A Transgenic Mouse Model of Heart Failure Using Inducible Gαq*

Receptors coupled to Gαq play a key role in the development of heart failure. Studies using genetically modified mice suggest that Gαq mediates a hypertrophic response in cardiac myocytes. Gαq signaling in these models is modified during early growth and development, whereas most heart failure in humans occurs after cardiac damage sustained during adulthood. To determine the phenotype of animals that express increased Gαq signaling only as adults, we generated transgenic mice that express a silent Gαq protein (GαqQ209L-hbER) in cardiac myocytes that can be activated by tamoxifen. Following drug treatment to activate Gαq Q209L-hbER, these mice rapidly develop a dilated cardiomyopathy and heart failure. This phenotype does not appear to involve myocyte hypertrophy but is associated with dephosphorylation of phospholamban (PLB), decreased sarcoplasmic reticulum Ca2+-ATPase activity, and a decrease in L-type Ca2+ current density. Changes in Ca2+ handling and decreased cardiac contractility are apparent 1 week after GαqQ209L-hbER activation. In contrast, transgenic mice that express an inducible Gαq mutant that cannot activate phospholipase Cβ (PLCβ) do not develop heart failure or changes in PLB phosphorylation, but do show decreased L-type Ca2+ current density. These results demonstrate that activation of Gαq in cardiac myocytes of adult mice causes a dilated cardiomyopathy that requires the activation of PLCβ. However, increased PLCβ signaling is not required for all of the Gαq-induced cardiac abnormalities.

Following cardiac injury, up-regulation of angiotensin II and catecholamines leads to activation of G␣ q and progression to heart failure. The signaling pathway that mediates G␣ q -induced cardiomyopathy is unclear. The presumption has been that activation of PLC␤ by G␣ q is responsible for the cardiac pathology, but this hypothesis has not been directly tested. Activation of PLC␤ leads to the release of Ca 2ϩ from inositol (1,4,5)-trisphosphate-sensitive stores, but the role of this signaling event in cardiac myocytes is unknown. Activation of PLC␤ also leads to the diacylglycerol-dependent activation of some protein kinase C (PKC) isozymes. Although there are a number of mouse models with genetically altered PKC function in cardiac myocytes, it remains unclear whether activation of PKC is responsible for G␣ q -induced cardiomyopathy (4). The presence of several PKC isozymes in the myocardium complicates the interpretation of animal studies that examine the cardiac effect of a particular PKC isozyme.
Results from genetically modified mouse models suggest that G␣ q signaling mediates a hypertrophic response in cardiac myocytes (5)(6)(7)(8). Targeted overexpression of wild-type G␣ q at supraphysiologic levels in cardiac myocytes using the ␣ myosin heavy chain (MHC) promoter resulted in cardiac hypertrophy and subsequent decompensation into heart failure (5). Because wild-type G␣ q can sequester ␤␥ subunits, its expression might interfere with signals initiated by ␤␥. Mutation of Gln 209 to Leu of G␣ q produces a mutant (G␣ q Q209L) that no longer hydrolyzes GTP, resulting in constitutive signaling to downstream effectors. G␣ q Q209L does not bind to ␤␥, so expression of this mutant does not interfere with ␤␥ signaling. Conventional ␣MHC-driven expression of G␣ q Q209L in cardiac myocytes also led to cardiac hypertrophy. Surprisingly, these mice developed heart failure with age even though the mutant protein became undetectable (6).
Standard transgenic and gene knock-out models impose a genetic alteration during embryonic and postnatal periods that can drastically affect the phenotype of the animals. This is a weakness for studying heart failure, which mostly occurs in humans subsequent to a cardiac injury suffered during adulthood. Indeed, when wild-type G␣ q was overexpressed in myocytes of adult mice using an inducible Cre-loxbased system, the animals did not develop cardiac hypertrophy or other phenotypes seen in the conventional transgenic animals (9). This dramatic difference in phenotype raises doubt regarding the role of G␣ q in causing cardiac dysfunction in adult animals.
To better investigate the role of G␣ q activation in the development of heart failure in adult animals, we generated transgenic mice that express recombinant G␣ q Q209L proteins whose activity can be turned on by injection of the drug tamoxifen. This inducible model allows us to investigate the consequences of G␣ q activation in myocytes after completion of normal development and growth. In contrast to the standard transgenic models, we found that activation of G␣ q in 8-week-old animals causes a dilated cardiomyopathy and heart failure that does not appear to progress through a hypertrophic stage. Use of an additional transgenic mouse line that expresses an inducible G␣ q Q209L protein that cannot activate PLC␤ demonstrates that PLC␤ activation is required for the development of heart failure but not for all of the G␣ q -induced cardiac abnormalities. Another advantage this inducible model has over standard transgenic models is that it allows us to investigate early biochemical events initiated by G␣ q activation. We found that two aspects of Ca 2ϩ handling in myocytes are markedly altered 1 week after activation of the exogenous G␣ q protein. We speculate that these changes in Ca 2ϩ handling may be responsible for the contractility defect seen at this time.
DNA Constructs-The 4-HT-activateable G␣ q constructs were made using G␣ q Q209L and G␣ q Q209L-AA (10) and a mutant hormonebinding domain of the murine estrogen receptor (hbER) obtained from Dr. M. McMahon (University of California San Francisco). Using primers 5Ј-GGGGGCGGCCGCGCCACCATGACTCTGGAGTCCATC-ATGGCG (forward) and 5Ј-GCGCGGATCCGCGACCAGATTGTA-CTCCTTCAGG (reverse), we amplified G␣ q Q209L and G␣ q Q209-L-AA with the stop codons replaced with a BamHI site and subcloned them in front of hbER in pBluescript. The fidelity of the PCR fragments was confirmed by DNA sequencing. G␣ q -hbER constructs were then subcloned into pcDNA3.1 (Invitrogen) for expression in mammalian cells. Akt-HA was described earlier (2).
