Effects of PPARs Agonists on Cardiac Metabolism in Littermate and Cardiomyocyte-Specific PPAR-γ –Knockout (CM-PGKO) Mice

Understanding the molecular regulatory mechanisms controlling for myocardial lipid metabolism is of critical importance for the development of new therapeutic strategies for heart diseases. The role of PPARγ and thiazolidinediones in regulation of myocardial lipid metabolism is controversial. The aim of our study was to assess the role of PPARγ on myocardial lipid metabolism and function and differentiate local/from systemic actions of PPARs agonists using cardiomyocyte-specific PPARγ –knockout (CM-PGKO) mice. To this aim, the effect of PPARγ, PPARγ/PPARα and PPARα agonists on cardiac function, intra-myocyte lipid accumulation and myocardial expression profile of genes and proteins, affecting lipid oxidation, uptake, synthesis, and storage (CD36, CPT1MIIA, AOX, FAS, SREBP1-c and ADPR) was evaluated in cardiomyocyte-specific PPARγ –knockout (CM-PGKO) and littermate control mice undergoing standard and high fat diet (HFD). At baseline, protein levels and mRNA expression of genes involved in lipid uptake, oxidation, synthesis, and accumulation of CM-PGKO mice were not significantly different from those of their littermate controls. At baseline, no difference in myocardial lipid content was found between CM-PGKO and littermate controls. In standard condition, pioglitazone and rosiglitazone do not affect myocardial metabolism while, fenofibrate treatment significantly increased CD36 and CPT1MIIA gene expression. In both CM-PGKO and control mice submitted to HFD, six weeks of treatment with rosiglitazone, fenofibrate and pioglitazone lowered myocardial lipid accumulation shifting myocardial substrate utilization towards greater contribution of glucose. In conclusion, at baseline, PPARγ does not play a crucial role in regulating cardiac metabolism in mice, probably due to its low myocardial expression. PPARs agonists, indirectly protect myocardium from lipotoxic damage likely reducing fatty acids delivery to the heart through the actions on adipose tissue. Nevertheless a direct non- PPARγ mediated mechanism of PPARγ agonist could not be ruled out.


Introduction
Triglycerides are normally stored in adipocytes while extra adipocyte storage (liver, and myocardium) is minimum and very tightly regulated. However, several-fold increase of triglyceride storage in cardiomyocytes has been observed in animal models of obesity and diabetes [1][2][3][4] and contributes to cardiomyocyte death by non oxidative and oxidative [5] metabolic pathways. In humans, myocardial lipid content was recently reported to increase with the degree of adiposity and contribute to cardiac dysfunction [6], suggesting myocardial lipid content to be a biomarker and putative therapeutic target for cardiac diseases in patients with Metabolic Syndrome [7].
Understanding the molecular regulatory mechanisms controlling for myocardial lipid metabolism is of critical importance for the development of new therapeutic strategies for cardiac diseases. To this regard, it has been previously demonstrated that interventions leading to reduction in myocardial lipid stores improve cardiac function in rodent models of diabetes [1,4,8].
Heart energy is primarily derived from oxidation of FAs, and this correlates with the relatively high levels of PPARa expression in the heart [9][10].Activation of these receptors induces expression of a number of genes encoding proteins involved in transmembrane transport and mitochondrial b-oxidation of fatty acids [10]. In contrast, PPARc is expressed at relatively low levels in the heart [11][12][13][14] and recent data have shown that tissue specific loss of PPARc have no effect on expression of gene controlling lipid and glucose metabolism [15]. Furthermore, exposure of myocytes to PPARa or a/c specific agonists or long-chain fatty acids led to a significant induction of known PPAR target genes involved in fatty acid uptake and oxidation whereas a PPAR agonist had no effect [12].
Due to low myocardial PPARc expression, in vivo effects of PPARc agonists on cardiac metabolism are generally thought to be indirect and secondary to their lipid lowering properties [12]. Direct regulation of cardiac metabolism by PPARc is a subject of considerable debate. Baranowsky et al [13] found that pioglitazone induced lipid accumulation in the heart of rats fed both the standard and high fat diet in spite of concurrent reduction in plasma FFA concentration thus suggesting a direct action of PPARc agonist on the heart. More recently it has shown that tissue-specific loss of PPARc alters heart function and induces myocardial hypertrophy [15,30] with mitochondrial oxidative damage [30] although no effect on gene expression controlling lipid and glucose metabolism at baseline was observed [15]. Indeed, myocardial fatty acids utilization and cardiac contraction were depressed in tamoxifen inducible short term cardiomyocyte restricted PPARc knockout mice [31].
Interestingly enough, such previous findings suggest the possibility that expression and activity of PPARc may be induced in the heart only under certain physiological or pathophysiological conditions. Further studies using cardiac-specific PPARc deletion are required to determine if PPARc regulation of cardiac gene expression is mediated by direct or indirect mechanisms.
Thus, to better assess the role of PPARc on cardiac lipid metabolism and myocardial function and differentiate local against systemic actions of PPARc agonists the effect of PPARc (rosiglitazone), PPARc/PPARa (pioglitazone) and PPARa (fenofibrates) agonists on lipid accumulation and cardiac function was evaluated in cardiomyocyte-specific PPARc knockout (CM-PGKO) mice and littermate control undergoing to standard and high fat diet.
Use of PPARc -knockout (CM-PGKO) and control mice will allow to: a) compare myocardial lipid metabolism and function in absence and presence of cardiomyocyte PPARc b) differentiate local from systemic actions of PPARs agonists b) verify if the effect of PPARc agonist at myocardial level is dependent or independent of cardiomyocyte PPARc receptor activation.

