Right ventricular dysfunction (RVD) is an independent determinant of mortality in several cardiovascular disorders of right ventricular (RV) pressure or volume overload such as pulmonary arterial hypertension (PAH), congenital heart disease, left heart failure (LHF), or tricuspid valve regurgitation (TR) ^[1]. Whereas the current understanding of and treatment options for LHF have continuously improved during recent years, no therapeutic options are available in today’s clinical routine that specifically target the right heart [2,3,4]^[2][3][4]. This implies that significant interventricular differences exist, which are responsible for unsuccessful pharmacological treatment. Besides differences in morphology and hemodynamics, distinct cellular and molecular processes and their regulation under pathophysiological conditions may significantly account for individual maladaptive mechanisms of the right and left ventricles [3,5]^[3][5]. Whereas the principal pathogenic processes in right heart failure (RHF) are the same as in LHF, their importance and order may be diverse. Importantly, structural remodeling may play a minor role in RHF compared to LHF [6,7,8]^[6][7][8].

Given that the heart is the most metabolically active organ in the body, it is particularly reliant on mitochondrial integrity. Thus, mitochondrial dysfunction is suggested to play a prominent role in heart failure.

2. The Powerhouse of the Heart

Mitochondria are oval organelles around 0.7 to 1 µm in size containing a double membrane structure with transversally oriented cristae formed by infolding of the inner membrane. They are tightly connected and form chain-like structures assembled in mitochondrial clusters. In the cardiomyocyte, mitochondria can be perinuclear, interfibrillar, and subsarcolemmal and possess different proteomic and physiological statuses [9,10]^[9][10]. Although mitochondria exhibit a dynamic characteristic, they are the source of energy generating the majority of the adenosine triphosphate (ATP) in the heart. The perpetual energy demand in the heart requires a mitochondrial volume of 23 to 32% of the total cardiomyocyte volume. The cardiac mitochondrial volume increases continuously from humans and dogs to rats, hamsters, and mice. Thus, the increased mitochondrial density correlates well with the increased heart rate and oxygen consumption in smaller animals ^[11]. In mitochondria, energy is generated via oxidative phosphorylation (OXPHOS). This complex process can be described as a sequential passage of electrons from negative (nicotinamide adenine dinucleotide, NADH, and flavin adenine dinucleotide, FADH₂) to positive (molecular oxygen O₂) redox potentials down the electron transport chain (ETC) in aerobic respiration. OXPHOS normalized to the number of mitochondria reflects the generation of energy in terms of ATP and is defined as a mitochondrial function. The ETC consists of four protein complexes located at the inner mitochondrial membrane. The main energy conversion reactions take place at complex I (NADH dehydrogenase), complex III (cytochrome c reductase), and complex IV (cytochrome c oxidase), which use the energy of the stepwise electron transfer for active pumping of hydrogen ions (H⁺) out of the mitochondrial matrix into the intermembrane space. Thereby, an electrochemical gradient is established across the inner mitochondrial membrane. The translocation of H⁺ back into the mitochondrial matrix through the F₁F_O ATPase (ATP synthase complex) is coupled to the phosphorylation of ADP to generate ATP. The reduced coenzymes NADH and FADH₂, the electron donors for the ETC, are produced in the tricarboxylic acid (TCA) cycle within the mitochondrial matrix. One enzyme of the TCA cycle, succinate dehydrogenase or complex II, is part of the ETC located in the inner mitochondrial membrane. Mitochondrial function can be determined by respirometry, which measures the O₂ consumption rate in permeabilized muscle fibers, cardiac tissue samples, or isolated mitochondria after the addition of saturating concentrations of metabolic substrates and ADP. Of note, even by using living cells, the information is limited to the general function of the ETC protein complexes, their connectivity and coupling to the ATPase, and the lack of data on the availability of NADH, FADH₂, ADP, or O₂.

