CHAPTER 6: MOLECULAR AND CELLULAR BIOLOGY OF THE HEART

Biology of the heart is meticulously orchestrated by various mechanisms, including receptor-mediated signal transduction, post-translational modification, protein degradation, and transcriptional regulation. To maintain the hemodynamic homeostasis, the heart uses these processes in a chronologically regulated manner. Because the time between heart beats is usually less than one second, instant changes in the body blood demand must be rapidly adapted. This is achieved by modifying β-adrenergic signaling and subsequent phosphorylation/dephosphorylation of proteins related to excitation–contraction coupling. In contrast, chronically altered hemodynamic status, such as hypertension, is adapted via network of signal transductions leading to reprogramming of the gene expression. Emerging evidence indicates that microRNAs, long noncoding RNAs, or circular RNAs are also critically involved in these transcriptional regulations. Sustained activation of certain pathways initially induces adaptive responses but eventually become maladaptive, causing pathological hypertrophy, contractile dysfunction, cell loss, and extracellular remodeling. These changes are clinically represented as heart failure and predispose patients to increased morbidity and mortality. With the rapid advances in the development of small molecule inhibitors and gene therapy approaches that can possibly target key pathological pathways in vivo, we are approaching the point of being able to readily manipulate these pathways in patients. Noncoding RNAs not only play crucial roles in orchestrating the genome and cellular pathways but also can be used as novel therapeutic targets and diagnostic markers. Therefore, understanding the molecular mechanisms governing cardiac pathophysiology is essential to understanding the signaling networks that regulate heart failure development. In this chapter, we will highlight the role of important regulators and biological molecular principles in cardiac homeostasis and disease.

Activation of cardiac β-adrenergic receptors (ARs) represents a powerful system to increase cardiac contractility and heart rate. ARs are a family of G protein–coupled receptors, and agonist binding to β-ARs induces dissociation of heterotrimeric G protein αβγ into active Gα and free heterodimer Gβγ. These subunits activate their respective downstream signal transductions.¹ Switching off the activated β-adrenergic signaling is mediated by phosphorylation of β-ARs through G protein–coupled receptor kinases (GRKs)¹ (Fig. 6–1).

Among the nine members of ARs—three α₁, three α₂, and three β (β₁, β₂, and β₃)—the mammalian heart expresses all three β-AR subtypes.² In the healthy heart in most species, the majority (ie, 60%–80%) of receptors is of the β₁-subtype, whereas the β₂-subtype accounts for a minor fraction of total β-ARs. A third β-AR subtype, the β₃-AR, was initially thought to be limited to adipose tissue, but was later detected also in the heart. This subtype has recently attracted attention for its potential cardioprotective effects in the heart failure setting,³ and its role and mechanism of action are actively investigated in experimental and clinical studies.⁴

β₁- and β₂-ARs are potent stimulators of cardiac contraction and relaxation in the human heart.^2,5 As direct effectors of the sympathetic nervous system, they serve to rapidly adapt cardiac performance to an increased hemodynamic demand. Both β₁ and β₂ receptor subtypes couple to the stimulatory G protein (G_s), thereby activating adenylyl cyclase and leading to conversion of ATP to cAMP. The formation of the second messenger cAMP then leads to activation of protein kinase A (PKA), which phosphorylates several key regulators of the cardiac excitation–contraction machinery. This includes phospholamban, the L-type Ca²⁺-channel, the ryanodine receptor, troponin T and I, myosin-binding protein C, and the small protein phosphatase inhibitor-1.⁵ These events lead to rapid (within seconds) changes of the cardiomyocyte Ca²⁺ transient and enhanced myofilament sensitivity for calcium, resulting in potent inotropic effects.

