Cardiac Hypertrophy Caused by Peroxisome Proliferator- Activated ...

0013-7227/07/$15.00/0 Printed in U.S.A.

Endocrinology 148(12):6047– 6053 Copyright © 2007 by The Endocrine Society doi: 10.1210/en.2006-1559

Cardiac Hypertrophy Caused by Peroxisome ProliferatorActivated Receptor-␥ Agonist Treatment Occurs Independently of Changes in Myocardial Insulin Signaling Sandra Sena,* Isaac R. Rasmussen,* Adam R. Wende, Alfred P. McQueen, Heather A. Theobald, Nicole Wilde, Renata Oliveira Pereira, Sheldon E. Litwin, Joel P. Berger, and E. Dale Abel Division of Endocrinology, Metabolism and Diabetes, and Program in Human Molecular Biology and Genetics (S.S., I.R.R., A.R.W., H.A.T., N.W., R.O.P., E.D.A.), University of Utah, School of Medicine, Salt Lake City, Utah 84112; Division of Cardiology (A.P.M., S.E.L.), University of Utah School of Medicine, Salt Lake City, Utah 84132; and Department of Metabolic Disorders (J.P.B.), Merck Research Laboratories, Rahway, New Jersey 07065 Peroxisome proliferator-activated receptor (PPAR)-␥ ligands are insulin sensitizers, widely used in the treatment of type 2 diabetes. A consistent observation in preclinical species is the development of cardiac hypertrophy after short-term treatment with these agents. The mechanisms for this hypertrophy are incompletely understood. Given the important role of insulin signaling in the regulation of myocardial size, we tested the hypothesis that augmentation of myocardial insulin signaling may play a role in PPAR-␥ ligand-induced cardiac hypertrophy. We treated mice with cardiomyocyte-restricted knockout of insulin receptors (CIRKO) and littermate controls (wild type) with 2-(2-(4-phenoxy-2-propylphenoxy) ethyl) indole-5-acetic acid (COOH), which is a non-thiazolidinedione PPAR-␥ agonist for 2 wk. Two weeks of COOH treatment increased heart weights by 22% in CIRKO mice and 16% in wild

type, and induced similar fold increase in the expression of hypertrophic markers such as ␣-skeletal actin, brain natriuretic peptide, and atrial natriuretic peptide in CIRKO and wild-type (WT) hearts. COOH treatment increased plasma volume by 10% in COOH-treated WT and CIRKO mice but did not increase systolic or diastolic blood pressure. Echocardiographic analysis was also consistent with volume overload, as evidenced by increased left ventricular diastolic diameters and cardiac output in COOH-treated CIRKO and WT mice. These data indicate that cardiac hypertrophy after PPAR-␥ agonist treatment can occur in the absence of myocardial insulin signaling and is likely secondary to the hemodynamic consequences of plasma volume expansion. (Endocrinology 148: 6047– 6053, 2007)

P

PPAR-␥ agonist treatment is increased edema, and in some patients, exacerbation of heart failure (6). Indeed, in preclinical studies a frequently observed consequence of PPAR-␥ agonist treatment was cardiac hypertrophy (7). The mechanisms responsible for cardiac hypertrophy after PPAR-␥ agonist treatment are incompletely understood but may occur independently of the expression of PPAR-␥ in the cardiomyocyte. Mice with cardiomyocyte-restricted knockout of PPAR-␥ developed cardiac hypertrophy after treatment with the PPAR-␥ ligand rosiglitazone (8). Therefore, it is likely that alternative mechanisms may play an important role in PPAR-␥ ligand-induced cardiac hypertrophy. Insulin is an important regulator of physiological cardiac growth. Postnatal deletion of insulin receptors in cardiomyocytes [cardiomyocyte-restricted knockout of insulin receptors (CIRKO mice)] is associated with reduced cardiomyocyte size that is similar to the effects observed in cardiomyocytes with reduced phosphoinositide 3 (PI3)-kinase signaling (9 – 11). Conversely, activation of IGF-I or PI3-kinase signaling in the heart leads to cardiac hypertrophy (12–15). Administration of insulin to isolated cardiac myocytes induces cellular hypertrophy (9), and exposure of cardiomyocytes to the PPAR-␥ agonist troglitazone increased the sensitivity of cultured cardiomyocytes to insulin-induced increases in cellular protein content (16). Moreover, hyperinsulinemia induced by chronic insulin administration to normal rats leads to cardiac hypertrophy (17, 18). These observations raise the

EROXISOME PROLIFERATOR-activated receptor (PPAR)-␥ agonists are important therapeutic tools in the treatment of type 2 diabetes, acting primarily by activating PPAR-␥ receptors in adipocytes and macrophages (1). They improve insulin sensitivity by a variety of mechanisms such as increased release of adiponectin, which may enhance lipid oxidation in liver and muscle (2), and mobilization of lipid from liver and muscle, while increasing storage in adipocytes (3). Activation of macrophage PPAR-␥ may also have potent antiinflammatory effects, which in the vessel wall may retard atherosclerosis (4, 5). A clinical consequence of First Published Online September 6, 2007 * S.S. and I.R.R. contributed equally to this project. Abbreviations: ANP, Atrial natriuretic peptide; BNP, brain natriuretic peptide; CIRKO, cardiomyocyte-restricted knockout of insulin receptors; COOH, 2-(2-(4-phenoxy-2-propylphenoxy) ethyl) indole5-acetic acid; ⫹dP/dt, peak rate of left ventricular pressure increase; EnaC, epithelial sodium channel; FA, fatty acid; HADH, hydroxyacyl coenzyme A dehydrogenase; IVS, interventricular septum; LV, left ventricular; LVEDP, left ventricular end diastolic pressure; LVIDd, left ventricular internal diameter in diastole; LVIDs, left ventricular internal diameter in systole; MTE-1, mitochondrial thioesterase-1; PDK4, pyruvate dehydrogenase kinase; PGC-1␣, PPAR-␥ coactivator1␣; PI3, phosphoinositide 3; PPAR, peroxisome proliferator-activated receptor; PW, posterior wall; TL, tibia length; TZD, thiazolidinedione; UCP3, uncoupling protein 3; WT, wild type. Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.