G␣ q Q209L-hbER Transgenic Mice-G␣ q Q209L-hbER and G␣ q Q209L-AA-hbER were subcloned into the ␣MHC promoter cassette (obtained from Dr. J. Robbins, University of Cinncinati) at the SalI site for construction of transgenic mice with cardiac-specific expression of the transgenes. After digestion with NotI to remove the bacterial elements, the constructs were injected into the pronucleus of single cell embryos from C57BL/6 mice in the Transgenic Mouse Facility at Stony Brook University. Transgenic animals were genotyped by PCR using primers that span the junction between G␣ q Q209L and hbER (5Ј-CCACTACAGGGATCATCGAATAC and 5Ј-GACGTAGCAAG-CAACATGTC). All animal-related experimental protocols were approved by the Institutional Animal Care and Use Committee.
Preparation of Tissue and Cell Lysates-Frozen mouse organs were homogenized with a PRO250 (Pro Scientific, Inc., Monroe, CT) in icecold lysis buffer (50 mM HEPES, 1% Triton X-100, 50 mM NaCl, 5 mM EDTA, 50 mM NaF, 10 mM sodium pyrophosphate, 1 mM sodium orthovandate, 0.5 mM phenylmethylsulfonyl fluoride, and 10 g/ml each of aprotinin and leupeptin, pH 7.5). Homogenates were centrifuged at 15,000 ϫ g for 30 min at 4°C. Cultured cells were rinsed twice with cold phosphate-buffered saline, scraped into lysis buffer, and centrifuged at 15,000 ϫ g for 15 min at 4°C. Protein concentrations of the supernatants were determined using the Bradford assay (Bio-Rad).
Quantitative PCR-Total RNA was extracted from hearts using Tri Reagent (Sigma) following the protocol provided by the manufacturer. cDNA was generated using the iScript cDNA Synthesis kit (Bio-Rad) following the protocol provided by the manufacturer. TaqMan Universal PCR Master Mix and TaqMan gene expression assays for natriuretic peptide precursor type B (BNP; Mm00435304_g1), ␤MHC (Mm00600555_m1), ␣MHC (Mm00440354_m1), and 18 S RNA (HS99999901_s1) were purchased from Applied Biosystems (Foster City, CA). Real-time PCR was performed on duplicate samples using a DNA Engine Opticon 2 System (MJ Research, Alameda, CA). The abundance of BNP, ␤MHC, and ␣MHC mRNA relative to 18 S RNA was determined using the equation r ϭ 2 (CT18 S RNA Ϫ CTgene of interest) , where C T is the number of cycles needed to achieve a preset threshold value of fluorescence. The results are expressed as fold change in r values as compared with wild-type (WT) samples.
Heart Membrane Preparation-Mouse ventricles were homogenized with a PRO250 in phosphate-buffered saline containing protease inhibitors (Sigma). The homogenate was centrifuged at 1500 ϫ g for 15 min at 4°C. The supernatant was centrifuged at 72,000 ϫ g for 15 min at 4°C, and the resulting pellet was suspended in 50 mM Tris-HCl, pH 7.4. Protein concentration was determined by a modified Lowry method (11).
Western Blotting-Membrane suspensions were mixed with a ureacontaining loading buffer (110 mM Tris-HCl, pH 6.8, 4% sodium dodecyl sulfate, 200 mM dithiothreitol, 0.2% bromphenol blue, 20% glycerol, and 8 M urea) without boiling, and proteins were separated by polyacrylamide gel electrophoresis. The loading buffer used for non-membrane Western samples contained no urea. Equal amounts of protein were loaded in each lane. After immunoblotting, signals were detected using either an enhanced chemiluminescence kit (PerkinElmer Life Sciences) or the Odyssey Infrared Imaging System (LI-COR Biosciences, Lincoln, NE).
SERCA Activity-Crude ventricular membranes were prepared by homogenizing the tissue as described above followed by centrifugation at 14,000 ϫ g for 20 min at 4°C. The supernatant was centrifuged at 72,000 ϫ g for 20 min at 4°C. The resulting pellet was suspended in 50 mM Tris-HCl, pH 7.4 plus 0.6 M KCl and kept on ice for 45 min. The suspension was centrifuged again at 72,000 ϫ g for 30 min at 4°C. The pellet was assayed for Ca 2ϩ -ATPase activity using a modification of an NADH-coupled assay (12). The standard assay medium contained 10 mM Tris, pH 7.5, 100 mM KCl, 5 mM MgCl 2 , 5 mM Na 2 ATP, 0.1 mM CaCl 2 , ϳ0.2 mM NADH, 1.5 mM trisodium phosphoenolpyruvate, 15 units/ml pyruvate kinase, and 36 units/ml lactate dehydrogenase. Additionally, the assay mix contained 10 mM NaN 3 and 0.1 mM ouabain to inhibit contaminating NADH oxidase and Na ϩ ,K ϩ -ATPase, respectively. Total activity (TA) was assayed by monitoring the rate of loss of A 340 after addition of the membrane preparation to a thermostatically controlled (37°C) cuvette in a Hewlett-Packard 8452A diode array spectrophotometer. Background oxidase activity (OA) was assayed in the absence of ATP. Total ATPase activity (TAA) was calculated as TA-OA. Ca 2ϩ -independent ATPase activity (CIA) (corrected for OA) was assayed in the presence of 10 mM EGTA instead of Ca 2ϩ . Ca 2ϩ -dependent ATPase activity was calculated as TAA-CIA and is expressed as nmol ATP hydrolyzed per mg of protein per min.
Hemodynamic Measurements of Cardiac Function-Measurements of heart rate, left ventricular end-diastolic and end-systolic pressure (LVEDP and LVESP, respectively), dP/dt, and ejection fraction were performed using the Millar ARIA TM pressure-volume conductance system (Houston, TX). A closed-chest approach similar to the one described by Lorenz and Robbins was used (13). After adequate anesthesia was attained with ketamine and fentanyl, the right carotid artery was isolated and cannulated with a 1.4 French Millar P-V catheter that was passed down the aorta into the left ventricle.