Ethics Statement
All the procedures performed on mice conform with the Directive 2010/63/EU of the European Parliament and have been approved from the University Committee on Animal Care and Use at Second University of Naples (Permit Number: 0009352/08/CB). All efforts were made to ameliorate suffering of animals. Animals a MHC-Cre mice (Cre driven by a -myosin heavy-chain [MHC] promoter) [32] (were bred to floxed PPARc) [33] mice to obtain cardiomyocyte-specific PPARc knockout (CM-PGKO) mice (homozygous floxed PPARc and a MHC-Cre positive). CM-PGKO mice were bred to homozygous floxed PPARc mice to generate both CM-PGKO mice and littermate control mice (homozygous floxed PPARc and a MHC-Cre negative).
All mice were housed in 12 h light/dark cycle (light on 7 a.m.), temperature 22uC, and given standard chow and water ad libitum. After 8 weeks acclimation CM-PGKO and littermate mice were randomly assigned to receive for 18 weeks: The dose of the drug was chosen according to previous studies in which similar dose was used in mice for investigations of the effect of pioglitazone and rosiglitazone on left ventricular remodelling [21,34] In vivo hemodynamic measurement Mice were weighed and anesthetized with diazepam (6 mg/kg sc) and Hypnorm TM (0.7 mg/kg fentanyl citrate and 20 mg/kg fluanisone im), and a tracheotomy was performed for artificial ventilation (Harvard Apparatus, AH 40-1000). The tidal volume of the respirator as set at 1 ml/min, with the rate set at 100 strokes/min, and was supplemented with 100% oxygen. Animals were closely monitored during the procedure to assure that they are maintained in the proper anaesthetic plane. In particular, respiration, colour of the mucous membranes, core body temperature and responses to pedal withdrawal reflex stimuli were continuously monitored Core body temperature was maintained at 37uC with the use of a rectal thermometer attached to a heating pad. The right carotid artery was cannulated with PE-10 tubing, and the mean arterial blood pressure (MABP) was recorded by a computerized data acquisition system (Power Lab Instrument). Heart rate (HR) was recorded during a ECG continuous for 15 s at 50 mm/s velocity. The Pressure rate index (PRI), an index of myocardial oxygen consumption, has been calculated as FC6PAM/1000 and expressed in mmHg/min/10 3 .

Plasma FFA and glucose levels
Blood from fasted (12 hours) mice was collected and the plasma stored at 220uC until assay. Blood glucose level and FFAs were measured using commercially available enzymatic kits (Wako Chemicals, Richmond, VA and Sigma, St. Louis, MO)

Cardiac Hypertrophy Estimation
Hearts were removed and ventricles were dissected for weighting. Ventricular weight to body weight ratio (VW/BW) (milligrams per gram) was used as indicators of cardiac size. Left ventricles were then snap-frozen in liquid nitrogen for further analyses.

Transcript analyses
Total RNA samples were extracted using a RNA extraction kit (Tristapure Eurogetec). QPCR (Quantitative real-time RT-PCR) analyses (QUANTICA sequence detection system-TECHNE) were carried out to determine transcript levels of target genes. The following genes were assayed: PPARc, PPARa, AOX, CPT1MIIA, CD36, FAS, ADRP, GLUT4 and SREBP1-c. SYBR green was used as a universal fluorescent probe and bACTIN was used as an endogenous control. QPCR results from each gene/ primer pair were normalized to bACTIN, and compared across conditions.
In order to validate the efficiency of the target mRNA amplification in relation to the endogenous reference bACTIN mRNA, first, a standard curve was generated with serial dilutions of the cDNA. After amplification, the change in the fluorescence was plotted against the cycle number, and the threshold cycle (CT) was calculated through the Quantica software. Plotting CT values vs log cDNA (ng) revealed efficiencies .96%, and therefore, proofing the suitability of this method for heart samples.

