The Peroxisome Proliferator-Activated Receptor-Gamma Coactivator-1α–Heme Oxygenase 1 Axis, a Powerful Antioxidative Pathway with Potential to Attenuate Diabetic Cardiomyopathy

Abstract

Significance: From studies of diabetic animal models, the downregulation of peroxisome proliferator-activated receptor-gamma coactivator-1α (PGC-1α)–heme oxygenase 1 (HO-1) axis appears to be a crucial event in the development of obesity and diabetic cardiomyopathy (DCM). In this review, we discuss the role of metabolic and biochemical stressors in the rodent and human pathophysiology of DCM. A crucial contributor for many cardiac pathologies is excessive production of reactive oxygen species (ROS) pathologies, which lead to extensive cellular damage by impairing mitochondrial function and directly oxidizing DNA, proteins, and lipid membranes. We discuss the role of ROS production and inflammatory pathways with multiple contributing and confounding factors leading to DCM.

Recent Advances: The relevant biochemical pathways that are critical to a therapeutic approach to treat DCM, specifically caloric restriction and its relation to the PGC-1α–HO-1 axis in the attenuation of DCM, are elucidated.

Critical Issues: The increased prevalence of diabetes mellitus type 2, a major contributor to unique cardiomyopathy characterized by cardiomyocyte hypertrophy with no effective clinical treatment. This review highlights the role of mitochondrial dysfunction in the development of DCM and potential oxidative targets to attenuate oxidative stress and attenuate DCM.

Future Directions: Targeting the PGC-1α–HO-1 axis is a promising approach to ameliorate DCM through improvement in mitochondrial function and antioxidant defenses. A pharmacological inducer to activate PGC-1α and HO-1 described in this review may be a promising therapeutic approach in the clinical setting.

Introduction

Obesity is a major risk factor for vascular dysfunction, diabetes, hyperlipidemia, and hypertension (62). The prevalence of obesity and diabetes effects are on the rise in the United States of America (119), have doubled in many developed countries, and continue to increase in most developing countries (30). Clinical and experimental evidence reveals the existence of specific cardiomyopathy in diabetic patient and animal models, termed “diabetic cardiomyopathy” (DCM) (38). DCM was first described as a distinct pathology, in a small cohort of diabetic patients with adverse myocardial structural changes at postmortem (156). Considerable effort was invested in identifying and characterizing this cardiomyopathy, bringing forth the awareness of diabetes as a risk factor for heart failure (HF). The Framingham Heart Study found that the risk of developing HF increases by two-fold in diabetic males, and by five-fold in diabetic females, independent of age, obesity, dyslipidemia, and coronary heart disease (83). Besides being a risk factor for HF (45, 128), diabetes is also a predictor of poor prognosis in patients with HF (76, 139).

DCM Phenotype

DCM is a complex pathology that develops due to a combination of several metabolic disorders: hyperglycemia, insulin resistance in type 2 diabetes mellitus (DM2), hypertension, and obesity (78, 112, 208). DCM phenotype caused by metabolic co-morbidities (obesity/diabetes and hypertension) is summarized in Figure 1.

FIG. 1.

The phenotype and characteristic of diabetic cardiomyopathy. Diabetic cardiomyopathy phenotype caused by metabolic co-morbidities (obesity/diabetes and hypertension) is characterized by left ventricle hypertrophy (wall thickening and smaller cavity), fibrosis, and diastolic dysfunction; this is defined as increased resistance to filling and is diagnosed by doppler echo when the E/A waves ratio is reduced (<1). Color images are available online.

Diastolic dysfunction

Diastolic HF is defined as increased resistance to the filling of one or both cardiac ventricles. Diastolic HF may be the result of structural abnormalities that increase resistance to ventricular inflow or physiologic abnormalities, resulting in impaired myocardial relaxation. In advanced myocardial hypertrophy, the increased resistance to ventricular diastolic inflow is due to both structural alterations (increased wall thickness and myocardial fibrosis) and impaired diastolic relaxation of the hypertrophied myocardium (60). Diastolic dysfunction is the main functional change occurring in DCM, and it precedes left ventricle (LV) remodeling and systolic dysfunction (40). In the clinical setting, defining and diagnosing diastolic dysfunction poses a challenge. As per the European Society of Cardiology guidelines for the diagnosis of diastolic dysfunction, the E wave velocity is reduced, resulting in E/A reversal (ratio <1.0) as measured by using Doppler echocardiography (136). Besides the impaired cardiac filling, diastolic HF symptoms must comprise: (i) congestive HF; (ii) normal or mild systolic dysfunction; and (iii) elevated levels of brain natriuretic peptide (BNP). The development of diastolic dysfunction with preserved/normal systolic was shown to be associated with: reduced insulin sensitivity (10), increased cardiomyocyte stiffness (122, 204), impaired glucose metabolism and increased fatty acid utilization (14), oxidative stress (67, 68, 152), and Ca2+ overload (63).

Cardiac hypertrophy

LV hypertrophy is a major phenotype in patients with diabetes and is a risk factor for myocardial infarction (MI), stroke, and HF (120). In diabetic patients, LV hypertrophy initially develops as an adaptive response to elevated hemodynamic stress but, in the case of pathological stimuli, it becomes maladaptive (116). This hypertrophy is characterized by increased LV posterior wall thickness and cardiac mass, in addition to a higher ratio of wall thickness to chamber diameter together with molecular and histopathological evidence of cardiomyocyte hypertrophy, fibrosis (40, 100), reduced cardiomyocyte contractility (203), and neurohormonal activation (107). Although LV hypertrophy is frequently associated with increased afterload in diabetic patients with hypertension (15), it can also occur independently of pressure overload (53, 198). The clinical phenotype diabetes-induced hypertrophy was replicated in animal models that displayed increases in a heart weight-to-body weight ratio, cardiomyocyte size, and elevated hypertrophic gene expression (e.g., β-myosin heavy chain, atrial natriuretic peptide [ANP], BNP) (19, 25, 69, 152, 153).