6047

6048

Endocrinology, December 2007, 148(12):6047– 6053

possibility that myocardial insulin sensitization after PPAR-␥ agonist treatment could lead to increased cardiomyocyte growth on the basis of increased myocardial insulin signaling Another mechanism that may lead to cardiac hypertrophy after PPAR-␥ ligand treatment is plasma volume expansion, leading to volume-overload cardiac hypertrophy (19, 20). Recent studies have suggested that one molecular mechanism by which PPAR-␥ ligands increase renal tubular sodium reabsorption is via increased expression of the amiloride-sensitive sodium channel [epithelial sodium channel (EnaC)]-␥ (21, 22). We have previously shown that whereas insulin signaling may play an important role in physiological cardiac hypertrophy as evidenced by the reduced cardiac size in CIRKO mice, there is an augmented hypertrophic response to either pressure overload or sustained catecholamine administration (23, 24). Thus, we reasoned that PPAR-␥ ligand treatment of CIRKO would allow us to test directly the role of the necessity for myocardial insulin signaling in PPAR-␥ ligand-induced cardiac hypertrophy and to evaluate the potential role of volume overload, which could lead to an augmented hypertrophic response. Therefore, we treated CIRKO and littermate control mice with a non-thiazolidinedione (TZD) PPAR-␥ ligand 2-(2-(4phenoxy-2-propylphenoxy) ethyl) indole-5-acetic acid (COOH), which is a potent insulin sensitizer (25, 26) that has also induced cardiac hypertrophy in rodents (27). Here, we show that PPAR-␥ ligand treatment of CIRKO mice is associated with a robust hypertrophic response that is associated with transcriptional changes that are characteristic of increased hemodynamic load. Plasma volume is increased, and geometric alterations in the heart develop that are consistent with volume overload, leading us to conclude that this represents the underlying mechanism for PPAR-␥ agonist induced cardiac hypertrophy. Materials and Methods Animals Studies were performed on mice with CIRKO and age-matched littermate controls. Treatment was initiated between 8 and 9 wk of age. The generation and characterization of these mice have been described in previous publications from our group (11). CIRKO mice and littermate controls were treated with COOH for 2 wk. The drug was administered in mouse pellets that were impregnated with 1.4 g COOH/5 kg chow. Two cohorts of animals were used for this study. The first cohort was treated with drug-impregnated chow or standard laboratory chow, respectively. Echocardiography was performed at the beginning and end of the period of treatment in the first set of animals. All other data in this manuscript were derived from the second cohort, which was treated for 2 wk with drug-impregnated chow or control (drug-free) chow with identical macronutrient and micronutrient composition. These mice were used to obtain heart weight, body weight, tibia length (TL), invasive left ventricular (LV) hemodynamic measurements, plasma volume, hematocrit, and quantitative PCR analysis of heart and kidney RNA. COOH treatment started at time 0. At 11 d, hematocrit and then plasma volumes were recorded. Finally, at 14 d, animals were fasted for 6 h before anesthesia (with 400 mg/kg ip chloral hydrate) for LV catheterization. Body weights were taken, and after data collection, while still sedated, blood and tissues were collected. Hearts were excised and weighed, and RNA was extracted for myocardial gene expression. Tibiae were also removed and their lengths measured. The Institutional Animal Care and Use Committee of the University of Utah approved all animal studies. Animals were cared for according to the “Guiding Principles for Research Involving Animals and Human Beings.”

Sena et al. • PPAR-␥ Agonist-Mediated Cardiac Hypertrophy

Assessment of in vivo cardiac function Echocardiography was performed before treatment with COOH and again after 14 d of treatment. The mice were lightly anesthetized with isoflurane and were imaged in the left lateral decubitus position with a linear 13-MHz probe (Vivid V echocardiograph; GE Healthcare, Tampa, FL). Digital images were obtained at a frame rate of 180 images per sec. Two-dimensional images were recorded in parasternal long- and shortaxis projections with guided M-mode recordings at the midventricular level in both views. LV wall thickness [interventricular septum (IVS) and posterior wall (PW) thickness] and internal dimensions at diastole (LVIDd) and systole (LVIDs) were measured. LV fractional shortening [(LVIDd ⫺ LVIDs)/LVIDd], relative wall thickness [(IVS thickness ⫹ PW thickness)/LVIDd], and LV mass [1.05 (IVS thickness ⫹ LVIDd ⫹ PW thickness)3 ⫺ LVIDd3] were calculated from the M-mode measurements. Cardiac output was calculated from pulsed-wave Doppler recordings of LV outflow tract velocities (23). Invasive LV hemodynamic measurements were performed with a temperature-calibrated 1.4-Fr micromanometer-tipped catheter (Millar Instruments, Houston, TX) inserted through the right carotid artery in mice anesthetized with choral hydrate (400 mg/kg) and analyzed as described by us (23, 24, 28). Measurement of plasma volume. Two independent approaches were used to estimate plasma volume. The fist method used an Evans blue dilution technique, and the second method was to determine the hematocrit. Evans blue dye dilution. One hundred microliters of a 0.08% (weight/ vol) solution of Evans Blue (Sigma-Aldrich, St. Louis, MO) diluted in sterile saline were injected into the tail vein (time zero). Blood samples (20 –30 ␮l) from incisions in the tail were drawn into microhematocrit tubes at 10 and 30 min after injection to measure disappearance kinetics. Tubes were centrifuged, and the hematocrit was recorded. Plasma was diluted 1:15 in water. Evans blue concentration in the plasma was measured as optical density at ␭ ⫽ 620 nm. A least square line was fitted to the data obtained at 10 and 30 min from each mouse. Extrapolation of the regression line to zero allowed plasma volume to be calculated as the quotient of the total quantity of injected Evans blue divided by the zero intercept. Plasma volume is expressed as milliliter per 10 g body weight to normalize for variations in body size. This method has been previously used by others to study plasma volume in humans and rodents (29, 30). Hematocrit. Before injecting Evans blue, blood samples (⬃20 –30 ␮l) from incisions in the tail were drawn into microhematocrit tubes. Tubes were centrifuged, and the total height of sample and the height of the red blood cell column were measured. The hematocrit is the ratio between the red blood cell column and total height expressed as a percentage. The hematocrit will decrease as plasma volume increases. RNA extraction and quantitative RT-PCR. Heart and kidney (cortex and inner medulla) total RNA was extracted using TRIzol reagent (Invitrogen Corp., Carlsbad, CA) and purified using the RNAeasy total RNA isolation kit (QIAGEN, Inc., Valencia, CA) according to the manufacturer’s instructions. After extraction, kidney inner medulla and cortex total RNA was precipitated by adding 10 ␮l glycogen as a carrier before solubilization. Three micrograms of total RNA were synthesized to cDNA using Superscript III Reverse Transcriptase (Invitrogen). Quantitative real-time PCR was performed with an ABI Prism 7900 HT RealTime PCR System (Applied Biosystems, Foster City, CA) using a 384well plate. SYBR green I fluorescent dye (Molecular Probes-Invitrogen) was used to quantify the relative mRNA levels. Transcript levels for the constitutive housekeeping gene product cyclophilin were also quantitatively measured in each sample. PCR data were initially calculated as the number of transcripts per number of cyclophilin molecules, and the data were then normalized to the mean wild-type (WT) value. Primer sequences used are shown in supplemental Table 1, which is published as supplemental data on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org.