Historically, β-adrenergic agonism was long taken as a surrogate parameter for beneficial effects in heart failure in the light of short-term experimental studies improving cardiac contractility. Emergence of transgenic mice models of the β-adrenergic system contributed enormously to reveal the detrimental effects of chronic β-AR stimulation and the beneficial effects of receptor blockade during activated sympathetic nervous system. β₁-AR transgenic mice display cardiomyocyte hypertrophy, followed by fibrosis and finally heart failure.⁶ Enhanced apoptosis, impaired calcium transients in cardiomyocytes, together with altered expression pattern of sarcoplasmic reticulum (SR) proteins were found in these animals.⁷ These data indicate that the β₁-AR system, which is ideally suited to provide short-term increases in cardiac contraction, causes marked structural and functional damage to the heart after extended periods of either receptor stimulation or overexpression.

Transgenic mice with ventricular overexpression of the human β₂-AR have led to a multitude of new insights about β₂-adrenergic signaling in the heart. In the initial report by Milano et al, a mouse model with more than a 1000-fold overexpression led to marked enhancement of basal cardiac contractility, which could not be further augmented by β-agonist treatment.⁸ This dramatic level of receptor density remarkably did not result in overt cardiac pathology⁹; thus β₂-AR gene therapy was proposed to enhance myocardial function of failing hearts. However, overexpression of the human β₂-AR in disease models complicated the interpretation. High-density overexpression of the β₂-AR worsened cardiac function after aortic banding, whereas it rescued some of the consequences of myocardial infarction.¹⁰ A study comparing different levels of β₂-AR overexpression on cardiac structure and function in transgenic mice showed that moderate (ie, 50-fold) overexpression was well tolerated, whereas very high densities of transgenic β₂-AR receptors (350-fold) induced cardiac pathology.¹¹ Thus, for the purpose of enhancing cardiac function by β₂-ARs, there seems to exist an optimum level of β₂-AR overexpression, with higher levels being detrimental.

Cardiac overexpression of the β₁- and the β₂-AR exhibits remarkable differences between the two β-receptor subtypes. In contrast to predominant coupling of β₁-AR to G_s, the dual sequential coupling of the β₂-AR to G_s and G_i (inhibitory) proteins has been reported (first G_s and then G_i).¹² In addition to coupling to G_i proteins, several “alternative” signaling pathways have been described to be activated directly by the β₂-AR, including a mitogen-activated protein kinase (MAPK) pathway and a phosphatidylinositol-3-kinase (PI3-K) pathway.^12,13 Additionally, distinct temporal and spatial compartmentation of the classical cAMP pathway may play a role in AR subtype functional differences.¹⁴

One of the key mechanisms that drives heart failure development on chronic stimulation of cardiomyocyte β-ARs is apoptosis.^2,5 Induction of cardiomyocyte apoptosis was found after stimulation of isolated rat cardiomyocytes with high doses of isoproterenol.¹⁵ In this follow-up study, it was demonstrated that β₁- and the β₂-AR subtypes present opposing effects on cardiomyocyte apoptosis, with the β₁-subtype being proapoptotic and the β₂-subtype mediating antiapoptotic effects.¹⁶ These results have been confirmed by several groups, including the studies of Xiao et al,¹⁷ who isolated adult cardiomyocytes from β₁-/β₂-receptor double knockout mice and reintroduced the individual receptor subtypes by adenoviral gene transfer, resulting in “pure” β₁- and β₂-receptor expressing myocytes. Several responsible intracellular signaling pathways for antiapoptotic effects mediated by the β₂-receptor have been proposed, including Akt-kinase via G_i, reactive oxygen species, and PKA-independent activation of Ca²⁺/calmodulin-dependent protein kinase (CaMK). Interestingly, a recent study suggests that there exists cardioprotective pathway in β₁-adrenergic signaling, and selective inhibition of proapoptotic pathway may represent an alternative therapeutic approach.¹⁸

Changes of β-Adrenergic Signal Transduction in Heart Failure

Desensitization of β-adrenergic signal transduction in heart failure was first discovered by Bristow et al.¹⁹ Central findings to explain this phenomenon of reduced responsiveness to β-adrenergic stimulation were a reduction of β1-AR density in failing human myocardium,¹⁹ a decrease of norepinephrine reuptake, subsequently a decrease in cardiac norepinephrine stores, and finally an increased G_i protein expression and GRK2-activity.¹ Although initially considered detrimental, the observed desensitization of β-ARs is now realized as an adaptation process to the highly increased levels of catecholamines in heart failure. Thus, the β-adrenergic signal transduction cascade is operative but with its sensitivity adjusted to high catecholamine levels, to minimize the detrimental effects of chronic stimulation of the myocardium.²