Ventricular Myocyte Isolation-Mice were euthanized by intraperitoneal injection of 100 mg/kg sodium pentobarbital. The heart was removed and rinsed in three changes of PS solution (112 mM NaCl, 5.4 mM KCl, 1.7 mM NaH 2 PO 4 , 1.63 mM MgCl 2 , 4.2 mM NaHCO 3 , 20 mM HEPES, 5.4 mM glucose, 4.1 mM L-glutamine, 10 mM taurine, minimal essential medium vitamins, and minimal essential medium amino acids, pH 7.4) containing 20 units/ml sodium heparin. The heart was cannulated through the aorta and perfused on a Langendorf apparatus with PS containing 1 mg/ml 2,3-butanedione monoxime for 10 min and with PS containing 12.6 g/ml Liberase Blendzyme 4 (Roche Applied Science) for 8 -12 min. After digestion, the heart was perfused to wash out collagenase with KB solution (74.56 mM KCl, 30 mM K 2 HPO 4 , 5 mM MgSO 4 , 5 mM pyruvic acid, 5 mM ␤-hydroxybutyric acid, 5 mM creatine, 20 mM taurine, 10 mM glucose, 0.5 mM EGTA, 5 mM HEPES, and 5 mM Na 2 ATP, pH 7.2). The heart was minced in KB solution and the cells gently shaken out. After removal of the supernatant, the cells were washed once with KB solution after settling.
Electrophysiology-Only clearly rod-shaped myocytes were studied. Whole-cell patch clamp recordings used 2-3 M⍀ borosilicate glass pipettes measured prior to sealing (Sutter Instrument), pCLAMP 8 software, the DigiData 1350 interface, and the Axopatch 1D amplifier (Axon Instruments). For the recording of L-type Ca 2ϩ current (I Ca,L ) (14), pipettes contained 111 mM CsCl, 20 mM tetraethylammonium chloride, 10 mM glucose, 14 mM EGTA, 10 mM HEPES, and 5 mM MgATP, pH 7.2 (adjusted by CsOH). I Ca,L was recorded in Na ϩ -free bath solution (137 mM tetraethylammonium chloride, 1 mM MgCl 2 , 2 mM CaCl 2 , 10 mM HEPES, and 10 mM glucose, pH 7.4, adjusted with tetraethylammonium hydroxide). The I Ca,L was recorded as a single voltage step every 4 s. In current/voltage (I-V) experiments, current stability was usually checked for 2 min before beginning the protocol. I-V curves were generated using 300-ms depolarizing voltage steps from Ϫ50 to ϩ50 mV in 10 mV increments (holding potential Ϫ50 mV). The membrane capacity was measured by applying a voltage step from a holding potential of Ϫ50 mV, and the current amplitudes are normalized to the cell capacitances. Perfusion solutions contained 1 M 4-HT throughout the experiment. All the experiments were performed at 22°C.
Calculation of Percent Extracellular Space-Non-overlapping images on hematoxylin and eosin (H&E)-stained sections were digitally captured to obtain RGB color images with 1280 ϫ 1024 resolution. The following algorithm was developed in ImageJ Version 1.34i (National Institutes of Health, Bethesda, MD) to quantifiy extracellular space. The green channel, having the greatest contrast between stained tissue and non-stained extracellular space, was isolated in each image. A high pass filter in the Fourier domain of the image was then used to smooth the shading caused by non-uniform illumination of the sample. The histogram of each image was then normalized such that the low intensity pixels were set to 0, and high intensity pixels were set to 255. Pixels having intensities above a standard threshold value in each image were counted as extracellular space. The ratio of extracellular space was then computed as EC ϭ p ec /p total , where p ec is the number of pixels considered extracellular space, and p total is the number of pixels in the image.
Statistical Analysis-Values represent means Ϯ S.E., and significance is defined as p Ͻ 0.05. Unless otherwise indicated, experimental groups were compared using Student's t test.

RESULTS
Conditionally Active G␣ q Q209L-hbER Constructs-The G␣ q proteins used in this study are G␣ q Q209L, which signals constitutively to all of its downstream effectors, and a G␣ q Q209L mutant in which Arg 256 and Thr 257 are changed to Ala. The latter protein (G␣ q Q209L-AA) cannot activate PLC␤ (15) but still inhibits PI3K signaling (10). To obtain inducible forms of G␣ q , we fused a mutant hbER to the carboxyl terminus of G␣ q Q209L and G␣ q Q209L-AA (Fig. 1A). This mutant hbER does not bind 17␤-estradiol but is responsive to 4-HT (16).
The function of G␣ q Q209L-hbER was checked by measuring its ability to activate PLC␤ and increase the level of intracellular inositol phosphates in HEK 293 cells. The basal level of inositol phosphates in cells transfected with G␣ q Q209L-hbER was similar to that in the control cells, but treatment with 4-HT caused inositol phosphates to increase more than 6-fold at 4 h (Fig. 1B). In contrast, the level of inositol phosphates in cells transfected with either vector or G␣ q Q209L-AA-hbER did not change in response to 4-HT treatment (Fig. 1B). To confirm that the inducible form of G␣ q Q209L-AA can still inhibit PI3K signaling, we FIGURE 1. Conditionally active G␣ q proteins. A, G␣ q Q209L and G␣ q Q209L-AA were fused to a mutant hbER. To generate transgenic mice with cardiac myocyte-specific expression of these fusion proteins, the constructs were placed under the control of an ␣MHC promoter with a bovine growth hormone polyadenylation signal. B, cells were transfected with G␣ q Q209L-hbER (QL), G␣ q Q209L-AA-hbER (QL-AA), or empty vector, labeled with myo-[ 3 H]inositol, and then treated with 1 M 4-HT for the indicated times. Accumulation of inositol phosphates was measured as described previously (2). Values shown are from three experiments performed in duplicate. C, cells were cotransfected with Akt-HA and empty vector, G␣ q Q209L-AA-hbER (QL-AA), or G␣ q Q209L-hbER (QL) and then treated with or without 1 M 4-HT for 4 h. Akt activity was measured in HA immunoprecipitates (2). Values are from three or four experiments and show percent Akt activity as compared with the control (empty vector not treated with 4-HT). measured the activity of Akt, which acts downstream of PI3K. HEK 293 cells were cotransfected with epitope-tagged Akt-HA and either vector or G␣ q Q209L-AA-hbER. Akt activity was inhibited 40% when the fusion protein was activated by addition of 4-HT to the cells (Fig. 1C). A similar level of inhibition was obtained using G␣ q Q209L-hbER (Fig. 1C).
