Constitutively Active Mutants of the α1a- and the α1b-Adrenergic Receptor Subtypes Reveal Coupling to Different Signaling Pathways and Physiological Responses in Rat Cardiac Myocytes*

Activation of α1-adrenergic receptors influences both the contractile activity and the growth potential of cardiac myocytes. However, the signaling pathways linking activation of specific α1-adrenergic receptor (AR) subtypes to these physiological responses remain controversial. In the present study, a molecular approach was used to identify conclusively the signaling pathways activated in response to the individual α1A- and α1B-AR subtypes in cardiac myocytes. For this purpose, a mutant α1a-AR subtype (α1a-S290/293-AR) was constructed based on analogy to the previously described constitutively active mutant α1b-AR subtype (α1b-S288–294-AR). The mutant α1a-S290/293-AR subtype displayed constitutive activity based on four criteria. To introduce the constitutively active α1-AR subtypes into cardiac myocytes, recombinant Sindbis viruses encoding either the α1a-S290/293-AR or α1b-S288–294-AR subtype were used to infect the whole cell population with >90% efficiency, thereby allowing the biochemical activities of the various signaling pathways to be measured. When expressed at comparable levels, the α1a-S290/293-AR subtype exhibited a significantly elevated basal level as well as agonist-stimulated level of inositol phosphate accumulation, coincident with activation of atrial natriuretic factor-luciferase gene expression. By contrast, the α1b-S288–294-AR subtype displayed a markedly increased serum response element-luciferase gene expression but no activation of atrial natriuretic factor-luciferase gene expression. Taken together, this study provides the first molecular evidence for coupling of the α1a-AR and the α1b-AR subtypes to different signaling pathways in cardiac myocytes.

the contractile activity and the growth potential of cardiac myocytes. However, despite intense investigation, the signaling pathways linking activation of specific ␣ 1 -AR subtypes to these particular physiological responses remain controversial (1). The situation is complicated by the diversity of ␣ 1 -AR subtypes. Three distinct ␣ 1 -AR subtypes have been identified by molecular cloning (2)(3)(4). Recently, the relationships between the cloned and native ␣ 1 -AR subtypes have been established by comparison of their affinity constants for a wide variety of ␣ 1 -AR subtype-selective antagonists (5,6). From this comparison, it has been suggested that the cloned ␣ 1b -AR represents the native ␣ 1B -AR subtype; the cloned ␣ 1a/c -AR 2 corresponds to the native ␣ 1A -AR subtype; and the cloned ␣ 1d -AR is considered to represent a novel ␣ 1D -AR subtype. With the recognition that multiple ␣ 1 -AR subtypes exist, the roles of the individual ␣ 1 -AR subtypes in mediating specific physiological effects need to be investigated further.
The assignment of particular physiological responses and signaling pathways first requires the elucidation of the specific ␣ 1 -ARs subtypes present in cardiac myocytes. In a previous study, we showed that all three ␣ 1 -AR subtypes are expressed at the mRNA level, but only the ␣ 1A -and ␣ 1B -AR subtypes are detectable at the protein level in neonatal rat cardiac myocytes (7). Based on a wide range of subtype-selective receptor antagonists, this study suggested that the ␣ 1A -AR subtype appeared to be largely responsible for the stimulation of the phosphatidylinositol hydrolysis pathway in response to agonist treatment of these cells, whereas activation of the mitogen-activated protein kinase (MAPK) pathway appeared to occur through a subtype whose pharmacological profile most closely resembled that of the ␣ 1B -AR subtype. Although intriguing, the limited selectivities of the currently available ␣ 1 -AR antagonists do not allow the definitive assignments of functional responses and signaling pathways to be made. Thus, for a given ␣ 1 -AR antagonist, the binding constants for inhibition of signaling responses show a range of differences among published reports that is almost as great as the range of putative differences among the cloned ␣ 1 -AR subtypes (8). Therefore, to circumvent the limitations of a pharmacologic approach, we sought to develop a molecular approach that could be used to identify conclusively which signaling pathways, and ultimately which functional responses, are activated in response to the individual ␣ 1A -and ␣ 1B -AR subtypes in cardiac myocytes. To this end, we constructed a constitutively active mutant of the ␣ 1a -AR subtype by analogy to a previously described constitutively active mutant of the ␣ 1b -AR subtype (9). Such constitutively active receptors have the advantage that their signaling properties can be examined in the absence of agonist. Following introduction of the individual constitutively active ␣ 1 -AR subtypes into the normal cellular context of cardiac myocytes in which the wild type receptor subtypes are expressed, we determined which signaling pathways are activated in response to each mutant receptor subtype in the absence of agonist. Construction and Subcloning of the Mutant ␣ 1a -S 290/293 -AR cDNA-The ␣ 1a -S 290/293 -AR cDNA was constructed by site-directed mutagenesis of the bovine ␣ 1a -AR cDNA, which was generously provided by Dr. Jon Lomasney, Northwestern University Medical School. The oligonucleotides 5Ј-CGCGAGCATAAAGCGCTCAAAACGCTG-3Ј and 5Ј-CGTTTTGAGCGCTTTATGCTCGCGGGA-3Ј were used to generate the appropriate mutations at Lys 290 3 His and Ala 293 3 Leu. The mutant ␣ 1a -S 290/293 -AR cDNA was verified by DNA sequencing.
For expression studies in COS-m6 cells, the cDNA encoding the mutant ␣ 1a -S 290/293 -AR subtype was subcloned into the BamHI site of the pBC12B1 expression vector. The cDNA encoding the mutant ␣ 1b -S 288 -294 -AR subtype in the pBC12B1 expression vector was kindly provided by Dr. Susanna Cotecchia, Lausanne, Switzerland. The ␣ 1b -S 288 -294-AR should be phenotypically identical to the ␣ 1a -S 290/293 -AR, since only substitutions at Lys 290 3 His and Ala 293 3 Leu were shown to be responsible for constitutive activity (9). The cDNAs encoding the wild type ␣ 1a -AR and ␣ 1b -AR subtypes in the pREP4 and pREP8 expression vectors, respectively, were generously provided by Dr. Kenneth Minneman (Emory University Medical School).