Statistical analysis Data are shown as mean ⫾ sem, and differences were analyzed by ANOVA and significance of differences between means determined using the post hoc Fisher’s protected least significance difference test or Student’s t test, using the StatView or JMP statistical analysis programs



6049

TABLE 1. Serum lipid concentrations WT

COOH No. of mice FFA (mmol/liter) TG (mmol/liter)

⫺ 10 0.68 ⫾ 0.12 2.58 ⫾ 0.22

CIRKO

⫹ 10 0.41 ⫾ 0.09a 4.87 ⫾ 0.83a

⫺ 10 0.65 ⫾ 0.06 2.71 ⫾ 0.33

⫹ 9 0.35 ⫾ 0.04a 4.60 ⫾ 0.95a

Data are mean ⫾ SEM. Serum was obtained after a 6-h fast in male animals between 9 and 11 wk of age on standard chow or after 2-wk COOH. FFA, Free fatty acid; TG, triglyceride. a P ⬍ 0.05 for mice of the same genotype. (SAS Institute Inc., Cary, NC). A probability less than 0.05 was considered significant.

Results Serum lipids and heart weights

Treatment with COOH reduced serum free fatty acid (FA) concentrations in WT and CIRKO animals, and led to a reciprocal increase in serum triglycerides (Table 1). Table 2 shows heart weights that are normalized either to body weight or TL. As previously described (11, 23, 24), hearts from nontreated CIRKO mice were significantly smaller than hearts from WT mice. After 2-wk COOH treatment, absolute heart weights and heart weights normalized to TL increased by approximately 23 and 17% in CIRKO and WT mice, respectively, and heart weights normalized to body weights increased by approximately 10% for each genotype. Transcriptional response to cardiac hypertrophy

Pathological cardiac hypertrophy is defined in part by a characteristic transcriptional response consisting of increased expression of ␣-actin, atrial natriuretic peptide (ANP), and brain natriuretic peptide (BNP) (31), and reduced expression levels of genes involved in FA metabolism (32– 34). As shown in Table 3, relative to nontreated mice, cardiac hypertrophy was associated with increased expression of ␣-actin and BNP in mice of both genotypes, and additionally with increased expression of ANP in COOH-treated CIRKO mice. It should also be noted that basal levels of these markers were increased in CIRKO hearts, as we have previously described (23, 24), but were increased further upon COOH treatment. The modest degree of LV hypertrophy in these mice was not associated with reduced expression of PPAR-␥ coactivator-1␣ (PGC-1␣). However, there was a coordinate down-regulation of many PPAR-␣ regulated transcripts such as pyruvate dehydrogenase kinase (PDK4), hydroxyacyl co-

enzyme A dehydrogenase (HADH), mitochondrial thioesterase-1 (MTE1), and uncoupling protein 3 (UCP3) in the hearts of COOH-treated CIRKO and WT hearts. Conversely to markers of hypertrophy or LV dysfunction, basal levels of these FA transcripts were for the most part lower in CIRKO mouse hearts before COOH treatment (Table 3). Although the reduction in expression of these FA-metabolism genes after COOH treatment could be consistent with pathological hypertrophy, it is also likely they reflect the fact that concentrations of FAs (a ligand for PPAR-␣) were also reduced by COOH treatment. Plasma volume expansion after COOH treatment

Our observations that absence of insulin receptors in cardiomyocytes did not attenuate the magnitude of the hypertrophic response to COOH treatment prompted us to evaluate the alternative hypothesis, namely that plasma volume expansion could contribute to the cardiac hypertrophy that we observed. Hematocrit was reduced by 11.4 and 8.4% in wild types and CIRKO, respectively, after COOH treatment (P ⬍ 0.05 vs. no drug; Fig. 1A). This was mirrored by a 10.4 and 9.8% increase in plasma volume in WT and CIRKO mice, respectively, as measured by Evans blue dilution, thereby excluding anemia as a cause for the lower hematocrit in COOH-treated mice (Fig. 1B). Because the se of the plasma volume measurements was greater than the hematocrit measurements, the differences in plasma volume (by dye dilution) in each respective genotype before and after COOH treatment did not achieve statistical significance. However, when data from WT and CIRKO mice were pooled, the increase in plasma volume in COOH treated vs. untreated mice was statistically significant (P ⬍ 0.04; Fig. 1C). It has been proposed that transcriptional up-regulation of the expression of EnaC-␥ expression by PPAR-␥ activation represents an important mechanism that is responsible for plasma