Statistical analysis
Statistical analysis was performed using SPSS-12. Continuous variable was expressed as means 6 standard deviations and compared by analysis of variance (ANOVA) with Bonferroni's test. Correlations between continuous variable will be calculate by regression coefficient. P less than 0,05 will be considered statistically significant. HFD alone resulted in significantly higher weight gain compared to the standard chow fed group in both littermate control (50.366 g; p = 0.001) and CM-PGKO mice (4963 g p = 0.001). In the HFD fed animals six weeks pioglitazone, rosiglitazone and fenofibrate treatment did not produce significant change in body weight.

General features of the experimental animals
HFD progressively elevated plasma free fatty acids (FFA) concentration in both control (from 290634 to 940652 nmol/ ml; P,0.05) and CM-PGKO mice (from 309624 to 912696 nmol/ml; P,0.05). Six weeks administration of pioglitazone, rosiglitazone and fenofibrate significantly decreased plasma FFA concentration in HFD animals (Fig. 1).

Expression of PPARs isoforms in the heart
At baseline, the protein levels of PPARc in littermate control were below the limit of detection with western blot analysis and the percent of positivity at immunoistochimical assay was very low (1%). Also results by quantitative RT-PCR showed low PPARc mRNA content compared to PPARa ones.
After 24 HFD weeks a progressive increase of PPARc mRNA expression and protein levels in littermate controls was found (Fig. 2, A and B).
No difference in baseline PPARa gene and protein expression was found between CM-PGKO and littermate controls. HFD increased gene and protein expression of PPARa in both CM-PGKO mice and in littermate controls and induced a progressive increase of PPARc in littermate controls ( fig. 2 A and B).
In HFD fed animals, six weeks administration of pioglitazone, rosiglitazone and fenofibrate significantly decreased PPARc expression in littermate controls and PPARa in both mice groups ( fig. 2 A and B).

Cardiac Glucose and Lipid Metabolism Gene Expression and Protein Levels
Genes involved in cardiac glucose and lipid metabolism were assessed by quantitative real-time PCR. At baseline, protein levels and mRNA expression of genes involved in lipid uptake, oxidation, synthesis, and accumulation (CD36, CPT1MIIA, AOX, FAS, SREBP1-c and ADPR) of CM-PGKO mice were not significantly different from those of their littermate controls (data not shown). No difference in GLUT4 gene expression and protein levels was also found (data not shown).
As expected, 24 weeks high fat diet induced a progressive increase in protein levels and expression of genes involved in lipid uptake, oxidation and accumulation and a decrease in GLUT4 gene expression in both CM-PGKO and littermate controls (Fig. 3 and Fig. 4).
In both CM-PGKO and littermate controls standard chow fed animals, six weeks of fibrates treatment increased CD36 and CPT1MIIA mRNA levels (Fig. 4).
In HFD fed animals, six weeks administration of pioglitazone, rosiglitazone and fenofibrate significantly decreased protein levels and expression of genes involved in lipid uptake, oxidation and accumulation and significantly increased level of cardiac GLUT4 expression in both groups (Fig. 4). In both groups, the increase in GLUT4 expression was significantly higher in pioglitazone and rosiglitazone group compared to fenofibrates one (respectively 1.36 and 1.27 fold in littermate control and 1.6 and 1.49 fold in CM-PGKO). No effect on ADPR mRNA expression was found.

Heart lipid accumulation
At baseline, no difference in myocardial lipid content was found between CM-PGKO and littermate controls. In both group, twenty four high fat feeding weeks induced a statistically significant progressive increase in the myocardial lipid accumulation. (Fig. 5  A and B) PPAR agonists did not induced statistically significant change of myocardial lipids in mice fed with SD (Fig. 5 B). On the contrary, in the HFD fed animals six weeks treatment with pioglitazone and rosiglitazone and fenofibrate, induced a significant decrease in lipid content in both CM-PGKO and littermate controls (Fig. 5 A  and B) The myocardial lipid content decrease significantly correlated with plasma FFA decrease (r = 0.68; p = 0.001)

