G-Protein Signaling Participates in the Development of Diabetic Cardiomyopathy
- Ian S. Harris1,
- Ilya Treskov1,
- Michael W. Rowley1,
- Scott Heximer2,
- Kevin Kaltenbronn2,
- Brian N. Finck1,
- Richard W. Gross13,
- Daniel P. Kelly1,
- Kendall J. Blumer2 and
- Anthony J. Muslin12
- 1Center for Cardiovascular Research, Department of Internal Medicine, Washington University School of Medicine, St. Louis, Missouri
- 2Department of Cell Biology and Physiology, Washington University School of Medicine, St. Louis, Missouri
- 3Division of Bio-Organic Chemistry, Washington University School of Medicine, St. Louis, Missouri
- Address correspondence and reprint requests to Dr. Anthony J. Muslin, Box 8086, 660 South Euclid Ave., St. Louis, MO 63110. E-mail:
Diabetic patients develop a cardiomyopathy that consists of ventricular hypertrophy and diastolic dysfunction. Although the pathogenesis of this condition is poorly understood, previous studies implicated abnormal G-protein activation. In this work, mice with cardiac overexpression of the transcription factor peroxisome proliferator–activated receptor-α (PPAR-α) were examined as a model of diabetic cardiomyopathy. PPAR-α transgenic mice develop spontaneous cardiac hypertrophy, contractile dysfunction, and “fetal” gene induction. We examined the role of abnormal G-protein activation in the pathogenesis of cardiac dysfunction by crossing PPAR-α mice with transgenic mice with cardiac-specific overexpression of regulator of G-protein signaling subtype 4 (RGS4), a GTPase activating protein for Gq and Gi. Generation of compound transgenic mice demonstrated that cardiac RGS4 overexpression ameliorated the cardiomyopathic phenotype that occurred as a result of PPAR-α overexpression without affecting the metabolic abnormalities seen in these hearts. Next, transgenic mice with increased or decreased cardiac Gq signaling were made diabetic by injection with streptozotocin (STZ). RGS4 transgenic mice were resistant to STZ-induced cardiac fetal gene induction. Transgenic mice with cardiac-specific expression of mutant Gαq, Gαq-G188S, that is resistant to RGS protein action were sensitized to the development of STZ-induced cardiac fetal gene induction and bradycardia. These results establish that Gq-mediated signaling plays a critical role in the pathogenesis of diabetic cardiomyopathy.
- ANF, atrial natriuretic factor
- LV, left ventricular
- mCPT-1, mitochondrial carnitine palmitoyltransferase 1
- MHC, myosin heavy chain
- PPAR, peroxisome proliferator–activated receptor
- RGS, regulator of G-protein signaling
- SERCA2a, sarcoplasmic and endoplasmic reticulum calcium exchanger 2a
- STZ, streptozotocin
Both insulin-dependent (type 1) and non–insulin-dependent (type 2) forms of diabetes in humans are accompanied by a greatly increased risk of cardiovascular death (1–3). Both types of diabetes are associated with cardiomyopathy that share many similar characteristics, including ventricular hypertrophy, decreased ventricular diastolic relaxation, and a reduced peak filling rate (4–7). Diabetic cardiomyopathy is distinct from ischemic cardiomyopathy because it is present in diabetic patients and animal models of diabetes in the absence of coronary artery disease.
Recently, rodent models of diabetes were used to uncover the pathogenesis of diabetic cardiomyopathy. One animal model of diabetic cardiomyopathy was developed that consists of cardiac-specific overexpression of the transcription factor myosin heavy chain (MHC)–peroxisome proliferator–activated receptor-α (PPAR-α) (8). PPAR-α is a ligand-activated transcription factor that regulates genes involved in cardiac fatty acid uptake and oxidation. PPAR-α is activated in the diabetic heart (8). MHC-PPAR transgenic mice develop a metabolic phenotype that is similar to the diabetic state in the heart but not in other tissues, with increased lipid uptake and oxidation and reduced glucose uptake and oxidation (8). Furthermore, MHC-PPAR mice develop a cardiomyopathy with ventricular hypertrophy, activation of gene markers of pathological hypertrophic growth, and transgene expression–dependent alteration in systolic ventricular dysfunction. It is unclear whether the cardiomyopathy seen in MHC-PPAR mice develops as a direct consequence of altered myocardial metabolism or rather because of abnormal intracellular signal transduction, which may occur secondarily to these metabolic changes.
