The role of mitochondria in metabolic disease: a special emphasis on the heart dysfunction

Abstract

Metabolic diseases (MetD) embrace a series of pathologies characterize by abnormal body glucose usage. The known diseases included in this group are metabolic syndrome, prediabetes and diabetes mellitus type 1 and 2, all of them are chronic pathologies that present metabolic disturbances and are classified as multi-organ diseases. Cardiomyopathy has been extensively described in diabetic patients without overt macrovascular complications. The heart is severely damaged during the progression of the disease, in fact, diabetic cardiomyopathies are the main cause of death in MetD. Insulin resistance, hyperglycemia, and increased free fatty acid metabolism promote cardiac damage through mitochondria. These organelles supply most of the energy that the heart needs to beat and control essential cellular functions, including Ca²⁺ signaling modulation, reactive oxygen species production, and apoptotic cell death regulation. Several aspects of the common mitochondrial functions have been described to be altered in diabetic cardiomyopathies include impairments of energy metabolism, compromises of mitochondrial dynamics, deficiencies in the Ca²⁺ handling, increases in ROS production, and a higher probability of mitochondrial permeability transition pore opening. Therefore, the mitochondrial role in MetD mediated heart dysfunction has been studied extensively to identify potential therapeutic targets for improving cardiac performance. Herein we review the cardiac pathology in metabolic syndrome, prediabetes, and diabetes mellitus, focusing on the role of mitochondrial dysfunctions.

Graphical Abstract.

Cardiac mitochondrial function in metabolic disease. Metabolic disease is characterized by a decreased glycolysis due to insulin-resistance and increased free fatty acids (FA) uptake that promotes FA oxidation for ATP generation. Excessive FA accumulation leads to increases in superoxide anion (O₂⁻) and hydrogen peroxide (H₂O₂) production over a threshold limit. High levels of reactive oxygen species (ROS) leads to uncoupling of the mitochondria electron transport chain (ETC), which reduces mitochondrial ATP production and eventually triggers mitochondrial permeability transition pore (mPTP) opening, leading to cardiomyocyte death. All these alterations promote diastolic and systolic dysfunction which leads to diabetic cardiomyopathy.

INTRODUCTION

Metabolic diseases (MetD), which result from disrupted normal glucose usage, have increased in incidence in recent years (World Health Organization, 2020). MetD are divided into three types—prediabetes, metabolic syndrome (MetS), and diabetes mellitus (DM)—according to the organ compromise and/or the severeness of the disease by itself.

MetD affect several organs across the human body and raise the risk for multiple conditions, including cardiovascular disease (CVD). In patients with DM, metabolism mediated heart damage can affect heart structure and function, leading to diabetic cardiomyopathy (DCM) (Rubler et al., 1972), without the need of other associated co-morbidities. DCM manifests as persistent cardiac dysfunction that frequently leads to heart failure (HF), the principal cause of death in DM (Boonman-de Winter et al., 2012; Vasiliadis et al., 2014; Lee & Kim, 2017).

Among the factors that affect MetD etiology, mitochondrial dysfunction has been shown to play a critical role. It is essential to elucidate the mitochondrial processes that lead to MetD progression, from the early stages to the most severe conditions. Improved understanding may guide the identification of potential therapeutic targets and eventually develops strategies to mitigate or even revert the DCM. The purpose of this review is to summarize the current knowledge about the mitochondrial role within several MetD, including MetS, prediabetes, and DM, with a focus on the mechanisms of cardiac dysfunction.

1. METABOLIC DISEASES

1.1. Prediabetes

Impaired glucose tolerance and impaired fasting glycemia are conditions that define prediabetes. Without changes in their diet and exercise habits, half of the patients with prediabetes will develop type 2 diabetes mellitus (T2DM) (American Diabetes Association, 2019). Diagnosis for prediabetes can be made using the same tests for T2DM (i.e., fasting glucose (FG), glycated hemoglobin test, and/or oral glucose tolerance test) but with a different cut-off (for prediabetes diagnosis: 100 mg/dL ≤ fasting glucose <126 mg/dL, 5.7% ≤ glycated hemoglobin < 6.4%, and/or 140 mg/dL ≤ Glucose after 2h tolerance test < 200 mg/dL). However, because prediabetes is clinically silent, and its detection is usually random, increasing the chances that the patient evolves to develop the more severe T2DM.

To study prediabetes, researchers use animal models in which glycemia is altered without other relevant risk factors (such as obesity and hypertension). Mouse and rat fed with a fructose-rich diet (FRD) are proved models of prediabetes (Alzugaray et al., 2009; Mellor et al., 2011; Sommese et al., 2016; Federico et al., 2017; Szűcs et al., 2019), as well as FRD + streptozotocin (STZ, to destruct pancreatic β-cell) (Koncsos et al., 2016), or a single dose of STZ (Mali et al., 2016).

1.2. Metabolic Syndrome

According to the American Heart Association, MetS is diagnosed when an individual shows at least three of the following risk factors: hyperglycemia, increased blood pressure, dyslipidemia, and abdominal obesity (Grundy et al., 2004; American Heart Association, 2016; Dommermuth & Ewing, 2018). The U.S. prevalence of MetS is around 35% (Moore et al., 2017), and the worldwide incidence is linearly associated with the degree of obesity and overweight (Saklayen, 2018).

In earlier studies of MetS, researchers used a transgenic mouse model ob/ob (Ingalls et al., 1950; Enser, 1972), which shows hyperinsulinemia, hyperglycemia, obesity, and associated cardiac complications. Currently, additional models are available to mimic MetS, including mice with deficient leptin receptor (db/db) (Hummel et al., 1966), transgenic mice that overexpress 11β hydroxysteroid dehydrogenase type 1 to develop increased visceral obesity (Masuzaki et al., 2001), and mice fed with a high-sucrose diet plus high-fat diet (HFD) (Surwit et al., 1995) among others [for full review see (Kennedy et al., 2010; Fellmann et al., 2013; Panchal & Brown, 2011)].