For expression studies in cardiac myocytes, the cDNA fragments encoding the mutant ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes were subcloned into the pSinRep5 vector (Invitrogen). The 1.5-kb cDNA fragment encoding the mutant ␣ 1a -S 290/293 -AR was released from the pBC12B1 vector by digestion with BamHI. After filling in the sticky ends with dNTPs and Klenow polymerase, this cDNA fragment was ligated into the StuI site of the pSinRep5 vector. Similarly, the 2-kb cDNA fragment encoding the mutant ␣ 1b -S 288 -294 -AR was released from the pBC12BI vector by digestion with EcoRI and Pml1. The 2-kb cDNA fragment was isolated and further digested with ApaLI to generate a 1.9-kb cDNA fragment. After filling in the sticky ends with dNTPs and Klenow polymerase, this cDNA fragment was ligated into the StuI site of the pSinRep5 vector.
Expression of the Wild Type and Mutant ␣ 1a -S 290/293 and ␣ 1b -S 288/ 294-AR Subtypes in COS-m6 Cells-Expression of the wild type and mutant receptors in COS-m6 cells was carried out by the DEAE-dextran transfection procedure (10). On day 0, COS-m6 cells were plated at a density of 2 ϫ 10 6 cells/100-mm dish in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin. On day 1, cells were transfected with varying concentrations of plasmid DNA and DEAE-dextran. The cells were harvested 48 h later for measurement of total or cell surface receptor binding. For total receptor binding, the cells were lysed with a Polytron homogenizer in ice-cold buffer containing 5 mM Tris-HCl, pH 7.4, and 5 mM EDTA, and the cell lysates were centrifuged at 50,000 ϫ g for 20 min at 4°C. The resulting membrane pellets were resuspended in buffer containing 50 mM Tris-HCl, pH 7.4, and 1 mM EDTA at a protein concentration of 1-3 mg/ml. Receptor expression was monitored by selective displacement of the ␣ 1 -AR antagonist, [ 3 H]prazosin, from these membranes. Briefly, 0.03 mg of membrane protein was incubated in a total volume of 90 l of assay buffer consisting of 50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1% bovine serum albumin, and [ 3 H]prazosin for 45 min at room temperature. In the competition binding experiments, membranes were incubated with 1 nM of the [ 3 H]prazosin in the presence or absence of varying concentrations of competing ligands. In the saturation binding experiments, membranes were incubated with 0.2-15 nM [ 3 H]prazosin in the presence or absence of 10 M phentolamine to determine nonspecific and total binding, respectively. The incubation was terminated by the addition of 1 ml of ice-cold assay buffer, followed by rapid filtration over Whatman GF/C glass fiber filters. After washing the tubes and filters three times with 1 ml of ice-cold assay buffer, membrane-bound [ 3 H]prazosin retained on the filters was counted by liquid scintillation spectrometry.
For cell surface receptor binding, the cells were washed 8 times with Dulbecco's phosphate-buffered saline to remove antagonists that were present in the growth media (see "Results" for details) and then gently released from the dishes by incubation in 1 ml of Dulbecco's phosphatebuffered saline containing 0.05% trypsin and 0.5 mM EDTA for 2 min at 37°C. The cell suspensions were treated with 5 ml of Dulbecco's modified Eagle's medium containing 10% serum to inactivate the trypsin and centrifuged at 1,500 ϫ rpm. The cell pellets were resuspended in assay buffer to give approximately 2.5 ϫ 10 6 cells/ml. The binding of the ␣ 1 -AR antagonist, [ 3 H]prazosin, to the resuspended cells was performed as described above in a final volume of 500 l containing approximately 1 ϫ 10 6 cells. An aliquot of cells was counted by trypan blue staining to ensure the intactness of the cells.
Expression of Mutant ␣ 1a -S 290/293 and ␣ 1b -S 288/294 -AR Subtypes in Rat Cardiac Myocytes-Cardiac myocytes were prepared from hearts of 1-2-day-old Harlan Sprague-Dawley rats. Briefly, the ventricles were removed, digested with a mixture of trypsin, chymotrypsin, and elastase in a Celstir apparatus at 37°C, and subjected to Percoll step gradients to obtain an enriched fraction of greater than 94% myocytes, as described previously (11). Myocytes were suspended in modified Eagle's medium (MEM) containing 5% newborn calf serum, 100 M 5-bromo-2Ј-deoxyuridine, 50 units/ml penicillin, and 50 g/ml streptomycin and plated at a density of 5 ϫ 10 5 /18-mm well, 1 ϫ 10 6 /35-mm well, and 2 ϫ 10 6 /60-mm dish. Following overnight incubation, the serum-containing medium was removed and replaced with a defined serum-free medium, as detailed previously (11). Myocytes were maintained in the defined serum-free medium for 24 h before being used for viral infection.
By standard gene transfer methods, cardiac myocytes are not easily transfected. Therefore, it was necessary to develop a viral infection procedure in order to be able to measure activation of signaling pathways in response to the introduction of a specific mutant receptor subtype in the whole cell population. In another study, we demonstrated the ability of a recombinant Sindbis virus to infect greater than 90% of the cardiac myocytes, as measured by positive ␤-galactosidase staining. 3 Therefore, in the present study, recombinant Sindbis viruses encoding the mutant ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes were used to infect cardiac myocytes. For construction of recombinant Sindbis viruses, the pSinRep5 vector containing the cDNA for either the mutant ␣ 1a -S 290/293 -AR or mutant ␣ 1b -S 288 -294 -AR was linearized with NotI. The pSinRep5 vector containing the cDNA for the LacZ was linearized with XhoI. The capped RNA transcripts were generated by in vitro transcription of the pSinRep cDNAs as well as the DH-BB helper virus DNA, as described by the manufacturer (Invitrogen). The recombinant Sindbis viruses were harvested from the medium of baby hamster kidney cells that had been electroporated 28 h earlier with the capped RNA transcripts.
For measurement of receptor expression, neonatal cardiac myocytes growing on 60-mm dishes were infected with recombinant Sindbis virus encoding either LacZ, mutant ␣ 1a -S 290/293 -AR, or mutant ␣ 1b -S 288 -294 -AR. The appropriate dilutions of recombinant Sindbis virus were determined empirically to yield comparable levels of mutant ␣ 1a -S 290/293 -AR and mutant ␣ 1b -S 288 -294 -AR expression. After the initial 1-h infection period, the medium was supplemented with the ␣ 1 -AR antagonist, WB-4101, to a final concentration of 1 M. The inclusion of WB-4101 in the medium was found to increase significantly the number of mutant receptors expressed on the cell surface. At 48 h post-infection, the cardiac myocytes were lysed in homogenization buffer consisting of 20 mM HEPES, pH 8.0, 2 mM MgCl 2 , 1 mM EDTA, 1 mM benzamidine, 1 mM AEBSF, and 10 g/ml pepstatin A by 12 passages through a 25-gauge needle. The cell lysates were centrifuged at 265,000 ϫ g for 30 min at 4°C to pellet the membrane fractions. The membrane pellets were resuspended in 100 l of 50 mM Tris-HCl, pH 7.4, and 1 mM EDTA, assayed for protein concentration, and used for measurement of receptor expression. Briefly, 0.01 mg of membranes were incubated with a saturating concentration of [ 125 I]HEAT (4 nM) in a total volume of 90 l of incubation buffer consisting of 50 mM Tris-HCl, pH 7.4, 1 mM EDTA, and 1% bovine serum albumin for 45 min at 25°C (7). The reactions were stopped by addition of 1 ml of ice-cold incubation buffer and rapid filtration of the mixture over Whatman GF/C filters. The tubes and filters were washed three times with ice-cold incubation buffer, and the filters were counted in a Beckman gamma counter. Nonspecific binding was determined by the inclusion of 1 M prazosin.