TABLE 2. Heart and body weight measurements WT

COOH treatment No. of mice HW (mg) Fold change in HW BW (g) HW/BW ratio Fold change in HW/BW TL (mm) HW/TL ratio Fold change in HW/TL

⫺ 10 96 ⫾ 3a 1.0 ⫾ 0.03 23.7 ⫾ 1.1a 4.088 ⫾ 0.127a 1.0 ⫾ 0.03 15.96 ⫾ 0.2 6.014 ⫾ 0.156a 1.0 ⫾ 0.03

CIRKO

⫹ 10 112 ⫾ 2a,b 1.16 ⫾ 0.02b 24.9 ⫾ 0.5 4.494 ⫾ 0.084a,b 1.10 ⫾ 0.02b 15.96 ⫾ 0.07 6.987 ⫾ 0.121a,b 1.16 ⫾ 0.02b

⫺ 10 69 ⫾ 3 1.0 ⫾ 0.04 21.2 ⫾ 0.9 3.280 ⫾ 0.058 1.0 ⫾ 0.02 15.75 ⫾ 0.19 4.356 ⫾ 0.153 1.0 ⫾ 0.04

Data are mean ⫾ SEM. BW, Body weight; HW, heart weight. a P ⬍ 0.05 vs. equivalently treated CIRKO mice (n ⫽ 9 for heart weight per TL in untreated CIRKO). b P ⬍ 0.05 for mice of the same genotype.

⫹ 9 85 ⫾ 4b 1.22 ⫾ 0.05b 23.5 ⫾ 0.8 3.619 ⫾ 0.150b 1.10 ⫾ 0.05b 15.93 ⫾ 0.14 5.306 ⫾ 0.201b 1.22 ⫾ 0.05b

6050



TABLE 3. Cardiac transcriptional responses and EnaC-␥ expression in renal cortex and inner medulla (fold change relative to untreated wild type) after COOH treatment WT

COOH treatment Acta-1 (␣-actin) nppb (BNP) nppa (ANP) PGC1-␣ MCAD PDK4 HADHa HADHb MTE1 UCP3 EnaC-␥ (renal cortex) EnaC-␥ (renal inner medulla)

⫺ 1.0 ⫾ 0.18 1.0 ⫾ 0.07 1.0 ⫾ 0.09 1.0 ⫾ 0.18 1.0 ⫾ 0.1 1.0 ⫾ 0.16 1.0 ⫾ 0.03 1.0 ⫾ 0.07 1.0 ⫾ 0.06 1.0 ⫾ 0.11 1.0 ⫾ 0.09 1.0 ⫾ 0.21

CIRKO

⫹ 1.52 ⫾ 0.49a 1.59 ⫾ 0.07a,b 1.66 ⫾ 0.13 1.18 ⫾ 0.18 0.9 ⫾ 0.06b 0.61 ⫾ 0.09a 0.74 ⫾ 0.02a 0.84 ⫾ 0.04a 0.45 ⫾ 0.05a 0.36 ⫾ 0.05a,b 1.09 ⫾ 0.1 1.53 ⫾ 0.22

⫺ 2.29 ⫾ 0.19 2.45 ⫾ 0.16a 7.88 ⫾ 0.75a,c 1.01 ⫾ 0.07 0.54 ⫾ 0.1a 0.74 ⫾ 0.09 0.72 ⫾ 0.03a 0.71 ⫾ 0.07a 0.53 ⫾ 0.07a 0.8 ⫾ 0.08 1.0 ⫾ 0.08 1.51 ⫾ 0.24

⫹ 3.49 ⫾ 0.49a– c 3.58 ⫾ 0.2a– c 10.50 ⫾ 0.62a– c 1.58 ⫾ 0.3 0.6 ⫾ 0.09a,c 0.55 ⫾ 0.13a 0.43 ⫾ 0.02a– c 0.58 ⫾ 0.03a,c 0.22 ⫾ 0.04a 0.33 ⫾ 0.07a,b 0.96 ⫾ 0.04 1.46 ⫾ 0.26

Data represent mean values ⫾ SEM of eight to 10 samples (run in duplicate) in which transcript copy number was normalized to that of cyclophilin to control for any differences in cDNA. Statistical differences are by ANOVA with post hoc Fisher’s protected least significance difference. MCAD, Medium chain acyl coenzyme A dehydrogenase; nppa, atrial natriuretic peptide; nppb, brain natriuretic peptide. a P ⬍ 0.05 vs. untreated wild type no drug. b P ⬍ 0.05 vs. CIRKO no drug. c P ⬍ 0.05 vs. treated wild type.

volume expansion after treatment with PPAR-␥ receptor ligands. Therefore, we measured the EnaC-␥ expression in mRNA isolated from CIRKO and WT renal cortex and inner medulla in COOH-treated WT and CIRKO mice. Although there was a tendency toward increased EnaC-␥ expression in the renal inner medulla of COOH-treated wild types, the differences did not achieve statistical significance (Table 3). Analysis of cardiac function in vivo

We performed invasive LV catheterization and echocardiography to evaluate in vivo cardiac function and to measure blood pressure.