1.3. Diabetes Mellitus

DM is a well-studied chronic disease that is classified into three different types: type 1 diabetes mellitus (T1DM), T2DM, and gestational diabetes (GD). DM is defined as “a chronic MetD characterized by elevated levels of blood glucose which occurs when the body becomes resistant to insulin or doesn't make enough insulin” (World Health Organization, 2020). T1DM can be an idiopathic disease or an autoimmune disease in which islet autoantibodies are produced against the structure of pancreatic β-cell and therefore, insulin production is defective. T2DM results from ineffective use of insulin and is often associated with modifiable factors like obesity and sedentarism (World Health Organization, 2020). An increase in blood pressure due to the impact of the underlying insulin resistance on the vasculature and kidney is also usually related to T2DM (Ferrannini & Cushman, 2012). GD is diabetes diagnosed during pregnancy for the first time and leads to an increased risk of developing DM in the future for both the mother and the child. Additionally, pregnancy and delivery complications are higher in individuals with GD than in non-diabetic people (Alberti & Zimmet, 1998; American Diabetes Association, 2019; World Health Organization, 2020). The DM diagnosis can be made by measuring fasting glucose (FG > 126 mg/dL), glycated hemoglobin test (> 6.4%), and/or oral glucose tolerance test (Glucose after 2h tolerance test > 200 mg/dL).

Prevalence of DM has been increasing in recent years and is currently 8.6% on average worldwide, although it varies widely by country (World Health Organization, 2019). Several non-biological factors (e.g., socioeconomic, demographics, environmental), as well as the increase in human population age and obesity (Unwin & International Diabetes Federation, 2009; International Diabetes Federation, 2019) contribute substantially to the increasing prevalence. The most frequent type of DM is T2DM, which is per se a risk for heart disease and can be running with other vascular co-morbidities such as increased blood pressure, microangiopathy, or kidney disease (among others), enhancing the vicious cycle to increase the risk for heart damage (Shah et al., 2012; Boonman-de Winter et al., 2012; Chen et al., 2018).

To study DM, several models are available. For T1DM, the most used is the induction of pancreatic β-cell destruction by STZ injection (McNeill, 2018). However, the potential of STZ to cause nonspecific effects has been the main criticism for this model. Other animal models are based on genetic manipulation, such as the diabetic BB-rat (Mordes et al., 2005), the non-obese diabetic mice (Li et al., 2008), and the Otsuka Long-Evans Tokushima fatty rat (Karakikes et al., 2009) [see for review (Yorek, 2016)]. The distinction among DM, MetS, and prediabetes could be difficult since they are closely related to DM as the common endpoint of the disease progression. For T2DM, several researchers adopt ob/ob and db/db mice since they present hyperinsulinemia and hyperglycemia (Han et al., 2017; Lee et al., 2018), however, both models present hyperlipidemia and obesity, worsening the cardiovascular risk. Another widespread model of T2DM is based on an HFD (Surwit et al., 1988; Namekawa et al., 2017; Li et al., 2020) where animals are obese and develop T2DM. Recently, HFD plus low doses of STZ have been used to mimic T2DM (Guo et al., 2018).

The genetic models of T2DM have an advantage in that they can be used at an early age, however, the HFD-models are more representative of the human disease. The decision of choosing which model to use for experiments should be made carefully, taking into account that T2DM and MetS present different features that could interfere with the interpretation of the results.

2. CARDIAC PATHOLOGY IN METABOLIC DISEASES

Heart disease is a major concern in MetD because it is the primary cause of death in these patients. An extensive study in patients from the Netherlands with T2DM showed that the prevalence of unknown HF was 27.7%, which was higher than in patients with increased body mass index and patients treated for arterial hypertension (Boonman-de Winter et al., 2012). The presence of prediabetes or MetS also enhances the probabilities to develop T2DM and its progression to DCM and HF (Grundy et al., 2004).

The triggers that lead to DCM include hyperglycemia, hyperlipidemia, and hyperinsulinemia, but the molecular mechanisms are not completely understood (Battiprolu et al., 2010). Several harmful processes occur together in DCM, including left ventricular hypertrophy, interstitial fibrosis, cell death, diastolic and systolic dysfunction, impaired contractility, changes in Ca²⁺ homeostasis, altered substrate utilization, myocardial lipotoxicity, increased reactive oxygen species (ROS) production, and several of them are consequences of mitochondrial dysfunction (Battiprolu et al., 2010). Not all of these deleterious alterations mentioned are developed at the same time and some of them are the cause or consequence of another. Indeed, to study the MetD model, it is critical not only the election of the model is but also the timing for the research. In normal conditions, cardiac mitochondria use fatty acids (FA) to generate approximately 70% of the ATP required by the working heart. In DCM, the decreases in glucose transporter type 4 (GLUT4) cause excessive mitochondrial FA uptake, which enhances ROS generation to the toxic levels leading to subsequent oxidative stress damage (Boudina & Abel, 2010). The cellular redox environment is one of the major posttranslational modulators of proteins’ activity, such as the Ca²⁺ handling proteins responsible for excitation-contraction coupling (ECC). Sarcoplasmic reticulum (SR) Ca²⁺ uptake and release proteins are subjected to oxidative modulations [for review see (Federico et al., 2020)]. For instance, oxidative conditions generally increase the ryanodine receptor 2 (RyR2) open probability, which can lead to cardiac arrhythmias and HF (Xu et al., 1998; Sun et al., 2008). Moreover, the kinases and phosphatases are also subjects of oxidation. Lastly, the decreased rate of glycolysis generates glucose accumulation and advanced glycation end products (AGEs). AGEs complexes can compromise several enzyme activities, altering cardiac contraction (Shao & Tian, 2015).

Taken together, the metabolic imbalance between FA oxidation and glycolysis is critical in the pathogenesis of MetD mediated cardiac pathology (Wang et al., 2006). The disproportionate mitochondrial FA uptake subsequently leads to disturbances in mitochondrial functions that have a direct impact on cardiac performance.