Measurement of Phosphatidylinositol Hydrolysis-In the studies employing COS-m6 cells, cells growing on 18-mm wells were transfected and then labeled for 24 h with medium containing myo-[ 3 H]inositol (2 Ci/ml). After 24 h labeling, the medium was replaced with medium containing 20 mM LiCl followed by addition of agonist or vehicle for another 45 min. The cells were stopped by the addition of ice-cold trichloroacetic acid (final concentration of 6%). The precipitated proteins were removed by centrifugation, solubilized in 0.25 M NaOH, 0.2% SDS, and protein concentrations were determined. The supernatants were extracted three times with 3 volumes of water-saturated diethyl ether and incubated at 55°C for 60 min, and the neutralized aqueous extracts were applied to 1-ml columns of Dowex AG-1-X8, formate form. Total inositol phosphates were eluted using a standard procedure (13) and counted by liquid scintillation spectrometry.
For the neonatal cardiac myocyte studies, cells growing on 18-mm wells were infected with the various recombinant Sindbis viruses. Twenty-four hours later, the cells were labeled with media supplemented with myo-[ 3 H]inositol (2 Ci/ml). At 48 h post-Sindbis virus infection, the cells were treated with 10 mM LiCl in the presence or absence of 100 M phenylephrine and 1 M propranolol for 30 min at 37°C. The cells were stopped by two rapid 1-ml washes of ice-cold phosphate-buffered saline followed by the addition of 1 ml of 6% ice-cold trichloroacetic acid. After scraping, the cells were centrifuged at 14,000 ϫ g for 10 min to remove the precipitated proteins. The supernatants were extracted three times with diethyl ether, and the extracts were applied to 0.5-ml columns of Dowex AG-1-X8, formate form. Inositol phosphates were eluted by a standard procedure (13) and counted by liquid scintillation spectrometry.
Measurement of MAPK Activation-Neonatal rat cardiac myocytes growing on 35-mm wells were treated in the presence or absence of 100 M phenylephrine in the presence of 1 M propranolol for 5 min at 37°C. Reactions were stopped by two washes with ice-cold phosphatebuffered saline followed by addition of 75 l of MAPK extraction buffer, consisting of 20 mM ␤-glycerophosphate, 20 mM NaF, 2 mM EDTA, 1 mM EGTA, 1% Triton X-100, 1 mM vanadate, 1 mM DTT, 10 mM benzamidine, 1 mM AEBSF, 25 g/ml leupeptin, and 5 g/ml pepstatin A. After scraping, the cell extracts were incubated on ice for 15 min and centrifuged at 14,000 ϫ g for 10 min at 4°C. The supernatant fractions were assayed for protein concentration using the Coomassie Plus Protein Assay (Pierce). Equal amounts of protein were electrophoresed on 11% SDS-polyacrylamide gels, and the proteins were transferred to Immobilon-P membrane (Millipore). MAPK activation was determined by immunoblotting with a polyclonal antisera that recognizes only the doubly phosphorylated forms of ERK1 and ERK2 (Anti-active MAPK, Promega, 1:10,000), which is the form of MAPK that possesses myelin basic protein phosphorylating activity. Western blots stripped and reprobed with a monoclonal pan-ERK1/ERK2 antibody (Promega) showed equivalent amounts of ERK1/ERK2 protein across the lanes. The amounts of active and total ERK1 and ERK2 were visualized by incubating the immunoblot with a goat anti-rabbit horseradish peroxidase conjugate (1:5000) followed by enhanced chemiluminescent detection using the SuperSignal Chemiluminescent Substrate Kit (Pierce).
Luciferase Gene Expression-Cultured cardiac myocytes were transfected with 1 g of plasmid DNA using either the SRE-luciferase reporter plasmid (Stratagene, Inc.) or the ANF promoter-luciferase reporter plasmid, which was the generous gift of Dr. J. H. Brown. An aliquot of 1 g of luciferase reporter plasmids (SRE or ANF) was mixed with 7 l of the "Plus" component of LipofectAmine Plus reagent (Life Technologies, Inc.) and 100 l of Opti-MEM medium (Life Technologies, Inc.) and incubated for 15 min at room temperature. In a separate tube, 7 l of the lipid component of LipofectAMINE Plus reagent was added to 100 l of Opti-MEM medium and incubated at room temperature for 15 min. The contents of the two tubes were then mixed together and incubated an additional 30 min at room temperature. The cardiac myocyte tissue culture plates (60 mm) were washed with Opti-MEM medium and then 1 ml of fresh Opti-MEM medium without antibiotics was added to each cell plate. The contents of the DNA/LipofectAMINE Plus mixture was then added to each dish of myocytes. After 5 h, the transfection media were removed and the myocytes maintained in Opti-MEM medium without antibiotics.
For luciferase gene expression assays, the cardiac myocytes were lysed in 0.5 ml of 1ϫ lysis buffer (Promega Corp.). The cell lysates were centrifuged at 15,000 ϫ g for 1 min at 4°C, and the supernatants were transferred to a new tube. Approximately 50 l of the supernatant was assayed in duplicate for SRE-and/or ANF-luciferase gene expression using a Luciferase Assay System with Reporter Lysis Buffer (Promega Corp.). Relative light units were measured utilizing a Lumat LB9507 EG&G Berthold luminometer with myocyte cell lysates. Following luciferase assays, ␤-galactosidase enzyme assays were performed in duplicate samples. The luciferase relative light unit numbers were divided by the ␤-galactosidase enzyme assay numbers to correct for the transfection efficiency in the myocytes. The cardiac cell lysate from the luciferase assays was assayed spectrophotometrically for the ␤-galactosidase enzyme activity in cardiac myocytes using the ␤-Galactosidase Enzyme Assay System with Reporter Lysis Buffer (Promega).
