Differential Coupling of α1-, α2-, and β-Adrenergic Receptors to Mitogen-activated Protein Kinase Pathways and Differentiation in Transfected PC12 Cells*

Three adrenergic receptor families that selectively activate three different G proteins (α1/Gq/11, α2/Gi, and β/Gs) were used to study mitogen-activated protein kinase (MAPK) activation and differentiation in PC12 cells. PC12 cells were stably transfected with α1A-, α2A-, or β1-adrenergic receptors (ARs) in an inducible expression vector, and subclones were characterized. Norepinephrine stimulated inositol phosphate formation in α1A-transfected cells, inhibited cyclic adenosine 3′5′-monophosphate (cAMP) formation in α2A-transfected cells, and stimulated cAMP formation in β1-transfected cells. Nerve growth factor activated extracellular signal-regulated kinases (ERKs) in all cell lines; however, norepinephrine activated ERKs only in α1A- and β1-transfected cells but not in α2A-transfected cells. Norepinephrine also activated c-Jun NH2-terminal kinase and p38 MAPK in α1A-transfected cells but not in β1- or α2A-transfected cells. Norepinephrine caused differentiation of PC12 cells expressing α1A-ARs but not those expressing β1- or α2A-ARs. However, norepinephrine acted synergistically with nerve growth factor in promoting differentiation of cells expressing β1-ARs. Whereas ERKs are activated by Gi- but not Gs-linked receptors in many fibroblastic cell lines, we observed the opposite in PC12 cells. The results show that activation of the different G protein signaling pathways has different effects on MAPKs and differentiation in PC12 cells, with Gq signaling pathways activating all three major MAPK pathways.

MAPKs are subdivided into three major pathways (17). Extracellular signal regulated kinases 1 and 2 (ERKs) are stimulated by growth factors and cytokines and stimulate growth and differentiation. The proto-oncogene c-ras and the cytoplasmic kinases c-Raf and MEK are known to play important roles in the activation of ERKs (18 -20). The other two MAPK pathways, c-Jun-NH 2 -terminal kinase (JNK) (also known as stressactivated protein kinase) and p38 MAPK, are generally activated by stresses such as inflammatory cytokines, osmotic shock, or UV irradiation and may be involved in inhibition of cell growth and/or apoptosis. The balance between these pathways may be critical in determining cell fate (21).
The mechanisms of activation of ERKs by GPCRs remain controversial. Responses to G q -linked receptors are thought to involve both the ␣and ␤␥-subunits of G proteins, although the ␣ q -dependent activation of protein kinase C is thought to play the predominant role (3,22). Response to both G i -linked (22,23) and G s -linked (12) receptors are thought to be due primarily to release of ␤␥-subunits, although other mechanisms have also been proposed (12,24). Similar mechanisms have been implicated in activation of JNK/SAPK by GPCRs (24,25), and GPCR activation of p38 MAPK was recently suggested to involve G␣ q -as well as ␤␥-subunits (26). ␤␥-Dependent activation of p38 MAPK is inhibited by coexpression of G␣ o in HEK 293 cells (26).
PC12 cells, derived from a rat pheochromocytoma, have been a primary model for studying mechanisms underlying neuronal differentiation (27). Nerve growth factor (NGF) acts on receptors with tyrosine kinase activity to differentiate these cells into a neuronal phenotype, through a Ras-dependent activation of ERKs (28). Stimulation of both bradykinin (G q/11 -coupled) and lysophosphatidic acid (G i -coupled) receptors also activates ERKs in PC12 cells, apparently through the tyrosine kinases Pyk2 (6) and Src (8) in a Ras-dependent manner. cAMP analogs also activate ERKs and potentiate NGF-induced neurite formation in PC12 cells (29). Thus, G q/11 -, G i -, and G s -linked receptors may all activate ERKs in this cell line.
This raises questions about signaling specificity. If all three types of G proteins can activate ERKs, albeit through different mechanisms, do they have similar functional consequences? To what extent are known second messengers involved in activation of the MAPK pathways? Are the functional consequences of G protein activation similar to those of tyrosine kinase receptor activation?
We wanted to directly address the specificity by which  (32) was generously provided by Dr. G. Tsujimoto (National Children's Hospital, Tokyo, Japan), the cDNA for the rat ␤ 1 -AR (33) was provided by Dr. Curtis A. Machida (Oregon Regional Primate Research Center, Beaverton, OR), and the cDNA for the human ␣ 2A -AR (34) was obtained from ATCC (Manassas, VA). PC12 cells were obtained from Cindy Miranti and Michael Greenberg (Harvard Medical School, Boston, MA). NGF was generously provided by David Ginty (Johns Hopkins, Baltimore, MD).