LV catheterization. As shown in Table 4, COOH treatment was not associated with any increase in blood pressure. In fact, diastolic blood pressures were significantly lower in COOHtreated WT and CIRKO mice, and the highest systolic blood pressures were observed in untreated wild types. These data exclude hypertension as a cause for the cardiac hypertrophy in these mice. CIRKO mice exhibited diastolic and systolic contractile dysfunction at baseline. Thus, LV end diastolic pressure (LVEDP) was significantly increased in CIRKO mice, and peak rate of LV pressure increase (⫹dP/dt) was significantly reduced. These contractile disturbances were not worsened by COOH treatment. Echocardiography. In COOH-treated WT mice, cardiac hypertrophy was accompanied by increased systolic and diastolic chamber size, increased stroke volume and cardiac output, with no change in fractional shortening. In COOH-treated CIRKO mice, similar changes were noted for diastolic chamber size and cardiac output, and a trend toward increased stroke volume (P ⫽ 0.09) was also present. In contrast to COOH-treated wild types, fractional shortening was increased in COOH-treated CIRKO hearts (Table 5). No changes in septal or PW thickness were discerned using echocardiography in treated WT mice. However, a modest thinning of the IVS accompanied cardiac hypertrophy and LV dilatation in COOH-treated CIRKO mice. The increased chamber size and increased cardiac output in mice of both genotypes are consistent with the presence of volume overload. Discussion

FIG. 1. Hematocrit and plasma volume after COOH treatment. A, Hematocrit in WT and CIRKO mice with or without COOH treatment. A reduction in hematocrit is consistent with plasma volume expansion. B, Plasma volume estimated using the Evans blue dilution technique in WT and CIRKO mice with or without COOH treatment. C, Plasma volume data from all treated and untreated mice (WT ⫹ CIRKO combined). *, P ⬍ 0.05 vs. no drug. WT no treatment (n ⫽ 7). WT COOH treated (n ⫽ 10). CIRKO no treatment (n ⫽ 7). CIRKO COOH treated (n ⫽ 9). BW, Body weight.

This study was performed to test the hypothesis that PPAR-␥ mediated cardiac hypertrophy requires an intact insulin signal transduction pathway, and this was achieved by treating mice with genetic absence of insulin signaling in cardiomyocytes with a non-TZD PPAR-␥ agonist, COOH. Although CIRKO hearts were always smaller than WT hearts, whether treated with COOH or not, the main finding in this study was that COOH treatment resulted in a 22%



6051

TABLE 4. Hemodynamic parameters obtained after LV catheterization WT

COOH No. of mice Aortic systolic pressure (mm Hg) Aortic diastolic pressure (mm Hg) LVSP (mm Hg) LVEDP (mm Hg) LV Dev P (mm Hg) Heart rate (min⫺1) ⫹dP/dt (mm Hg.msec⫺1) ⫺dP/dt (mm Hg.msec⫺1)

⫺ 9 106 ⫾ 3a 79 ⫾ 3 98 ⫾ 4a 13 ⫾ 2a 95 ⫾ 5a 468 ⫾ 18 7221 ⫾ 600a ⫺5354 ⫾ 526

CIRKO

⫹ 7 94 ⫾ 2b 69 ⫾ 2b 93 ⫾ 3 15 ⫾ 2 87 ⫾ 3 449 ⫾ 8 7012 ⫾ 153a ⫺5445 ⫾ 166

⫺ 9 94 ⫾ 3 74 ⫾ 3 86 ⫾ 4 24 ⫾ 3 76 ⫾ 3 492 ⫾ 25 5042 ⫾ 241 ⫺4485 ⫾ 249

⫹ 7 88 ⫾ 3 66 ⫾ 3b 84 ⫾ 3 19 ⫾ 3 78 ⫾ 2 457 ⫾ 32 5477 ⫾ 304 ⫺4761 ⫾ 326

Data are mean ⫾ SEM. Data shown are from studies in which aortic valve function was normal. If aortic regurgitation was present, then data were excluded from analysis (n ⫽ 1 WT ⫺ COOH; n ⫽ 3 WT ⫹ COOH; n ⫽ 1 CIRKO ⫺ COOH; and n ⫽ 2 CIRKO ⫹ COOH). ⫺dP/dt, peak rate of LV pressure decline; LV Dev P, left ventricle developed pressure; LVSP, LV systolic pressure. a P ⬍ 0.05 vs. equivalently treated CIRKO mice. b P ⬍ 0.05 for mice of the same genotype.

increase in cardiac size in CIRKO hearts, which was as robust as the hypertrophic response of WT hearts with an intact insulin signaling pathway. Moreover, CIRKO hearts exhibited gene expression changes that are characteristic of those seen in pathological forms of cardiac hypertrophy. These changes were associated with echocardiographic changes such as increased diastolic chamber dimensions and increased cardiac output that were strongly suggestive of the existence of increased plasma volume, which was confirmed by direct measurement of plasma volume by two independent approaches. Furthermore, there was no increase in blood pressure, which leads us to conclude that the basis for COOH-mediated hypertrophy in these murine models is volume overload cardiac hypertrophy. We have previously shown that insulin signaling is an important regulator of postnatal physiological cardiac growth, and that this is mediated in part via insulinmediated activation of Akt (9, 11). Other studies have confirmed that intact signaling via PI3-kinase and Akt1 are required for physiological cardiac growth (10, 12, 14). In contrast, pathological cardiac hypertrophy, as occurs in response to pressure and volume overload, can occur independently of intact insulin signaling or PI3-kinase-Akt1 signaling. Indeed, the response of insulin receptor, PI3K or Akt1 deficient hearts to pressure overload hypertrophy is for an exaggerated hypertrophic response that ultimately leads to impaired cardiac function (12, 14, 23, 24). Thus, the robust hypertrophic response observed in COOH-treated CIRKO hearts is

consistent with the previously observed augmented hypertrophic response to hemodynamic challenges, such as pressure overload or catecholamine stimulation. Additional evidence to support the hypothesis that the cardiac hypertrophy that was observed in CIRKO hearts was the result of hemodynamic changes as opposed to direct effects on the heart come from the analysis of the transcriptional responses of these hearts to COOH-induced cardiac hypertrophy. Specifically, classical markers of “pathological” cardiac hypertrophy and heart failure were markedly induced in CIRKO hearts relative to WT hearts after COOH treatment. It should be noted that these genes were also increased at baseline in CIRKO hearts, confirming previous observations in this model, which we believe might be an indication of increased wall stress in these smaller hearts in vivo (23, 24). Indeed, CIRKO mice exhibited evidence of contractile dysfunction at baseline (increased LVEDP and decreased ⫹dP/dt), which are consistent with increased expression of ANP, BNP, and ␣-actin. Thus, it is possible that the presence of baseline cardiac dysfunction could sensitize CIRKO hearts to manifest an exaggerated hypertrophic response to a hemodynamic stress. However, it is important to note that the degree of hypertrophy that developed in response to COOH treatment is relatively modest and was not associated with any worsening of contractile function in CIRKO hearts. Recent studies have also suggested that whereas physiological cardiac hypertrophy is associated with an increase in