The desensitization pattern of the two main cardiac β-AR-subtypes, β₁ and β₂, is also different. Although the β₁-subtype shows prominent downregulation (by about 50%), the β₂-subtype is nearly unaffected in its density.²⁰ The selective loss of the β₁-AR-subtype in human heart failure leads to a β₁:β₂ ratio of approximately 1:1, thus enhancing the relative contribution of the β₂-AR-subtype. It is to be taken into account, however, that the relevant ligand in vivo, norepinephrine, which is released by sympathetic nerve endings and the adrenal gland, is significantly β₁-selective (about 10- to 20-fold), and norepinephrine levels are enhanced in many heart failure patients.⁵ Thus, the overall signal will still be dominated by the β₁-subtype. Recently, altered spatial distribution pattern of β₂-AR has been reported, and abnormal cAMP compartmentation was suggested as a potential contributor to the heart failure phenotype.²¹

After the first description of elevated GRK activity and GRK2 (synonymous for βAR kinase1 [βARK1]) expression in heart failure, numerous studies have confirmed elevated GRK2 activities in heart failure.¹ Elevated GRK activity is often regarded as a major factor contributing to the progression of heart failure through desensitization of the β-adrenergic signal transduction pathway. Evidence for this argument comes mainly from studies using a C-terminal peptide of GRK2 (termed βARKct) that not only has been demonstrated to effectively block desensitization of the β-AR in vitro²² but that also proved to be an effective therapeutic agent.²³ The βγ-scavenging properties of the βARKct-peptide result in reduced translocation of GRK2 to the membrane, thereby reducing receptor-phosphorylation.²² This inhibition of β-AR desensitization is regarded as the protective mechanism, through increasing inotropy and preventing the internalization and coupling to “alternative” signaling mechanisms such as the activation of src-dependent pathways.²⁴ Despite the deregulated GRK2 expression in heart failure, it is interesting to find that GRK2-transgenic mice do not show obvious signs of overt cardiac pathology or heart failure.²⁵ This suggests that, although it is evident that GRK2 is critically involved in heart failure pathogenesis, GRK2 may not be the primary promoter of heart failure but rather collaborates with other signals that can be modified by interfering Gβγ-related pathways.

β-Blockers as THERAPEUTICS in Heart Failure

β-AR blockade is now regarded as the single most effective therapy in heart failure treatment.^2,26 Despite the discovery of antagonists for the β-AR more than half a century ago, it took a long time for the β-blockers to be introduced into the clinical treatment of heart failure. After several large clinical trials, a marked benefit of this therapeutic approach in heart failure therapy was realized.^27,28 Because angiotensin-converting enzyme (ACE) inhibitors had been already introduced into heart failure therapy, the effect of β-blocker treatment in heart failure had to be tested not versus placebo but versus the patients already treated with established treatment regime, including ACE inhibitors. Nevertheless, β-blockade led to impressive reductions of mortality (ranging from 35% to 60%^27,28,29), with larger effects than ACE inhibitors tested versus placebo. In contrast to ACE inhibitors, β-blockers not only reverse remodeling, but also increase systolic function after several months of treatment.⁵

Several large clinical trials with carvedilol, metoprolol, and bisoprolol have demonstrated a significant benefit in large placebo-controlled trials.^27,28,29 On the contrary, two β-blockers (xamoterol and bucindolol) with intrinsic sympathomimetic activity as a result of partial β-AR agonism have failed to reduce mortality or even increased mortality.^30,31 These drugs can increase the peripheral vascular resistance and may offset the beneficial effects; thus, they are currently not recommended for use in patients with heart failure.