Preparation of Expression Vectors-The full-length AR sequences were cloned into the multiple cloning site of the operator vector (pOPRSVICAT) of the inducible Lac-Switch system. The NotI fragment of pOPRSVICAT containing the chloramphenicol acetyltransferase reporter gene was replaced with the multiple cloning site of pBluescript KSϩ (where an additional NotI site had been inserted 5Ј to the XhoI site) to facilitate insertion of the gene of interest (35).
Cell Culture-Rat pheochromocytoma PC12 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) horse serum, 5% fetal bovine serum, 10 mg/ml streptomycin, and 100 units/ml penicillin at 37°C in a humidified atmosphere with 5% CO 2 . Confluent cells were subcultured in a 1:3 ratio. Where indicated, transfected cells were treated with 1 mM IPTG for various time periods to induce receptor expression.
Transfection-PC12 cells were co-transfected with the LacSwitch repressor (p3ЈSS) and operator vectors by calcium phosphate precipitation and propagated for several weeks in the presence of 250 g/ml hygromycin and 500 g/ml Geneticin to obtain resistant cells. Subclones expressing each of the different ARs were obtained by screening for cell lines that exhibited low constitutive and high inducible receptor levels by radioligand binding. Cells for radioligand binding and second messenger measurements were plated at lower ( 1 ⁄10) antibiotic concentrations.
Radioligand Binding-Confluent 100-mm plates of cells were washed in phosphate-buffered saline (20 mM NaPO 4 , 154 mM NaCl, pH 7.6) and harvested by scraping. Cells were homogenized with a Polytron, and membranes collected by centrifugation at 30,000 ϫ g for 10 min, washed, and resuspended by homogenization. Receptor density was determined by saturation analysis of specific antagonist radioligands. Western Blots-Confluent cells were serum-starved for 2 h at 37 o C before treatment. Agonists were generally added for 15 min, and cells were washed twice with ice-cold phosphate-buffered saline and lysed in Nonidet P-40 lysis buffer containing 137 mM NaCl, 20 mM Tris-Cl (pH 8), 1 mM MgCl 2 , 1 mM CaCl 2 , 1% Nonidet P-40, 2 mM phenylmethylsulfonyl fluoride, 20 mM sodium orthovanadate, 20 mM leupeptin, and 10 g/ml aprotinin. 20 g of total protein was subjected to SDS-polyacrylamide gel electrophoresis, and proteins were transferred to a nitrocellulose membrane. Activation of ERK1 and 2, JNK/SAPK, or p38 MAPK was detected by blotting the membrane with phosphospecific ERK, JNK/SAPK, or p38 MAPK antibodies (New England Biolabs) that specifically recognize the activated, threonine, and tyrosine dually phosphorylated forms. Blots were stripped and probed with nonphosphospecific antibodies to the enzymes to control for protein loading. Proteins were visualized using a horseradish peroxidase-conjugated goat antirabbit IgG and by ECL (Amersham Pharmacia Biotech).

Lack of Endogenous ARs in Parental PC12
Cells-We used radioligand binding and functional approaches to determine whether any endogenous ARs are expressed in PC12 cells. Low levels of endogenous ␣ 2 -ARs have been reported in PC12 cells (39), but we found no detectable levels of any AR subtype in our PC12 cells. No specific binding was detected in membrane preparations using the antagonist radioligands [ 125 I]BE (␣ 1 -AR), [ 3 H]rauwolscine (␣ 2 -AR), or [ 125 I]cyanopindolol (␤-AR) (Ͻ5 fmol/mg of protein, data not shown). There was also no detectable stimulation of cAMP by the ␤-AR agonist isoproterenol, no inhibition of forskolin-stimulated cAMP by the ␣ 2 -AR agonist UK 14,304 (see also below), and no stimulation of [ 3 H]InsP formation by the ␣ 1 -AR agonists phenylephrine or NE in parental PC12 cells (data not shown). These data suggest that this is one of the few cell lines that do not express measurable levels of any AR subtype.