Discussion
To better assess the role of PPARc on cardiac lipid metabolism and myocardial function and to better differentiate local/from systemic actions of PPARc agonists the effect of PPARc (rosiglitazone), PPARc/PPARa (pioglitazone) and PPARa (fenofibrates) agonists on lipid accumulation and cardiac function was evaluated in cardiomyocyte-specific PPARc -knockout (CM-PGKO) mice and littermate control undergoing to standard and high fat diet. Use of PPARc -knockout (CM-PGKO) and control mice allowed us: a) to compare myocardial lipid metabolism and function in absence and presence of PPARc; b) to verify if the effect of PPARc agonist at myocardial level was dependent or independent of cardiomyocyte PPARc receptor activation.
Our study demonstrate that: a) at baseline, PPARc does not play a crucial role in regulating cardiac metabolism in mice. When the cardiac FFA afflux increase, (such as in high fat diet, diabetes or metabolic syndrome), myocardial lipid uptake, oxidation and accumulation increase independently of PPARc receptor activation. Such results, showing a minor effect of PPARc on myocardial metabolism, do not exclude that PPARc may have other physiological roles in cardiomyocytes; in fact, different myocardial size/weight and resting heart rate between CM-PGKO mice and littermate control has been found; b) the effect of PPARc agonist at myocardial level seems to be independent of PPARc receptors. PPARs agonists, indirectly protect myocardium from lipotoxic damage probably reducing the delivery of fatty acids to the heart via actions on adipose tissue although a direct action in the heart through not PPARc mediated mechanisms could not be also ruled out.
The role of PPARc as a key regulator of adipose development and metabolism has been extensively characterized [36][37]. However, the exact mechanisms by which PPARc exerts its effects on cardiac metabolism and function are unclear. Previous studies in mice have generally shown low levels of PPARc in heart and myocytes and suggested that most of the cardiac effects of PPARc may be mediated by indirect mechanisms [11][12][13][14]. However, several recent studies have indicated that tissue-specific loss of PPARc alters heart function [30] and that cardiac over expression of PPARc leads to a dilated cardiomyopathy associated with increased lipid and glycogen stores [29].
Our study confirmed low PPARc mRNA expression and protein levels in standard diet littermate control mice heart while higher levels were observed in HFD mice. High fat diet was associated with a progressive increase in protein levels and expression of genes involved in lipid uptake, oxidation and accumulation and a decrease in glucose uptake in both CM-PGKO and littermate controls.
Interestingly, at baseline and after 24 weeks HFD, any difference in vacuolated myocytes, protein levels and mRNA expression of genes involved in lipid uptake, oxidation, synthesis, and accumulation between CM-PGKO mice and littermate controls was found, despite the marked difference in myocardial PPARc expression after an HFD. Such results suggest that PPARc does not play a crucial role in regulating cardiac lipid metabolism in mice and seem to support the hypothesis that in vivo beneficial effects on cardiac metabolism of selective PPARc activator, thiazolidinediones (TZDs),observed in mice, are indirect and secondary to their lipid lowering properties [27]. Nevertheless, it should be pointed out that rosiglitazone increased myocardial mRNA levels of fatty acids (FAs) oxidation and uptake genes exacerbating heart dysfunction [29] in transgenic mice over expressing cardiac PPARc and that pioglitazone induced accumulation of myocardial FFA and DAG in the heart of rats fed both the standard and the high-fat diet [13] thus suggesting a direct actions of PPARc agonist in the heart.
To better differentiate a local against systemic actions of PPARc agonists the effect of rosiglitazone and pioglitazone agonists on lipid accumulation and cardiac function was evaluated in cardiomyocyte-specific PPARc -knockout and littermate controls and compared with the effect of fenofibrates a selective PPARa activators. In standard condition we found that pioglitazone and rosiglitazone treatment have no effect on myocardial metabolism while, fenofibrate treatment significantly increases CD36 and CPT1MIIA gene expression, both involved in lipid uptake as also reported by others [9]. Indeed, in both CM-PGKO and control mice undergoing to HFD, six weeks of treatment with rosiglitazone, fenofibrate and pioglitazone was associated with a reduction of myocardial lipid accumulation and a with a shift of myocardial substrate utilization towards greater contribution of glucose in both groups, as evidenced by decreased in fatty acid oxidation and increased GLUT4 mRNA and protein expression. In both groups, the increase in GLUT4 expression was significantly higher in pioglitazone and rosiglitazone group compared to fenofibrates one. Such results indicate the effect on cardiac lipid metabolism to be independent of cardiomyocyte PPARc. Due to the significant correlation between decrease in myocardial lipid content and fatty acid oxidation and decrease in plasma FFA, we hypothesize that PPARc agonists may indirectly protect myocardium from lipotoxic damage probably reducing the delivery of fatty acids to the heart. Indeed, a direct action in the heart through not PPARc mediated mechanisms could not be also ruled out. To this regard, it has been demonstrated that rosiglitazone activates the expression of adiponectin and its receptors in cultured cardiomyocytes [38] and increases adiponectin levels through a direct activation of the adiponectin promoter [39]. Indeed, adiponectin role in modulating glucose and lipid metabolism via activating AMP-activated protein kinase (AMPK) activities is well known [40].
In conclusion, in baseline condition PPARc does not play a crucial role in regulating cardiac metabolism in mice, likely due to its low myocardial expression. PPARs agonists, indirectly protect myocardium from lipotoxic damage probably reducing fatty acids delivery to the heart through the actions on adipose tissue, although a direct non-PPARc mediated mechanism of PPARc agonist could not be fully excluded.