Cardiac fibrosis

Hyperglycemia is sufficient to activate pro-growth (230), promoting cardiac fibroblast and vascular smooth muscle cell proliferation (13, 175). As a result, a major characteristic feature of the diabetic myocardium is extensive fibrosis (176). Tissue fibrosis is a result of abnormally elevated extracellular matrix (ECM) deposition, composed of collagen, elastin, laminin, and fibronectin. The role of the ECM is to provide a scaffold for cardiomyocytes to maintain myocardial structure, shape, and LV wall thickness (37). Elevated myocardial levels and gene expression of ECM proteins (particularly collagen) were observed in animal models of DM2 (122), and they were shown to be related to the impairments in LV diastolic filling, increased myocardial stiffness, and impaired LV relaxation [reviewed in ref. (158)]. Elevated myocardial content and gene expression of ECM proteins (particularly collagen) were observed in experimental models of DM2 (122), including an increase in the insoluble form (176). Importantly, an increase in Collagen was observed in the human diabetic heart (9). The accumulation of ECM is controlled by the degrading enzymes, matrix metalloproteinases (MMP) and their inhibitors (tissue inhibitors of metalloproteinases). They have been suggested as a mechanism underlying ECM accumulation in the diabetic heart (205, 218). MMP-9 knock out (KO) mice are protected from diabetes-induced contractile dysfunction (142). In addition to an increase in ECM, upregulation of transforming growth factor β1 (TGF-β1), its receptor TGF-β receptor II, and its downstream target, connective tissue growth factor promotes fibroblast proliferation in the diabetic heart (12, 158, 212, 231).

Apoptosis, Autophagy, and Mitophagy in the Diabetic Heart

Apoptosis, a form of programmed cell death, is essential for maintaining homeostasis under normal physiological conditions (52, 102). In the diabetic heart, increased apoptosis is involved in the transition from compensated to decompensated hypertrophic state (52) and plays an important role in the development of DCM (18, 49, 52, 94). Apoptosis in the diabetic heart is also associated with pathological structural changes, including increased fibrosis and cardiomyocyte hypertrophy (52, 68, 152). In the diabetic heart, apoptosis is triggered by hyperglycemia-induced caspase-3 activation (18).

Autophagy is another form of programmed cell death, whose role is to remove defective cells and organelles, as well as to recycle damaged proteins (46, 58, 151). The autophagosome is a double-membrane vacuole containing cell components or proteins marked for degradation, before vacuole–lysosome fusion. Well-established markers to detect autophagy include the microtubule-associated protein light chain 3 (LC3) puncta, the ratio of active LC3-II isoform to the inactive LC3-I isoform levels, visualizing double-membrane vesicles on electron microscopy, and increased levels of the autophagy-related genes Atg5, Atg7, Beclin-1, and p62 (43, 46, 151, 171, 228). Autophagy is mainly a cell survival mechanism, regulating the turnover proteins and protecting cells during starvation and other cell stressors (43, 46, 58, 151, 229). However, dysregulated autophagy can result in excessive cell death, and it has been implicated in various diseases, including cancer, neurodegenerative disorders, and diabetes (43, 46, 118).

In the animal model of fructose-fed, insulin-resistant mice exhibited, there is an increase in autophagy but not apoptosis, accompanied by elevated oxidative stress and fibrosis (117). In contrast, in diabetic OVE26 mice, reduced cardiomyocyte autophagy is observed, with reduced adenosine monophosphate-activated protein kinase (AMPK) activity and cardiac dysfunction (222). In view of this conflicting evidence of excessive or insufficient autophagy in the pathophysiology of cardiomyocyte death and cardiac dysfunction (58, 151, 170), the contribution of the autophagic response to the pathogenesis of DCM requires further elucidation.

Mitochondrial function is essential to maintain cellular homeostasis and metabolism. When energy is abundant, excess mitochondria generate excessive reactive oxygen species (ROS) and can start consuming ATP (215). Damaged or unstable mitochondria release ROS and apoptosis-promoting factors that cause damage to neighboring mitochondria and the entire cell (35). Therefore, autophagy-mediated regulation of mitochondria numbers termed “mitophagy,” in response to metabolic demands, is an important mechanism to prevent mitochondria-related cellular damage. The fact that mitophagy occurs after apoptosis (225), and in the absence of apoptosis (226), reinforces the idea that mitophagy's primary role is to preserve the cell when there is mitochondrial dysfunction (57, 199). Mitophagy is activated in cardiomyocytes at baseline and in response to stress to degrade both damaged and excess mitochondria (70, 179). Together with the process of mitochondrial fission, fusion, and mitochondrial biogenesis, mitophagy is one of the most important steps that are necessary to regulate mitochondrial function (171). Previous studies have shown that autophagy, which nonselectively degrades mitochondria, and mitophagy are either downregulated (74, 79, 172, 222, 224) or upregulated (117, 159) in the heart with metabolic syndrome. The occurrence of mitophagy in the heart during the development of DCM and its underlying mechanism is not unequivocally documented. General autophagy, evaluated with LC3II in the whole-cell fraction, is activated by high-fat diet (HFD) feeding, peaking at 6 weeks, but declining thereafter. Interestingly, both LC3II in the mitochondrial fraction and mitophagy were increased. LC3II found predominantly in the mitochondrial fraction and the majority (>75%) of GFP-LC3 ring-like structures colocalized with mitochondria. Thus, activation of autophagy may take place primarily in mitochondria during HFD consumption. These data suggest that the activities of autophagy in the whole-cell fraction and in mitochondria are differentially regulated. Alternatively, additional mechanisms of mitophagy, such as LC3-independent mitophagy, may be activated after conventional mechanisms of autophagy are inactivated. Mitophagy after HFD is attenuated in the absence of Atg7- or Parkin, suggesting that Atg7- or Parkin-dependent mitophagy is activated in the heart during HFD consumption. Cardiac hypertrophy, diastolic dysfunction, and fibrosis are exacerbated in the atg7-cKO mice. These results suggest that Atg7-dependent activation of autophagy/mitophagy acts as an adaptive mechanism to protect the heart against DCM (200).