Characterization of Stably Transfected PC12 Subclones-PC12 cells were co-transfected with the lac repressor vector and the lac operator vector containing either human ␣ 1A -, human ␣ 2A -, or rat ␤ 1 -AR coding sequences. Subclones expressing each receptor were screened for low constitutive expression and inducibility by IPTG. Saturation analysis of specific radioligand binding was used to measure receptor density. Several subclones were isolated with inducible expression of ␣ 1A -or ␣ 2A -ARs; however, we were unable to obtain subclones showing inducible expression of ␤ 1 -ARs. Several subclones with constitutive expression of ␤ 1 -ARs were isolated and used for further studies. Constitutive and IPTG-induced receptor density in selected subclones is summarized in Table I. ␣ 1A -AR Expression and Induction-Three subclones of PC12 cells expressing ␣ 1A -ARs were extensively characterized ( Fig.  1). Saturation analysis of [ 125 I]BE binding showed that each subclone exhibited different levels of constitutive and IPTGinduced (1 mM, 48 h) receptor expression, with ␣ 1A -3 showing the highest degree of induction. The effect of NE on [ 3 H]InsP formation was studied to ensure that the expressed receptors were functional. Basal [ 3 H]InsP formation was similar in each subclone, and was not affected by treatment with IPTG ( Fig. 1). NGF caused small increases in [ 3 H]InsP formation in each subclone, and this response was unaffected by treatment with IPTG. On the other hand, NE increased [ 3 H]InsP formation in each subclone, and this response was substantially increased by induction of receptor expression with IPTG. NE-stimulated [ 3 H]InsP formation was highly correlated with receptor den-sity, being highest in subclone 3, which expressed the highest density of ␣ 1A -ARs and lowest in subclone 9, which expressed the lowest density of ␣ 1A -ARs (Fig. 1). ␣ 1A -AR activation had no effect on cAMP levels (data not shown), showing an absence of cross-talk with G i or G S .  Fig. 2. IPTG caused about a 7-fold increase in receptor expression with an EC 50 around 5-10 M. Inhibition of forskolin-stimulated cAMP accumulation in these cells by the ␣ 2 -AR-selective agonist UK 14,304 was used to ensure that the expressed receptors were functional. As expected, there was no effect of IPTG-induced receptor expression on forskolin-stimulated cAMP accumulation (Fig. 2). However, the potency of UK 14,304 in inhibiting forskolin-stimulated cAMP accumulation was enhanced 5-7-fold by IPTG pretreatment (1 mM, 48 h). Interestingly, the maximal inhibition of the forskolin response by UK 14,304 was not affected by IPTG, suggesting that the density of ␣ 2A -ARs in uninduced cells is sufficient for maximal inhibition. ␣ 2A -AR activation did not affect InsP formation (data not shown), showing an absence of cross-talk with G q/11 . ␤ 1 -AR Expression-All ␤ 1 -AR-expressing PC12 subclones that we isolated showed constitutive receptor expression. Each subclone showed a receptor density around 200 fmol/mg of protein, which was not significantly altered by treatment with IPTG (1 mM, 48 h). Because this receptor density is in the range of expression of the ␣ 1A -AR-and ␣ 2A -AR-expressing subclones, we used the constitutive expression of subclone ␤ 1 -3 to study ␤ 1 -AR responses in PC12 cells. Because the ␤-AR is not activated in the absence of ligand, we did not make further attempts to isolate an inducible ␤-AR PC12 cell line. Stimulation with forskolin (30 M) caused about a 10-fold increase in cAMP accumulation in both parental PC12 cells and the ␤ 1 -3 subclone (Fig. 3). Stimulation with the ␤-AR agonist isoproterenol (10 M) had no effect on cAMP accumulation in parental PC12 cells but caused a significant 50 -100% increase in the ␤ 1 -3 subclone. UK 14,304 had no effect on forskolin-stimulated cAMP accumulation in either parental PC12 cells or in the ␤ 1 -3 subclone (Fig. 3), confirming the absence of endogenous ␣ 2 -ARs. ␤ 1 -AR activation also did not affect InsP formation (data not shown), showing an absence of cross-talk with G q/11 . Activation of ERKs-We studied the effect of NE on ERK  activation in PC12 cells expressing the different AR subtypes. Exposure to NGF (100 ng/ml) caused activation of ERK 1 and 2 phosphorylation in parental PC12 cells (Fig. 4), as well as in PC12 cells expressing each of the AR subtypes (␣ 1A -3, ␣ 2A -5, and ␤ 1 -3). Exposure to NE (100 M) had no effect in parental PC12 cells but caused activation of ERKs in the ␣ 1A -3 PC12 cells (Fig. 4). As expected, the degree of activation of ERKs by NE was increased by increasing ␣ 1A -AR expression with IPTG. NE also caused ERK activation in ␤ 1 -3 PC12 cells, although this effect was not increased by IPTG, which does not increase ␤ 1 -AR expression in these cells. Surprisingly, NE had no effect on ERK activation in ␣ 2A -5 PC12 cells, even after increasing receptor density by IPTG exposure (Fig. 4). Blotting for total ERK protein showed equivalent sample loading for each condition (Fig. 4). To confirm the lack of ␣ 2A -AR-mediated ERK activation in PC12 cells, the effect of NE was also tested on the ␣ 2A -2 subclone. Again, NGF increased ERK phosphorylation in this cell line, whereas NE had no effect either with or without exposure to IPTG (data not shown).
