell Culture-C2C12 cells were from Dr. E. S. Ralston (NINDS, NIH), which were a subclone (24) of the C2 cells originally derived from mouse thigh muscle by Yaffe and Saxel (25). The C2C12 cells were maintained as undifferentiated myoblasts in a nutrient-rich growth medium containing Dulbecco's modified Eagle's medium supplemented with 20% fetal bovine serum, 0.5% chicken embryo extract, 2 mM L-glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin at 37°C in an atmosphere of 5% CO 2 and 95% humidity. The C2C12 cells were split at about 70% confluence, using 0.05% trypsin, 0.02% EDTA in saline. To induce myotube formation, myoblasts were cultured for 48 h in differentiation medium (DM), Dulbecco's modified Eagle's medium supplemented with 4% horse serum and 2 mM L-glutamine.
Transfection Procedures-The DNA constructs including p⑀3500-nlacZ and p␣-lacZ were gifts from Drs. G. Chu and J. Sanes (Washington University). To establish stable C2C12 cell lines carrying these transgenes, myoblasts were cotransfected with 20 g of p⑀3500-nlacZ or p␣-lacZ and 1 g of a neomycin-resistant plasmid (pcDNA3, Invitrogen) by the calcium phosphate technique (26). G418-resistant myoblast colonies were initially selected in 800 g/ml G418 48 h after transfection for 3 weeks and maintained at 400 g/ml G418 for another 2 weeks. ␤-Galactosidase activity assay was used to screen for clones of the stable transfectants which responded to the ARIA stimulation.
In some experiments C2C12 myoblasts at 50% confluence were transiently transfected with p⑀3500-nlacZ or p␣-nlacZ with or without various plasmid DNA by the calcium phosphate technique (26). Forty-eight hours after transfection, cells were switched to DM to induce myotube formation.
Stimulation of C2C12 Cells with ARIA-Forty-eight hours after switching to DM, the C2C12 cells formed myotubes under our conditions. Unless otherwise specified, the C2C12 myotubes were treated immediately after being switched to DM with ARIA or other chemicals for 72 h to induce AChR gene expression. In experiments studying inhibitory effects, chemicals were added into the culture medium 30 min prior to ARIA treatment and remained in the medium for the entire ARIA stimulation period. The volume of solvent vehicles was kept equal to or less than 0.1% of the culture medium which did not significantly change the biological outcome. Culture medium was changed every 24 h to keep myotubes healthy.
␤-Galactosidase Assays-Monolayers of C2C12 myotubes were collected in a microcentrifuge tube after being rinsed with phosphatebuffered saline (PBS). After being washed three times with PBS, the cells were resuspended in 0.25 M Tris-HCl, pH 7.8, and disrupted by three cycles of freezing in liquid nitrogen and thawing at 37°C. After centrifugation, the supernatant of the cell lysates was used for the ␤-galactosidase assay which was carried out according to published methods (26). ␤-Galactosidase activity was determined by hydrolysis of o-nitrophenyl-␤-D-galactopyranoside. Briefly, an aliquot of cell lysate was incubated with 1 mM MgCl 2 , 45 mM ␤-mercaptoethanol, 0.88 mg/ml o-nitrophenyl-␤-D-galactopyranoside, 67 mM sodium phosphate, pH 7.5, in a volume of 300 l at 37°C for 30 min or until a faint yellow color developed. The reaction was stopped by adding 500 l of 1 M Na 2 CO 3 , and absorbance at 420 nm was measured using a spectrophotometer. Unless otherwise specified, ␤-galactosidase activity was expressed as A 420 nm /mg of protein/12 h. The basal ␤-galactosidase activity was A 420 nm 0.25 Ϯ 0.043 in the parental C2C12 myotubes, A 420 nm 0.30 Ϯ 0.056 in the ⑀3500-nlacZ-possessing myotubes, and A 420 nm 0.28 Ϯ 0.048 in the ␣-nlacZ-possessing myotubes.