Mitochondrial Function and Oxidative Stress

Despite the clinical challenges in diagnosing and treating DCM, the specific cellular and molecular mechanisms driving the adverse cardiac remodeling and diastolic dysfunction have not been fully elucidated. However, it is frequently attributed and associated with increased oxidative stress.

Molecular oxygen is inert (174). But the enzyme-driven addition of electrons to oxygen molecules increases their reactivity, producing different types of ROS molecules. The oxidizing potential of ROS is used by aerobic cells to regulate the function and activity of cell signaling molecules. But this process must be tightly regulated by antioxidant molecules to prevent oxidative damage (114, 174, 196).

The endogenous superoxide dismutase (SOD) enzymes degrade ^•O₂⁻ to the more stable forms, H₂O₂ and O₂. Then, H₂O₂ is further broken down to H₂O by glutathione peroxidase and catalase. Under pathological states, H₂O₂ also generates ^•OH, which is a highly reactive molecule. In addition, the interaction between ^•O₂⁻ and NO^• produces Peroxynitrite (ONOO⁻) (173).

Excessive ROS production is a significant factor in the development of various pathologies, due to the oxidizing capacity of DNA, proteins, and lipid membranes (219) (Fig. 2), and to activate stress pathways that regulate cellular damage such as endoplasmic reticulum stress (132, 157). Figure 2 describes the role of oxidative stress formation and its consequences in the diabetic heart. Studies linked diabetes with an increase in ROS and a reduction in antioxidant molecules (26, 41, 147). In obese and diabetic patients, elevated levels of the oxidative DNA damage marker 8-hydroxy 2′-deoxyguanosine were detected and they positively correlated with the body mass index (5). Nicotinamide adenine dinucleotide phosphate (NADPH) oxidases and uncoupled nitrite oxide synthases (NOS) are a major cellular source of ^•O₂⁻ generated from the mitochondrial activity that contribute to cardiac ROS (193).

FIG. 2.

Oxidative stress formation and consequences in the diabetic heart. Increase in NADH oxidation, mitochondrial leakage leading to increased O₂ and XO production, and eNOS uncoupling promote the formation of ROS and via iNOS, RNS. ROS and RNS are responsible for cellular damage through direct DNA damage, protein glycation, nitration, and peroxidation. eNOS, endothelial nitric oxide synthase; iNOS, inducible nitric oxide synthase; RNS, reactive nitrogen species; ROS, reactive oxygen species; XO, xanthine oxidase.

Under normal conditions, the endogenous antioxidant system is sufficient to break down the small amount of ROS produced during mitochondrial respiration to water. But the impaired antioxidant capacity observed in HF increases oxidative stress (133, 194). Another source of oxidative stress in the cardiovascular system is the uncoupling of endothelial nitric oxide synthase (eNOS) (85). Under physiological conditions, eNOS functions as a homodimer enzyme to produce NO^• to transfer electrons from NADPH (via the) to the heme molecule in the oxygenase domain, using flavin adenine dinucleotide and flavinmononucleotide as cofactors. Finally, the substrate l-arginine is oxidized to l-citrulline and NO^•; therefore, uncoupling of the eNOS complex is critical for the formation of NO, ROS, and reactive nitrogen species (51). When eNOS is in its uncoupled state, electrons transferred from the reductase to the oxygenase domain are donated to molecular oxygen instead of l-arginine, producing ^•O₂⁻. Therefore, the regulation of eNOS activity is crucial in preventing the uncoupling of O₂ reduction from NO^• generation (47, 98, 197). HMG-CoA reductase was demonstrated as a regulator of eNOS uncoupling and its inhibition normalized endothelial function and reduced oxidative stress in diabetic rats by inhibiting the activation of NADPH oxidase and preventing eNOS uncoupling (217). Alterations in the function of NOS, predominantly the endothelial isoform eNOS, are generally accepted as responsible for the diabetes-induced dysfunction of NO signaling, production, and bioavailability (28, 108). Increased ROS promote LV hypertrophy via several mechanisms, including the elevated release of vasoconstrictor peptides such as Angiotensin II (Ang II) and ET-1, activation of protein kinases such as c-jun N-terminal protein kinase and p38 mitogen-activated protein kinase (MAPK), and increased mechanical pressure (59, 160, 213, 233, 234). Treating with antioxidant compounds such as butylated hydroxyanisole vitamin E, catalase, probucol, tempol, and N-acetylcysteine was beneficial in preventing tumor necrosis factor (TNF)-α and Ang II-induced cardiomyocyte hypertrophy, de novo protein synthesis, and the elevation in ANP expression (125, 126). Treating mouse papillary muscles with hydroxyl radical for 2 min was sufficient to impair both developed and diastolic force, indicative of systolic and diastolic dysfunction (65). Exposure of cardiomyocytes to H₂O₂, which is a less reactive species, is sufficient to impair cell contractility (96). These direct effects of ROS on cardiac function in isolated cardiomyocytes also occur in vivo, as evidenced by the favorable effects of antioxidant compounds and overexpression of protein with antioxidative activity in models of HF and DCM (75, 106, 127). Increased systematic activation of xanthine oxidase (XO) has been reported in rats and mice (66), including the heart (146). The inhibition of XO attenuates most pathological alteration of DCM and improves both systolic and diastolic dysfunctions in type I diabetic mouse (146) and rat (54) hearts. In patients with ischemic heart disease, treatment with the XO inhibitor allopurinol reduced left ventricular mass and improved clinical symptoms (195). These studies suggest that increased ROS production plays an important role in the pathophysiology and progression of cardiomyocyte hypertrophy. Thus, reduction of ROS generation and restoration of redox balance, using antioxidative treatment, is a promising approach to prevent and treat myocardial fibrosis and hypertrophy in the failing heart.