G␣ q Q209L-hbER and G␣ q Q209L-AA-hbER Transgenic Mice-G␣ q transgenic mice were generated using an ␣MHC promoter construct to selectively express these proteins in cardiac myocytes (Fig. 1A). Three G␣ q Q209L-hbER founders (lines 5, 7, and 11) were identified by PCR analysis of tail DNA ( Fig. 2A). Mice from line 7 express the G␣ q Q209L-hbER protein at a higher level than mice from lines 5 or 11, as demonstrated by Western blot analysis of heart lysates (Fig. 2B). Although all three lines of transgenic mice develop heart failure when injected with tamoxifen (see below), transgenic mice from line 7 developed this phenotype most rapidly. Therefore, this line was selected for further study. Cardiac-specific expression of the recombinant protein was confirmed in line 7 animals (Fig. 2C). The transgenic mice are born at the expected Mendelian ratio and breed normally.
Five G␣ q Q209L-AA-hbER founders (lines 27, 41, 56, 59, and 66) were identified by PCR analysis of tail DNA (data not shown). Expression of the recombinant protein in four of these lines was very low (Fig. 2D). However, line 66 G␣ q Q209L-AA-hbER mice express the protein at a high level comparable with that seen in line 7 G␣ q Q209L-hbER mice (Fig. 2D), so they were selected for all subsequent experiments. This allows us to directly compare the phenotypes of mice expressing either G␣ q protein. It is important to note that expression of endogenous G␣ q is not affected in any of the transgenic mouse lines (Fig. 2D).
To confirm that tamoxifen activates G␣ q /PLC␤ signaling in the G␣ q Q209L-hbER mice but not in WT or G␣ q Q209L-AA-hbER animals, we examined the membrane localization of PKC␣. Phosphorylated PKC␣ translocates from the cytosol to membranes and becomes active in response to the PLC␤-catalyzed production of diacylglycerol and elevated intracellular Ca 2ϩ (17). Western blotting showed that phospho-PKC␣ was undetectable in heart membranes isolated from transgenic and WT mice before tamoxifen injection, but after 7 days of drug treatment the protein was enriched in the membrane fraction from G␣ q Q209L-hbER animals only (Fig. 3A, top panel). Similar results were obtained using an antibody that recognizes total PKC␣ (Fig. 3A, middle  panel). The total amount of PKC␣ in unfractionated lysates did not change in any of the three groups after drug treatment (Fig. 3A, bottom  panel). Thus, G␣ q /PLC␤ signaling is elevated as early as 1 week after tamoxifen injection in the G␣ q Q209L-hbER hearts. This signaling pathway remains active for up to 28 days of drug treatment, as PKC␣ was still enriched in membranes from G␣ q Q209L-hbER hearts at this time point (Fig. 3B). The amount of PKC␦, but not PKC⑀, was also increased in the membrane fraction prepared from G␣ q Q209L-hbER as compared with WT hearts (data not shown).
Peripheral Edema, Heart Enlargement, and Increased Extracellular Space in G␣ q Q209L-hbER Mice-Starting at 8 weeks of age, transgenic and WT mice were injected intraperitoneally daily with 1 mg of tamoxifen, which is converted to 4-HT in animals. Approximately two-thirds of the male G␣ q Q209L-hbER mice developed peripheral edema between 21 and 28 days of tamoxifen injection, suggestive of heart failure (Fig. 4A). The average body weight of these edematous mice increased 33% over preinjection levels (Fig. 4B). Mice that did not develop severe peripheral edema exhibited a distended abdomen, suggesting the presence of ascites. These mice then lost weight and developed a disheveled appearance and decreased mobility. This phenotype was more common for the female transgenic mice. Approximately onethird of all tamoxifen-injected G␣ q Q209L-hbER mice died within 28 days, although not all of these animals had an edematous or ill appearance prior to death. Without tamoxifen treatment, they displayed no FIGURE 3. Tamoxifen activates PLC␤ signaling in G␣ q Q209L-hbER transgenic mice. Starting at 8 weeks of age, WT and G␣ q transgenic mice were either sacrificed (t ϭ 0) or injected intraperitoneally daily with 1 mg of tamoxifen. Western blots of heart membranes prepared at the indicated times were probed with antibodies to detect phospho-Ser 657 PKC␣ or total PKC␣. The bottom panel in A shows the total amount of PKC␣ in unfractionated heart lysates. The experiments were repeated on another set of animals with similar results. QL, G␣ q Q209L-hbER; QL-AA, G␣ q Q209L-AA-hbER. FIGURE 2. Transgenic mice with cardiac-specific expression of G␣ q fusion proteins. A, three G␣ q Q209L-hbER founders (11, 5, and 7) were identified by PCR analysis of tail DNA. The positive control (ϩ) uses G␣ q Q209L-hbER in pcDNA3.1 as template and the negative control (Ϫ) does not contain template DNA. B, cardiac expression of G␣ q Q209L-hbER in lines 11, 5, and 7 was confirmed by Western blotting using an anti-estrogen receptor antibody that detected a 75-kDa band that corresponds to the fusion protein. C, heartspecific expression of G␣ q Q209L-hbER protein in an F1 offspring of founder 7 was demonstrated by Western blotting using the anti-estrogen receptor antibody. D, cardiac expression of G␣ q Q209L-AA-hbER (QL-AA) in lines 27, 41, 56, 59, and 66 is compared with expression of G␣ q Q209L-hbER (QL) in line 7 on a Western blot probed with antibody to G␣ q .
phenotypic abnormalities up to 2 years of age, indicating that the G␣ q Q209L-hbER protein is essentially inactive in the absence of drug. This result is consistent with a recent report by Syed et al. (9) showing that overexpression of wild-type G␣ q in adult mice does not have a pathological consequence. One would expect that wild-type G␣ q in the absence of receptor activation is in the inactive state.