3. MITOCHONDRIAL ROLE IN MetD MEDIATED CARDIAC DYSFUNCTION

The heart obtains most of its energy from FA oxidation (FAO) and switches to the glycolysis pathway under pathological conditions (Stanley et al., 2005; Shao & Tian, 2015). In MetD, the heart is forced to use FA almost exclusively for generating ATP and this overburdens mitochondria and subjected them to oxidative stresses and injury (Christoffersen et al., 2003; Hall et al., 2014). On the contrary, it has also been shown that FA can regulate mitochondrial biogenesis by modulating the activity of the peroxisome proliferator-activated receptor-gamma (PPARγ) and PPARγ-coactivator 1 α (PGC1α) (Lehman et al., 2000; Finck et al., 2002; Arany et al., 2005).

The morphology of mitochondria is directly related to their functions including ATP and ROS production. It has been described that there are subpopulations of mitochondria defined by their spatial location: subsarcolemmal mitochondria (SSM), placed immediately beneath the plasma membrane, perinuclear mitochondria, surrounded to the nucleus, and intermyofibrillar (IMF) mitochondria, embedded within the myofibrillar networks and being the most abundant population (Palmer & Hoppel, 1977). The IMF mitochondria in the adult heart are densely compacted between sarcomeres, but their morphology can be changed by the fission/fusion process (Kane & Youle, 2010) and tunneling connection (Lavorato et al., 2017). Although mitochondrial fission/fusion processes in adult cardiomyocytes are infrequent in physiological conditions, it has been described in pathological situations such as MetD (Galloway & Yoon, 2015). Furthermore, mitochondria distribution is altered in MetS and prediabetes (Federico et al., 2017; Yuan et al., 2018) affecting SR-mitochondrial communication, impairing mitochondrial Ca²⁺ and ADP exchange. Therefore, not only mitochondria distribution is important to warrant the adequate traffic of molecules from and to SR, but also the expression of several key proteins that tether both organelles and maintain the optimal distance to ensure privileged signal transduction (e.g. Ca²⁺) between them (Seidlmayer et al., 2019). Since three enzymes (2-oxoglutarate dehydrogenase, pyruvate dehydrogenase, and NAD⁺-isocitrate dehydrogenase) of the tricarboxylic acid (TCA) cycle are regulated by Ca²⁺ in the mitochondrial matrix, it is important to maintain proper Ca²⁺ communication between these two organelles to have an efficient excitation-contraction-bioenergetic (ECB) coupling (Brookes et al., 2004).

Taken together, there are multifaceted changes in mitochondrial energy metabolism, shape, Ca²⁺ signaling, connections to SR, ROS generation, and quality control that can contribute to the pathogenesis of the MetD mediated heart dysfunction.

3.1. Mitochondrial energy metabolism in MetD mediated heart dysfunction

To carry out blood-pumping activities, the human heart requires 6 kilograms of ATP per day (Neubauer, 2007). To meet this high energy demand, the cardiomyocytes possess a higher number of mitochondria when compared to other cell types. ATP production is coupled to the O₂ consumption at the electron transport chain (ETC). H⁺ is pumped to the intermembrane space during the electron (e⁻) transfer process, which generates an inwardly directed proton motive force that will be used later on by the ATP synthase to produce ATP (Mitchell, 1972).

The alterations in O₂ consumption and ATP production rate of MetD heart remains controversial and may depend on the degree of the disease progression. Pham et al. have demonstrated that in the rat STZ-model, the decrease in O₂ consumption is associated with a decrease in ATP production, either stimulating complex I with glutamate/malate/pyruvate or complex II with succinate. The authors conclude that diabetic hearts have an overall depression of respiration capacity and ATP production with a significantly decreased P/O ratio (ATP production per O₂ consumed) (Pham et al., 2014). As mentioned above, mitochondria can be divided into different subpopulations according to their location. In patients with T2DM where the O₂ consumption was decreased in states 3 and 4 in SSM mitochondria either with glutamate/malate or FA/malate as substrates, without changes in IMF mitochondria.

Additional experiments showed a decrease in ATP production rate in rat models with one dose of STZ (Bombicino et al., 2017) or HFD+STZ (Fang et al., 2018). Mitochondria oxidative phosphorylation (OXPHOS) alterations were also studied by measuring the activity of enzymatic complexes from ETC, TCA enzyme activity, and proteomics. The analyzed data showed that several of the essential proteins required for a normal mitochondrial function, such as PGC1α, Complex I, II, III, and IV, among others were downregulated in MetD (Yan et al., 2013; Szűcs et al., 2019; Wang et al., 2020).

However, How et al. showed increased O₂ consumption in isolated mitochondria from db/db mice in state 3 when the substrate was palmitoyl-carnitine and detected no changes when pyruvate was used, suggesting an enhanced FAO and a decreased glucose-oxidation. The increased O₂ consumption did not translate into the enhancement of cardiac output, indicating inefficiency in cardiac performance and may contribute to contractile dysfunction in the diabetic heart (How et al., 2006).

These changes in energy metabolism observed in diabetic animal models are also found in humans. It has been shown that mitochondria in atrial tissue of T2DM patients show a decrease in respiration with glutamate and FA as substrates. Furthermore, this atrial tissue from diabetic patients show increased mitochondrial H₂O₂ emission and decreased glutathione (GSH) levels. These data support the role of mitochondrial dysfunction and oxidative stress in the pathogenesis of HF in diabetic patients (Anderson et al., 2009). Another study also shows decreases in complex I and IV activity in mitochondria isolated from right atrial appendages of diabetic patients compared to non-diabetic (Croston et al., 2014). Finally, Montaigne et al. reported that heart tissue from patients with T2DM has reduced complex II and III activity, and decreased state 3 respiration, supported by FA, pyruvate, or succinate. In contrast, heart tissue from obese patients, associated with less pronounced contractile dysfunction than T2DM, did not show any significant perturbation of mitochondrial function or oxidative stress (Montaigne et al., 2014). From these results can be concluded that the worsening intrinsic myocardial contraction in the transition from obesity to DM is likely related to the impairment of cardiac mitochondrial function (Montaigne et al., 2014).