Carvedilol, as an unselective β-blocker with multiple other pharmacological effects, has been compared to the β₁-selective metoprolol in a landmark clinical trial, COMET.³² Carvedilol showed a 6% higher reduction in overall mortality (17% higher relative to metoprolol).²⁸ In addition, carvedilol led to a significantly greater reduction of heart rate (short-term) and blood pressure (long-term) than did metoprolol—likely caused by the α1-AR blockade.²⁶

Mutations in Adrenergic Receptor Genes Leading to Altered Signaling Properties

In addition to the general principles of receptor regulation and dysregulation in health and disease discussed above, there is increasing evidence that genetic polymorphisms of key proteins of the β-adrenergic signaling cascade can significantly influence the signaling properties in individual patients. Both the α₂-ARs, as the principal presynaptic regulators of norepinephrine-release, and the β-ARs, as the principal effectors, have been extensively studied, and numerous polymorphisms with potential significant clinical impact have been identified (Table 6–1).^33,34 Although the β₁-AR variants were significantly associated with worse clinical outcomes, that of β₂-AR variants seems less apparent. Among the presynaptic α₂-receptors, a deletion mutant of the α_2C-subtype has been shown to be significantly associated both with enhanced norepinephrine release and the risk of developing heart failure.³⁴ Thus, the clinical data from these variants of the β-AR signal transduction system are consistent with the experimental evidence that chronic β₁-adrenergic signaling is detrimental and β₂-AR signaling is protective. With the emerging gene sequencing technologies, it is reasonable to expect that our picture of genetic variations in receptors and other key proteins of the β-adrenergic signaling cascade will soon be refined, with more variants to be discovered and more reliable clinical data to be obtained.

In the mammalian heart, depolarization of the cell membrane leads to the opening of voltage-gated L-type Ca²⁺ channels, located in the T-tubular regions of the myocytes, allowing the trans-sarcolemmal influx of Ca²⁺ into the cell³⁵ (Fig. 6–2). T-tubules invaginate the plasma membrane, rendering L-type Ca²⁺ channels in close proximity to the Ca²⁺ release channels, known as ryanodine receptors (RyRs). RyRs are located on the SR, the major Ca²⁺ storing organelles in the cardiomyocyte. Ca²⁺ entering the cells through a single L-type Ca²⁺ channel triggers the opening of one or a cluster of RyRs, inducing the local release of Ca²⁺ from the SR.^35,36 During membrane depolarization, a large number of L-type Ca²⁺ channels are opened, triggering a large release of Ca²⁺ from the SR, raising cytosolic Ca²⁺ from 0.1 to 0.2 μM to 2 to 10 μM, namely the Ca²⁺ spark.³⁵ The distance of the cleft between the cell plasma membrane, where L-type Ca²⁺ channel exists, and the SR membrane, which is roughly 12 nm in normal cardiomyocytes, is very critical for such a coupling and signal transduction. Extension of the distance between them by pulling the plasma membrane impairs the signaling reliability.³⁷ The activity of individual RyRs and the number of the RyRs recruited during a spark play an important role in the spark morphology. The greater the activity of the RyRs, the greater the release of Ca²⁺ and, as a consequence, an increased Ca²⁺ spark size is observed. SR Ca²⁺ content also regulates Ca²⁺ sparks, because stored Ca²⁺ plays a critical role in regulating RyR by increasing the activity of the RyRs and by sensitizing the threshold of the RyR for activation to the stimulus.

The release of Ca²⁺ into the myofibrillar space, in turn, results in cross-bridge formation and contraction^35,38 (Fig. 6–3). The force–frequency relationship also contributes to the regulation of contractile strength in the heart. In a normal heart, increasing heart rate results in enhanced trans-sarcolemmal Ca²⁺ influx secondary to high-frequency–induced recruitment of Ca²⁺ currents, resulting in more Ca²⁺ release from the SR and stronger contraction. In addition, length-dependent activation of the myofilaments, known as the Frank-Starling mechanism, serves as a largely Ca²⁺-independent mechanism for the short-term regulation of the strength of contraction.