A second rodent model of type 1 diabetes is based on the systemic administration of streptozotocin (STZ), a compound that contains a glucose molecule with a nitrosourea side chain that has cytotoxic action (9). The glucose portion of STZ directs the agent to pancreatic β-cells, which are then destroyed by the highly reactive nitrosourea side chain. STZ-induced diabetic rats develop a cardiomyopathy that is characterized by decreased left ventricular (LV) contractility, diminished ventricular compliance with markedly abnormal diastolic function, and decreased inotropic and chronotropic responses to certain ligands (9,10). The abnormalities in diastolic function in diabetic rats are manifested by prolonged isovolumic relaxation time, increased atrial contribution to diastolic filling, and elevated in vivo LV end-diastolic pressure (10). STZ-induced diabetic rats do not develop atherosclerosis or hypertension, so the cardiomyopathy is presumably caused by a direct effect of diabetes on the cardiac myocyte. The cardiomyopathy in STZ-induced diabetic rats may be caused, in part, by abnormalities in G-protein–mediated signal transduction. In particular, β-adrenergic receptor density is decreased in STZ-induced diabetic rat hearts, and there is reduced inotropic and chronotropic responsiveness to β-adrenergic agonists (11,12). In addition, Gαq protein content and protein kinase Cε activity is increased in STZ-induced diabetic rat cardiac tissue (13–16). Interestingly, there is an increase in diacylglycerol levels in the myocardium of diabetic rats that presumably reflects increased phospholipase Cβ-activity. Furthermore, treatment of diabetic rats with angiotensin type 1 receptor antagonists blocks myocardial protein kinase C activation (15).
Regulator of G-protein signaling (RGS) proteins are GTPase-activating proteins that promote the deactivation of heterotrimeric G-proteins (17,18). We previously demonstrated that RGS family members are expressed in the heart and that overexpression of RGS4, a GAP for Gq and Gi, inhibited ligand-induced cardiomyocyte hypertrophy in culture (19,20). Furthermore, we demonstrated that transgenic mice with cardiac-specific overexpression of RGS4 were resistant to pressure overload–induced cardiac hypertrophy (21). Cardiac overexpression of RGS4 also reversed the cardiac dysfunction caused by overexpression of the Gαq subunit (22). In the present study, we evaluated the role of Gq-mediated signal transduction in the pathogenesis of diabetic cardiomyopathy by use of transgenic mice with increased or decreased cardiac Gq activity.
RESEARCH DESIGN AND METHODS
Generation of RGS4 transgenic mice.
Transgenic mice with cardiac-specific overexpression of RGS4 were generated as previously described (21). In brief, two transgenic RGS4-myc lines were established: RGS4–5x and RGS4–8x. Transgenic mice were identified by amplification of the α-MHC–RGS4-myc construct from mouse tail DNA by PCR (21,22). The RGS4–5x line in the C57BL/6 genetic background (MHC-RGS4) was used in all experiments in this study because it exhibited higher RGS4 protein levels in cardiac tissue (21).
Generation of MHC-PPAR × MHC-RGS4 compound transgenic mice.
MHC- PPAR transgenic mice in the C57BL/6 × CBA/J genetic background were generated as previously described (8). Positive offspring in the highest expressing line (404-3) were bred over six generations with C57BL/6 nontransgenic mice to generate hemizygous offspring in the C57BL/6 strain.
MHC-PPAR (404-3) hemizygous transgenic mice in the C57BL/6 genetic background were bred with MHC-RGS4 hemizygous transgenic mice to generate compound transgenic mice. Compound and single transgenic mice were identified by amplification of α-MHC–PPAR and α-MHC–RGS4-myc constructs from mouse tail DNA by PCR (8,21,22). Sex- and age-matched mice were compared for each genotype (nontransgenic, MHC-RGS4, MHC-PPAR, and MHC-PPAR × MHC-RGS4), and 50% of mice for each genotype were male.
Generation of MHC-G188S transgenic mice.