Therefore, using different models of MetD, the majority of studies show an altered O₂ consumption (Table.1) suggesting the importance of mitochondrial energy metabolism in the pathogenesis of MetD mediated heart dysfunction. Results obtained from human heart tissues (Anderson et al., 2009; Montaigne et al., 2014) were similar to that from animal models (Yan et al., 2013; Wang et al., 2020), endorsing the translation of animal studies to humans.

Table 1.

O₂ consumption rate in different models.

O2 Consumption	Substrate	Model	Disease	References
↓↓	FA/Malate	Human	Diabetic patients (right atrial appendages)	(Croston et al., 2014)
↑	FA	db/db	Genetic model of T2DM	(How et al., 2006)
↔	Pyruvate	db/db	Genetic model of T2DM
↓↓	FA/Glutamate	Human	Diabetic patients (right atrial appendages)	(Anderson et al., 2009)
↓↓	FA/Pyruvate/Succinate	Human	Diabetic patients (right atrial appendages)	(Montaigne et al., 2014)
↓↓	Pyruvate/Malate	OVE26	Genetic model of T1DM	(Shen et al., 2004)
↔	Glutamate/Malate/Succinate	HFD	Diet model of T2DM	(Koncsos et al., 2016)

FA: Fatty acids, HFD: High-fat diet, T2DM: type 2 diabetes mellitus, T1DM: type 1 diabetes mellitus. Note: The references mentioned in this table are cited among the main text.

3.2. Mitochondrial dynamics in MetD mediated heart dysfunction

Mitochondrial dynamics encompasses fusion, fission, selective degradation, and transport processes (Chan, 2020). Fission and fusion are in balance in physiological conditions and maintain normal mitochondrial mass, shape, network, biogenesis and turnover. The fusion process generates a bigger mitochondrion from two smaller ones, merging the content of both original mitochondria, which helps to mitigate the mitochondrial damage in addition to establish the mitochondrial network. On the opposite, fission creates two new mitochondria from a single one, contributing to mitochondrial turnover and facilitating apoptosis during high levels of cellular stress (Kane & Youle, 2010).

Fission and fusion are also related to the mitochondria OXPHOS. Stressed mitochondria or defective OXPHOS promotes mitochondrial fragmentation (Sauvanet et al., 2010). The fission process is regulated by several proteins, including dynamin-related protein 1 (DRP1) (Herskovits et al., 1993), Fis-1, which connects DRP1 to the outer mitochondria membrane (OMM), and other described adaptors such as MFF, MiD49, and MiD 51 (Yoon et al., 2003; Otera et al., 2010; Palmer et al., 2011). Mitochondrial fusion is controlled by two major proteins: Optic atrophy protein (OPA1), located in the inner mitochondrial membrane (IMM), and mitofusin 1 and 2 (Mfn1/2), located in the OMM (Cipolat et al., 2004, 2006). The balance between mitochondrial fusion and fission is essential in mammals, and even mild defects in mitochondrial dynamics are associated with disease etiology. In general, a tip of balance toward fission is usually associated with deleterious processes and toward fusion is associated with compensatory mechanisms (Chen et al., 2003; Ishihara et al., 2009).

In the adult myocardium, fission events are challenging to detect in part due to the abundance and immobility of mitochondria, therefore, the data available related to the mitochondrial dynamics in MetD mediated heart dysfunction is limited. Transmission electron microscopy images of the heart tissue have shown that in the early stages of cardiac MetD pathogenesis, such as prediabetes induced by FRD, mitochondria are smaller and more spherical in shape compared with control animals (Federico et al., 2017), indicating that fission processes might occur. Koncsos et al. similarly described a decrease in area, perimeter, and sphericity of mitochondria in the prediabetic rat model of HFD and STZ treatments (Koncsos et al., 2016). Flow cytometry showed that mitochondrial size and cristae complexity were decreased in diabetic IMF in mice models with one dose of STZ (Williamson et al., 2010). In another paper, the same research group reported that cardiac mitochondria density was increased and mitochondria area was decreased, showing unbalanced mitochondrial dynamics towards the fission processes in this diabetic model (Dabkowski et al., 2010).

Montaigne et al. explore the role of impaired mitochondrial dynamics in myocardial contractile dysfunction in patients with T2DM without obesity. The mitochondria of the heart tissues from these patients, harvested during cardiopulmonary bypass, showed no difference in density but a significant decrease in size. The authors also analyzed the amount of the proteins related to the mitochondrial dynamics such as Mfn2, Mfn1, OPA1, DRP1, and Fis1. They only found a large decrease in the expression of the mitochondrial fusion related protein Mfn1, which may account for mitochondrial fragmentation. These changes in mitochondrial morphology were associated with impaired complex I, II, and III activity decreased respiratory control ratio (RCR) and increased oxidative stress further confirming the close relationship between mitochondrial shape and function (Montaigne et al., 2014).

3.3. Mitochondrial calcium signaling in MetD mediated heart dysfunction

The mitochondria are juxtaposed with the SR in the cardiomyocytes and participate in uptaking a fraction of Ca²⁺ during each heart-beat (Beutner et al., 2005). Localized Ca²⁺ released from the SR creates a high Ca²⁺ microdomain in the SR mitochondria contact sites to stimulate mitochondrial Ca²⁺ uptake. MCU at the IMM, a highly Ca²⁺ selective ion channel (Kirichok et al., 2004), is responsible for the bulk of Ca²⁺ uptake from microdomains to the mitochondrial matrix (Csordás et al., 2006; de Brito & Scorrano, 2008). In addition to MCU, other mitochondrial Ca²⁺ influx mechanisms such as mitochondrial ryanodine receptor 1 (Beutner et al., 2005) and the rapid mode of Ca²⁺ uptake (RaM) (Buntinas et al., 2001) have also been identified. The extrusion of Ca²⁺ from the matrix to the cytosol is carried out by the Na⁺/Ca²⁺/Li⁺ exchanger (NCLX) (Li et al., 1992) and the Ca²⁺/H⁺ exchanger (Gunter et al., 1991).