All G␣ q Q209L-AA-hbER and WT mice were alive after 28 days of tamoxifen injection and appeared healthy and free of edema (Fig. 4A). These animals exhibited body weight increases of 9% (WT) and 12% (G␣ q Q209L-AA-hbER) over the course of the experiment (Fig. 4B). Transgenic or WT mice injected with vehicle suffered no ill effects.
When sacrificed after 28 days of drug treatment, the G␣ q Q209L-hbER hearts were found to be enlarged, with ventricular dilatation (Fig.  4C). The heart weight/tibial length ratio was significantly increased (1.65-fold) in G␣ q Q209L-hbER mice with edema as compared with WT mice injected with tamoxifen for 28 days (Fig. 4D). G␣ q Q209L-hbER mice as a group, including non-edematous animals that were injected for 28 days, exhibited a 1.89-fold increase in heart weight/tibial length ratio as compared with WT mice (12.97 Ϯ 0.74 mg/mm, n ϭ 10). Although most G␣ q Q209L-AA-hbER hearts did not exhibit architectural changes after 28 days of tamoxifen injection (Fig. 4C), modest ventricular dilatation was seen in ϳ10% of these animals (data not shown). The heart weight/tibial length ratio was 1.3-fold higher in G␣ q Q209L-AA-hbER mice than in WT animals (Fig. 4D), an increase that is not statistically significant.
Transgenic mice expressing exogenous proteins in cardiac myocytes often exhibit increased expression of fetal genes such as ␤MHC in the heart, with no increase in the expression level of ␣MHC. Up-regulation of fetal genes is associated with cardiac hypertrophy and/or heart failure in many, but not all, of these animal models. In humans, increased circulating BNP levels are indicative of heart failure. Quantitative PCR analysis of RNA extracted from hearts after tamoxifen injection showed a large increase in expression of BNP and ␤MHC mRNA in both G␣ q Q209L-hbER and G␣ q Q209L-AA-hbER hearts as compared with WT (Fig. 4E). This result was surprising, since G␣ q Q209L-AA-hbER mice do not develop enlarged hearts or heart failure after tamoxifen injection. In addition, ␣MHC levels decreased in both transgenic animals (Fig. 4E). Thus, these genes do not represent useful markers for distinguishing between these phenotypes in the two transgenic G␣ q lines studied here.
The myocytes in heart sections from G␣ q Q209L-hbER mice injected for 28 days appeared normal, with no obvious evidence of hypertrophy, apoptosis, or necrosis (Fig. 5A, upper panel). Cleaved forms of caspase 3 or poly(ADP-ribose) polymerase were not detected on Western blots of heart extracts prepared after 7 days of tamoxifen treatment, indicating that apoptosis was also not occurring at this early time point (data not shown). Trichrome staining of heart sections showed that the interstitial space between myocytes was increased in G␣ q Q209L-hbER hearts as compared with G␣ q Q209L-AA-hbER and WT hearts, although only a minimal increase in interstitial fibrosis was evident (Fig. 5A, lower  panel). The percent extracellular space was quantified using a computer algorithm that analyzes H&E-stained heart sections (see "Experimental Procedures"). The extracellular space in G␣ q Q209L-hbER mice exhibiting edema was significantly increased (2.45-fold) as compared with FIGURE 4. G␣ q Q209L-hbER transgenic mice develop peripheral edema and enlarged hearts. Except where noted in B and D, all mice were injected intraperitoneally with 1 mg of tamoxifen daily for 28 days. QL, G␣ q Q209L-hbER; QL-AA, G␣ q Q209L-AA-hbER. A, some of the QL mice developed decompensated heart failure leading to severe whole body edema. B, increase in body weight was determined at the end of the drug treatment (n ϭ 8 for all groups). QL mice were injected until they developed edema. They were weighed and sacrificed after 21-28 days of tamoxifen injection. C, hearts from WT littermates and QL and QL-AA mice were fixed overnight in 4% paraformaldehyde buffered with phosphate-buffered saline (top). Hearts were paraffin-embedded, sectioned, and then stained with H&E (bottom). D, the heart weight/tibial length ratio was determined after completing the drug treatment (n ϭ 5 for all groups). QL animals were treated as described for B. E, quantitative real-time PCR using total RNA extracted from hearts was performed to determine relative changes in mRNA expression as compared with WT values (n ϭ 3 for each group). In B and D, * denotes a statistically significant difference between WT and QL. Data were analyzed by one-way ANOVA and pairwise comparisons were obtained using post-hoc Tukey tests.
WT mice injected for 28 days (Fig. 5B). G␣ q Q209L-hbER mice as a group, including non-edematous animals that were injected for 28 days, exhibited a 3.13-fold increase in extracelluar space as compared with WT mice (25.7 Ϯ 0.74%, n ϭ 22). The difference in calculated extracellular space between WT and G␣ q Q209L-AA-hbER heart sections (1.61fold) was not statistically significant.
The enlarged hearts seen in G␣ q Q209L-hbER animals could be due to cellular hypertrophy. It was surprising that cell hypertrophy was not apparent upon histological examination of heart sections from these mice, as this phenotype was reported in other G␣ q mouse models (5,6). Cell capacitance as measured by whole-cell patch clamping is proportional to surface area, and thus can be used as a measure of cell size. We measured capacitance in WT and G␣ q Q209L-hbER myocytes isolated from animals after different times of tamoxifen treatment. Although some G␣ q Q209L-hbER myocytes exhibited relatively high capacitance, especially after 14 days of drug treatment (Fig. 6), there was no statistically significant difference between the mean values for WT versus G␣ q Q209L-hbER cells at any time point (TABLE ONE). Thus, it does not appear that cell hypertrophy is responsible for the cardiac enlargement seen in G␣ q Q209L-hbER animals.