Summary of AR-mediated Activation of MAPK Pathways in PC12 Subclones- Fig. 6 shows a summary of the effects of NE and NGF on activation of ERKs, JNK/SAPK, and p38 MAPK in Time Course of NE-stimulated ERK Activation in ␣ 1A and ␤ 1 -AR PC12 Cells-Because activation of either ␣ 1A -or ␤ 1 -ARs activated ERKs in PC12 cells, we compared the time course of the two responses. Fig. 7 shows that NE activation of ␣ 1A -ARs caused a large and sustained activation of ERKs, which was highly dependent on receptor induction by IPTG. NE activation of ␤ 1 -ARs also caused sustained ERK activation.
Differentiation of AR-expressing PC12 Cells-Exposure of ␣ 1A -3 PC12 cells to either NE or NGF caused differentiation of the cells within 36 -48 h after exposure (Fig. 8). The extent of NE-induced differentiation of ␣ 1A -3 PC12 cells was dependent on the level of receptor expression. Cells expressing high levels of ␣ 1A -ARs (ϳ300 fmol/mg of protein following induction with IPTG) displayed NE-induced differentiation similar to that observed with NGF alone, whereas cells expressing lower levels of ␣ 1A -ARs (ϳ40 fmol/mg of protein, not induced with IPTG) showed NE-induced differentiation only slightly higher than untreated cells. Exposure of IPTG-induced ␣ 1A -3 cells to both NE and NGF caused differentiation levels (size and number of neurites) greater than those caused by either agonist alone (Fig. 8), suggesting that differentiation in response to the two agonists is additive. Exposure of uninduced ␣ 1A -3 cells to both FIG. 4. Activation of MAPK by NE and NGF in PC12 cell subclones. Cells were treated with (ϩ) or without (Ϫ) 1 mM IPTG for 48 h; serum-starved for 2 h; treated with vehicle (C), NE (100 M), or NGF (100 ng/ml) for 15 min; lysed; and harvested as described. 20 g of protein from each sample was subjected to SDSpolyacrylamide gel electrophoresis, transferred to nitrocellulose, and blotted for activated phosphorylated MAPK (P-MAPK) or total MAPK (MAPK) as described. Data from parental PC12 cells and subclones transfected with each of the three AR subtypes are shown. Data are representative of seven or more separate experiments. NE and NGF caused differentiation similar to that of cells exposed to NGF alone, confirming that the effects of NE and NGF are additive, not synergistic.
NE treatment had no significant effect on the differentiation state of either of the ␣ 2A -AR-expressing PC12 subclones (subclones 2 and 5), either with or without induction of receptor expression with IPTG (1 mM, 48 h), in the presence or absence of NGF, or at any time point examined (data not shown).
Exposure of ␤ 1 -3 PC12 cells to NE for 24 -36 h had little visible effect on differentiation (Fig. 9). However, NE had a synergistic effect on differentiation caused by NGF. Exposure to both NGF and NE for just 24 h caused differentiation of cells to levels higher than that seen in cells exposed to either NGF or NE alone (Fig. 9).

DISCUSSION
In this report, we directly compare the effects of G␣ q/11 , G␣ i , and G␣ S -coupled ARs on mitogenic responses and differentiation of PC12 cells. ARs affect growth and differentiation of many cells, although the subtypes and mechanisms involved are not yet clear. PC12 cells have been widely used to study the events involved in cellular differentiation (27). NGF causes differentiation of PC12 cells through a sustained activation of the Ras/Raf/ERK pathway (21,28). We took advantage of the fact that the three AR families selectively couple to three major G protein families (␣ 1 /G q/11 , ␣ 2 /G i , and ␤/G S ) to examine the specificity with which these receptors activate MAPK in PC12 cells. Because receptor density is critically important in signaling specificity, we used an IPTG-inducible vector system to control receptor expression within a range normally observed in many tissues. We wanted to directly compare AR-mediated activation of G q/11 , G i , and G S on MAPK and differentiation responses in the same cellular phenotype. PC12 cells also allow direct comparison of signals generated by GPCRs and tyrosine kinase receptors involved in growth, differentiation, and apoptosis.