Data Analysis-Values represent the mean Ϯ S.E. Saturation binding and competition binding curves were analyzed using the nonlinear, least squares regression analysis program GraphPad Prizm (GraphPad Software). Estimates of ligand binding affinity (K D ) and receptor density (B max ) were obtained from saturation isotherms by fitting the data to a rectangular hyperbola. IC 50 values from competition binding experiments were converted to K i values using the Cheng and Prusoff equation (14): is the concentration of ligand used and K D was determined from the saturation binding studies.

RESULTS
Construction and Characterization of the Constitutively Active ␣ 1a -S 290/293 -AR-To address whether constitutively active mutants of the ␣ 1a -AR and the ␣ 1b -AR subtypes could be used to probe their subtype-specific coupling to signaling pathways, a constitutively active mutant of the ␣ 1a -AR needed to be constructed. A constitutively active mutant ␣ 1b -AR subtype has already been constructed and characterized by Dr. Cotecchia (9). Accordingly, the ␣ 1a -AR subtype was mutated at Lys 290 and Ala 293 (␣ 1a -S 290/293 -AR) to mirror the substitutions previously found to result in maximal constitutive activation of the mutant ␣ 1b -AR subtype (9). Properties that have been shown to be characteristic of constitutive activation include the following: 1) an increased affinity of a receptor for agonist but not for antagonist; 2) an elevated basal level of signaling; and 3) an increased agonist-stimulated level of signaling. To determine whether the mutant ␣ 1a -S 290/293 -AR subtype exhibited any or all of these properties, the mutant ␣ 1a -S 290/293 -AR subtype was expressed and then characterized in COS-m6 cells. Since COS-m6 cells do not express wild type ␣ 1 -ARs, these cells provide a convenient background in which to study the properties of the mutant ␣ 1a -S 290/293 -AR subtype.
Increased Affinity for Agonist-COS-m6 cells transfected with either the wild type ␣ 1a -AR or the mutant ␣ 1a -S 290/293 -AR were characterized in terms of their affinity for various agonists and antagonists by monitoring the displacement of [ 3 H]prazosin through agonist or antagonist competition binding assays. Phenylephrine and norepinephrine, both full ␣ 1 -AR agonists (15), displaced [ 3 H]prazosin from membranes expressing the ␣ 1a -S 290/293 mutant with approximately 40-fold higher affinities than those expressing the wild type receptor ( Fig. 1; Table I). The partial imidazoline agonist, oxymetazoline (15), also demonstrated a higher affinity for the ␣ 1a -S 290/293 -AR ( Fig.  1), although the difference in affinity, in this case, was only 6-fold (Table I). The antagonist, 5-methylurapidil, demonstrated similar affinity for both the wild type and mutant receptors ( Fig. 1; Table I). The agonist and antagonist competition curves fitted best to a single-site model, where Hill coefficients were between 0.7 and 1.0. These data demonstrate that the mutant ␣ 1a -S 290/293 -AR showed an increase in binding affinity for agonist but not for antagonist compared with its wild type counterpart. These data provide the first confirmation that the mutant ␣ 1a -S 290/293 -AR possesses properties reminis-cent of other constitutively active receptors (9). Furthermore, these data suggest that the difference in agonist binding affinity observed between the wild type and mutant ␣ 1a -ARs may be related to the intrinsic activity of the agonist (i.e. a larger difference is observed for the full agonist, phenylephrine, compared with the partial agonist, oxymetazoline), and this observation is in agreement with that previously described for a constitutively active mutant of the ␤ 2 -AR (16).
Enhanced Basal and Agonist-stimulated Phosphatidylinositol Hydrolysis-Previously, it was shown that all ␣ 1 -AR subtypes couple to the phosphatidylinositol hydrolysis pathway when overexpressed at high, "non-physiological" levels (e.g. Ͼ1500 fmol of receptor/mg of protein) (17)(18)(19). In the present report, we decided to study the signaling properties of both the constitutively active and wild type ␣ 1 -ARs when they were expressed at more physiological levels in order to get a better idea of how these receptors operate in vivo. By altering the amount of plasmid DNA used in the transfections, COS-m6 cells expressing various levels of receptor were obtained. In particular, cells expressing receptor in the 300 fmol/mg protein range were studied, since this density seems comparable with many intact cellular systems that normally express ␣ 1 -ARs. Examples of B max values for various tissues are as follows: 256 fmol/mg protein in rat cardiac myocytes (20); 158 fmol/mg protein in rat cerebral cortex (21); and 233, 313, and 690 fmol/mg protein in rat hippocampus, vas deferens, and liver, respectively (22).
The second confirmation that the mutant ␣ 1a -S 290/293 -AR exhibits properties characteristic of other constitutively active receptors was a reproducible and marked increase in the basal level of phosphatidylinositol hydrolysis (i.e. in the absence of agonist). As shown in Fig. 2, COS-m6 cells expressing the mutant ␣ 1a -S 290/293 -AR displayed a high basal level of inositol phosphate accumulation at the lowest receptor density examined of 199 fmol/mg protein. By contrast, cells expressing the wild type ␣ 1a -AR showed no detectable basal level of inositol phosphate accumulation over a wide range of receptor densities from 308 to 2098 fmol/mg protein. These data demonstrate that the mutant ␣ 1a -S 290/293 -AR had an increased ability to interact with G protein to stimulate inositol phosphate accumulation in the absence of agonist, which is indicative of constitutive activity. Moreover, since the mutant ␣ 1a -S 290/293 -AR was expressed at a lower level than the wild type ␣ 1a -AR, these data likely underestimate the greater constitutive activity of the mutant ␣ 1a -S 290/293 -AR.
Also, as shown in Fig. 2, COS-m6 cells expressing the mutant ␣ 1a -S 290/293 -AR displayed a higher agonist-stimulated level of inositol phosphate accumulation compared with cells expressing the wild type ␣ 1a -AR. In fact, the wild type ␣ 1a -AR had to be expressed at a 10-fold higher level than the mutant ␣ 1a -S 290/ 293-AR before their maximal responses were comparable. These data are consistent with previous studies of wild type and mutant ␤ 2 -ARs, which showed that greatly elevated expression of the wild type receptor was able to increase agonist-stimulated cAMP accumulation to a level comparable to that achieved with the constitutively active mutant receptor (19).