The G188S point mutant form of Gαq, Gαq-G188S, is unable to bind to RGS proteins (23). Gαq-G188S, in contrast to the Gαq-Q209L mutant form, is not constitutively active when expressed in cultured cells, but it is resistant to deactivation by RGS proteins after ligand-mediated activation. We generated transgenic mice with cardiac-specific expression of Gαq-G188S. The α-MHC promoter was linkered to the coding region of Gαq-G188S (24). The transgenic construct had nucleotide substitutions at codon 188: TCG instead of GGG, thereby encoding serine instead of glycine. Two transgenic lines were generated in the C57BL/6 strain: one with genomic integration of three copies of the transgene (MHC–G188S-3x) and a second with genomic integration of seven copies of the transgene (MHC–G188S-7x). Positive offspring were bred with C57BL/6 nontransgenic mice to generate hemizygous offspring. Transgenic mice were identified by amplification of the α-MHC–Gαq-G188S construct from mouse-tail DNA by PCR (21,22). The MHC–G188S-7x transgenic line was used in all experiments in this study because it had higher Gαq-G188S protein levels in cardiac tissue.
The 12-week-old C57BL/6 nontransgenic, MHC-RGS4 transgenic, or MHC–G188S-7x transgenic mice were injected with 200 mg/kg i.p. STZ or vehicle (8). Random tail blood glucose measurements were performed at 7-day intervals after STZ injection. Animals were considered to be diabetic if they had a random blood glucose measurement of >250 mg/dl (8). Mice with tail blood glucose concentrations of <250 mg/dl were reinjected with STZ (200 mg/kg), and the tail blood glucose was measured again 7 days later. All other physiological, anatomic, biochemical, and genetic studies were performed on animals 28 days after STZ injection. Mice were age matched in each case but were not sex matched.
Live anaesthetized mice were evaluated by echocardiography 28 days after injection with STZ or vehicle to determine cardiac mass and function (21,22). We used a Sequoia cardiac echocardiography system (Acuson, Mountain View, CA) that was equipped with a 15-MHz linear transducer.
Live anesthetized mice were evaluated by cardiac catheterization 28 days after injection with STZ or vehicle to determine intracardiac pressures and cardiac function as previously described (21). A 1.4-F Millar catheter was used, and continuous aortic pressure and LV systolic and diastolic pressures were recorded.
Quantitative real-time PCR.
RNA was purified from quick-frozen cardiac tissue by use of Trizol reagent (Sigma). The TaqMan Gold RT-PCR kit (PE Biosystems) was used according to the manufacturer’s instructions. Quantitative PCR was performed by use of real-time detection technology and analyzed on a model 7700 Sequence Detector (Applied Biosystems) with specific primers and fluorescent probes for atrial natriuretic factor (ANF), β-MHC, mitochondrial carnitine palmitoyltransferase 1 (mCPT-1), sarcoplasmic and endoplasmic reticulum calcium exchanger 2a (SERCA2a), and GLUT4. mRNA levels were compared at various time points after correction by use of concurrent glyceraldehyde-3-phosphate dehydrogenase message amplification.
Murine ventricular cytosolic extracts were generated as previously described, and proteins were separated by SDS-PAGE (21). Proteins were electrophoretically transferred to nitrocellulose. Filters were blocked in Tris-buffered saline containing 1% Tween 20 and 2% nonfat dried milk. Filters were washed and incubated with primary antibody. Primary antibodies used included murine monoclonal (M5) anti-FLAG antibody (Sigma), rabbit polyclonal anti-p38 mitogen-activated protein kinase antibody, and rabbit polyclonal Gαq/Gα11 (C-19; Santa Cruz Biotechnology). Filters were extensively washed in Tris-buffered saline containing 1% Tween 20 and then incubated with horseradish peroxidase–conjugated anti-mouse or anti-rabbit secondary antibody (Amersham, Piscataway, NJ). Bands were visualized by use of the enhanced chemiluminescence system (Amersham) (21).
Ventricular tissue was fixed in formalin and embedded in paraffin (21). Embedded mid-LV tissue was sectioned with a microtome, deparaffinized, stained with Masson’s trichrome or hematoxylin and eosin, and examined by compound microscopy.
All data are reported as the means ± SD. Differences between values were evaluated for statistical significance by use of a nonpaired Student’s t test or one-way ANOVA with the Fisher’s post hoc procedure when appropriate. Significance was defined as P < 0.05.
MHC-PPAR × MHC-RGS4 compound transgenic mice.