The development of DCM requires the occurrence of one or more comorbidities that lead to the upregulation of inflammation ROS, impair NO signaling responses and passive tension at both the systemic and the endothelial level. These impact the adjacent cardiomyocytes to induce hypertrophy (6, 135).

In a novel “2 hits” mouse model of DCM, combining HFD and inhibition of constitutive nitric oxide synthase using Nω-nitro-l-arginine methyl ester increased the activity of inducible nitric oxide synthase (iNOS), and S-nitrosylation was observed. Importantly, in patients with similar diastolic dysfunction and HF, NOS2 was elevated as well as increased protein nitrosylation was observed. Pharmacological and genetic suppression of iNOS ameliorated the phenotype (167).

In summary, maintaining the balance between the generation of ROS and their removal by the antioxidant system is critical for normal cardiac function. In the diabetic heart, the excess ROS that are not eliminated by the endogenous antioxidant system cause irreparable damage to cellular organelles and proteins, affecting cardiomyocyte contractility and promoting hypertrophy and fibrosis.

Activation of the Antioxidant Defense Mechanism by Caloric Restriction in the Diabetic Heart

There are several modes of therapy for DCM and lifestyle modification comprising caloric restriction (CR), and physical activity is considered the main therapy. Ultimately, glycemic control over time ameliorates the hazardous effects of obesity and glucose on atherosclerosis and cardiac muscle (17). Physiological studies demonstrated a similar energy intake in HF patients and higher energy expenditure in HF patients, which leads to a negative energy balance associated with increased catecholamines and proinflammatory cytokines levels that are manifested in the development of cachexia and increased mortality (7, 36, 104, 166). On the other side, obesity increases the risk of HF and contributes to its pathophysiology (87).

CR is a regimen that reduces the caloric intake in about 30%–50% below the energy demand that is required to maintain normal body weight and adiposity but still maintains essential nutrients. CR is well known for its ability to increase longevity and to delay or slow the progression of multiple age-related diseases in many laboratory animal models (31, 115, 190). Studies following individuals who practice long-term CR demonstrate that it favorably affects chronic disease risk factors and alleviates oxidative stress. Beyond the direct metabolic response, CR also influences the activity of transcription factors and gene expression (214).

In the aging cardiovascular system, CR has profound beneficial effects, likely related to a reduction in inflammation and oxidative stress (214). CR also protects against endothelial dysfunction, arterial stiffness and attenuates atherogenesis by reducing metabolic risk factors (214).

Adiponectin levels rise under CR, concomitant with a significant decline in circulating insulin and glucose levels. In the heart, CR promotes a shift from carbohydrate to fat metabolism and reduces oxidative stress. Several metabolic regulators/pathways are believed to mediate the CR effect with a focus on the insulin-like growth factor insulin signaling, AMPK, sirtuin (SIRT)1, and peroxisome proliferator-activated receptor-gamma coactivator-1α (PGC-1α). After CR, adiponectin and phospho-AMPK increase in the heart and are involved in cardioprotection (44, 92, 154, 177, 178). The beneficial effects on the aging cardiovascular system of CR include inhibition of nuclear nuclear factor Kappa-light-chain-enhancer of activated B cells (NFκβ) activation and the subsequent inflammatory response, protection against fibrosis, cardiomyocyte apoptosis, and preservation of left ventricular diastolic function (214). The generalized cellular mechanism of CR is illustrated in Figure 3. Figure 3 illustrates the cardiometabolic protective pathway activated by CR. A unique metabolic case of spontaneously reduced food intake and cardioprotection is depicted by transgenic alpha MUPA mice carrying the urokinase-type plasminogen activator (uPA) (105, 121) and is caused by increased levels of leptin, an anorectic satiety hormone. These mice show a set of metabolic changes compared with their control wild-type mice, including reduced food intake when fed at libitum, resistance to obesity, cardioprotection, increased lifespan, and several other benefits seen in CR mice. In contrast, leptin is strongly reduced after experimentally imposed CR, leading to sustained hunger. Leptin/JAK2/PI3K/AKT and leptin/JAK2/STAT3 cascades appear to mediate the leptin-induced cardioprotection in aMUPA mice, possibly by activating myocardial survival pathways (105, 121). In diabetic mice, CR attenuates the development of cardiomyopathy and improves metabolic markers. CR in angiotensin II-stressed-diabetic mice reduces body weight, improves the metabolic profile, and attenuates the cardiomyopathy phenotype. CR ameliorates both fibrosis and inflammation. CR was associated with increased levels of cardiac peroxisome proliferator-activated receptor (PPAR)α followed by improved fatty acids (FAs) utilization, and a reduction in the inflammatory markers TNF-α, toll-like receptor (TLR)2, and TLR4 (29).