Direct staining for the ␤-galactosidase in the transfected C2C12 cells was performed as described (11,27). Briefly, the stable transfectants were fixed in 2% paraformaldehyde in PBS for 60 min at room temperature after being rinsed twice with PBS. Then the cells were incubated at 37°C for 16 -20 h in a solution containing 2 mM 5-bromo-4-chloro-3indolyl-␤-D--galactoside, 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, and 2 mM MgCl 2 in PBS. The cells were stored at 4°C in 2% paraformaldehyde in PBS until being photographed.
Western Blot Analysis-C2C12 cells were lysed in RIPA buffer containing 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, 50 mM Tris-HCl, pH 7.5, 1 M pepstatin, 1 g/ml leupeptin, 0.2 mM phenylmethylsulfonyl fluoride, 2 g/ml aprotinin, and 2 mM sodium vanadate. The lysates or immunocomplexes after immunoprecipitation were subjected to SDS-polyacrylamide gel electrophoresis (PAGE) and transferred to nitrocellulose (Schleicher & Schuell) by electroblotting (30 V, overnight) at 4°C. The blots were incubated with a blocking buffer of 0.1% Tween 20 in 50 mM Tris-buffered saline with 1% bovine serum albumin for anti-MAP kinase antibodies (TR10) and horseradish peroxidase-conjugated anti-phosphotyrosine antibody RC20 (Signal Transduction Laboratories), or 1% non-fat dry milk for anti-erbB2 an-tibodies (Oncogene Sciences), at room temperature for 1 h. The blot was then incubated for 1 h in the blocking buffer containing a 1:1,000 dilution of TR10 antiserum, or recommended concentrations of antiphosphotyrosine antibodies or anti-erbB2 antibodies. After washing five times for 5 min each with the blocking buffer, the blots were incubated with horseradish peroxidase-conjugated donkey anti-rabbit IgG (1:3,000, Amersham) at room temperature for 1 h. The blots were then washed three times with the blocking buffer, and twice with Tris-buffered saline. Immunoactive proteins were visualized with enhanced chemiluminescence (Amersham).
MAP Kinase Assay-The MAP kinase assay was performed essentially as described previously (28). Briefly, the C2C12 cells were washed three times with cold PBS and lysed in HO lysis buffer containing 50 mM HEPES, pH 7.5, 100 mM NaCl, 2 mM EDTA, 1 M pepstatin, 1 g/ml leupeptin, 0.2 mM phenylmethylsulfonyl fluoride, 2 g/ml aprotinin, 40 mM p-nitrophenylphosphate, and 2 mM sodium vanadate. The lysates were then centrifuged at 16,000 ϫ g for 10 min at 4°C. An aliquot (100 g of protein) of the supernatant was incubated at 4°C for 2 h with 40 l of 50% protein A-agarose beads (Sigma) which had been preloaded with anti-MAP kinase antiserum TR10. The beads were washed twice with the lysis buffer and twice with 10 mM HEPES, 10 mM magnesium acetate. This bead-antibody-antigen complex was resuspended in the phosphorylation buffer (40 l) containing 10 mM HEPES, pH 7.6, 10 mM magnesium acetate, 5 Ci of [ 32 P-␥]ATP (6,000 cpm/pmol), 50 M ATP, and 40 g of myelin basic protein (MBP). After incubation at 30°C for 30 min, the reaction was stopped by the addition of 20 l of SDS-PAGE sample buffer with dithiothreitol. Phosphorylated MBP was resolved on 15% SDS-PAGE and exposed to an x-ray film. Radioactivity of MBP in sliced gel pieces was measured with Econo-Safe counting mixture (Research Products International Corp.) in a Beckman scintillation counter.