Increased Receptor Expression in the Presence of Antagonist-We observed that for any given plasmid DNA concentration used for transfection, the level of expression of the mutant ␣ 1a -S 290/293 -AR was considerably lower than that of the wild type receptor. A similar finding was recently reported for a constitutively active mutant of the ␤ 2 -AR (23). Moreover, inclusion of antagonist in the post-transfection period was shown to increase the level of expression of the constitutively active mutant of the ␤ 2 -AR. To determine whether this property was shared by the constitutively active mutant of the ␣ 1a -S 290/293 -AR, inclusion of a variety of antagonists in the post-transfection period was studied. These antagonists were removed just prior to collection of the cells by exhaustive washing (8 times) with Dulbecco's phosphate-buffered saline. As shown in Fig. 4, the ␣ 1a -AR-selective antagonists, 5-methylurapidil and WB-4101, in the post-transfection period led to a significant increase in the number of mutant ␣ 1a -S 290/293 -ARs on the cell surface. The ␣ 1d -AR selective antagonist, BMY 7378, resulted in a similar increase. Interestingly, the non-selective ␣ 1 -AR antagonist, prazosin, was unique in that it did not elevate the density of mutant ␣ 1a -S 290/293 -AR. These data suggest that the ability of antagonists to enhance receptor expression is a property shared by constitutively active mutants of the ␤ 2 -AR and ␣ 1a -S 290/293 -AR. The mechanism by which antagonists result in higher receptor expression is not yet clear but may relate to stabilization of receptor conformation (23). In this regard, the finding that not all antagonists display this property is interesting in that it might support the existence of multiple receptor conformations of the mutant ␣ 1a -S 290/293 -AR.
Comparison of the Signaling Properties of the Constitutively Active ␣ 1a -S 290/293 and ␣ 1b -S 288/294 -AR Subtypes in COS-m6 Cells-By having shown that the mutant ␣ 1a -S 290/293 -AR possesses all of the characteristics of a constitutively active receptor, the second aim of this study was to compare the signaling properties of the constitutively active mutant ␣ 1a -S 290/293 -AR subtype with that of the analogous constitutively active mutant ␣ 1b -S 288 -294 -AR subtype developed by Dr. Cotecchia (9), since it is their ability to activate signaling pathways in the absence of agonist that makes them useful tools for linking a specific receptor subtype to a particular signaling pathway. Activation of both the phospholipase C and MAPK signaling pathways has been demonstrated upon addition of ␣ 1 -AR agonists to cardiac myocytes (1,7,27,30). Fig. 5 shows a comparison of basal and agonist-stimulated phosphatidylinositol hydrolysis in COS-m6 cells expressing either the mutant ␣ 1a -S 290/293 -AR subtype or the ␣ 1b -S 288 -294 -AR subtype. As can be seen, the mutant ␣ 1a -S 290/293 -AR subtype displayed an elevated basal activity and a higher agonist-stimulated phosphatidylinositol hydrolysis activity compared with the mutant ␣ 1b -S 288 -294 -AR subtype. In fact, the mutant ␣ 1b -S 288 -294 -AR subtype had to be expressed at a five times greater receptor density in order to observe similar levels of basal and agonist-stimulated activity as the mutant ␣ 1a -S 290/293 -AR subtype, suggesting that the mutant ␣ 1b -S 288 -294 -AR subtype had a lower potency in activating the phosphatidylinositol hydrolysis signaling pathway. These data indicate that, even though they are closely related structurally, the mutant ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes differ functionally in terms of their relative abilities to stimulate the phosphatidylinositol hydrolysis signaling pathway. Although interesting, meaningful interpretation of these results is difficult since COS-m6 cells do not normally express the ␣ 1a -and ␣ 1b -ARs and, thus, are unlikely to possess the physiologically appropriate G proteins and downstream signaling components. To circumvent this problem, we decided to study the signaling properties of the mutant ␣ 1a -S 290/293 -and ␣ 1b -S 288 -294 -AR subtypes in another type of cells in which these receptor subtypes are normally expressed. For this purpose, we selected neonatal rat cardiac myocytes.
Comparison of the Signaling Properties of the Constitutively Active ␣ 1a -S 290/293 and ␣ 1b -S 288/294 -AR Subtypes in Rat Cardiac Myocytes-In a previous study, we showed that both the ␣ 1A -and ␣ 1B -AR subtypes are expressed in neonatal rat cardiac myocytes (7). Activation of both the phospholipase C and MAPK signaling pathways has been demonstrated upon addition of ␣ 1 agonists to cardiac myocytes (1,7,27,30). Moreover, several studies have demonstrated that the combined activation of these receptor subtypes is associated with stimulation of several signaling pathways, including stimulation of phosphatidylinositol hydrolysis (19) and activation of ERK (24), JNK (25), and p38 kinase (26). However, the definitive assignment of these signaling pathways to the activation of the individual ␣ 1A -and ␣ 1B -AR subtypes has been hampered by the lack of pharmacological tools with the requisite subtype specificity. With the development of constitutively activated ␣ 1a -AR and ␣ 1b -AR subtypes, we now have the molecular tools to examine their respective signaling properties in the appropriate cellular context in which the wild type ␣ 1 -AR receptors are normally expressed, but in the absence of agonist to stimulate the various wild type ␣ 1 -AR subtypes.
Whereas standard gene transfer methods were effective for introducing luciferase reporter plasmids into cardiac myocytes, they were ineffective for expression of the constitutively active ␣ 1 -AR subtypes in myocytes. However, we recently demonstrated the first successful use of recombinant Sindbis viruses to express G protein ␤ subunits in cardiac myocytes. 3 This viral infection procedure was found to provide a rapid and efficient method to introduce genes into cardiac myocytes, with greater than 90% of cardiac myocytes being successfully targeted. Thus, activation of signaling pathways in response to the introduction of the individual constitutively activated receptor subtypes can be measured on the whole cell population. Therefore, to utilize this method, we constructed recombinant Sindbis viruses encoding the constitutively activated ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes. To control for any nonspecific effects of Sindbis virus infection, a recombinant Sindbis virus containing the bacterial lacZ gene or pSinRep5 vector was utilized. The expression of the constitutively activated ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes was quantitated by radioligand binding. As shown in Fig. 6, uninfected cardiac myocytes and myocytes infected with the LacZ Sindbis virus expressed similar numbers of ␣ 1 -ARs (B max ϭ 270 and 178 Ϯ 16 fmol/mg protein for control and LacZ virus-infected cells, respectively). By contrast, cardiac myocytes infected with either the ␣ 1a -S 290/293 -AR Sindbis virus or the ␣ 1b -S 288 -294 -AR Sindbis virus expressed approximately 10-fold higher receptor densities than myocytes infected with the LacZ Sindbis virus (B max ϭ 2120 Ϯ 219 and 1518 Ϯ 350 fmol/mg protein for ␣ 1a -S 290/293 -AR virus-infected and ␣ 1b -S 288 -294 -AR virusinfected cells, respectively).