FIG. 3.

The cardiometabolic protective pathway activated by CR. After CR, SIRT1 and AMPK are activated. SIRT1, through deacetylation, increases PGC-1α translocation to the nucleus and through binding to NRF2 activates the expression of HO-1 and UCP. Ultimately, CR attenuates the inflammatory response (by preventing NFκB activation), inhibits AKT overactivation, and reduces oxidative stress. Further, the reduction in ROS prevents the degradation of circulating adiponectin levels and through the adiponectin receptors the activation of this protective pathway is enhanced. AMPK, adenosine monophosphate-activated protein kinase; CR, caloric restriction; HO-1, heme oxygenase 1; NFκB, nuclear factor Kappa-light-chain-enhancer of activated B cells; PGC-1α, peroxisome proliferator-activated receptor gamma coactivator-1α; SIRT, sirtuin; UCP, uncoupling protein. Color images are available online.

Sirtuins

SIRTs are class III histone deacetylases that are nicotinamide adenine dinucleotide dependent for their deacetylation reaction (86, 123). In mammalians, there are seven SIRT homologs (SIRT1–SIRT7). These proteins are energy sensors that regulate a major of biological processes, including: glucose metabolism (55), gene expression (206), aging and longevity (16, 61), and cellular stress response (134). It was first demonstrated in yeast in which increased expression of SIRT2 extended the lifespan by 50%, whereas its deletion reduced lifespan. In mice, the deletion of SIRT6 showed the worst phenotype of premature aging and severe metabolic impairment (48). On the other hand, overexpression of SIRT6 in mice on HFD prevented the metabolic dysfunction and the development of comorbidities associated with cardiomyopathy (71, 81, 82, 235) and extended lifespan similarly to CR (81). Similarly, SIRT1 activation also promotes longevity and mimics CR (16, 220). The cellular protection afforded by overexpressing SIRTs includes activation of the AMPKα pathway, inhibition of apoptosis, induction of autophagy, attenuation of inflammation through the inhibition of NFκβ, a decrease in ROS production, and a reduction in the activation of protein kinase B (AKT). All these pathways protect cardiomyocytes from metabolic stress and increase cell survival and longevity (113).

Heme oxygenase 1

Heme oxygenase 1 (HO-1) is the HO-inducible isoform that degrades heme to biliverdin to release CO and iron (1, 2). Biliverdin is then reduced to form bilirubin (4, 84). HO-1 and HO-2 are similar in terms of mechanism, cofactor, and substrate requirements. HO-1 is competitively inhibited by synthetic metalloporphyrins in which the central iron atom is replaced by tin, zinc, cobalt, or chromium (3). HO-1 expression is regulated by various transcription factors, including nuclear factor erythroid-2-related factor 2, and several intracellular signaling molecules, such as MAPK and AKT. HO-1 is induced by an extraordinarily wide variety of drugs, including statins, aspirin, niacin, prostaglandins, eicosanoids, and metallic compounds (1, 161, 162). The degradation products of HO-1 iron, bilirubin, and CO have important cellular functions of their own. Iron is vital for the synthesis of hemoglobin and ferritin, and HO-1 deficiency leads to anemia (140). HO-1 deficiency leads to reduced stress defense (141) and accelerates the formation of arterial thrombosis (201). Importantly, activating the HO-1 signaling was demonstrated to reduce inflammation (99, 129) and fibrosis (33, 88) and to improve both adiposity and endothelial cell dysfunction. (137). HO-1 induction in diabetic mice reduces oxidative stress and inflammation after MI and improves heart function. In cardiomyocytes, an increase in HO-1 levels improved mitochondria membrane potential after hypoxia (73) and reduced ROS production in cardiomyocytes exposed to high glucose levels whereas inhibition of heme degradation promoted ROS production (211). In diabetic rats, increased levels of HO-1 activated AMPK and improved insulin sensitivity (129). The protective role of HO-1 against stress and metabolic dysfunction makes it a very interesting therapeutic target in the battle against obesity and its related disorders.

Peroxisome proliferator-activated receptor-gamma coactivator-1α

PGC-1α is a transcriptional coactivator of nuclear receptors and transcription factors that regulates and enhances their activity (50, 91). PGC-1α regulates oxidative metabolism, adipogenesis, mitochondrial function, and mitochondrial biogenesis (97, 143). PGC-1α was first discovered as a coactivator of the PPARs in a yeast two-hybrid system designed to identify the proteins that regulate the differentiation of brown adipose tissue (97, 144). Another metabolically important target includes nuclear respiratory factor-1 and -2 (207, 221). The induction of PGC-1α decreases lipid accumulation and mitochondrial ROS and increases adiponectin (97, 216). However, PGC-1α KO impairs mitochondrial biogenesis, reduces FA oxidation, as well as increases glucose and insulin resistance (91). The main physiological inducers are CR, fasting (103), exercise (77), and thyroid hormone (56). The activation of PGC-1α is either through deacetylation by SIRT1 or through phosphorylation by AMPK, leading to its translocation to the nucleus (97). Reduction in the expression of PGC-1α has been found to be associated with cardiac dysfunction (180). The regulation of mitochondrial biogenesis by PGC-1α is integral to normal cardiac function, and models of PGC-1α KO or deficiency have significant alterations in cardiac function and structure (8). Data from both pharmacological and genetic intervention, targeting PGC-1α and its downstream targets, support its role in maintaining mitochondrial function and the critical role of mitochondrial dysfunction in the pathogenesis of cardiomyopathy (180).