The kinetics of mitochondrial Ca²⁺ uptake and extrusion in the beating adult cardiomyocytes is still under debate (De la Fuente & Sheu, 2019). Some authors describe that the amount of Ca²⁺ taking up by mitochondria in each heartbeat is modest and gradually accumulated inside the mitochondrial matrix throughout the heartbeats until a new steady state is reached, in which the uptake and extrusion are balanced (Miyata et al., 1991). Other authors propose that mitochondria can follow the cytosolic Ca²⁺ oscillations and take up and release Ca²⁺ on a beat-to-beat basis (Murgia et al., 2009; Andrienko et al., 2009). It has been proposed that mitochondria may function as a Ca²⁺ buffer due to their capacity in taking up a large amount of Ca²⁺ through MCU, as such they can modulate the amplitude of cytosolic Ca²⁺ transients (Drago et al., 2012). This will require that mitochondria can take up Ca²⁺ on a beat-to-beat basis, which is still unclear presently. The most accepted role for the mitochondrial Ca²⁺ in cardiomyocytes is associated with the regulation of cardiac energy production. The mitochondrial Ca²⁺ regulates the activity of the TCA cycle, as mentioned above (Duchen, 1992; Kohlhaas et al., 2017; De la Fuente & Sheu, 2019). Interestingly, this function has been recently challenged due to the lack of energetic phenotype in the normal beating heart of germline MCU-KO mouse model (Pan et al., 2013). However, under intense β-adrenergic stimulation, it was reported that Ca²⁺ influx through MCU is a requisition for the fight or flight response (Wu et al., 2015).

In pathological conditions, the mitochondrial Ca²⁺ uptake and overload have detrimental consequences on energy production and possibly promote DCM complications (Federico et al., 2017; Wu et al., 2019). Several studies have attempted to monitor Ca²⁺ signaling in isolated cardiomyocytes from different models of MetD. In cardiomyocytes isolated from 5-week STZ-induced diabetic rats, diastolic Ca²⁺ concentration and Ca²⁺ sparks frequency significantly increased in comparison to age-matched control rats. The amplitude of Ca²⁺ transients was significantly decreased and the duration was prolonged (Yaras et al., 2005). In cardiomyocytes isolated from a prediabetic model, increased spontaneous Ca²⁺ oscillations in the cytosol were associated with spontaneous contractions, which were prevented by KN-93, a Ca²⁺ calmodulin kinase II (CaMKII) inhibitor, or the addition of Tempol, a ROS scavenger, to the diet (Sommese et al., 2016). The increases in Ca²⁺ sparks frequency and spontaneous Ca²⁺ transients indicate that these animals are prone to cardiac arrhythmias. In another study, the increment in the frequency of sparks, which also depends on ROS and CaMKII, promoted apoptosis that was linked to increased mitochondrial swelling and decreased mitochondrial membrane potential (Federico et al., 2017).

A small number of studies have measured mitochondrial Ca²⁺ regulation in the cardiomyocytes of MetD animal models. Suarez et al. observed decreased MCU expression, low glucose usage and high FAO in the heart of an STZ mouse model, which led to a decrease in RCR, ATP production and mitochondria membrane depolarization (Suarez et al., 2018). The mitochondrial Ca²⁺ concentration was monitored with a mitochondria-targeted Ca²⁺ probe Pericam and was found to be decreased in diabetic heart. Moreover, when the MCU was restored by AVV9-MCU injection, the mitochondria metabolism was recovered to normal levels (Suarez et al., 2018). Also, mitochondrial Ca²⁺ was found to be decreased in neonatal cardiomyocytes (Suarez et al., 2008) and adult cardiomyocytes exposed to HG (Diaz-Juarez et al., 2016). Besides, Diaz-Juarez et al. showed decreased expression of MCU in HG condition, and recovered MCU expression improved mitochondrial Ca²⁺ handling (Diaz-Juarez et al., 2016), proving that optimal MCU expression reverts the mitochondrial metabolic and functional changes in MetD.

Despite the importance of mitochondrial Ca²⁺ in regulating heart function, very few studies have monitored mitochondrial Ca²⁺ dynamics in MetD hearts (Yaras et al., 2005; Sommese et al., 2016; Federico et al., 2017; Suarez et al., 2018). Several additional mitochondria-targeted Ca²⁺ probes are currently available, such as genetically expressed CEPIA and MitoCam (Lu Xiyuan et al., 2013; Kanemaru et al., 2020), and chemical dyes such as Rhod-2AM (Chen Yun et al., 2012; Fernandez-Sanz et al., 2014), for measuring spatiotemporal aspects of the mitochondrial Ca²⁺ signaling in the heart of MetD models. These measurements will help in our understanding of the role of mitochondrial Ca²⁺ dynamics in MetD mediated heart dysfunction.

3.4. ROS in MetD mediated heart dysfunction

The mitochondrion is the principal organelle involved in ROS production (Jensen, 1966). ROS are free radical oxidants, such as superoxide (O₂⁻) and hydroxyl radical (^●OH), and non-radical oxidants, such as singlet oxygen (¹O₂) and hydrogen peroxide (H₂O₂). ROS generation is tightly related to ATP production, O₂ consumption, and mitochondrial Ca²⁺ signaling (Brookes et al., 2004). A net increase of ROS sustainably will eventually damage the cell (Wang et al., 2008; Zorov et al., 2014; Nickel et al., 2014; Korge et al., 2017). Therefore, both ROS generating and eliminating systems exist in cells. Under physiological conditions, mitochondria produce O₂⁻ from O₂ oxidation by complex I or complex III, which is dismutated to H₂O₂ by manganese-dependent superoxide dismutase (MnSOD). The H₂O₂ will be further eliminated in the mitochondrial matrix by antioxidant systems, glutathione peroxidase and peroxiredoxin (PRX). These systems are coupled with NADPH production. Thus, increased cytosolic ROS can be a cause of increased production and/or decreased elimination (Nickel et al., 2014). Zorov et al., described a process named ROS-induce ROS-release where the ROS produced by a single mitochondrion can be transferred to an adjacent mitochondrion, leading to a chain reaction where many consecutive mitochondrion produces a massive amount of ROS, which cause subsequent cell injury (Zorov et al., 2000, 2014).