By having confirmed receptor expression, we examined which signaling pathways were activated in cardiac myocytes expressing either the constitutively active ␣ 1a -S 290/293 -AR subtype or the ␣ 1b -S 288 -294 -AR subtype. Fig. 7 shows a comparison of basal and phenylephrine-stimulated phosphatidylinositol hydrolysis in cardiac myocytes expressing similar levels of constitutively activated ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes. Under the basal condition (30 min with 10 mM LiCl in the absence of agonist), cells expressing the constitutively activated ␣ 1a -S 290/293 -AR subtype displayed a 2.4-fold increase in inositol phosphate accumulation compared with the LacZ cells, which showed no enhancement of inositol phosphates accumulation during the 30 min of incubation with 10 mM LiCl. On the other hand, cells expressing the constitutively activated ␣ 1b -S 288 -294 -AR subtype showed no increase in inositol phosphates accumulation during the 30 min of incubation with 10 mM LiCl, which makes them indistinguishable from the LacZ cells (Fig.  7A). Under the agonist-stimulated condition, phenylephrine produced a 6-fold increase in inositol phosphates accumulation in the LacZ cells. Cells expressing the constitutively active ␣ 1a -S 290/293 -AR subtype showed a further 2-fold increase in phenylephrine-stimulated inositol phosphate accumulation compared with the LacZ cells, whereas the phenylephrinestimulated activity in the cells expressing the constitutively activated ␣ 1b -S 288 -294 -AR subtype was indistinguishable from that observed in the LacZ cells (Fig. 7B). These data clearly demonstrate that the constitutively activated ␣ 1a -S 290/293 -AR subtype couples to the phosphatidylinositol hydrolysis signaling pathway in both the absence and presence of agonist in cardiac myocytes, which confirms and extends previous pharmacological data from our own laboratory (7). By contrast, the constitutively activated ␣ 1b -S 288 -294 -AR subtype was not able to couple to the phosphatidylinositol hydrolysis signaling pathway in cardiac myocytes.

Comparison of the MAPK Signaling Pathways of the Constitutively Active ␣ 1a -S 290/293 and ␣ 1b -S 288/294 -AR Subtypes in
Rat Cardiac Myocytes-Next, we examined the MAPK signaling pathway. Fig. 8 shows a comparison of basal and phenylephrine-induced activation of MAPK activity in cardiac myocytes expressing similar levels of the activated ␣ 1a -S 290/ 293-AR and ␣ 1b -S 288 -294 -AR subtypes. Under the basal condition (in the absence of agonist), cardiac myocytes expressing the constitutively activated ␣ 1a -S 290/293 -AR subtype showed a slight activation (1.5-fold), with the level of active MAPK being similar to that observed in LacZ cells (Fig. 8A). By contrast, cells expressing the constitutively activated ␣ 1b -S 288 -294 -AR subtype displayed marked activation, with the level of active MAPK being 5.8-fold above that observed in LacZ cells. Under the agonist-stimulated condition, phenylephrine produced a modest increase in the level of active MAPK in the LacZ cells, which was less than the level of active MAPK observed in the ␣ 1b -S 288 -294 -AR subtype in the absence of agonist. In cells expressing the constitutively activated ␣ 1a -S 290/293 -AR subtype, phenylephrine stimulation increased the level of active MAPK only marginally (1.3-fold) compared with that in the phenylephrine-stimulated LacZ cells. In contrast, cells expressing the constitutively activated ␣ 1b -S 288 -294 -AR subtype showed a further 3.3-fold increase in the agonist-stimulated level of active MAPK compared with that in the phenylephrinestimulated LacZ cells (Fig. 8B). These results clearly demonstrate that the constitutively activated ␣ 1b -S 288 -294 -AR subtype effectively couples to the MAPK signaling pathway both in the absence and presence of agonist in cardiac myocytes, whereas the constitutively activated ␣ 1a -S 290/293 -AR subtype shows little or no ability to couple to the MAPK signaling pathway in cardiac myocytes. Not only was the strict and specific coupling of the ␣ 1b -S 288 -294 -AR subtype to the MAPK kinase signaling pathway, and the ␣ 1a -S 290/293 -AR subtype to the phosphatidylinositol hydrolysis pathway unexpected, this specificity was observed even though the mutant receptor subtypes were being expressed at 10-fold higher levels than the wild type receptor subtypes.
Comparison of Gene Activation by the Signaling Pathways of the Constitutively Active ␣ 1a -S 290/293 and ␣ 1b -S 288/294 -AR Sub-types in Rat Cardiac Myocytes-We next examined the effect of the two signaling pathways on downstream events at the level of gene regulation in cardiac myocytes. In order to assess the impact of the two divergent signaling pathways activated by the constitutively activated ␣ 1a -AR and ␣ 1b -AR subtypes, we chose to measure the downstream effect on ANF and c-fos cardiac gene transcription. Both of these genes are activated during norephinephrine-induced cardiac hypertrophy in the heart, and as such represent important targets for the ␣ 1A -AR and ␣ 1B -AR subtype signaling pathways. The ANF gene is activated by the ␣ 1 -AR receptor agonist, phenylephrine, through multiple promoter elements (AP-1, SP1, A/T, and SRE, see Refs. [27][28][29][30][31], and is linked to activation of PI hydrolysis (32)(33)(34)(35)(36). Moreover, previous reports have suggested that the ␣ 1A -AR subtype mediates the activation of ANF gene expression (33,37). The early response genes (c-fos, c-jun, and c-myc) are activated by stimulation of ␣ 1 -ARs (36, 38 -43). The c-fos gene is up-regulated during cardiac hypertrophy of myocytes (44 -45). Therefore, we investigated the affect of the constitutively activated ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes on c-fos gene regulation in cardiac myocytes.