Activation of AChR Subunit Gene Expression in C2C12
Myotubes-To establish an in vitro model to study the signal transduction pathway by which ARIA induces activation of AChR gene expression, C2C12 myoblasts were transfected with p⑀3500-nlacZ or p␣-nlacZ together with the neomycin-resistant gene containing plasmid pcDNA3. The plasmid p⑀3500-nlacZ contains the mouse AChR ⑀ subunit gene sequence extending from nucleotide Ϫ3500 to ϩ82 and the reporter gene nlacZ (11,27). The plasmid p␣-lacZ contains the 850-base pair chicken ␣ subunit promoter fragment and nlacZ. From each transfection, more than 100 colonies were obtained which were G418-resistant, of which 20 clones were screened for expression of ␤-galactosidase, the product of the reporter gene nlacZ, in response to recombinant ARIA (rHRG ␤1 177-244). Seven clones were found to stably possess the ⑀3500-nlacZ transgene, and four clones stably possessed the ␣-nlacZ transgene. We have further characterized two lines of the transfected cells: ⑀1 and ␣1, possessing the ⑀3500-nlacZ or ␣-nlacZ transgene, respectively, which demonstrated maximal response to ARIA stimulation. Both ⑀1 and ␣1 C2C12 myoblasts differentiated into myotubes when the culture medium was changed to the differentiation medium. Treatment of myotubes of either cell line with recombinant ARIA increased the reporter gene expression as detected by measuring ␤-galactosidase activity in intact cells histochemically or in cell lysates. The ARIA-stimulated ␤-galactosidase activity appeared to be higher in ⑀1 than in ␣1 myotubes.
Most nuclei in the ⑀1 C2C12 myotubes treated with ARIA for 72 h demonstrated a strong nlacZ staining although some nuclei failed to stain (Fig. 1). Occasionally, a few nuclei of the transfected myotubes stained with ␤-galactosidase activity in the absence of ARIA (Fig. 1). Similar staining pattern was obtained with the ␣1 C2C12 myotubes. ARIA at subnanomolar concentrations was able to increase ␤-galactosidase activity in the transfected C2C12 myotubes, with an EC 50 around 0.2 nM for both ⑀ and ␣ subunit promoter activation (Fig. 2). These values were comparable with those from previous observations in muscle cells cultured from transgenic mice or cell lines (10,11,21). Interestingly, although they stably possessed the transgene, ⑀1 C2C12 myoblasts showed little, if any, response to ARIA (Fig. 1). Activation of the AChR ␣ and ⑀ subunit gene expression appeared to be specific for ARIA. Insulin, known to stimulate tyrosine phosphorylation in muscle (29), did not affect the basal or ARIA-increased ␤-galactosidase activity in the ␣1 and ⑀1 myotubes (Fig. 3). Moreover, the ␤-galactosidase activity in these myotubes was unaffected by stimulation with EGF ( Fig. 3) or 10% fetal bovine serum (data not shown).
Tyrosine Phosphorylation of erbB Proteins-ARIA induces tyrosine phosphorylation of 185-kDa proteins in the muscle cells (10,14,21) which are believed to be receptors for ARIA. In ⑀1 myotubes, ARIA at 1 nM also induced tyrosine phosphorylation of 185-kDa proteins, consistent with previous observations. Tyrosine phosphorylation of the 185-kDa proteins occurred within 15 s of ARIA treatment with its peak at 2-5 min (Fig. 4A). ARIA seemed to have little effect on expression of erbB2 proteins in C2C12 cells as assessed by Western blot analysis (data not shown). In a separate experiment, the erbB2 protein was immunoprecipitated from ARIA-treated (1 nM, 5 min) ⑀1 myotube lysates using anti-erbB2 antibodies and blotted using either anti-erbB2 antibodies or anti-phosphotyrosine antibodies. Again, ARIA stimulated tyrosine phosphorylation of the erbB2 protein (Fig. 4B). ARIA stimulated tyrosine phosphorylation of the erbB proteins in the ␣1 C2C12 cells which followed a similar time course (data not shown). Treatment of the ␣1 and ⑀1 myotubes with vanadate, a well characterized inhibitor of protein tyrosine phosphatases, increased the basal ␤-galactosidase activity and potentiated the ARIA-induced activation of the AChR ␣ and ⑀ subunit gene expression (Fig. 3). Interestingly, ARIA treatment of C2C12 myoblasts possessing either transgene also stimulated tyrosine phosphorylation of the erbB proteins (data not shown).