MHC-PPAR transgenic mice model diabetic cardiomyopathy and are characterized by enhanced cardiac fatty acid oxidation rates with reduced cardiac glucose uptake (8). These mice (line 404-3) develop cardiac hypertrophy, reduced systolic ventricular function, and “embryonic” cardiac gene expression (8). MHC-PPAR transgenic mice develop cardiac hypertrophy and dysfunction by an uncertain mechanism. We speculated that altered metabolism leads to the local release of ligands, such as endothelin-1 or angiotensin II, that activate G-proteins or other signaling intermediaries to promote cardiac hypertrophy and gene induction. To test this possibility, we bred MHC-PPAR mice with MHC-RGS4 transgenic mice to generate compound transgenic mice in the C57BL/6 genetic background.
The cardiac function and gene expression phenotype of MHC-PPAR × MHC-RGS4 compound transgenic mice were compared with MHC-PPAR and MHC-RGS4 single transgenic mice and with nontransgenic mice at 12 weeks of age. MHC-PPAR × MHC-RGS4 compound transgenic and MHC-PPAR single transgenic mice had identical cardiac PPAR-α protein levels, as measured by immunoblotting of ventricular cytosolic lysates (Fig. 1A).
Echocardiographic analysis demonstrated that MHC-PPAR single transgenic mice had markedly reduced systolic function at 12 weeks of age. The fractional shortening was 44.0 ± 3.9% in MHC-PPAR mice, compared with 64.9 ± 3.7% in nontransgenic C57BL/6 mice. MHC-PPAR × MHC-RGS4 compound transgenic mice exhibited an intermediate phenotype with a fractional shortening of 51.5 ± 0.3%. The fractional shortening in MHC-PPAR × MHC-RGS4 compound transgenic mice was significantly greater than that observed in MHC-PPAR single transgenic mice (P = 0.038) (Fig. 1B).
MHC-PPAR single transgenic mice had LV hypertrophy at 12 weeks of age by echocardiography with a calculated LV mass of 83.5 ± 13.8 mg compared with 57.4 ± 5.8 mg for nontransgenic mice (P < 0.05). MHC-PPAR × MHC-RGS4 compound transgenic mice exhibited a trend toward less LV mass by echocardiographic analysis than MHC-PPAR single transgenic mice (67.6 ± 9.8 mg, P = 0.1) (Fig. 1C).
Morphometric analysis confirmed that MHC-PPAR single transgenic mice, but not MHC-PPAR × MHC-RGS4 compound transgenic mice, had cardiac hypertrophy at 12 weeks of age. The biventricular weight–to–body weight ratio in MHC-PPAR single transgenic mice was 6.5 ± 0.62 mg/g, but it was only 5.3 ± 0.31 mg/g in MHC-PPAR × MHC-RGS4 compound transgenic mice (P = 0.024) (Fig. 1D). Similarly, the LV weight–to–body weight ratio was 4.9 ± 0.35 mg/g in MHC-PPAR single transgenic mice, but it was only 4.2 ± 0.36 in MHC-PPAR × MHC-RGS4 compound transgenic mice (P = 0.027) (Fig. 1E).
Analysis of gene expression by quantitative real-time RT-PCR revealed that MHC-PPAR single transgenic mice had elevated cardiac mCPT-1 gene expression, a PPAR-α target gene involved in fatty acid uptake and utilization known to be upregulated in the diabetic heart (8). MHC-PPAR × MHC-RGS4 compound transgenic mice had a similar elevation in mCPT-1 mRNA (Fig. 2). MHC-PPAR single transgenic mice had reduced cardiac glucose transporter GLUT4 gene expression, a gene inhibited by PPAR-α (8), and MHC-PPAR × MHC-RGS4 mice had a similar reduction in GLUT4 mRNA (Fig. 2). MHC-PPAR single transgenic mice had reduced SERCA2a gene expression, a marker of cardiac hypertrophy. In contrast, SERCA2a gene expression was normalized in MHC-PPAR × MHC-RGS4 compound transgenic mice (Fig. 2). Furthermore, MHC-PPAR single transgenic mice had increased ANF and β-MHC gene expression that was normalized in MHC-PPAR × MHC-RGS4 compound transgenic mice (Fig. 2).
Effect of STZ injection on cardiac function and gene expression.