As the c-fos gene is regulated by the (SRE) element in the promoter, we utilized luciferase gene expression reporter constructs coupled to the SRE promoter element. To measure ANF and c-fos gene activation, we measured the activation of the SRE-and the ANF-luciferase gene reporter constructs in cardiac myocytes co-transfected with the activated ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes. Fig. 9 shows a comparison of luciferase gene activity in transfected cardiac myocytes expressing similar levels of the activated ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes. In the absence of agonist, cells expressing the constitutively activated ␣ 1a -S 290/293 -AR subtype and the ANF-luciferase reporter construct showed a 2.4-fold increase in the level of luciferase activity (Fig. 9). By contrast, cells expressing the constitutively activated ␣ 1b -S 288 -294 -AR subtype and the ANF-luciferase reporter construct displayed no significant increase in the level of luciferase activity compared with control. In similar experiments in the absence of agonist, cells expressing the constitutively activated ␣ 1b -S 288 -294 -AR subtype and the SRE-luciferase reporter plasmid displayed a marked 4.7-fold increase in the level of luciferase activity. By contrast, the constitutively activated ␣ 1a -S 290/293 -AR and the SRE-luciferase construct showed no significant increase in the level of luciferase activity compared with control. Taken together, these data have several important ramifications. First, the ␣ 1a -S 290/293 -AR subtype preferentially activates ANF-luciferase activity compared with SRE/c-fos- luciferase activity. This is consistent with the ability of ␣ 1a -S 290/293 -AR to activate PI hydrolysis. Second, the lack of ␣ 1b -S 288 -294 -AR subtype stimulation of ANF-luciferase activity is consistent with the inability of the ␣ 1b -S 288 -294 -AR subtype to activate PI hydrolysis (Fig. 7). Third, the ␣ 1b -S 288 -294 -AR subtype activation of SRE/c-fos-luciferase activity compared with ANF-luciferase correlates with its ability to activate MAPK activity. Taken together, these results indicate that the ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes couple to distinct effector pathways and physiological responses in rat cardiac myocytes. These data confirm and extend previous pharmacological results that reached a similar conclusion (7). DISCUSSION Constitutively Activated Receptors-Constitutively activated receptors demonstrate various properties that set them apart from their wild type counterparts, but it is their ability to activate signaling pathways in the absence of agonist that makes them valuable molecular tools. Particularly in the case of the ␣ 1 -ARs, where pharmacological tools do not have the requisite specificity to allow the individual subtypes to be studied in isolation, the use of constitutively activated receptor subtypes provides a strategy to study the signaling properties of the individual receptor subtypes. In the present study, we constructed a constitutively active ␣ 1a -S 290/293 -AR subtype by making the analogous mutations in the COOH-terminal end of the third cytoplasmic loop that produced the previously described constitutively active ␣ 1b -S 288 -294 -AR subtype by Dr. Cotecchia (9). We did not characterize the properties of the constitutively active ␣ 1b -S 288 -294 -AR subtype as it has already been reported (9). Characterization of the functional properties of this mutant ␣ 1a -S 290/293 -AR subtype in COS-m6 cells revealed that it possesses the three criteria that define constitutive activation as follows: 1) an increased affinity of the mutant receptor for agonist but not for antagonist; 2) an elevated basal level of signaling; and 3) an increased agonist-stimulated level of signaling (9,23). Clearly, these data indicate that mutations in the analogous region of the third cytoplasmic loop previously reported to induce constitutive activation of the ␣ 1b -AR subtype also evoke constitutive activation of the ␣ 1a -AR subtype. Furthermore, an additional feature of the mutant ␣ 1a -S 290/293 -AR subtype was observed, namely the ability of antagonists to enhance the level of receptor expression. Since this feature has also been observed in the case of the constitutively active ␤-AR (23), it may represent a fourth property characteristic of constitutively active receptors. With regard to this property, it is interesting that only some antagonists increase the level of receptor expression, suggesting multiple conformational states may exist that can be distinguished by the various antagonists. Future studies will be necessary to elucidate the mechanism by which various antagonists can increase receptor density.
The Mutant ␣ 1a -S 290/293 -AR and ␣ 1b -S 288/294 -AR Subtypes Couple to Different Signaling Pathways and Gene Expression in Cardiac Myocytes-With the development of constitutively activated ␣ 1 -AR subtype probes, we applied these molecular tools to cardiac myocytes in order to decipher which signaling pathways and physiological responses (i.e. gene transcription) were activated by each of the endogenous ␣ 1 -AR subtypes. Rat cardiac myocytes express comparable levels of the endogenous ␣ 1A -AR and ␣ 1B -AR subtypes (138.7 and 147.3 fmol/mg protein, respectively, see Ref. 7). Therefore, they must also possess the appropriate complement of heterotrimeric G proteins, effector molecules, and other ancillary proteins that constitute physiologically relevant ␣ 1 -AR stimulus-response pathways. Thus, following the introduction of either the constitutively activated ␣ 1a -S 290/293 -AR or ␣ 1b -S 288 -294 -AR subtypes in these cells, the recombinant activated receptor subtype should compete with its native receptor subtype to couple to its appropriate signaling pathway(s) in a physiologically relevant manner. Since the constitutively activated receptor subtypes do not require the presence of agonist, the signaling properties of the individual ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes should be evident as an enhanced basal level of stimulation of the signaling pathways (i.e. in the absence of agonist).
A novel finding of this study is that the ␣ 1b -S 288 -294 -AR subtype is responsible for activation of SRE/c-fos-luciferase gene expression and MAPK signaling pathway. The SRE from the c-fos promoter has been shown to be the point of integration of MAPK signaling pathways (47)(48)(49)(50)(51). The Ras-Raf-MEK-ERK signaling pathway activates transcription factors Elk-1 and Sap-1a, which bind to the SRE of the c-fos promoter (47)(48)(49)(50)(51)(52)(53)(54)(55)(56)(57)(58)(59)(60)(61). Interestingly, the SRE from the ANF promoter is different than that from the c-fos promoter in that it does not contain the sequences necessary for the binding of Elk-1 transcription factor (62). Our finding that the ␣ 1b -S 288 -294 -AR subtype is responsible for SRE/c-fos gene activation in cardiac myocytes is in contrast with the previous report of Deng et al. (40). They demonstrated, using pharmacological methods, that the ␣ 1a -S 290/293 -AR subtype was responsible for c-fos gene activation in cardiac myocytes (40). However, our results are in agreement with the previously reported pharmacological studies in vascular smooth muscle cells, where the ␣ 1b -S 288 -294 -AR subtype was responsible for c-fos gene activation (39). The discrepancy in findings may be due to the limitations of the pharmacological approach, where the ␣ 1 -AR subtype agonist and antagonists are not sufficiently selective.
These data, as well as previously published pharmacological data (7), indicate that the ␣ 1b -S 288 -294 -AR subtype couples to the MAPK/SRE/c-fos signaling pathway in rat cardiac myo-cytes. In Fig. 8, we observed the ␣ 1b -S 288 -294 -AR subtype preferentially activates of the MAPK signaling pathway in cardiac myocytes. Finally, these data are consistent with another study in NIH-3T3 cells that reported differential coupling of the ␣ 1aand ␣ 1b -AR subtypes to the MAPK signaling pathways (63).