Cumulatively, these pathways lead to metabolic adaptation to glucose and energy deficiency by promoting gluconeogenesis; increasing mitochondrial function and FA oxidation; and attenuating oxidative stress (24, 101), thus playing a crucial role in maintaining cardiac function.

Ang II-Induced Hypertension, Oxidative Stress, and PGC-1α–HO-1 Signaling

Bartter's syndromes/Gitelman's syndromes (BS/GS) are rare diseases caused by genetic mutations in kidney transporters and ion channels. The patients suffer from hypokalemia, sodium depletion, and activation of the renin–angiotensin–aldosterone system. But surprisingly, these patients have normal blood pressure, reduced vascular resistance and are not responsive to pressor agents (124). In BS/GS patients, SIRT1 and HO-1 are elevated compared with patients with hypertension and healthy subjects, pointing toward a role for reduced SIRT1 and HO-1 in the endothelial dysfunction caused by hypertension (39). A link between oxidative stress mediated by Ang II and PGC-1α was shown in PGC-1α KO mice that exhibited vascular dysfunction and inflammation during chronic Ang II infusion by increasing mitochondrial ROS production (93). However, endothelial-specific PGC-1α gain of function suppressed Ang II-induced hypertension (34). Mechanistically, Ang II stimulates the phosphorylation of PGC-1 by Akt at Ser570. Sequentially, this phosphorylation is required for the binding of general control nonderepressible 5 and the lysine acetylation of PGC-1α. This chain of events disrupts the PGC-1α forkhead transcription factor 1 (FoxO1) complex bound to the FoxO1 response element in the catalase promoter, downregulating the expression of catalase, increasing ROS levels, and promoting hypertrophy (202, 223). In conclusion, hypertension and increased oxidative stress caused by Ang II induction can be attenuated by PGC-1α and HO-1 signaling.

Activation of PGC-1α and HO-1 in Response to Exercise As Antifatigue and Antioxidant Mechanism

The production of ROS and nitric oxide plays a critical role in the controlling contractile force in skeletal muscle (148), and it is evident that physical exercise increases ROS production [reviewed in ref. (130)]. When muscle cells are not stressed, ROS and nitric oxide favor a more pro-oxidant state that is necessary for optimal contractility. Conversely, the treatment of an unfatigued muscle with antioxidants can lead to a suppression of force production (32, 149). However, the production of excessive ROS decreases maximal force production and increases muscle fatigue (187). Therefore, a strenuous exercise that produces high levels of ROS negatively affects contractility and contributes to the development of fatigue (150). To counteract this, exercise training elicits several adaptive mechanisms in skeletal muscle that improve cellular metabolism: It increases the transcription of uncoupling protein 3, pyruvate dehydrogenase kinase 4, HO-1, and PGC-1α during the recovery period (138, 186). The antioxidant mechanism involving the activation of PGC-1α and HO-1 after exercise might be of similar importance for cardiomyocyte contractility.

The Role of PGC-1α–HO-1 Axis in the Attenuation of DCM

In the hearts of diabetic animals, HO-1 levels are reduced (95, 211), contributing to an increase in O₂⁻ production (95). In mice fed an HFD, HO-1 overexpression, specifically in adipose tissue, leads to weight loss; reduced blood glucose, blood pressure, and inflammatory cytokines. It also improved insulin sensitivity, adiponectin levels and increased vascular relaxation in response to acetylcholine. (20). In diabetic mice (induced by streptozotocin) treated with cobalt protoporphyrin (CoPP) to induce HO-1 expression, the increase in HO-1 was linked to an increase in adiponectin, pLKB1, pAKT, pAMPK, pGSK-3, and peNOS levels; a decrease in myocardial superoxide and 3-nitrotyrosine levels; and reduced coronary resistance (95). HO-1 ablation in adipose tissue reduced PGC-1α expression, promoted mitochondrial dysfunction, and contributed to an increase of pro-inflammatory visceral fat (182). In 3T3–L1 preadipocytes subjected to adipogenesis, induction of the PGC-1α signaling cascade prevented the differentiation of the pre-adipocyte cells into large and inflamed adipocytes, maintaining mitochondrial function and increasing HO-1 expression (209). Human studies aimed at investigating the levels of signaling molecules and inflammatory adipokines that regulate cardiovascular risk in epicardial fat from patients with HF who underwent coronary artery bypass surgery. The study revealed a decrease in mitochondrial signaling of HO-1–PGC-1α and an increase of the inflammatory adipokine cellular communication network factor family member 3 levels. Induction of HO-1 in epicardial fat attenuated cardiometabolic dysfunction and improved LV fractional shortening in obese mice, supporting the role of HO-1 signaling in cardiac metabolism and function (184). Individuals with shorter (GT)n repeats in the HO-1 gene, when compared with individuals with longer ones, have a higher transcriptional activity and thus higher HO-1 levels. The shorter (GT)n repeats polymorphism was associated with higher oxidative stress and increased risk of developing coronary artery disease and atherosclerosis (27, 227). Individuals with shorter (<27) (GT)n repeats, despite having significant risk factors (hyperlipidemia, diabetes, and smoking) for coronary artery disease, had better prognosis and reduced risk (80). These data show that polymorphisms correlated with a higher expression of HO-1 are protective against cardiovascular disease in the clinical setting.