Inducible Heart Failure in G␣ q Q209L-hbER Mice-Hemodynamic measurements performed using a pressure-volume conductance system confirmed our clinical observation of heart failure in G␣ q Q209L-hbER mice. Maximum and minimum dP/dt values were already significantly altered after only 7 days of tamoxifen treatment, indicating the presence of a contractile defect that affects both the contraction and relaxation phases of the heart beat (TABLE TWO). The mice progressed into overt heart failure with continued drug treatment, so that all of the hemodynamic measurements were significantly altered in G␣ q Q209L-hbER mice as compared with WT after 28 days (TABLE   TWO). LVEDP was elevated 21-fold, and LVESP and ejection fractions were depressed. The heart rate was also paradoxically depressed, as observed in other mouse models of heart failure. DP/dt measurements were also significantly altered in the G␣ q Q209L-hbER mice as compared with the WT mice. By contrast, no hemodynamic measurements in G␣ q Q209L-AA-hbER mice were significantly different from WT after 28 days of tamoxifen treatment (TABLE TWO).
Changes in PLB and SERCA-2 in G␣ q Q209L-hbER Mice-Heart failure is often associated with decreased myocyte contractility. Since myocyte contractility is regulated by changes in intracellular Ca 2ϩ concentration generated with each heart beat, we examined two major pathways for Ca 2ϩ handling in myocytes. One major regulator of Ca 2ϩ concentration in cardiac myocytes is SERCA-2. Entry of extracellular Ca 2ϩ through the L-type Ca 2ϩ channel induces a large release of Ca 2ϩ from the sarcoplasmic reticulum (SR) that is sequestered back into the SR by the SERCA-2 pump. Over time, inhibition of SERCA-2 activity decreases the amount of Ca 2ϩ stored in the SR and leads to decreased contractile force. SERCA-2 is regulated by PLB, which occurs as monomeric and pentameric forms. Monomeric PLB binds to SERCA-2 to inhibit its activity. The pentameric form of PLB is phosphorylated on Ser 16 , which blocks PLB from binding to SERCA-2. Using a phosphospecific antibody, we found that pentameric PLB in membranes prepared from hearts of G␣ q Q209L-hbER mice becomes dephosphorylated in response to tamoxifen treatment (Fig. 7B, upper panel). The decrease in Ser 16 phosphorylation occurred gradually over 7 days and remained low for the duration of the experiment. In contrast, tamoxifen injection did not affect PLB phosphorylation in either WT or G␣ q Q209L-AA-hbER hearts (Fig. 7, A and C, upper panels).
Using a general PLB antibody, we detected an increase in the monomeric form of PLB in G␣ q Q209L-hbER hearts after tamoxifen injection FIGURE 5. Increased extracellular space in G␣ q Q209L-hbER hearts. A, WT and transgenic mice were injected intraperitoneally with 1 mg of tamoxifen daily for 28 days. Hearts were paraffinembedded, sectioned, and then stained with H&E (top) or Masson's trichrome stain (bottom). Views are 400ϫ magnification. B, H&E-stained sections of paraffin-embedded hearts were prepared from WT and G␣ q Q209L-AA-hbER mice injected with tamoxifen for 28 days and from male G␣ q Q209L-hbER mice that developed peripheral edema after tamoxifen injection (see legend to Fig. 4B). Percent extracellular space was calculated as described under "Experimental Procedures." n ϭ 11 for WT and G␣ q Q209L-hbER; n ϭ 9 for G␣ q Q209L-AA-hbER. * denotes a statistically significant difference between WT and G␣ q Q209L-hbER. Data were analyzed by one-way ANOVA, and pairwise comparisons were obtained using Fisher's post-hoc tests. QL, G␣ q Q209L-hbER; QL-AA, G␣ q Q209L-AA-hbER.
for 7-28 days (Fig. 7B, lower panel). Furthermore, after 7-28 days, the PLB pentamers from G␣ q Q209L-hbER mice migrated faster during polyacrylamide gel electrophoresis, indicating that the protein contains less phosphate than PLB from earlier time points (Fig. 7B, lower panel). Tamoxifen injection did not affect the PLB migration pattern or increase the amount of monomeric PLB in either WT or G␣ q Q209L-AA-hbER mice (Fig. 7, A and C, lower panels).
Finally, SERCA activity was measured by assaying Ca 2ϩ -dependent ATPase activity in heart membranes prepared from mice that were injected with tamoxifen for 14 days. SERCA specific activity was significantly decreased by 65% in G␣ q Q209L-hbER hearts as compared with WT hearts (Fig. 8A). In contrast, SERCA activity in the G␣ q Q209L-AA-hbER hearts was not reduced as compared with WT (Fig. 8A). The drop in activity in G␣ q Q209L-hbER hearts was not due to a lower level of SERCA-2 protein, as tamoxifen injection for up to 28 days did not affect the amount of SERCA-2 protein in any of the three groups of animals (Fig. 8B). These results suggest that activation of G␣ q Q209L-hbER by tamoxifen induces dephosphorylation of PLB and the release of PLB monomers that bind tightly to SERCA-2 to inhibit its activity. This response is probably PLC␤-dependent, as it is not seen in the G␣ q Q209L-AA-hbER hearts.
Decreased I Ca,L in Transgenic Myocytes-Next we examined the major pathway for Ca 2ϩ entry into myocytes, I Ca,L , using the whole-cell patch clamp technique. Isolated myocytes from transgenic animals were always paired with cells from WT littermates. In the first experiment, myocytes were isolated from animals at different times after tamoxifen injection. The membrane was held at Ϫ50 mV and depolarized for 300 ms to ϩ10 mV to measure I Ca,L . I Ca,L density in G␣ q Q209L-hbER myocytes was significantly reduced by 32% as compared with WT after only 7 days of drug treatment (TABLE ONE). The decrease was even larger at 14 days (72%) and 21 days (61%). After 28 days of tamoxifen treatment, I Ca,L density in G␣ q Q209L-hbER myocytes was still 58% Cell capacitance and I Ca,L density in WT and G␣ q Q209L-hbER myocytes Cardiac myocytes were isolated from G␣ q Q209L-hbER (QL) mice and WT littermates after injection with tamoxifen for the indicated number of days. Membrane capacitance and I Ca,L density at ϩ10 mV were determined by whole-cell patch clamping as described under "Experimental Procedures" and "Results."