In the early stages of MetD, such as prediabetes and hyperglycemia, morphological and functional changes in mitochondria have been associated with an increase in oxidative stress. Koncsos et al. showed an increased H₂O₂ and nitrotyrosine production in DCM (Koncsos et al., 2016). As mentioned above, excessive Ca²⁺ traffic from SR/ER to the mitochondria can causes mitochondrial Ca²⁺ overload that results in ETC uncoupling and excessive ROS production. Koncsos et al. attributed the increase in ROS production to an enhanced SR/ER-mitochondria connection through Mfn2 due to their reported overexpression of the Mfn2 without any changes in other fusion proteins (Koncsos et al., 2016). In a prediabetic model, we found an increase of ROS as well as lipid peroxidation in cardiac homogenates, which promoted CaMKII mediated arrhythmias and apoptosis. These effects could be prevented either by ROS scavenging or by CaMKII inhibition (Sommese et al., 2016; Federico et al., 2017). Phosphorylation of RyR2 by CaMKII enhances the open probability of the channel increasing SR Ca²⁺ leak. Therefore, preventing the RyR2 activation by CaMKII avoids not only mitochondria swelling but also mitochondria membrane depolarization induced by prediabetes. We also reported a decreased distance between SR and mitochondria in the prediabetic heart, which would further augment ROS-mediated CaMKII activation (Federico et al., 2017).

Additionally, in advanced stages of MetD, an increase in ROS, malondialdehyde (MDA), or 4Hydroxy-2-nonenal has been reported, as well as changes in O₂ consumption, MnSOD activity, and/or NADPH oxidase activity (Santos et al., 2003; Csont et al., 2007; Rajesh et al., 2010; Suzuki et al., 2015). Furthermore, increases in mitochondrial superoxide flashes and ROS generation have been described in the STZ model (Ni et al., 2015).

Either increased production or decreased antioxidant capacity can cause net increased ROS levels. Anderson et al. showed human atrial tissue from T2DM had enhanced H₂O₂ production and decreased GSH/GSSG ratio (Ghosh et al., 2005; Anderson et al., 2009). Dabkowski et al. showed the ob/ob model had increased MDA and 4-hydroxyalkenal (both products of oxidation of polyunsaturated FA), coupled with decreased PRX-V (Dabkowski et al., 2010). Accordingly, Shen et al. showed that intensifying the ROS scavenger systems such as overexpression of MnSOD were beneficial in preventing DCM (Shen et al., 2006).

Taken together, ROS imbalance appears to be one of the most damaging factors in cardiometabolic pathologies (Shen et al., 2006; Suzuki et al., 2015). Therefore, antioxidant treatments could be a reasonable approach to deter the harmful ROS effects on the heart (Qin Fuzhong et al., 2012; Fang et al., 2018). Mito-TEMPO, a scavenger of mitochondrial ROS, has been shown to prevent mitochondrial ROS-mediated damages and mitigate the diastolic dysfunction in DCM (Ni et al., 2016). The search for effective candidates, include SOD mimetics, ROS scavengers, among others, to relief oxidative stress in MetD is an ongoing research field (see Kiyuna et al., 2018). However, only limited studies have proven the benefits of the antioxidant approach in humans. Coenzyme Q10 (CoQ10) has been reported to improve cardiac function in patients with DM and HF (Mortensen et al., 2014). CoQ10 is a component of the ETC, mediating the e- transport from Complexes I and II to complex III. Ubiquinol is a reduced form of CoQ10 that acts as an antioxidant inside mitochondria (Kelso et al., 2001). Since optimal ROS concentrations are critical in carrying out the physiological signaling mechanisms, future studies will be needed to identify compounds that will lessen the pathological oxidative stresses while preserving physiological redox signaling.

3.5. Mitochondrial permeability transition pore (mPTP) in MetD mediated heart dysfunction

The mPTP is a non-selective pore in the mitochondria membrane that allows any solute up to 1,5KDa to pass through (Hunter & Haworth, 1979). The mPTP has multi-conductance that suggests the molecular nature is a multi-subunit complex oligomerizing to varying degrees (Hunter et al., 1976). The molecular identity of the mPTP has been studied for years and is still a matter of debate. One of the first models proposed that the mPTP is composed of Bcl-2 associated-X-protein (Bax), VDAC, the peripheral benzodiazepine receptor (TSPO), and Hexokinase II (HKII) in the OMM; mitochondrial creatine kinase (mtCK) in the inter-membrane space; adenine nucleotide transporter (ANT) in the IMM; and mitochondrial cyclophilin D (CypD) bound to ANT in the matrix (Halestrap & Davidson, 1990; Kinnally et al., 1993; Beutner et al., 1997; Marzo et al., 1998). The phosphate carrier model proposed the following composition of mPTP: Bax, VDAC, TSPO, HKII, mtCK, ANT, and CypD, bound to the phosphate inorganic carrier (PiC) (Kokoszka et al., 2004). Recent studies have provided new insights about several potential candidates for the molecular identity of mPTP, which include multiple subtypes of ANT (Bround et al., 2020), F-ATP synthase c-subunit (Mnatsakanyan & Jonas, 2020), and the dimer (tetramer) of F-ATP synthase (Carraro et al., 2020). The idea that more than one protein may act as mPTP offers a rational clarification for current disagreements in the field. Intriguingly, these candidate proteins are already well-known for their role in catalyzing ATP generation. Therefore, mPTPs appear to have two opposite functions: controlling cell life and death through their participation in energy metabolism and apoptosis/necrosis, respectively. The mPTP can be regulated or inhibited by different compounds. The most well known is the cyclosporin A (CsA), which blocks CypD (Fournier et al., 1987). Other inhibitors include Mg²⁺ and Mn²⁺ (by competing off Ca²⁺ to bind), adenine nucleotides, and matrix acidic pH. The primary activator of the mPTP is mitochondrial Ca²⁺overload (Baumgartner et al., 2009), nevertheless, ROS (Seidlmayer et al., 2015), reactive nitrogen species, mitochondrial morphology, and inorganic phosphate can also modulate its activity (Hurst et al., 2017).