Activation of MAP Kinase in C2C12 Cells by ARIA-Many extracellular signals including growth factors induce rapid, transient activation of MAP kinase via phosphorylation on tyrosine and threonine residues (30,31). We proposed that activation of MAP kinase may be involved in the signaling pathway of ARIA-induced activation of AChR gene expression. Since phosphorylation of MAP kinase is accompanied by a decrease in mobility which could determined by SDS-PAGE, we first examined whether ARIA treatment changed the electrophoretic mobility of MAP kinase in transfected C2C12 myotubes. Fig. 5A shows immunoblotting using anti-MAP kinase antibodies of the lysates of ARIA-treated ⑀1 C2C12 myotubes. MAP kinase appeared as a single band at 42 kDa in the absence of ARIA but as two bands in the presence of ARIA (Fig. 5A). The mobility shift appeared at 2 min in the presence of ARIA and peaked at 8 min. To determine if this mobility shift represented an increase in enzyme activity, MAP kinase was immunoprecipitated using anti-MAP kinase antibodies and its activity directly assayed in vitro using MBP as a substrate (28). The basal MAP kinase activity was 25,800 Ϯ 1,700 cpm/assay in myoblasts and 48,200 Ϯ 9,200 cpm/assay in myotubes (p Ͻ 0.01), suggesting that differentiation of C2C12 cells regulates MAP kinase activity. As shown in Fig. 5B, MAP kinase activity in ⑀1 C2C12 myotubes was indeed increased by stimulation with ARIA at concentrations as low as 1 pM. The ARIA-induced MAP kinase activation was both concentration-and time-dependent. Maximal stimulation was 9-fold above the basal level and occurred at 1 nM ARIA. The EC 50  early as 30 s and peaked at 5-8 min (Fig. 5C). These data demonstrated that the MAP kinase activation occurred immediately after tyrosine phosphorylation of erbB proteins, consistent with the hypothesis that MAP kinase mediates the ARIAinduced activation of AChR gene expression. Although ARIA did not appear to induce AChR gene expression, MAP kinase activity was increased by 2.2-fold in ARIA (1 nM)-treated transfected C2C12 myoblasts.
Dependence of ARIA-induced Activation of AChR Gene Expression on MEK and Raf-Signaling pathways of many growth factors have been studied extensively. By analogy with other growth factor receptor tyrosine kinases, ARIA receptors may stimulate Ras by forming complexes with Shc, GrB2, and Ras activator (30,31). Once it is recruited to the plasma membrane, the serine/threonine kinase Raf becomes activated and phosphorylates and stimulates MEK, which in turn phosphorylates and activates MAP kinase (30,31). We have examined whether the ARIA-induced activation of AChR gene expression is dependent on functional MEK or Raf.
The PD98059 compound is a selective inhibitor of MEK (23). It has been demonstrated that selective inhibition of MEK by PD98059 prevents activation of MAP kinase (29,32). The transfected C2C12 myotubes were incubated for 30 min in the presence or absence of PD98059 prior to the treatment with or without ARIA. PD98059 had little effect on basal MAP kinase activity (data not shown). ARIA-increased MAP kinase activity in ⑀1 myotubes was inhibited by PD98059 in a concentrationdependent manner (Fig. 6). The IC 50 value of PD98059 in inhibiting ARIA-stimulated MAP kinase activation was 12 M (Fig. 6B), a concentration at which PD98059 specifically inhibits MEK (29). This compound also inhibited ARIA-stimulated MAP kinase activation in ␣1 myotubes (data not shown). However, the interaction between ARIA and its receptors, and its mediation of tyrosine phosphorylation of the erbB proteins, appeared to be unaffected by PD98059 (Fig. 6A). We next examined the effects of PD98059 on the ARIA-induced activation of AChR gene expression in the ␣1 and ⑀1 C2C12 myotubes. The ␤-galactosidase activity induced by 1 nM ARIA was decreased dramatically by pretreating myotubes with PD98059 (Fig. 6C). The maximal inhibitory effects of PD98059 on ␤-galactosidase were achieved with 20 M in both the ⑀3500-nlacZand ␣-nlacZ-possessing myotubes, with IC 50 values of 5 and 3 M for the ⑀3500-nlacZ and ␣-nlacZ transgenes, respectively. The similar IC 50 values of PD98059 in inhibiting the ARIAinduced MAP kinase activation and AChR gene induction support the hypothesis that the MEK and MAP kinase activation is required in the ARIA induction of the AChR genes.