Hemodynamic measurements in WT and G␣ q transgenic animals
Starting at 8 weeks of age, WT, G␣ q Q209L-hbER (QL), and G␣ q Q209L-AA-hbER(QL-AA) mice were injected intraperitoneally daily with 1 mg of tamoxifen.
After the indicated number of days, hemodynamic parameters were measured using a pressure-volume conductance system (see "Experimental Procedures"). lower than in WT cells (TABLE ONE). To our surprise, I Ca,L density was also significantly reduced by 51% in myocytes from G␣ q Q209L-AA-hbER animals injected with tamoxifen for 28 days (5.8 Ϯ 0.8 pA/pF (n ϭ 14) as compared with 11.8 Ϯ 1.0 pA/pF for WT (n ϭ 15)). We also constructed the peak inward I-V relations for myocytes from WT and transgenic animals injected with tamoxifen for 28 days. Fig. 9, A and B, illustrate the patch clamp records, while Fig. 9, C and D, plot the peak inward current density for each data set at each potential. The I Ca,L density is decreased throughout the entire voltage range in myocytes from both transgenic mouse lines. This reduction in I Ca,L density was not due to a negative shift in the voltage dependence of inactivation nor to a difference in autonomic regulation (data not shown). These results demonstrate that G␣ q affects the L-type Ca 2ϩ current at least in part through a signaling pathway that is independent of PLC␤ activation. We did not detect consistent differences in protein or mRNA levels of Ca v 1.2 Ca 2ϩ channels in WT, G␣ q Q209L-hbER, or G␣ q Q209L-AA-hbER hearts (data not shown). Thus, activation of G␣ q leads to severe defects in two major Ca 2ϩ -handling proteins that could explain the contractile defect in G␣ q Q209L-hbER mice.

DISCUSSION
Previous studies suggested that activation of G␣ q induces cardiac hypertrophy and a contractile defect that is a consequence of this hypertrophic response. However, G␣ q signaling in the mice used in earlier studies was altered prior to adulthood, and it is becoming clear that the timing of genetic manipulations can have a profound effect on the phenotypic outcome (9). Here we used G␣ q Q209L-hbER fusion proteins to activate G␣ q signaling selectively in cardiac myocytes of adult mice. hbER has been fused to a wide variety of partners to make tamoxifencontrolled fusion proteins. Inducible forms of the transcription factor myc (18) and the recombinase Cre (19) have been successfully used in transgenic mice to study heart phenotypes. Our transgenic mouse model is the first application of this system to regulate the activity of G␣ proteins. We have found that this approach works for a variety of G␣ proteins and small GTPases (data not shown). Our results indicate that it might be possible to utilize hbER fusion proteins to investigate the function of other signaling molecules in a temporally controllable manner in a variety of tissues.
Using the inducible G␣ q Q209L-hbER model, we found that activation of G␣ q in adult mice causes a dilated cardiomyopathy that rapidly results in heart failure, apparently without progressing through a hypertrophic stage. It is possible that the rapid onset of cardiac failure precluded the animals from developing compensatory hypertrophy. This study also demonstrates that activation of G␣ q in adult mouse cardiac myocytes has early effects on Ca 2ϩ homeostasis and cardiac contractility. At least two aspects of Ca 2ϩ handling are markedly affected: the entry of Ca 2ϩ into myocytes through the L-type Ca 2ϩ channel is attenuated, and SERCA-2 activity is inhibited as a consequence of PLB dephosphorylation. Use of the G␣ q Q209L-AA-hbER construct shows that G␣ q regulates the two Ca 2ϩ -handling components via separate effector pathways: G␣ q -induced PLB dephosphorylation is mediated by the canonical PLC␤ pathway, whereas the effect on I Ca,L does not require PLC␤ activation. Cumulatively, these two effects on Ca 2ϩ homeostasis would be expected to severely depress myocyte contractility and in combination are probably responsible for G␣ q -induced heart failure.
Similar to our observations in G␣ q Q209L-hbER mice, SERCA activity FIGURE 7. PLB dephosphorylation in G␣ q Q209L-hbER hearts. Membranes were prepared from hearts of WT (A), G␣ q Q209L-hbER (B), and G␣ q Q209L-AA-hbER (C) mice after injection with 1 mg of tamoxifen daily for the indicated number of days. Immunoblotting was performed with antibodies that recognize phospho-Ser 16 PLB or the pentameric and monomeric forms of PLB (PLB(5) and PLB (1), respectively). The experiment was repeated twice with similar results. FIGURE 8. Decreased SERCA activity in G␣ q Q209L-hbER hearts. A, SERCA activity was measured in heart membranes prepared from mice that were injected for 14 days with tamoxifen. n ϭ 5 for WT; n ϭ 8 for G␣ q Q209L-hbER; and n ϭ 2 for G␣ q Q209L-AA-hbER. * denotes a statistically significant difference between WT and G␣ q Q209L-hbER (Student's t test). B, membranes were prepared from hearts of WT and transgenic mice after injection with 1 mg of tamoxifen daily for the indicated number of days. Immunoblotting was performed with antibody to SERCA-2. The experiment was repeated twice with similar results. QL, G␣ q Q209L-hbER; QL-AA, G␣ q Q209L-AA-hbER.