Although MHC-PPAR mice are an important model of the metabolic, structural, and functional abnormalities seen in diabetic cardiomyopathy, these mice do not develop systemic hyperglycemia. To induce diabetes in nontransgenic mice, we injected the pancreatic β-cell toxin STZ. We injected 12-week-old nontransgenic C57BL/6 mice with STZ (or vehicle) at a dose of 200 mg/kg, and random tail blood glucose measurements were obtained every 7 days after STZ injection. Mice with a random blood glucose level >250 mg/dl were defined as diabetic. Of the nontransgenic mice, ∼90% developed diabetes within 14 days of the initial STZ injection by this protocol (Table 1). All but one STZ-injected animal survived for at least 28 days after injection, despite the finding that several animals had blood glucose >500 mg/dl. Most STZ-injected animals exhibited substantial whole-body weight loss after 28 days (Table 1).
Echocardiographic analysis of STZ- or vehicle-injected animals was performed 28 days after injection, and this demonstrated that cardiac function was largely unchanged in diabetic animals. Fractional shortening, a measure of cardiac systolic function, was 64 ± 4% in vehicle-injected C57BL/6 animals, and it was 61 ± 7% in STZ-injected animals (Table 2). The heart rate in anaesthetized diabetic mice was mildly reduced at 639 ± 45 bpm compared with 655 ± 17 in vehicle-injected mice, but this difference was not statistically significant (P = NS). Analysis by invasive cardiac catheterization also revealed that cardiac function was unchanged in diabetic animals (Table 3). The calculated dP/dtmax (maximal change in pressure per unit time) was 3,558 ± 467 in vehicle-injected mice and 3,735 ± 877 in diabetic animals. Morphometric analysis of diabetic hearts revealed that the biventricular weight of diabetic mice was less than that of vehicle-injected mice, but that the biventricular weight–to–body weight ratio was unchanged because of a reduction in body weight (Table 1).
Analysis of gene expression by quantitative real-time PCR revealed that ANF and β-MHC expression, marker genes of cardiac hypertrophy and failure, were markedly induced 28 days after STZ injection. Indeed, ANF gene expression was induced by 14.1 ± 7.4-fold in STZ-injected animals, and β-MHC gene expression was induced by 13.5 ± 7.0-fold in STZ-injected animals when compared with vehicle-injected mice (Fig. 3).
Induction of diabetes in MHC-RGS4 transgenic mice.
Although nontransgenic diabetic C57BL/6 mice did not develop overt cardiac dysfunction, they did exhibit abnormalities in gene expression. To determine whether G-protein–mediated signaling was involved in these phenotypic abnormalities, we injected MHC-RGS4 transgenic mice with STZ. MHC-RGS4 mice were injected with 200 mg/kg of STZ or vehicle. MHC-RGS4 mice developed diabetes at the same rate as nontransgenic animals (Table 1). Echocardiographic analysis 28 days after STZ or vehicle injection demonstrated that MHC-RGS4 diabetic animals had preserved systolic function, with a fractional shortening of 60 ± 2% when compared with vehicle-injected RGS4 mice (59 ± 2% fractional shortening). In addition, MHC-RGS4 diabetic animals did not have a reduced heart rate when compared with vehicle-injected animals. Indeed, the heart rate was slightly increased at 665 ± 20 bpm in STZ-injected MHC-RGS4 mice, and it was 653 ± 14 bpm in vehicle-injected MHC-RGS4 mice (Table 2). Morphometric analysis of diabetic hearts revealed that the biventricular weight of diabetic MHC-RGS4 mice was less than that in vehicle-injected mice, but that the biventricular weight–to–body weight ratio was unchanged because of a reduction in body weight (Table 1).
Analysis of gene expression revealed that MHC-RGS4 transgenic animals were resistant to STZ-stimulated ANF and β-MHC induction. ANF gene expression was induced by 14.1-fold in STZ-injected nontransgenic C57BL/6 animals, but it actually declined by 50 ± 40% in STZ-injected MHC-RGS4 animals when compared with vehicle-injected MHC-RGS4 mice (Fig. 3). β-MHC gene expression was induced by 13.5-fold in STZ-injected nontransgenic C57BL/6 animals, but it declined by 16% in STZ-injected MHC-RGS4 animals when compared with vehicle-injected MHC-RGS4 animals (Fig. 3).