In rats on HFD, the induction of HO-1 using CoPP not only improved cardiac function but also prevented myocardial and perivascular fibrosis. These beneficial effects were accompanied by increased levels of cardiac adiponectin levels, and the activation of AMPK, eNOS, and iNOS. Inhibition of HO activity by tin mesoporphrin abolished these beneficial effects, confirming the role of the HO-1 in the cardioprotection (22).

In summary, by examining induction/overexpression or inhibition/knocking down approaches to target HO-1, it is clear that HO-1 and HO activity is central for preventing and even reversing cardiac complications caused by the metabolic syndrome and vascular diseases [reviewed in ref. (42)].

In db/db mice, the induction of PGC-1α reduced the levels of fasting blood glucose; improved cardiac function; increased O₂; elevated pro-inflammatory adipokines; activated the nephroblastoma overexpressed (NOV) signaling, HO-1 levels, and insulin receptor phosphorylation; and improved mitochondrial function. Importantly, these beneficial effects were reversed by the deletion of PGC-1α (21, 183). Significantly, the deletion of PGC-1α resulted in inhibition of HO-1, suggesting that PGC-1α regulates the expression of HO-1 (185). In a model of DCM of db/db mice treated with Ang II, reduced levels of cardiac PGC-1α were observed and were associated with increased oxidative stress. CR attenuated the cardiomyopathy and increased SIRT1 activity and PGC-1α levels. The increased activity of the PGC-1α in CR elevated the expression of cardiac HO-1 and reduced the oxidative stress (210). Inhibition of HO activity promoted the hypertrophic response in diabetic mice, abolished the protective effects of CR on the diabetic heart, increased ROS production, and dramatically reduced PGC-1α levels (211), indicating a crucial role of HO-1 and HO activity in maintaining metabolically “healthy” cardiomyocytes.

These findings suggest the critical role of PGC-1α HO-1 and their interaction, in regulating obesity, diabetes, and heart function, and that targeting this axis constitutes a novel and promising therapy for DCM. The role of HO-1 and PGC-1α in maintaining mitochondrial function is illustrated in Figure 4.

FIG. 4.

The role of HO-1 and PGC-1α in maintaining mitochondrial function. Under normal diet, high levels of adiponectin secreted for “healthy” adipocytes increase the activity of HO-1 and PGC-1α that are important to mitochondrial and, subsequently, cardiac function. Under high-fat diet condition, ROS and RNS levels are increased, and inflamed dysfunctional adipocytes secrete less adiponectin, resulting in reduced levels and activity of HO-1 and PGC-1α, reduced level of mitochondrial fusion proteins, and impaired mitochondrial function, culminating in cardiac metabolic dysfunction, inflammation, and fibrosis and the development of diabetic cardiomyopathy. Color images are available online.

Epoxyeicosatrienoic acids As Inducers of PGC-1α and HO-1

Arachidonic acids, a group of polyunsaturated fatty acids that are part of the membranal phospholipids membranes, are degraded by the enzyme cytochrome P450 to form a group of monooxygenase metabolites called “epoxyeicosatrienoic acids” (EETs) and two isoforms of hydroxyeicosatrienoic acids (HETEs). The metabolism of arachidonic acid to EETs and its physiological function is shown in Figure 5 (169). HETEs and EETs have a contrasting effect on the vascular tone, whereas HETEs function primarily as vasoconstrictors in the microcirculation. EET functions as a vasodilator (64). Soluble epoxide hydrolase (sEH) metabolizes EET to dihydroxyeicosatrienoic acid, a metabolite that has only 20% the activity of EETs (191, 232). sEH inhibition or deletion increases cellular and circulating EETs (mainly 5,6-EET, 8,9-EET, 11,12-EET, and 14,15-EET) (131, 192); it results in a significant decrease in adipocyte size, inflammatory adipokines NOV and TNF-α, while increasing adiponectin. The differentiation into beige-like fat, and the expression of mitochondrial and thermogenic genes are promoted. Overexpression of HO-1 further increased the levels of EETs, suggesting that the antioxidant HO-1 system protects EETs from degradation by ROS. This also suggests that EETs and HO-1 act synergistically and their effects on adipocytes and vascular function are greater than the effects of sEH deletion alone. (110). Treatment with EET agonists prevented both vascular dysfunction (72) and adiposity in vitro (90) and in vivo in mice fed an HFD (189) and increased antioxidant levels, including HO-1 (3, 11, 188). EET-mediated increase in HO-1 upregulates Akt signaling cascade and the activity of eNOS (23, 89), resulting in vasodilation (163, 164). In addition, EET induction reduced the expression of adipogenic markers, affecting adipocyte differentiation (89). This is highlighted by the evidence in obese and diabetic mice where treatment with EET analog (EET-A [(S)-2-(11-(nonyloxy) undec-8(Z)-enamido) succinic acid]) diminished fatty acid accumulation, fibrosis, improved renal function, and attenuated non-alcoholic fatty liver disease through upregulation of HO-1 and PGC-1α (145, 168). This was mirrored by a similar metabolic improvement by HO-1 induction (165). EET agonist as well as overexpression of CYP2J2 attenuated cardiomyopathy in diabetic mice (111, 188). These results suggest that EET has the potential to have a profound effect on hypertension, metabolic syndrome, and cardiac failure. In the 3T3 cells model of adipocyte differentiation, an EET-A induced the PGC-1α expression and protein levels, whereas silencing of PGC-1α abolished EET-mediated elevation of HO-1 levels (181, 209). In mice fed an HFD treated with EET-A, levels of PGC-1α and HO-1 were elevated in adipose tissue, an increase that was abolished in the absence of PGC-1α. The EET-PGC1α-mediated reduction in adiposity in mice fed an HFD involved upregulated mitochondrial function, demonstrated by an increase in MnSOD and SIRT3, as well as in three fusion mediators, Mitofusin 1 and 2, and OPA1, regulating adipocyte cell differentiation, blood pressure and improving metabolism (185). In db/db mice, EET-A ameliorated cardiomyopathy and improved mitochondrial function. Similar to the finding in adipocytes, EET-A's beneficial effects against DCM were mediated by PGC-1α (21, 183). These studies demonstrate that both in vitro and in vivo EET is upstream of the PGC-1α signaling. Thus, an EETs agonist can serve as an inducer of PGC-1α and its downstream signaling. Further, the increase in the activity of HO-1 by EET is dependent on PGC-1α.