Previously, differential coupling to the PI signaling pathway has been shown to exist between the wild type ␣ 1a -and ␣ 1b -AR subtypes when heterologously expressed in COS cells (19). This differential coupling to the PI signaling pathway was also observed in the present study (Fig. 2) when heterologously expressed in COS cells. Moreover, both the wild type and constitutively active mutant ␣ 1a -and ␣ 1b -AR subtypes showed this differential coupling indicating the difference is a property of the receptor subtypes themselves and not a characteristic of the constitutive activating mutation. Finally, this is the first conclusive demonstration of differential coupling of the ␣ 1aand ␣ 1b -AR subtypes in a primary cell type (i.e. cardiac myocytes) where the two receptor subtypes are normally expressed. These data indicate that the two receptor subtypes must possess differences in their abilities to interact with downstream components of these pathways. The basis for these differences may involve the selective interaction of each ␣ 1 -AR subtype with a different heterotrimeric G protein to produce distinct bifurcating signals in the form of G␣ and G␤␥ subunits. Since the phenylephrine-mediated stimulation of the phosphatidylinositol hydrolysis pathway in cardiac myocytes is insensitive to pertussis toxin (64), it is likely that the ␣ 1A -AR subtype couples though a member of the G q/11 protein family to regulate phospholipase C-␤. The predominant phospholipase C-␤ isoform in rat neonatal cardiac myocytes is phospholipase C-␤3 (65), which can be regulated by either the ␣ or the ␤ subunits of the G q/11 protein family in vitro (66). Whether the ␣ or the ␤ subunits of the G q/11 protein family are responsible for in vivo regulation will be the subject of future investigations.
The underlying mechanism for activation of the MAPK pathway in rat cardiac myocytes has been controversial. The present study sheds new insights on this mechanism. Since agonist stimulation of the MAPK pathway is insensitive to pertussis toxin in cardiac myocytes (67), this suggests that the ␣ 1B -AR subtype associates with a member of the G q/11 or G 12/13 protein family rather than the G i/o protein family. However, other than the activation of phospholipase C-␤, the downstream components regulated by the ␣ or ␤␥ subunits of G q/11 or G 12/13 have yet to be conclusively identified. Studies by Thorburn and colleagues (68) suggest that agonist-induced activation of the MAPK pathway is mediated by Raf-1 kinase in cardiac myocytes. Since Raf-1 kinase can be activated by protein kinase C (24, 69 -71), which, in turn, can be activated by products of the PI hydrolysis pathway (72), this could provide a mechanism for activation of the MAPK pathway. However, this mechanism is difficult to reconcile with the results of the present study showing the ␣ 1b -S 288 -294 -AR subtype potently stimulates SRE-luciferase gene expression without any activation of the PI hydrolysis pathway or ANF gene expression. This argument is further supported by the fact that protein kinase C has been shown to activate ANF gene expression in myocytes (29 -30). Additionally, the recent finding that transgenic hearts overexpressing the G q ␣ subunit exhibited marked stimulation of the PI hydrolysis pathway but no activation of the MAPK pathway also argues against this mechanism (73). Alternatively, the ␤␥ subunits rather than the ␣ subunits of G q/11 or G 12/13 could be involved in stimulation of the MAPK pathway, since it has been shown that the ␤␥ subunits released from G i can activate a Ras-dependent pathway leading to stimulation of Raf-1 kinase (74 -76). Whether the ␤␥ subunits released from G q/11 or G 12/13 can similarly activate a Ras-dependent pathway in cardiac myocytes will be the subject of future investigations.
The Mutant ␣ 1a -S 290/293 -AR and ␣ 1b -S 288/294 -AR Subtypes May Mediate Different Physiological Responses in the Heart-The demonstration that the ␣ 1A -AR and ␣ 1B -AR subtypes couple to different signaling pathways may explain the wide variety of contractile and cell growth processes that are altered upon addition of ␣ 1 agonists to cardiac myocytes. Activation of both the phospholipase C and MAPK signaling pathways has been demonstrated upon addition of ␣ 1 agonists to cardiac myocytes (1,7,27,30). Interestingly, transgenic hearts overexpressing the wild type G q ␣ subunit showed stimulation of the phosphatidylinositol hydrolysis pathway but no activation of the MAPK pathway (73). Stimulation of the phosphatidylinositol hydrolysis pathway was associated with severe contractile defects as well as an increased cell size with enhanced expression of ANF, ␤-myosin heavy chain, and ␣-skeletal actin. In addition, the expression of the constitutively activated G q ␣ subunit in the heart caused cardiac hypertrophy, which was followed by apoptosis of myocytes through an increase in p38 and JNK activities (76).
The results of the present study would predict that myocytes overexpressing the constitutively activated ␣ 1a -S 290/293 -AR subtype through its activation of the PI hydrolysis pathway should produce similar physiological effects. This is the topic of ongoing investigations. By contrast, myocytes overexpressing the constitutively activated ␣ 1b -S 288 -294 -AR subtype through its activation of a different effector signaling pathway should produce a different repertoire of physiological effects. Previous studies on this point are controversial. On the one hand, Milano and colleagues (77) reported that transgenic hearts overexpressing the constitutively activated ␣ 1b -S 288 -294 -AR subtype showed a modest degree of cell hypertrophy, although stimulation of the phospholipase C and MAPK signaling pathways was not examined. Investigators (78) have demonstrated that the overexpression of ␣ 1B -adrenergic receptors in the heart induced left ventricular dysfunction but not activation of ␤-myosin heavy chain and/or myosin light chain-2v gene expression. On the other hand, Akhter and colleagues (12) showed that transgenic hearts overexpressing the wild type ␣ 1b -AR subtype did not develop hypertrophy despite an activation of ANF gene expression. Although the MAPK signaling pathway was not examined, these investigators reported an elevated diacylglycerol content that was attributed to stimulation of the PI hydrolysis pathway. However, in view of the results of the present study, additional analysis of these transgenic hearts will be needed to determine whether the elevated diacylglycerol content was in fact due to stimulation of a phospholipase C pathway or perhaps due to activation of a phospholipase D pathway. Finally, the results of the present study showing the constitutively activated ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes couple to different signaling pathways raise the possibility that activation of both receptor subtypes may be necessary to reproduce all the features of the ␣ 1 -AR-induced hypertrophic phenotype. Future experiments employing constitutively activated ␣ 1a -S 290/293 -AR and ␣ 1b -S 288 -294 -AR subtypes should provide a definitive assessment of the roles of these receptor subtypes in mediating ␣ 1 -ARinduced cardiac hypertrophy.