We found dramatic differences in the ability of the three AR families to promote MAPK activation and differentiation of PC12 cells. Stimulation of both ␣ 1A -and ␤ 1 -ARs caused ERK activation, whereas stimulation of ␣ 2A -ARs did not. The activation of ERKs by ␣ 1A -and ␤ 1 -ARs was sustained and elevated even 1 h after stimulation. Only in ␣ 1A -AR transfected cells did NE cause significant activation of the JNK/SAPK and p38 MAPK pathways. In addition, only the ␣ 1A -AR subtype caused PC12 differentiation in the absence of other stimuli. Most surprisingly, activation of ␣ 2A -ARs at either low or high density had no effect on ERK activation in PC12 cells, despite previous studies showing activation of this pathway through the endogenous lysophosphatidic acid receptor in these cells (6,8).
Activation of ␣ 1A -ARs by NE caused a substantial activation of ERKs in PC12 cells, and this effect was increased by IPTG exposure, suggesting that it was proportional to receptor density. ERK activation by ␣ 1A -ARs corresponded with a NE-induced differentiation of these cells into a neuronal-like phenotype. In cells expressing high levels of ␣ 1A -ARs, NE caused differentiation indistinguishable from that caused by NGF. Differentiation of ␣ 1A -AR cells exposed to both NGF and NE was no more than additive and occurred on the same time scale as that caused by NGF alone (36 -48 h).
NE also caused a large activation of JNK/SAPK and a smaller activation of p38 MAPK in ␣ 1A -transfected PC12 cells, and these responses were also increased by IPTG exposure. ␣ 1 -AR-mediated activation of JNK/SAPK has been reported in cardiomyocytes (40); however, activation of p38 MAPK by ␣ 1 -ARs has not been reported previously. p38 MAPK has been shown to be activated by G q/11 -coupled m1 muscarinic receptors (26) and by activated forms of ␣ q (14). Activation of JNK/SAPK and p38 MAPK has previously been associated with stress responses (17,21); however, recent data in cardiomyocytes suggest a role for JNK in cell growth (40). In most cases, ERK, JNK/SAPK, and p38 MAPK pathways are activated by different stimuli (17), and ␣ 1A -transfected PC12 cells are unusual in activation of these pathways by a single stimulus.
GPCRs coupling through ␣ i (including ␣ 2 -ARs) were among the first GPCRs shown to activate ERKs (1,2,4). Other ␣ icoupled receptors, such as lysophosphatidic acid receptors, have also been reported to activate ERKs in PC12 cells (8). We were surprised to find that stimulation of ␣ 2A -ARs did not activate ERKs in PC12 cells, even at high expression levels. Studies on inhibition of forskolin-stimulated cAMP accumulation showed that the expressed ␣ 2A -ARs were functional, and receptor density in the presence of IPTG was higher for ␣ 2A -ARs than either ␣ 1A -or ␤ 1 -ARs (Table I). Although lysophosphatidic acid often acts via ␣ i , it has also been reported to activate ␣ q in PC12 cells (41,42), and it could be causing at least some of its effects via ␣ q in these cells.
The fact that ␣ 2A -ARs activate ERKs in fibroblastic cell lines such as Rat1a cells (2,4) in a PTX-sensitive manner but do not activate ERKs in PC12 cells indicates that there are important mechanistic differences in signaling between the cell types. Activation of ERKs by G i -linked receptors appears to be mediated by ␤␥-subunits (3,22,23,43), and recent work suggests that ␤␥ signaling is impaired in the presence of either ␣ t -or ␣ o -subunits (3,44,45). ␣ o is selectively expressed in brain and many neuronal cell lines, including PC12 cells (39,46,47). Fibroblastic cell lines, such as NIH3T3 and Rat1a, express ␣ i2 and ␣ i3 but not ␣ o (39). Because ␣ o is expressed selectively in PC12 cells, it may play a similar suppressive role in ␤␥-mediated MAPK activation by ␣ 2A -ARs. Regardless, our results show that ERK activation is not a universal response to activation of G i -coupled receptors in PC12 cells.