To establish further the involvement of the Raf/MEK/MAP kinase pathway in ARIA-induced AChR gene expression, we expressed a dominant negative mutant of human MEK1 (33), in which serine residues 218 and 222 were replaced with alanine (pCMV-MEK1m) and/or a dominant negative mutant of human Raf1 (pEBG-RafC4, which encodes only the aminoterminal 1-257 amino acid residues without the kinase do-main, 34), in C2C12 cells together with either the ⑀3500-nlacZ or ␣-nlacZ transgene. In the cells transiently transfected with p⑀3500-nlacZ or p␣-nlacZ and an empty vector (pCMV), ARIA was able to induce AChR gene expression (Fig. 7). Expression of either mutant inhibited the ARIA-induced increase in ␤-ga- were run out on 8% SDS-PAGE, blotted onto nitrocellulose, and probed with anti-phosphotyrosine antibodies (top panel). Phosphorylated MBP from a typical experiment was resolved on 15% SDS-PAGE and exposed to x-ray film (bottom panel). Panel B, effects of PD98059 on ARIAinduced activation of MAP kinase in ⑀1 C2C12 myotubes. MAP kinase activity was assayed using MBP as a substrate. The control (100%) MAP kinase activity was 205,000 Ϯ 26,300 cpm/assay. Panel C, effects of PD98059 on ARIA-induced expression of the ⑀3500-nlacZ and ␣-nlacZ transgenes. The A 420 nm values/mg of protein/12 h represent the difference between ␤-galactosidase activity in the presence of ARIA and the basal level (in the absence of ARIA). The basal levels of ␤-galactosidase activity of each cell line are stated under "Experimental Procedures." Open circles, ␣-nlacZ; closed circles, ⑀3500-nlacZ. Shown are the mean Ϯ S.D. of two or three independent experiments. When not visible, the S.D. is smaller than the symbol. lactosidase activity. When both MEK1 and Raf1 mutants were expressed in the C2C12 myotubes, the ARIA-induced increase in ␤-galactosidase activity was completely abolished (Fig. 7). Consistent with Raf and MEK functioning as the upstream activator of the MAP kinase, expression of their mutants also inhibited ARIA-induced MAP kinase activation (data not shown). These results clearly demonstrated that ARIA-induced AChR subunit gene expression was dependent on functional Raf and MEK in C2C12 cells, in support of the hypothesis that the Raf/MEK/MAP kinase pathway mediates in ARIA-induced AChR subunit gene expression. We also transiently expressed in C2C12 cells constitutively active Raf1 (pEBG-BXBRaf) (34) and MEK1 (pcDNA3-MEK1) (35). 2 Expression of active MEK1 or Raf1 with the ␣-nlacZ transgene appeared to increase ARIAinduced ␤-galactosidase expression in C2C12 myotubes (Fig.  7). Constitutively active MEK1 greatly potentiated the effect of ARIA on ⑀3500-nlacZ expression, but no significant effect of active Raf was observed. It remains to be determined whether the discrepancy between the mouse ⑀ subunit and chick ␣ subunit transgene activation following overexpression of MEK1 and Raf represents a species-specific activation of AChR genes.