The transient opening of the mPTP may have a physiological role by serving as a releasing valve for Ca²⁺ efflux to prevent the mitochondrial Ca²⁺ overload. Under pathological situations, the pore is opened more sustainably which leads to the dissipation of proton force, loss of metabolites, mitochondrial swelling, and cytochrome C-release. All of these processes will eventually lead to cell death (Halestrap, 2009a, 2009b).

Studies using the STZ model have reported that Ca²⁺ retention capacity (CRC) is decreased, which is a sign of increased propensity for mPTP opening, coupled to a decreased O₂ consumption (Oliveira et al., 2003; Ma et al., 2016). Similar results were found when the cardiac myoblast cell line H9c2 was exposed to HG (Diao et al., 2019). The mPTP opening can also be determined by Ca²⁺-induced swelling of isolated cardiac mitochondria. Some studies have shown increased mitochondrial swelling in mitochondria isolated from prediabetic or diabetic animals, demonstrating susceptibility to the Ca²⁺ overload (Federico et al., 2017; Guo et al., 2018). Finally, when Anderson et al. tested the mPTP opening by CRC assays in human atria fibers, they found, once more, an increase in mPTP opening sensitivity. The authors noted that during prolonged metabolic changes and oxidative stress (as happens in DM), the components that can activate the mPTP opening, such as CypD, may be overexpressed (Anderson et al., 2011).

3.6. Mitochondrial biogenesis in MetD mediated heart dysfunction

Mitochondrial biogenesis has been described as a process that includes mitochondrial division and growth. The mitochondrion has its own DNA (mtDNA) which encodes 13 subunits of ETC proteins (Robin & Wong, 1988; Dorn et al., 2015). Initially, the nDNA control mtDNA quality and biogenesis by regulating mtDNA replication. The critical regulator of mitochondrial biogenesis is the PGC1α (Wu et al., 1999; Ventura-Clapier et al., 2008). PGC1α can regulate nuclear factors 1 and 2 (NRF1 and NRF2), as well as transcriptor factor A mitochondrial (TFAM) (Wu et al., 1999). It has been described that when PGC1α level is down, there is a loss of mtDNA, and when it is overexpressed, there is an increase in mitochondrial biogenesis, OXPHOS, FAO, and glycolysis (Lehman et al., 2000; Arany et al., 2005; Lin et al., n.d.). In organs that require high energy for their functional performances, such as skeletal or heart muscle, PGC1α level is higher in comparison with other tissues (Garnier et al., 2003). Multiple mechanisms regulate PGC1α, including epigenetic regulation, post-transcriptional modifications, and post-translational modifications (Duncan et al., 2007; Oka et al., 2020).

When mitochondrial biogenesis declines, heart function is eventually compromised. Bombicino et al. showed in an STZ-diabetic rat model, there was an enhancement in PGC1α expression. The authors proposed that the H₂O₂ and nitric oxide are responsible for the PGC1α activation (Bombicino et al., 2017). Similar results were shown previously by Finck et al., where PPARγ and PGC1α were increased in an STZ model (Finck et al., 2002). In contrast, Yan et al. described decreased PGC1α function in the heart in the ob/ob model, due to acetylation by AMPK, which is activated by dephosphorylation for adiponectin activity (Yan et al., 2013).

In an OVE26 mouse model of T1DM, mRNA levels of TFAM and two mitochondrial encoded proteins were increased, suggesting that mitochondrial biogenesis was augmented (Shen et al., 2004). Using HFD as a model of T2DM, it was found that mtDNA, PGC1α expression, and NRF expression were all decreased (Fang et al., 2018). The same result was confirmed by Duncan et al. in an insulin-resistance model (Duncan et al., 2007).

Even though several studies have measured amounts of PGC1α in the MetD mouse models, the data obtained is still controversial. The different MetD models used may underlie these discrepancies, where some studies detect a decrease in PGC1α while others see an increase. The expression of PGC1α has never been correlated with the PPARγ expression and mtDNA changes in MetD. Therefore, divergent outcomes from these studies show that the regulation of the expression or PGC1α activity is multifaceted. Further studies would be required to fully understand the whole process of mitochondrial biogenesis in MetD (Nisoli Enzo et al., 2007; Ren et al., 2010).

CONCLUSIONS

The prevalence of MetD has been increasing in recent years. Several research groups around the world have focused on determining the mechanisms of disease pathogenesis with a common goal to prevent, delay, or revert MetD. The impact of MetD on the heart is highly relevant to human health due to its propensity to cause life-threatening cardiac arrhythmias and HF.

The heart has extremely high energy demands. Heart mitochondria consume glucose and FA to produce the necessary ATP (Figure 1, left), however, since the usage of glucose is reduced in MetD, FAO by mitochondria is increased. This leads to a shift in the balance of mitochondrial energy metabolism favoring cardiac dysfunction.

Figure 1. Mitochondrial metabolism in the physiological condition vs metabolic disease.

In the physiological condition (left), the cardiomyocytes fuel up mainly with free fatty acid (FA) but also supplement from glycolysis. The FA oxidation (FAO), the tricarboxylic acid (TCA) cycle, the electron transport chain (ETC), and ATP synthase produce the ATP required for muscle contraction. The ATP production is coupled with O₂ consumption and superoxide anion (O₂⁻) and hydrogen peroxide (H₂O₂) production in healthy amounts. Metabolic disease (right) presents elevated free FA and insulin resistance. Therefore, decreased expression of GLUT-4 favors FA uptake and oxidation. FAO requires higher levels of O₂ than glucose to produce an equal amount of ATP, decreasing the efficiency of energy production. Under this condition, the reactive oxygen species (ROS) production increases up to pathological levels, triggering mitochondrial permeability transition pore (mPTP) opening and oxidation of other proteins for normal cell function. The whole process decreases the cardiomyocyte's performance and favors the pathogenesis of diabetic cardiomyopathy that finally leads to heart failure.