3. Redox-Optimized ROS Balance in the Heart

The ETC releases the superoxide radical (O₂^•−), a reactive oxygen species (ROS), even under physiological conditions, given that a small percentage of electrons prematurely reduce O₂ at complex I and complex III. This production of mitochondrial ROS (mitoROS) is balanced by an antioxidant system located in the mitochondrial matrix ^[12]. The manganese-dependent superoxide dismutase (mitochondrial isoform: SOD2) converts O₂^•− to O₂ and hydrogen peroxide (H₂O₂), which is then further reduced to water by glutathione peroxidase (GPX) and peroxiredoxin (mitochondrial isoforms: PRX3 and PRX5). For the regeneration of GPX and PRX, the pyridine nucleotide NADPH is needed, whose reduction in mitochondria depends on TCA cycle activity. Therefore, TCA cycle enzyme integrity is essential for ATP production and the regeneration of antioxidative pathways. Pathological conditions can lead to dysfunction of the TCA cycle and OXPHOS in mitochondria, which effects ATP production and the generation of mitoROS due to an altered redox balance (NAD⁺/NADH, FAD/FADH_2, NADP⁺/NADPH ratios). Under physiological conditions, the emission of mitoROS in the ETC is low, allowing the antioxidant system to balance ROS levels sufficiently with adequate NADPH availability for the regeneration of the antioxidant enzyme systems. However, under pathophysiological conditions, the ETC becomes uncoupled, resulting in a decreased membrane potential and less ATP generation. An extreme oxidized state, probably due to altered metabolic flows and less TCA cycle activity and therefore lower restoration of reduced coenzymes NADH and FADH₂, depletes H₂O₂ scavenging capacity and the net ROS emission is enhanced despite eventually lower ROS production (Figure 1). In addition to the generation of O₂^•− or H₂O₂ in mitochondria, there are other intracellular sources of ROS production: monoamine oxidase (MAO), xanthine oxidase (XO), uncoupled nitric oxide synthase (NOS) and NADPH oxidase (NOX). MAOs are located in the outer mitochondrial membrane deputed to inactivate neurotransmitters, like serotonin, thereby producing H₂O₂ ^[13]. Endothelial NOS generates nitric oxide (NO) in the vascular endothelium to regulate vascular tone but is also expressed in cardiomyocytes. Uncoupled dysfunctional NOS produces O₂^•−, contributing to increased ROS ^[14]. The NOX isoforms are membrane-bound, ROS-producing enzymes. While most NOX isoforms produce O₂^•−, NOX2 and NOX4, both expressed in cardiomyocytes, generate H₂O₂. During the discovery of NOX, the physiological role of ROS was highlighted, taking into account that many redox-dependent biological processes are fine-tuned by ROS production and antioxidant enzyme systems: SOD, GPX, PRX, coupled NOS, and catalase (CAT) ^[15]. An unbalanced action of ROS, which is not buffered by the antioxidant system, can result in DNA, protein, and lipid modifications, a process which is defined as oxidative stress (Figure 1). DNA exposed to ROS can be modified to guanine, allowing it to pair with cytosine and adenine leading to double-strand breaks and genetic instability. In general, this can affect nuclear and mitochondrial DNA (mtDNA). However, due to the circular nature of mtDNA and cell-cycle-independent replication as well as the high copy number, mtDNA is more prone to oxidative damage. mtDNA fragmentation can be detected by quantitative real time PCR via assessment of the ratio of stable amplified long and shorter fragments of mtDNA ^[16]. Oxidative DNA modification, such as 8-hydroxy-2′-deoxyguanosin (8-OHdG), is an accepted marker of ROS and can be measured by various standard techniques such as immunohistochemistry or high-pressure liquid chromatography (HPLC). Oxidative post-translational modifications of proteins are often associated with functional inactivation or even damage of the affected proteins. These include protein carbonylation, detected by 1,3-dinitrophenylhydrazine (DNPH)-based Oxyblots ^[17], dityrosine, nitrotyrosine, and nitrocysteine formation, measured with specific antibodies or mass spectrometry, oxidation of thiol residues such as glutathionylation, and the formation of disulfides or sulfenic, sulfinic, and sulfonic acids, which can be detected by native gel electrophoresis and specific antibodies. For the quantitative detection of oxidative protein modifications in excised myocardial tissue, it needs to be considered that some oxidative protein modifications are reversible, like disulfid formation, glutathionylation, and sulfenic acid formation, and might be reduced during sample preparation processes. Likewise, ex vivo oxidative modifications can develop during tissue processing and will result in false positive results. Nitrotyrosine and dityrosine formation, carbonylation, and the formation of sulfinic and sulfonic acids from thiol groups of proteins are irreversible. In addition, there are several mass-spectrometry-based techniques available to specifically detect oxidative modification of proteins in comprehensive screening approaches [18,19]^[18][19]. Furthermore, the detection of lipid peroxidations, like 4-hydroxynonenal (4-HNE) or malondialdehyde (MDA), is a widely used method to indirectly detect increased ROS formation in tissues. Of note, all the explained methods detect modifications, which occur as a consequence of increased cellular ROS. Direct measurements of O₂^•− and H₂O₂ are possible in tissue or isolated mitochondria by using, for example, fluorescence-dye-based techniques. However, it needs to be considered that mitoROS release in excised cardiac tissue does not absolutely reflect the in vivo situation. To conclude, ROS occur naturally in the myocardium and are important signal mediators. In contrast, oxidative stress impairs cellular function and is associated with pathological conditions.