was found to be decreased in human myocardium obtained from patients with end-stage heart failure (20). Schwinger et al. (21) also found that PLB Ser 16 phosphorylation was reduced in failing as compared with control hearts. These investigators did not find significant changes in the amount of SERCA-2 or PLB protein (20), but controversy still exists as to whether levels of these proteins are altered in human heart failure. In transgenic mice overexpressing wild-type G␣ q , SERCA-2 protein levels were decreased, and PLB protein was markedly increased and dephosphorylated (22,23). In our heart failure model, we did not detect changes in the expression of SERCA-2 or PLB. Instead, PLB became dephosphorylated, monomeric PLB appeared, and SERCA activity was reduced. These G␣ q effects are probably mediated by the PLC␤ pathway that leads to the activation of PKC isozymes including PKC␣. Interestingly, mice overexpressing PKC␣ in cardiac myocytes also show decreased PLB phosphorylation on Ser 16 (24). Although our results suggest that inhibition of SERCA-2 plays an important role in the development of G␣ q -induced cardiomyopathy, this effect alone is probably not sufficient to induce heart failure. The PKC␣ transgenic mice mentioned above develop contractile defects and cardiac hypertrophy but not heart failure, even though dephosphorylation of PLB should lead to inhibition of SERCA-2 (24). SERCA-2 Ϫ/ϩ heterozygous mice have a 35% reduction in SERCA activity but do not develop heart failure or cardiac hypertrophy (25). SERCA-2 haploinsufficiency does, however, cause a decrease in cardiac contractility (25) and myocyte contractility (26). Transgenic mice that overexpress PLB 2-fold in cardiac myocytes do not develop heart failure despite a decrease in left ventricular fractional shortening (27). Similarly, expression of PLB mutants L37A and I40A that persist in the inhibitory monomeric form reduces dP/dt and fractional shortening but does not cause heart failure (28).
The lack of a severe cardiac phenotype in the animals with decreased SERCA-2 activity discussed above as compared with our G␣ q Q209L-hbER mice may be due to the fact that I Ca,L is not concurrently inhibited. In fact, overexpression of a superinhibitory PLB V49G mutant in transgenic mice caused a significant compensatory increase in the I Ca,L density in myocytes (29). G␣ q may have additional effects on other components of the Ca 2ϩ -cycling machinery such as the ryanodine receptor and the Na ϩ /Ca 2ϩ exchanger that contribute to heart failure. However, we did not detect changes in expression of the exchanger or the ryanodine receptor or its phosphorylation in G␣ q Q209L-hbER hearts (data not shown). It is interesting to note that PLB ablation in mice overexpressing wild-type G␣ q did not improve global cardiac function, suggesting that G␣ q -induced heart failure might involve more than just SERCA-2 inhibition (22).
Since Ca 2ϩ influx through the L-type Ca 2ϩ channel is essential for triggering SR Ca 2ϩ release, a reduction in I Ca,L density would be expected to lead to depressed cardiac contractility. However, the contribution of altered L-type Ca 2ϩ channel expression or function to the development of heart failure is controversial. Measurements in myocytes isolated from patients with end-stage heart failure have generally not detected a reduction in basal FIGURE 9. I Ca,L activation in transgenic and WT myocytes. Representative traces of I Ca,L activation in myocytes from WT and G␣ q Q209L-hbER (QL) mice (A) and WT and G␣ q Q209L-AA-hbER (QL-AA) mice (B). The currents were elicited by voltage steps from Ϫ50 mV to ϩ50 mV (300-ms duration) in 10-mV increments from a holding potential of Ϫ50 mV. The peak I-V relationships of I Ca,L for myocytes from G␣ q Q209L-hbER mice (C; E) and matched WT littermates ((C; •) (n ϭ 17 for both) and G␣ q Q209L-AA-hbER (D; E) and matched WT (D; •) mice (n ϭ 14 for both). I Ca,L density (30 -32). However, a study by Ouadid et al. (33) did report a reduction in I Ca,L density in atrial and ventricular myocytes from patients undergoing cardiac transplantation. Most studies in animal models of heart failure have shown either unchanged or reduced L-type current density (34). Both the G␣ q Q209L-hbER and G␣ q Q209L-AA-hbER mice studied here exhibit a significant reduction in I Ca,L density after tamoxifen injection, but only the G␣ q Q209L-hbER animals developed a contractile defect and heart failure. While these results suggest that inhibition of I Ca,L alone is not sufficient to induce heart failure, they do not rule out the possibility that a reduction in I Ca,L density contributes to G␣ q -induced cardiomyopathy.
The suppression of I Ca,L density in G␣ q Q209L-AA-hbER mice indicates that G␣ q inhibits the L-type Ca 2ϩ channel independently of PLC␤. Studies are ongoing in our laboratory to identify the signaling pathway that mediates this effect. We recently demonstrated that G␣ q inhibits the PI3K p110␣ isoform (2) and binds directly to the enzyme in vitro. 4 G␣ q inhibition of PI3K may play a significant role in causing the inhibition of I Ca,L . Indeed, we have found that injection of the second messenger phosphatidylinositol (3,4,5)-trisphosphate into G␣ q Q209L-AA-hbER myocytes can reverse the inhibition of I Ca,L (37). This hypothesis can be further tested in vivo by producing transgenic mice that express a G␣ q mutant that cannot inhibit PI3K but that retains the ability to activate PLC␤.
Studies in end-stage heart failure patients treated with left ventricular assist devices have shown that mechanical unloading improves myocyte function and reverses structural remodeling of the failing heart (35,36). This therapy also normalizes Ca 2ϩ transients and increases SERCAmediated Ca 2ϩ uptake. These findings support the concept that some aspects of myocyte dysfunction in the failing heart are reversible. Preliminary studies using the G␣ q Q209L-hbER mice support this view. We have found that terminating the tamoxifen injections in edematous mice allows the gradual reversal of the overt signs of heart failure, heart enlargement, ventricular dilation, and interstitial space expansion. 5 Additional studies are needed to determine whether the biochemical and hemodynamic alterations in these animals are also reversible. Future studies that utilize these transgenic mice to examine genetic changes associated with the onset and reversal of heart failure may yield fruitful information.
Results from our study suggest that activation of G␣ q -coupled receptors by elevated levels of neuroendocrine hormones found in heart failure will lead to dampening of the Ca 2ϩ transient and depression of cardiac contractility. Development of pharmacological interventions that block G␣ q or PLC␤ signaling in myocytes may prove useful as a strategy to improve contractile function in heart failure.