TABLE 5. Echocardiographic measurements WT

COOH No. of mice LVDd (mm) LVDs (mm) IVSd (mm) LVPWd (mm) FS (%) HR (bpm) Stroke volume (ml) CO (ml/min)

⫺ 9 3.58 ⫾ 0.08 2.20 ⫾ 0.13b 0.96 ⫾ 0.09 0.91 ⫾ 0.007 38.72 ⫾ 2.98b 365 ⫾ 17 0.026 ⫾ 0.003 9.2 ⫾ 1.0

CIRKO

⫹ 8 4.10 ⫾ 0.10a 2.61 ⫾ 0.10a 0.94 ⫾ 0.07b 0.84 ⫾ 0.006 35.76 ⫾ 1.32 454 ⫾ 22a 0.032 ⫾ 0.001a 14.6 ⫾ 1.1a

⫺ 8 3.80 ⫾ 0.10 2.65 ⫾ 0.12 0.83 ⫾ 0.06 0.85 ⫾ 0.007 30.85 ⫾ 1.73 413 ⫾ 27 0.025 ⫾ 0.003 9.8 ⫾ 0.5

⫹ 8 4.15 ⫾ 0.13a 2.60 ⫾ 0.08 0.73 ⫾ 0.04 0.80 ⫾ 0.006 37.30 ⫾ 0.91a 427 ⫾ 32 0.030 ⫾ 0.001 13.0 ⫾ 1.4a

Data are mean ⫾ SEM. CO, Cardiac output; FS, fractional shortening; HR, heart rate; IVSd, interventricular septum thickness measured in diastole; LVDd, LV diastolic diameter; LVDs, LV systolic diameter; LVPWd, LV posterior wall thickness measured in diastole. a P ⬍ 0.05 for mice of the same genotype. b P ⬍ 0.05 vs. equivalently treated CIRKO mice.

6052


mitochondrial FA oxidative capacity in the heart (35, 36), pressure overload hypertrophy is associated with a reduction in FA oxidation in hearts, due in part to transcriptional repression of the FA oxidation gene regulatory program (32, 34). For this reason we determined the expression of representative genes in this pathway. Of interest, COOH treatment was associated both in WT and control mice, with a reduction in the expression levels of a number of PPAR-␣ regulated transcripts, such as PDK4, HADH, MTE1, and UCP3. However, in contrast to pressure overload hypertrophy (32), the expression of PGC-1 ␣ was not reduced by the modest degree of cardiac hypertrophy that was induced by COOH. Therefore, this raises the possibility that the reduced expression of PPAR-␣ regulated transcripts might not be secondary to hypertrophy related transcriptional repression of PGC-1 ␣ but rather to the reduction in serum FA concentrations that resulted from activation of PPAR-␥ receptors in adipocytes. Reduced expression of PDK4 would be predicted to increase flux through pyruvate dehydrogenase and increase glucose use, and the reduced expression of ␤-oxidation genes would reduce FA use. This was recently shown to be the case in a study in which treatment of normal rats with a potent PPAR-␥ agonist increased myocardial glucose use and decreased FA use in vivo (37). This switch in myocardial substrate use could potentially account for the modest increase in triglyceride concentrations in light of studies that suggest that circulating triglycerides represent the major source of metabolizable FAs for the heart and that impairment of myocardial FA oxidation, e.g. in mice with cardiomyocyte-deficiency of lipoprotein lipase, results in hypertriglyceridemia (38 – 40). Many clinical studies have shown that treatment of human subjects with PPAR-␥ agonists is associated with evidence of volume overload such as edema, and in patients with underlying heart disease, may precipitate heart failure or worsen existing heart failure (6). Recent studies have suggested that one potential molecular mechanism for volume overload that develops in response to TZD treatment is a PPAR-␥ dependent up-regulation in the expression of the amiloride-sensitive sodium channel EnaC-␥ in inner medullary collecting ducts, which would lead to increased renal tubular sodium absorption (21, 22). The present study confirmed that PPAR-␥ agonist treatment increased plasma volume in mice, and this was supported by our echocardiographic analyses, which suggested a modest increase in plasma volume as evidenced by increased cardiac chamber diameters and increased cardiac output. Our findings may also explain the recently reported observations of worsening of cardiac hypertrophy that occurs after TZD treatment in mice with cardiomyocyte-specific deletion of PPAR-␥ (8). We observed a trend toward an increase in the expression of EnaC-␥ in inner medullary mRNA isolated from COOHtreated WT mice that did not achieve statistical significance and saw no increase in EnaC-␥ expression in COOH-treated CIRKO mice. The previous report that demonstrated the PPAR-␥ mediated increase in EnaC-␥ expression detected this in primary cultures of inner medullary collecting ducts (22). Thus, the possibility exists that the isolation of mRNA from gross sections of the inner renal medulla (as performed in the present study) did not result in a sufficiently enriched


population of inner medullary collecting ducts. Therefore, although we cannot exclude a role for increased EnaC-␥ expression in the increased plasma volume that we observed, it is also possible that additional mechanisms might exist by which PPAR-␥ activation could lead to an expansion in plasma volume. Together, these studies suggest that the effect of PPAR-␥ activation on cardiac function and cardiac hypertrophy that have long been observed in preclinical and clinical studies do not represent direct myocardial effects of these agents in terms of activation of cardiac PPAR-␥ receptors or increased myocardial insulin sensitivity but are likely secondary to the ability of these agents to increase plasma volume. Acknowledgments We thank Vlad G. Zaha for advice and assistance with the gene expression analyses. Received November 22, 2006. Accepted August 30, 2007. Address all correspondence and requests for reprints to: E. Dale Abel, Division of Endocrinology, Metabolism and Diabetes, Program in Human Molecular Biology and Genetics, 15 North 2030 East, Building 533, Room 3410B, Salt Lake City, Utah 84112. E-mail: dale.abel@ hmbg.utah.edu. This study was supported by an unrestricted research grant from Merck and Company and by the National Institutes of Health (NIH) Grant RO1 HL070070. E.D.A. is an established investigator of the American Heart Association. A.R.W. was supported by NIH Grant 5T32 HL007576 and N.W. by a medical student fellowship from the American Heart Association. Disclosure Statement: E.D.A. received grant support from Merck and Co. (May 12, 2003–April 30, 2005). J.P.B. is an employee of Merck and Co. All other authors have nothing to declare.