4. Mitochondrial Integrity Is Impaired in Right Heart Failure

In order to assess the significance of mitochondrial function in the development of RHF, the experimental model system needs to be considered critically. As RHF is a leading cause of mortality in PAH patients, many experimental studies have used rodent models of PAH, induced either by injection of the vascular endothelial growth factor (VEGF) inhibitor, Sugen 5416, followed by hypoxia (SuHx) or the injection of monocrotaline (MCT), a pneumotoxic pyrrolizidine alkaloid. Both model systems are characterized by pulmonary arterial (PA) remodeling and constriction with increased PA pressure and pulmonary vascular resistance leading to compensated RV hypertrophy and eventually RHF. In contrast to the PAH rodent model systems, the model of PA banding (PAB) is widely used to study RV remodeling independently of pulmonary vascular disease. Of note, different mouse wild-type strains might react differently to RV pressure overload, either developing compensated RV dysfunction or failure ^[6], which was also observed by Shults et al. [29]^[20] in SuHx Sprague Dawley and Fisher rats. It might be obvious that the degree of afterload triggered by the degree of PA stenosis in PAB rodent models is associated with the extent of RV dysfunction and structural remodeling [6,37]^[6][21]. However, to assess the molecular mechanisms mediating the transition of compensated RV hypertrophy (RVH) and RVD to RHF, it is essential to determine the signs of RHF in experimental rodent models. Signs of congestion reliably reflect the presence of RHF, as presented in experimental studies, i.e., liver weight [31^{[22][23][24][25]},32,33,34], macroscopic signs of a nutmeg-colored liver ^[6], and ascites. In addition, impaired exercise capacity [34]^[25] and/or mortality [6,24,29]^[6][20][26] can reflect RHF. A quantitative parameter to determine signs of RHF might be the morphometric assessment of congestion-modified areas in haematoxylin/eosin-stained liver sections. The extent of hepatic congestion showed a strong positive correlation with right atrial (RA) area and a strong negative correlation with systolic RV function assessed by echocardiography in PAB mice ^[6]. To understand RHF, it is important to consider that already under physiological conditions, the RV differs from the left ventricle (LV) in loading conditions and morphology and displays fundamental differences on the cellular and molecular levels. There is accumulating evidence that the RV is more susceptible to oxidative stress probably because of a less adaptable antioxidant enzyme system. In healthy rats, Nagendran et al. [38]^[27] measured a lower mitochondrial membrane potential in RV cardiomyocytes and RV tissue compared to the LV, which might provide evidence for a lower activity of the ETC complexes in RV compared to LV tissue. Furthermore, the membrane potential rose with disease severity in a mouse model of PAH induced by MCT [38]^[27], indicating adaptation of RV cardiomyocytes to increasing workloads. Interestingly, Phillips et al. detected an identical expression pattern and relative amount of proteins, the same amount of post-translational protein modifications, and an identical amount of mitochondria in RV and LV tissue of healthy rabbits and pigs [39]^[28]. Given the lower oxygen consumption and lower ATP generation rate [40,41]^[29][30] of the resting RV, the RV should hold larger metabolic reserve capacities than the LV. This should imply lower mitoROS generation and a less vulnerable RV under an increasing workload. But in contrast to the LV showing enhanced antioxidative capacity under the progression of pathological conditions [22^[31][32][33],23,36], the antioxidant enzyme system is unchanged or is even decreased in the RV myocardium of mice after RV pressure overload [6^[6][25],34], rats with chronic NO deficiency [36]^[33], and in