Effects of Rapamycin and Wortmaninn on ARIA-induced Activation of AChR Gene Expression-It is believed that heregulin/NDF/GGF/ARIA plays an important role in regulation of cell proliferation, differentiation, and growth by activating several protein kinases. Various isoforms of heregulin/NDF/ARIA have been found to activate the S6 protein kinase in HC11 murine epithelial cells or T47D breast cancer cells (36) and PI 3-kinase in mouse fibroblasts (37,38). To determine whether activation of these other kinases might contribute to the ARIAinduced activation of AChR gene expression in C2C12 myotubes, we examined effects of rapamycin, a S6 kinase inhibitor (39), and wortmannin, a PI 3-kinase inhibitor (37), on ARIAstimulated expression of ␤-galactosidase in ␣1 and ⑀1 C2C12 myotubes. Myotubes were pretreated with rapamycin or wortmannin for 30 min prior to ARIA stimulation. Rapamycin exhibited no apparent effect on expression of ␤-galactosidase in ␣1 or ⑀1 C2C12 myotubes, at concentrations up to 3 M, which was more than 100-fold higher than that required for specific inhibition of S6 kinase (40) (Fig. 8). At a concentration of 100 nM where wortmannin specifically targets PI 3-kinase (37), it did not affect ARIA-induced activation of the AChR ␣ or ⑀ promoter (Fig. 8). It was observed, however, that wortmannin at 10 M appeared to decrease the ␤-galactosidase activity in ARIA-treated ␣1 but not ⑀1 C2C12 myotubes. This inhibition may be nonspecific since it occurred only at a very high concentration. These results suggest that the activation of S6 protein kinase and PI 3-kinase may not be involved in the ARIA signaling pathway that leads to the AChR gene induction.

DISCUSSION
To study mechanisms by which ARIA activates AChR gene expression, we have generated stable C2C12 cells which possess the ⑀3500-nlacZ or ␣-nlacZ transgene. We believe that the C2C12 cells carrying these transgenes are an excellent in vitro model to study the molecular mechanism of AChR gene induction for several reasons. First, the C2C12 muscle cells have been used successfully as an expression system to study transcription of muscle genes and have the unusual capability to activate muscle-specific genes upon fusion (24). Second, recent studies demonstrated that the C2C12 muscle cells have the essential elements required for the ARIA-induced activation of AChR gene expression (10). In addition, expression of the ⑀3500-nlacZ and ␣-nlacZ transgene is spatially restricted in muscle nuclei beneath the motor end plate in transgenic mice (27), suggesting that these promoters contain the elements necessary to respond to natural cues released from motoneurons. In vitro, ARIA treatment of muscle fibers isolated from transgenic mice carrying ⑀3500-nlacZ transgene induces the AChR ⑀ subunit gene expression (11). We show here that in the C2C12 myotubes stably carrying these transgenes, ARIA was able to induce the AChR ␣ and ⑀ gene expression (Figs. 1 and 2). Interestingly, we found that ARIA treatment of the C2C12 myotubes resulted in activation of MAP kinase (Fig. 5) which followed in time the tyrosine phosphorylation of erbB proteins (Fig. 4). Moreover, the ARIA-induced expression of AChR ␣ and ⑀ subunit genes was dependent on activation of MAP kinase.  7. Dependence of ARIA-induced expression of the ⑀3500-nlacZ and ␣-nlacZ transgene on Raf and MEK. C2C12 myoblasts were transiently transfected with 10 g of p⑀3500-nlacZ (left group of columns) or p␣-nlacZ (right group of columns) plus 10 g of pCMV, pcDNA3-MEK1 (MEK1), pCMV-MEK1m (MEK1m), pEBG-BXBRaf (Raf), pEBG-RafC4 (RafC4), or pCMV-MEK1m and pEBG-RafC4 (see "Results" for description of plasmid constructs). Also included in all the transfection reactions was 5 g of pEBG encoding glutathione S-transferase. Forty-eight hours after transfection, myoblasts were incubated with the DM to induce myotube formation. After another 48 h, myotubes were stimulated with 1 nM ARIA for 72 h, and ␤-galactosidase activity (␤-GAL) was assayed. Transfection efficiency was evaluated by Western blot analysis of glutathione S-transferase expression in transfected myotubes. The control (100%) was calculated as the difference of the ARIA-stimulated and basal ␤-galactosidase activity, which was A 420 nm 0.11 Ϯ 0.03 for the ⑀3500-nlacZ transgene and A 420 nm 0.082 Ϯ 0.02 for the ␣-nlacZ transgene. Shown are the mean Ϯ S.D. of two or three independent experiments. *, p Ͻ 0.05; **, p Ͻ 0.01.