References 1. Berger J, Moller DE 2002 The mechanisms of action of PPARs. Annu Rev Med 53:409 – 435 2. Combs TP, Wagner JA, Berger J, Doebber T, Wang WJ, Zhang BB, Tanen M, Berg AH, O’Rahilly S, Savage DB, Chatterjee K, Weiss S, Larson PJ, Gottesdiener KM, Gertz BJ, Charron MJ, Scherer PE, Moller DE 2002 Induction of adipocyte complement-related protein of 30 kilodaltons by PPAR␥ agonists: a potential mechanism of insulin sensitization. Endocrinology 143:998 –1007 3. Moller DE 2001 New drug targets for type 2 diabetes and the metabolic syndrome. Nature 414:821– 827 4. Li AC, Binder CJ, Gutierrez A, Brown KK, Plotkin CR, Pattison JW, Valledor AF, Davis RA, Willson TM, Witztum JL, Palinski W, Glass CK 2004 Differential inhibition of macrophage foam-cell formation and atherosclerosis in mice by PPAR␣, ␤/␦, and ␥. J Clin Invest 114:1564 –1576 5. Collins AR, Meehan WP, Kintscher U, Jackson S, Wakino S, Noh G, Palinski W, Hsueh WA, Law RE 2001 Troglitazone inhibits formation of early atherosclerotic lesions in diabetic and nondiabetic low density lipoprotein receptordeficient mice. Arterioscler Thromb Vasc Biol 21:365–371 6. Nesto RW, Bell D, Bonow RO, Fonseca V, Grundy SM, Horton ES, Le Winter M, Porte D, Semenkovich CF, Smith S, Young LH, Kahn R 2003 Thiazolidinedione use, fluid retention, and congestive heart failure: a consensus statement from the American Heart Association and American Diabetes Association. October 7, 2003. Circulation 108:2941–2948 7. Moller DE, Greene DA 2001 Peroxisome proliferator-activated receptor (PPAR) ␥ agonists for diabetes. Adv Protein Chem 56:181–212 8. Duan SZ, Ivashchenko CY, Russell MW, Milstone DS, Mortensen RM 2005 Cardiomyocyte-specific knockout and agonist of peroxisome proliferator-activated receptor-␥ both induce cardiac hypertrophy in mice. Circ Res 97:372– 379 9. Shiojima I, Yefremashvili M, Luo Z, Kureishi Y, Takahashi A, Tao J, Rosenzweig A, Kahn CR, Abel ED, Walsh K 2002 Akt signaling mediates postnatal heart growth in response to insulin and nutritional status. J Biol Chem 277: 37670 –37677 10. Shioi T, Kang PM, Douglas PS, Hampe J, Yballe CM, Lawitts J, Cantley LC, Izumo S 2000 The conserved phosphoinositide 3-kinase pathway determines heart size in mice. EMBO J 19:2537–2548 11. Belke DD, Betuing S, Tuttle MJ, Graveleau C, Young ME, Pham M, Zhang D, Cooksey RC, McClain DA, Litwin SE, Taegtmeyer H, Severson D, Kahn


12.

13.

14. 15. 16. 17. 18.

19. 20. 21.

22.

23.

24.

25.

CR, Abel ED 2002 Insulin signaling coordinately regulates cardiac size, metabolism, and contractile protein isoform expression. J Clin Invest 109:629 – 639 McMullen JR, Shioi T, Zhang L, Tarnavski O, Sherwood MC, Kang PM, Izumo S 2003 Phosphoinositide 3-kinase(p110␣) plays a critical role for the induction of physiological, but not pathological, cardiac hypertrophy. Proc Natl Acad Sci USA 100:12355–12360 McMullen JR, Shioi T, Huang WY, Zhang L, Tarnavski O, Bisping E, Schinke M, Kong S, Sherwood MC, Brown J, Riggi L, Kang PM, Izumo S 2004 The insulin-like growth factor 1 receptor induces physiological heart growth via the phosphoinositide 3-kinase(p110␣) pathway. J Biol Chem 279:4782– 4793 DeBosch B, Treskov I, Lupu TS, Weinheimer C, Kovacs A, Courtois M, Muslin AJ 2006 Akt1 is required for physiological cardiac growth. Circulation 113:2097–2104 Wilkins BJ, Dai YS, Bueno OF, Parsons SA, Xu J, Plank DM, Jones F, Kimball TR, Molkentin JD 2004 Calcineurin/NFAT coupling participates in pathological, but not physiological, cardiac hypertrophy. Circ Res 94:110 –118 Bell D, McDermott BJ 2000 Troglitazone does not initiate hypertrophy but can sensitise cardiomyocytes to growth effects of serum. Eur J Pharmacol 390: 237–244 Holmang A, Yoshida N, Jennische E, Waldenstrom A, Bjorntorp P 1996 The effects of hyperinsulinaemia on myocardial mass, blood pressure regulation and central haemodynamics in rats. Eur J Clin Invest 26:973–978 Samuelsson AM, Bollano E, Mobini R, Larsson BM, Omerovic E, Fu M, Waagstein F, Holmang A 2006 Hyperinsulinemia: effect on cardiac mass/ function, angiotensin II receptor expression, and insulin signaling pathways. Am J Physiol Heart Circ Physiol 291:H787–H796 Pickavance LC, Tadayyon M, Widdowson PS, Buckingham RE, Wilding JP 1999 Therapeutic index for rosiglitazone in dietary obese rats: separation of efficacy and haemodilution. Br J Pharmacol 128:1570 –1576 Arakawa K, Ishihara T, Aoto M, Inamasu M, Kitamura K, Saito A 2004 An antidiabetic thiazolidinedione induces eccentric cardiac hypertrophy by cardiac volume overload in rats. Clin Exp Pharmacol Physiol 31:8 –13 Zhang H, Zhang A, Kohan DE, Nelson RD, Gonzalez FJ, Yang T 2005 Collecting duct-specific deletion of peroxisome proliferator-activated receptor gamma blocks thiazolidinedione-induced fluid retention. Proc Natl Acad Sci USA 102:9406 –9411 Guan Y, Hao C, Cha DR, Rao R, Lu W, Kohan DE, Magnuson MA, Redha R, Zhang Y, Breyer MD 2005 Thiazolidinediones expand body fluid volume through PPAR␥ stimulation of ENaC-mediated renal salt absorption. Nat Med 11:861– 866 McQueen AP, Zhang D, Hu P, Swenson L, Yang Y, Zaha VG, Hoffman JL, Yun UJ, Chakrabarti G, Wang Z, Albertine KH, Abel ED, Litwin SE 2005 Contractile dysfunction in hypertrophied hearts with deficient insulin receptor signaling: possible role of reduced capillary density. J Mol Cell Cardiol 39: 882– 892 Hu P, Zhang D, Swenson L, Chakrabarti G, Abel ED, Litwin SE 2003 Minimally invasive aortic banding in mice: effects of altered cardiomyocyte insulin signaling during pressure overload. Am J Physiol Heart Circ Physiol 285: H1261–H1269 Sell H, Berger JP, Samson P, Castriota G, Lalonde J, Deshaies Y, Richard D 2004 Peroxisome proliferator-activated receptor ␥ agonism increases the ca-