Generation and preliminary characterization of MHC-G188S transgenic mice.
RGS proteins deactivate heterotrimeric G-proteins by stabilizing the transition state between the GTP- and GDP-bound forms of the α-subunit (17,18). We generated a point mutant form of Gαq, Gαq-G188S, that is unable to bind to RGS proteins (23). The G188S point mutant form of Gαq, in contrast to the Q209L mutant form, is not constitutively active when expressed in cultured cells, but it is resistant to deactivation by RGS proteins after ligand-mediated activation. We generated transgenic mice with cardiac-specific expression of Gαq-G188S. The α-MHC promoter was linked to the coding region of Gαq-G188S. Two transgenic lines were generated in the C57BL/6 strain: one with genomic integration of three copies of the transgene (MHC–G188S-3x) and a second with genomic integration of seven copies of the transgene (MHC–G188S-7x). Both lines have increased total Gαq protein in ventricular lysates (Fig. 4A). Mice from both lines appear normal and live normal life spans. The basal cardiac structure and function of these mice is normal at 12 weeks of age in the MHC–G188S-3x (data not shown) and MHC–G188S-7x lines, as determined by echocardiographic and histological examination (Figs. 4B–C).
The phenotype of the MHC–G188S-3x and MHC–G188S-7x transgenic mice is much milder than that observed in the Gαq-40 mice previously generated by another group, and this is likely a consequence of the lower cardiac Gαq protein levels achieved (25,26). In Gαq-40 mice, significant LV systolic dysfunction and bradycardia are observed at 4 weeks of age (22).
Response of MHC-G188S transgenic mice to STZ injection.
MHC–G188S-7x animals were evaluated by echocardiography 28 days after STZ or vehicle injection. MHC–G188S-7x diabetic mice had preserved systolic function when compared with vehicle-injected controls. The fractional shortening was 57 ± 3% in MHC–G188S-7x animals 28 days after STZ injection, and it was 53 ± 4% in vehicle-injected MHC–G188S-7x control animals. Interestingly, MHC–G188S-7x mice became bradycardic in response to STZ-injection. The heart rate was 606 ± 68 bpm in STZ-injected MHC–G188S-7x mice, but it was 680 ± 31 bpm in vehicle-injected animals (P = 0.01) (Fig. 4E). Although both the biventricular weight and the whole body weight was reduced in STZ-injected MHC–G188S-7x animals, the biventricular weight–to–body weight ratio was higher in diabetic mice (0.0043 ± 0.0002) than in vehicle-injected transgenic mice (0.0037 ± 0.0004) (Fig. 4D).
Analysis of gene expression revealed that MHC–G188S-7x transgenic animals were highly sensitized to STZ-stimulated ANF and β-MHC induction. ANF and β-MHC gene expression levels were similar in nondiabetic C57BL/6 and MHC–G188S-7x transgenic animals (data not shown). ANF gene expression was induced by 33-fold in STZ-injected MHC–G188S-7x animals, but it was induced by 14.1-fold in STZ-injected nontransgenic animals when compared with vehicle-injected controls (Fig. 3). β-MHC gene expression was induced by 218.7-fold in STZ-injected MHC–G188S-7x animals, but it was induced by 13.5-fold in STZ-injected nontransgenic animals when compared with vehicle-injected controls (Fig. 3).
Diabetic cardiomyopathy is a condition characterized by cardiac hypertrophy, decreased ventricular compliance, diastolic dysfunction, and, occasionally, systolic dysfunction (1–3). The underlying mechanism by which diabetes results in cardiomyopathy is unclear. It is unlikely that diabetic cardiomyopathy occurs as a result of increased insulin production or decreased target-organ insulin signaling because this condition occurs in type 1 as well as type 2 diabetes. Diabetes is often associated with increased production of angiotensin II in many tissues, and it is possible that hyperglycemia contributes to the production of this ligand. In diabetic nephropathy, considerable evidence suggests that increased angiotensin II production and signaling contributes to this disorder. Indeed, large clinical trials have demonstrated that ACE inhibitor therapy or angiotensin II receptor blocking therapy reduce the incidence of nephropathy in patients with either type 1 or type 2 diabetes (27–30).