Inhibition of MEK or Raf in C2C12 myotubes blocked activation of AChR gene expression by ARIA. These results suggest a machinery for the ARIA-induced activation of AChR gene expression which begins with stimulation of the erbB receptor tyrosine kinases and requires activation of the Raf/MEK/MAP kinase pathway.
ARIA induces tyrosine phosphorylation of 185-kDa erbB proteins in chicken (14,21) and C2C12 myotubes (Fig. 4, 10). Within seconds of ARIA stimulation, erbB proteins became tyrosine-phosphorylated (Fig. 4). Treatment of chicken myotubes with erbstatin, a specific inhibitor of protein tyrosine kinase, inhibits ARIA-elicited increase in ␣ subunit mRNA (21). Furthermore, vanadate, an inhibitor of protein tyrosine phosphatases, potentiates ARIA-induced AChR gene activation in chicken or C2 myotubes (Ref. 21 and Fig. 3). The close correlation of ARIA-induced tyrosine phosphorylation of the erbB proteins with activation of AChR gene expression suggests that tyrosine phosphorylation is a key step in the ARIA signaling pathway. NDF/heregulin (or ARIA) was initially identified by its ability to activate erbB2; however, it does not interact directly with the erbB2 protein (41). Its receptor appears to be erbB3 (42,43) and/or erbB4 (44,45). Since erbB3 has little, if any, tyrosine kinase activity (46), signaling through erbB3 has been postulated to require ligand-stimulated dimerization of erbB3 with another erbB protein or EGF receptor. In human embryonic kidney 293 cells and mammary carcinoma-derived MCF-7 cells, erbB2 forms heterodimers with erbB3 upon heregulin stimulation (47). The erbB2 and EGF receptor can associate into an active heterodimeric tyrosine kinase (48). The C2 cells express erbB2 and erbB3 but not erbB4 (10,20), suggesting that the ARIA-triggered signaling in C2 cells is initiated by erbB3. Wright et al. (49) have shown that EGF can stimulate a kinase-dead EGF receptor to activate MAP kinase via association with erbB2. In C2C12 cells, presumably, ARIA binds to erbB3 which has little kinase activity and stimulates MAP kinase by interacting with erbB2 or erbB2-related tyrosine kinase.
The activation of receptor tyrosine kinases by ligands initiates signal transduction cascades essential for cell proliferation and differentiation. The 21-kDa Ras proteins function as signaling mediators for activated receptor tyrosine kinases (50). Binding of adaptor proteins such as Grb2 and/or Shc to tyrosine-phosphorylated intracellular domains recruits guanine nucleotide exchanger proteins to the receptor complex, which promotes exchange of GTP for GDP and thus activates Ras proteins. There are several consensus binding sites for Grb2 and Shc in erbB2, erbB3, and erbB4 (43). In fact, erbB2 and erbB3 have been shown to bind to Grb2 (51,52) and Shc (17, 51, 52). Overexpression of constitutively active erbB2 in NIH3T3 cells activates Ras by forming complex with Shc/Grb2/ Sos (53). NDF has been recently found to stimulate growth and proliferation of mammary epithelial cells, which was accompanied by activation of MAP kinase (36). We found that ARIA activated MAP kinase via the Raf/MEK pathway in terminally differentiated C2C12 myotubes, and such an activation is required for ARIA-induced AChR gene expression. This suggests an important role of MAP kinase even in highly differentiated cells. In addition to Raf, Ras also stimulates many signaling proteins including PI 3-kinase, protein kinase C , and MEK kinase (50). Stimulation of cells with EGF results in the tyrosine phosphorylation of erbB3 and the recruitment of PI 3-kinase to this receptor (38). Moreover, NDF activates the p70/p85 S6 kinase pathway in mammary epithelial cells (36). However, the inability of wortmannin and rapamycin, well characterized inhibitors of PI 3-kinase and S6 kinase, respectively, to affect the ARIA-induced activation of AChR gene expression in C2C12 myotubes (Fig. 8) suggests that activation of these enzymes is probably not involved in this event.