26.

27. 28.

29. 30. 31. 32. 33. 34. 35. 36.

37.

38.

39.

40.

6053

pacity for sympathetically mediated thermogenesis in lean and ob/ob mice. Endocrinology 145:3925–3934 Laplante M, Sell H, MacNaul KL, Richard D, Berger JP, Deshaies Y 2003 PPAR-gamma activation mediates adipose depot-specific effects on gene expression and lipoprotein lipase activity: mechanisms for modulation of postprandial lipemia and differential adipose accretion. Diabetes 52:291–299 Jones AB 2001 Peroxisome proliferator-activated receptor (PPAR) modulators: diabetes and beyond. Med Res Rev 21:540 –552 Buchanan J, Mazumder PK, Hu P, Chakrabarti G, Roberts MW, Jeong Yun U, Cooksey RC, Litwin SE, Abel ED 2005 Reduced cardiac efficiency and altered substrate metabolism precedes the onset of hyperglycemia and contractile dysfunction in two mouse models of insulin resistance and obesity. Endocrinology 146:5341–5349 Baron AD 1994 Hemodynamic actions of insulin. Am J Physiol 267(2 Pt 1):E187–E202 Salas SP, Marshall G, Gutierrez BL, Rosso P 2006 Time course of maternal plasma volume and hormonal changes in women with preeclampsia or fetal growth restriction. Hypertension 47:203–208 Kong SW, Bodyak N, Yue P, Liu Z, Brown J, Izumo S, Kang PM 2005 Genetic expression profiles during physiological and pathological cardiac hypertrophy and heart failure in rats. Physiol Genomics 21:34 – 42 Arany Z, Novikov M, Chin S, Ma Y, Rosenzweig A, Spiegelman BM 2006 Transverse aortic constriction leads to accelerated heart failure in mice lacking PPAR-␥ coactivator 1␣. Proc Natl Acad Sci USA 103:10086 –10091 Garnier A, Fortin D, Delomenie C, Momken I, Veksler V, Ventura-Clapier R 2003 Depressed mitochondrial transcription factors and oxidative capacity in rat failing cardiac and skeletal muscles. J Physiol 551(Pt 2):491–501 Lehman JJ, Kelly DP 2002 Transcriptional activation of energy metabolic switches in the developing and hypertrophied heart. Clin Exp Pharmacol Physiol 29:339 –345 Allard MF 2004 Energy substrate metabolism in cardiac hypertrophy. Curr Hypertens Rep 6:430 – 435 Burelle Y, Wambolt RB, Grist M, Parsons HL, Chow JC, Antler C, Bonen A, Keller A, Dunaway GA, Popov KM, Hochachka PW, Allard MF 2004 Regular exercise is associated with a protective metabolic phenotype in the rat heart. Am J Physiol Heart Circ Physiol 287:H1055–H1063 Edgley AJ, Thalen PG, Dahllof B, Lanne B, Ljung B, Oakes ND 2006 PPARgamma agonist induced cardiac enlargement is associated with reduced fatty acid and increased glucose utilization in myocardium of Wistar rats. Eur J Pharmacol 538:195–206 Augustus A, Yagyu H, Haemmerle G, Bensadoun A, Vikramadithyan RK, Park SY, Kim JK, Zechner R, Goldberg IJ 2004 Cardiac-specific knockout of lipoprotein lipase alters plasma lipoprotein triglyceride metabolism and cardiac gene expression. J Biol Chem 279:25050 –25057 Augustus AS, Buchanan J, Park TS, Hirata K, Noh HL, Sun J, Homma S, D’Armiento J, Abel ED, Goldberg IJ 2006 Loss of lipoprotein lipase-derived fatty acids leads to increased cardiac glucose metabolism and heart dysfunction. J Biol Chem 281:8716 – 8723 Augustus AS, Kako Y, Yagyu H, Goldberg IJ 2003 Routes of FA delivery to cardiac muscle: modulation of lipoprotein lipolysis alters uptake of TG-derived FA. Am J Physiol Endocrinol Metab 284:E331–E339

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.