In addition to angiotensin II, it is possible that additional secreted ligands, such as endothelin-1, are involved in the pathogenesis of diabetic cardiomyopathy (31). Diabetic patients have elevated serum endothelin-1 levels (31). In addition, treatment of STZ-injected rats with bosentan, an endothelin-1 receptor antagonist, resulted in improved cardiac diastolic and systolic function (32). These findings suggest that angiotensin II inhibition may not completely block the development of diabetic cardiomyopathy.
In this work, we investigated the role of Gq proteins in the pathogenesis of diabetic cardiomyopathy. Both angiotensin II and endothelin-1 activated seven transmembrane receptors that are coupled to Gq. We first tested a robust genetic model of diabetic cardiomyopathy, MHC-PPAR transgenic mice (8). These mice exhibit abnormal cardiac metabolism, although they are systemically euglycemic, and they spontaneously develop ventricular hypertrophy, systolic contractile dysfunction, and “fetal” gene induction. The observation that MHC-PPAR mice are not hyperglycemic is a limitation that makes them an imperfect model of human diabetic cardiomyopathy. We hypothesized that G-protein signaling might play a role in the pathogenesis of MHC-PPAR cardiomyopathy. MHC-PPAR mice were bred with MHC-RGS4 transgenic mice because RGS4 antagonizes Gq signaling in the heart (22). MHC-PPAR × MHC-RGS4 compound transgenic mice exhibited an intermediate phenotype with near-complete rescue of the ventricular hypertrophy and partial rescue of the contractile dysfunction. In addition, embryonic gene induction was markedly reduced in compound transgenic mice when compared with MHC-PPAR single transgenic animals. However, RGS4 did not reverse the direct transcriptional effects of PPAR-α on the regulation of fatty acid metabolism gene expression. These results suggest that the effects of PPAR-α overexpression on metabolic gene expression do not inexorably lead to cardiac hypertrophy and dysfunction.
We next used a murine model of type 1 diabetes that involved the intraperitoneal injection of the pancreatic β-cell toxin STZ (8). We induced diabetes in nontransgenic animals and in transgenic animals with decreased (MHC-RGS4) and increased (MHC–G188S-7x) Gq signaling. Nontransgenic animals developed diabetes after a single injection of STZ, and when analyzed at 28 days, they exhibited mild bradycardia, intact systolic contractile function, and marked induction of embryonic marker genes, including ANF and β-MHC. In the present study, we demonstrated that MHC-RGS4 mice were resistant to STZ-induced bradycardia and embryonic gene induction.
We used MHC–G188S-7x transgenic mice as a model system of mildly increased Gq-mediated cardiac signaling. We thought that these mice would be superior to Gαq-40 mice for this purpose because they have a less dramatic cardiac phenotype in the absence of provocative stimulation (26,27). As predicted, MHC–G188S-7x mice were sensitized to STZ-induced bradycardia and embryonic gene induction. However, MHC–G188S-7x mice did not develop ventricular systolic dysfunction 28 days after STZ injection.
The cardiac phenotype observed in nontransgenic mice after STZ injection was much milder than that observed in rats, and this difference may be a result of the profound cachexia and dehydration that develops in diabetic mice. Indeed, STZ-injected mice did not develop the diastolic dysfunction that is characteristic of human diabetic cardiomyopathy and that is also observed in STZ-injected rats. Whether the profound cachexia and dehydration observed in STZ-injected mice prevented the development of diastolic dysfunction is not known, and further study is required to address this issue. It is possible that subtherapeutic insulin administration to STZ-injected mice—by reducing cachexia and volume depletion without completely restoring euglycemia—may better model human diabetic cardiomyopathy. Taken together, these experiments demonstrate that increased or decreased Gq-mediated signaling affect the development of diabetic cardiomyopathy, and they suggest that medical therapy targeting this signaling pathway may benefit patients with diabetic cardiomyopathy and may prevent the development of this condition in patients with diabetes.
This work was supported by National Institutes of Health Grants PO1 HL057278 (to A.J.M., R.W.G., and D.P.K.) and RO1 HL61567 (to A.J.M.). A.J.M. is a recipient of the Burroughs Wellcome Fund Clinical Scientist Award in Translational Medicine.
The authors thank Attila Kovacs and Mike Courtois of the Mouse Cardiovascular Phenotyping Core Facility of the Center for Cardiovascular Research at Washington University for assistance with the echocardiography and cardiac catheterization studies.