Activated MAP kinase is believed to translocate into nuclei and activate gene expression by phosphorylating respective transcriptional factors (30). It is worth noting that ARIA-induced activation of AChR ␣ and ⑀ gene expression in the C2C12 cells appeared to be dependent on the state of cell differentiation. Treatment with ARIA had little, if any, effect on the reporter gene expression in ⑀1 myoblasts (Fig. 1) and in ␣1 myoblasts (data not shown). Only after differentiation into myotubes did the C2C12 cells respond to ARIA with increased expression of the reporter gene. However, ARIA was able to elicit tyrosine phosphorylation of the erbB proteins and activation of MAP kinase in C2C12 myoblasts. Moreover, ARIAactivated MAP kinase appeared to be able to translocate into nuclei in C2C12 myotubes. 3 These results suggest that downstream signaling component(s) are regulated in a manner dependent on the myogenic program.
Following formation of neuromuscular junctions, in addition to the ARIA pathway, motor neurons control the AChR gene transcription also by eliciting AChR-evoked muscle depolarization, which represses AChR gene expression in extrasynaptic nuclei (54; for review, see Refs. 2 and 4). The electric activitydependent repression of AChR gene transcription in extrasyn-3 J. Si and L. Mei, unpublished data.
FIG. 8. Effects of rapamycin and wortmannin on ARIA-induced activation of the ⑀3500-nlacZ and ␣-nlacZ transgene expression. The ␣1 or ⑀1 C2C12 myotubes were treated without or with rapamycin and wortmannin at various concentrations 30 min prior to ARIA (1 nM) stimulation. After 72 h, myotubes in 100 ϫ 20-mm dishes were lysed and assayed for ␤-galactosidase activity. The controls (100%) were calculated as the difference between ARIA-stimulated and basal ␤-galactosidase activity in untreated myotubes, which were A 420 nm 1.00 Ϯ 0.12 and 0.60 Ϯ 0.03 for the ⑀3500-nlacZ transgene in experiments with rapamycin and wortmannin, respectively, and 0.84 Ϯ 0.11 and 0.56 Ϯ 0.04 for the ␣-nlacZ transgene in experiments with rapamycin and wortmannin, respectively. Shown are the mean Ϯ S.D. of two to three independent experiments. *, p Ͻ 0.05. aptic nuclei probably involves activation of protein kinase C, a serine/threonine kinase (55). Interestingly, it was reported recently that treatment of chick myotubes with phorbol 12-myristate-13-acetate, a stimulator of protein kinase C, blocks heregulin/NDF (or ARIA)-activated AChR ␣ subunit expression, presumably via inhibiting the heregulin-dependent tyrosine phosphorylation of the erbB2 and erbB3 proteins (21). These data suggest a coordination among different regulatory mechanisms of AChR gene expression by the ARIA pathway, the myogenic factors, and the electric activity-dependent regulation.
In summary, we have shown that ARIA activates MAP kinase in C2C12 myotubes which is required for ARIA-induced activation of AChR gene expression. Activation of MAP kinase occurs after tyrosine phosphorylation of ARIA receptors. Inhibition of MAP kinase activation blocks ARIA-dependent stimulation of AChR gene expression. We propose a pathway for ARIA-induced activation of AChR gene expression which begins with stimulation of the erbB protein tyrosine kinases and requires activation of the Raf/MEK/MAP kinase pathway. The C2C12 cells stably possessing AChR transgenes should prove to be a good model to further elucidate regulatory mechanisms of AChR gene expression.