human end-stage HF patients secondary to ischemic HF or idiopathic dilated cardiomyopathy [21,22,23]^[31][32][34]. To detect the status of the antioxidative system in the RV, several preclinical studies have measured SOD expression on mRNA levels [6,34,36]^[6][25][33] with differentiation between Sod1, an isoform mainly expressed in the cytosol, and Sod2, the mitochondrial isoform. In studies using human RV myocardium, CAT activity was determined [22,23]^[31][32] in the tissue of end-stage LHF patients. Progressive destruction of the mitochondrial network, which may result in impaired formation of ATP due to decreased OXPHOS under aerobic conditions, is associated with the development of RHF (Figure 1). A decreased amount of mitochondria has been shown in experimental PAH [25,32,42]^[23][35][36] and RV pressure overload [6,34]^[6][25] by detecting decreased activity of citrate synthetase (CS), the pace-maker enzyme of the TCA cycle in mitochondria, or decreased amounts of mtDNA. Experimental data could be confirmed in RV tissue samples of congenital heart disease patients divided into two groups of compensated RVH and RHF based on postoperative cardiac magnetic resonance imaging (MRI) [43]^[37]. While CS and succinate dehydrogenase activities were maintained in RVH, they were significantly decreased in RHF. The mtDNA copy number progressively decreased during the transition from RVH to RHF and inversely correlated with RV systolic pressure [43]^[37]. In contrast with reduced mitochondrial biogenesis, reflected by decreased Pparagc1a (PGC1α) mRNA levels, reported in SuHx rats [25]^[35], Karamanlidis et al. showed significantly increased Pparagc1a (PGC1α) mRNA expression in human RVH and again normalized mRNA levels in human RHF [43]^[37]. This might indicate mitochondrial biogenesis as a highly dynamic, compensatory effect during the progression of RVH/RVD to RHF. The mitochondrial architecture is also influenced by impaired mitochondrial fission and fusion (Figure 1), a process by which mitochondria divide or fuse to maintain their integrity and function [44]^[38]. Of note, increased mitochondrial fission as well as enhanced fusion are reported in experimental PAH [31,45,46,47]^{[22][39][40][41]}. The fission/fusion ratio and thereby mitochondrial biogenesis might also depend on disease severity, as Hwang et al. [34]^[25] showed significantly increased mitochondrial fusion in compensated RVH upon RV pressure overload by protein expression of mitochondrial dynamin-like GTPase (OPA1). Both OPA1 isoforms were significantly decreased in RHF upon RV pressure overload, while the mitochondrial fission protein, dynamin-related protein (DRP1), was significantly increased in isolated RV mitochondria in RHF upon PAB [34]^[25]. The same expression pattern, which ultimately leads to an impaired fission/fusion ratio in RVH, was also shown by Wüst et al. [42]^[36] in MCT rats. After injection of low-dose MCT resulting in compensated RVH, increased fusion and decreased fission were observed, while rats injected with a higher MCT dosage suffered from RHF and showed disrupted fission as well as fusion [42]^[36]. In accordance, data from transmission electron microscopy (TEM) of RV tissue showed an elevated mitochondrial area and increased mitochondrial size and higher amounts of mitochondrial clusters with a decreased number of mitochondria per cluster in the RV of SuHx rats at early time points. In contrast, at later time points, the mitochondrial area significantly decreased and mitochondrial integrity was disrupted, which was associated with increased ROS generation and reduced energy production in the RV tissue of SuHx rats [29]^[20]. An impaired mitochondrial ultrastructure was also shown in other PAH rodent models [25^[23][35][36],32,42], models of RV pressure overload [25,34]^[25][35], and in human RV tissue of congenital heart disease patients [43]^[37].