Release of ␤␥-subunits is also proposed to be important for ERK activation by ␣ S -linked GPCRs, although the specificity of these interactions in PC12 cells is not yet clear (12,43,48). The effects of increased levels of cAMP on differentiation of PC12 cells and activation of ERKs have been well studied (29,49,50). In most other cell lines, increases in cAMP generally inhibit ERK activation (15,16), and PC12 cells are relatively unusual in that forskolin-induced increases in cAMP activate ERKs (29, 49 -51). This study is one of the few examples of ERK activation FIG. 7. Time course for NE activation of ERKs in PC12 subclones. ␣ 1A -3 PC12 cells were treated with (q, Induced) or without (E, Control) 1 mM IPTG for 48 h, and ␤ 1 -3 cells (Ⅺ) were cultured without IPTG. Cells were serum-starved for 2 h, treated with NE (100 M) for 0 -120 min, lysed, and harvested as described. 20 g of protein from each sample was subjected to SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and blotted for activated phosphorylated MAPK as described. Data are representative of two separate experiments. The degree of stimulation is expressed in arbitrary units and is relative to basal intensity. The densitometer was adjusted for maximal sensitivity to emphasize responses.
FIG. 8. Effects of NE and NGF on differentiation of the ␣ 1A -3 subclone of PC12 cells. Cells were plated on collagen-coated plates and treated without (A, C, E, and G) or with (B, D, F, and H)  in PC12 cells caused by stimulation of an ␣ s -linked GPCR (rather than direct increases in cAMP), which would involve signaling from both ␣ S -and ␤␥-subunits. We observed sustained activation of ERKs upon stimulation of ␤ 1 -ARs in PC12 cells without observable differentiation in the absence of NGF. However, ␤ 1 -AR activation strongly potentiated NGF-induced differentiation, causing the appearance of neurites within 24 h after addition of both NE and NGF. This is consistent with previous studies suggesting that large increases in cAMP can alone cause differentiation of PC12 cells, but smaller increases in cAMP only potentiate growth factor-induced differentiation (52). Intracellular nonmitochondrial Ca 2ϩ pools have been shown to be necessary for the synergistic effects of NGF and cAMP analogs on PC12 cell differentiation (53).
Previous studies in which PC12 cells were transfected with activated forms of G protein ␣-subunits showed that activated forms of ␣ q alone, but not of ␣ i or ␣ o , were capable of differentiating PC12 cells (14). Differentiation by ␣ q coincided with activation of JNK, but ERK activation was not seen (14). The results using activated ␣ S are less clear. One group reported that expression of activated ␣ S caused proliferation of PC12 cells and constitutive activation of cAMP dependent pathways (54), whereas another group found that expression of activated ␣ S caused differentiation of PC12 cells (55). Transfection of G protein subunits into Cos7 cells showed that ERK activation may be due to signaling from ␤␥ rather than either ␣ q , ␣ i , or ␣ S (3).
These results have some similarities to studies in cardiac and smooth muscle, in which both ␣ 1 -and ␤-ARs are involved in growth and differentiation (56). In both cases, ␣ 1 -ARs dominate, causing rapid and divergent activation of MAPK pathways and transcription (40, 56 -60). Because stimulation of ␣ 1A -ARs activates ERKs, JNK/SAPK, and p38 MAPK and promotes differentiation of PC12 cells, it will be interesting to compare the signaling pathways involved with those in myocytes (13,60,61). Studies in cardiac and smooth muscle cells are generally performed in primary cultures or in vivo, and PC12 cells may be a useful alternative in defining the transcriptional effects of receptor activation and how they relate to growth and differentiation.
This report shows that activations of ␣ 1 -, ␣ 2 -, and ␤-ARs in transfected PC12 cells have different effects on MAPK pathways and differentiation, suggesting a clear specificity in activation of MAPK pathways. The marked stimulation of all three MAPK pathways and differentiation by ␣ 1A -AR activation may provide a useful system for studying the mechanisms by which GPCRs control cellular growth and differentiation and the relationship of these pathways to those activated by tyrosine kinase receptors. the treated for 24 -30 h with vehicle (A), 100 M NE (B), 100 ng/ml NGF (C), or NE ϩ NGF (D). Fresh NE was added every 24 h. A representative field of cells is shown in each case.
FIG. 9. Effects of NE and NGF on differentiation of the ␤ 1 -3 subclone of PC12 cells. Cells were plated on collagen-coated plates. Before treatment with agonists, the medium was replaced with Dulbecco's modified Eagle's medium containing 1% horse serum. Cells were