In this review, we have summarized what is known about the mitochondrial changes occurring in heart tissue during the progression of MetD. We have described the mechanisms by which metabolic stress causes mitochondrial dysfunction, including disturbances in energetics, changes in dynamics, Ca²⁺ signaling impairment, increased oxidative stress, mPTP opening, and mitochondrial biogenesis (Figure 1, right). The mitochondrial alterations describe before are not exclusive for MetD and DCM. Similar phenotypes have been also seen in the pathogenesis of other heart diseases like hypertension, coronary artery disease, and hypertrophy. As in MetD these mitochondrial alterations can also lead to heart failure in many cases (Graham Delyth et al., 2009; Ardanaz Noelia et al., 2010; Bhatt et al., 2011; Hollander et al., 2014; Ait-Aissa et al., 2019). However, the origin of mitochondria dysfunction is different and pivotal to design therapeutic strategies, i.e. for MetD the fuel for ATP production is the initial mitochondria signal to disturb its function. A confluence point in several heart pathologies in which mitochondria are compromised, is the increased ROS generation, the most important component for mitochondria dysfunction. As briefly examples, Graham et al. showed that inhibition of ROS production by mitoQ prevents hypertension development (Graham Delyth et al., 2009). Similarly, the use of resveratrol avoids NO generation and hypertension (Bhatt et al., 2011). Finally, in coronary arterial disease, altered mitochondrial dynamics due to alteration in the DRP1 levels as well as alterations in ETC complexes activity have been reported. Therefore general mitochondrial function and metabolism are affected in coronary arterial disease(Ait-Aissa et al., 2019).

There are still many questions and challenges that need to be addressed to further understand the role of mitochondria in MetD mediated heart dysfunction. For instance, there is a lack of reliable direct measurements of mitochondrial functional parameters such as ATP, fission and fusion events, Ca²⁺ and ROS concentrations in live cardiomyocytes isolated from the DCM heart to elucidate the dynamic and specific role of these parameters in the pathogenesis of MetD mediated heart dysfunction. These measurements will also provide information about the mechanisms of crosstalk signalings among these interconnected functions such as the role of mitochondrial Ca²⁺ signaling in the regulation of mitochondrial energy metabolism and ROS homeostasis. Future advancements in experimental technologies and theoretical concepts will help to resolve these challenges. One such progress is the recent discovery of multiple molecular identities in the formation of mPTP complex, which provides a new opportunity for using genetic approaches to decipher the role mPTP opening in MetD linked cell injury and death.

All of the MetD animal models and heart samples from MetD patients show increased ROS production. Thus, ROS appear to be a key player in the development of the MetD mediated heart dysfunction, although the mechanisms of ROS-mediated downstream effect are still a field to be elucidated. It is plausible that the ROS-mediated signaling pathways are potential drug targets for the treatments of MetD mediated heart dysfunction.

Findings of MetD-related changes in heart mitochondrial biogenesis are still contradictory. While some studies show an increase in PGC1α, others show no changes in this protein. Furthermore, other components involved in mitochondrial biogenesis have been barely studied in DCM. Therefore, additional experiments are required to elucidate the specific role of other mitochondrial biogenesis regulators on MetD.

Finally, the comparative studies between animal models and the heart samples from human patients (including induced pluripotent stem cells differentiated into cardiomyocytes) will be useful for translating the basic mechanisms into clinical practices. Collectively, this new knowledge will be useful for the development of new and effective therapeutic interventions for treating these devastating disorders.

Abbreviations

AGEs

advanced glycation ends-products

ANT

adenine nucleotide transporter

Bax

Bcl-2 associated-x-protein

CaMKII

Ca²⁺-calmodulin kinase II

CoQ10

coenzyme Q10

CsA

cyclosporine A

CVD

cardiovascular disease

CypD

cyclophilin D

DCM

diabetic cardiomyopathy

DM

diabetes mellitus

DRP1

dynamin-related protein 1

ECB

excitation-contraction-bioenergetics

ECC

excitation-contraction coupling

ETC

electron transport chain

FA

fatty acids

FAO

fatty acids oxidation

FRD

fructose-rich diet

GD

gestational diabetes

GLUT-4

glucose transporter type 4

GSH

glutathione

H₂O₂

hydrogen peroxide

HF

heart failure

HFD

high fat diet

HG

high glucose

HKII

hexokinase II

IMF

intermyofibrillar

IMM

inner mitochondrial membrane

MCU

mitochondrial Ca²⁺ uniporter

MDA

malondialdehyde

MetD

metabolic disease

MetS

metabolic syndrome

Mfn1/2

mitofusin 1 and/or 2

MnSOD

manganese superoxide dismutase

mPTP

mitochondrial permeability transition pore

mtCK

mitochondrial creatine kinase

NRF 1/2

nuclear factors 1 and 2

O₂⁻

superoxide

OMM

outer mitochondrial membrane

ONOO−

peroxynitrite

OPA1

optic protein atrophic 1

OXPHOS

oxidative phosphorylation

PGC1α

peroxisome proliferator-activated receptor gamma (PPARγ) coactivator 1 α

PiC

phosphate inorganic carrier

PRX

peroxiredoxin

RCR

respiratory control ratio

ROS

reactive oxygen species

RyR2

ryanodine receptor type 2

SERCA2a

sarco/endoplasmic reticulum

ATPase

2a

SR

sarcoplasmic reticulum

SSM

subsarcolemmal

STZ

streptozocin

T1DM

type 1 of diabetes mellitus

T2DM

type 2 of diabetes mellitus

TCA

tricarboxylic acid

TFAM

transcription factor A mitochondrial

TSPO

mitochondrial benzodiazepine receptor

VDAC

voltage-dependent anion channel