FIG. 5.

The metabolism of arachidonic acid to EETs and its physiological function. Arachidonic acid through CYP epoxygenases is metabolite to EETs or through CYP hydroxylases to HETEs. EETs that can have local effects or endocrine effects improve kidney and endothelial function, attenuate inflammation and oxidative stress. On the other hand, HETEs promote hypertension inflammation and ROS production. EET, epoxyeicosatrienoic acid; HETE, hydroxyeicosatrienoic acid.

To summarize, EETs agonists or sEH inhibitors emerge as a promising strategy for pharmacological treatment in metabolic disorders and some of them are already in clinical trials. EETs, through PGC-1α (acting as an inducer of PGC-1α), upregulate HO-1, to reduce inflammation, improve mitochondrial function, and increase vasodilation (109, 155). As there is no known direct pharmacological inducer for PGC-1α, its induction by EETs agonists is a novel and promising therapeutic approach.

Summary and Future Perspectives

Diabetes mellitus is posing a significant clinical challenge and burden around the globe, mainly due to the increased risk of cardiovascular diseases. Diabetic patients have a higher risk of developing diastolic dysfunction and atherosclerosis and some develop DCM. Despite extensive research, there are no specific and effective therapies for DCM. Therefore, unraveling the complex molecular and pathophysiological mechanisms that define the disease process, mainly the precise mechanisms linking chronic metabolic dysfunction observed in diabetes with the unique structural and functional changes, is important to the development of effective treatments. Clinical and preclinical models indicate that the combination of metabolic and biochemical stressors promotes heart dysfunction present in diabetic patients. It has become apparent that increased oxidative/nitrosative stress and inflammation are central triggers to the pathological processes associated with DCM. Despite the high prevalence and mortality of patients with DCM, the precise factors that trigger DCM remain unclear; therefore, research focusing on metabolic pathways that are disturbed in the diabetic heart to develop personalized and precise medicine is needed (78).

CR has profound beneficial effects on the aging or injured cardiovascular system, by reducing oxidative stress and inflammation in the vascular system and the heart as well as preventing adiposity and insulin resistance. From studies of diabetic models, the downregulation of the PGC-1α—HO-1 axis seems to be a crucial event in the development of obesity and DCM. Thus, targeting the PGC-1α–HO-1 axis is a promising approach to ameliorate DCM through improvement in mitochondrial function and antioxidant defenses. Since CR is hard to employ and maintain, a pharmacological inducer to activate PGC-1α and HO-1 and gain the beneficial effects of CR and/or exercise, as this review suggests, may be a promising alternative in the clinical setting as shown in Figure 6.

FIG. 6.

Inducers of the PGC-1α pathway improve mitochondrial function. CR, EETs, or exercise promotes the deacetylation and activation of PGC-1α. Deacetylated PGC-1α translocates to the nucleus and together with the transcription factor NRF2 induces the expression of mitochondrial and antioxidant genes such as HO-1 and UCP. Targeting this pathway has a therapeutic potential to improve mitochondrial function and prevent or attenuate the development of diabetic cardiomyopathy. NRF2, nuclear factor erythroid-2-related factor 2. Color images are available online.

Abbreviations Used

AMPK

adenosine monophosphate-activated protein kinase

Ang II

angiotensin II

ANP

atrial natriuretic peptide

BNP

brain natriuretic peptide

BS

Bartter's syndromes

CoPP

cobalt protoporphyrin

CR

caloric restriction

DM2

type 2 diabetes mellitus

ECM

extracellular matrix

EET

epoxyeicosatrienoic acid

EET-A

EET analog

eNOS

endothelial nitric oxide synthase

FoxO1

forkhead transcription factor 1

GCN5

general control nonderepressible 5

GS

Gitelman's syndromes

HETE

hydroxyeicosatrienoic acid

HF

heart failure

HFD

high-fat diet

HO-1

heme oxygenase 1

iNOS

inducible nitric oxide synthase

KO

knock out

LC3

microtubule-associated protein light chain 3

LV

left ventricle

MAPK

mitogen-activated protein kinase

MFN

mitofusin

MMP

matrix metalloproteinases

MI

myocardial infarction

NADPH

nicotinamide adenine dinucleotide phosphate

NFκB

nuclear factor Kappa-light-chain-enhancer of activated B cells

NOS

nitrite oxide synthases

NRF2

nuclear factor erythroid-2-related factor 2

NOV

nephroblastoma overexpressed

PGC-1α

peroxisome proliferator-activated receptor gamma coactivator-1α

PPAR

peroxisome proliferator-activated receptor

ROS

reactive oxygen species

sEH

soluble epoxide hydrolase

SIRT

sirtuin

SOD

superoxide dismutase

TGF-β1

transforming growth factor β1

TLR

toll-like receptor

TNF

tumor necrosis factor

UCP

uncoupling protein

XO

xanthine oxidase

Funding Information

This work was supported by the National Institutes of Health (grant 1R56HL139561; N.G.A.).