Relaxation occurs when calcium detaches from troponin C and is sequestered back into the SR by the cardiac isoform of the Ca²⁺ ATPase pump (sarcoplasmic reticulum Ca²⁺-ATPase [SERCA2a]) or extruded extracellularly by the sarcolemmal Na⁺/Ca²⁺ exchanger (NCX) (see Fig. 6–2). The contribution of each of these mechanisms for lowering cytosolic Ca²⁺ varies among species. The relative contribution of SERCA2a to diastolic Ca²⁺ removal seems to be significantly lower in larger mammals compared to rodents. In humans, approximately 75% of the Ca²⁺ is removed by SERCA2a, and approximately 25% by the NCX and the role of the exchanger becomes more important in failing hearts.³⁵ SERCA2a transports Ca²⁺ back to the luminal space of the SR against a Ca²⁺ gradient by an energy-dependent mechanism (one molecule of ATP is hydrolyzed for the transport of two molecules of Ca²⁺), where it binds to Ca²⁺-buffering proteins.

The Ca²⁺ pumping activity of SERCA2a is negatively regulated by phospholamban.^38,39 In its unphosphorylated state, phospholamban inhibits the SERCA2a, whereas phosphorylation of phospholamban by cAMP-dependent PKA and by calcium-calmodulin kinase II (CaMKII) reverses this inhibition. Consistently, gene therapy approaches applying inhibition of phospholamban have remarkably ameliorated the failure phenotype in vitro and in animal models of heart failure.^40,41 Mutation in phospholamban that inhibits the phosphorylation of wild type allele has been found in humans.⁴² Interestingly, a heterozygous mutation (thus more inhibition of SERCA2a by phospholamban) has been demonstrated to cause heart failure in a family with hereditary cardiomyopathy.⁴² In contrast, a human phospholamban mutation coding for a premature stop-codon also caused heart failure both in the heterozygous and the homozygous state.⁴³ At present, it is unclear how these findings can be integrated with the data obtained in animals, except the notion that any form of altered phospholamban function might be detrimental in human.

Recent evidence supports an active role of phosphatases in cardiomyocytes excitation–contraction coupling.³⁹ Heart failure is accompanied by an increase in global protein phosphatase (PP) activity. Increased dephosphorylation is expected to aggravate the imbalance between phosphorylation and dephosphorylation, resulting in a net decrease in steady-state phosphorylation of target proteins. Protein phosphatase inhibitor-1 (I-1) is an endogenous negative regulator of PP1, which is activated by PKA and attenuates PP1 activity.³⁹ I-1 exerts an amplifier role in β-adrenergic signaling in cardiomyocytes.⁴⁴ Both transgenic mice overexpressing PP1 as well as I-1 knockout mice were associated with depressed basal cardiac function with blunted response to β-AR stimulation.⁴⁵ Additional evidence for the critical role of PP1 and I-1 in cardiac contractility comes from a study in protein kinase C-alpha (PKC-α), a negative regulator of I-1.⁴⁶ Overexpression of PKC-α reduced the I-1 interaction with PP-1, resulting in increased PP-1 activity and reduced cardiac contractility.

Changes in Heart Failure

In cardiomyocytes isolated from patients with heart failure, contraction and relaxation exhibit a similar set of abnormalities regardless of the etiology. These can be attributed to alterations in cellular processes within the myocytes, including the sarcolemmal ionic channels, receptors, the intracellular pumps, the myofilaments, and the metabolic pathways that generate ATP. Abnormal Ca²⁺ dynamics is a key causal component in the failing cardiomyocyte.^36,47 Cardiac cells isolated from failing human hearts exhibit three important characteristics: (1) a decrease in systolic Ca²⁺, (2) an increase in diastolic Ca²⁺, and (3) a prolonged relaxation phase.³⁵ In addition, a critical characteristic of a failing cardiomyocyte is the negative frequency response: that is, a decrease in contraction and Ca²⁺ release with increasing frequency.

Because Ca²⁺ handling is regulated by the previously described pivotal proteins, including voltage-dependent L-type Ca²⁺ channel, RyR, the SERCA2a–phospholamban complex, and the NCX, studies have focused on the relative changes in these proteins between heart failure and normal hearts. The coupling between the L-type Ca²⁺ channels and the RyRs from patients with dilated cardiomyopathy has been shown to be defective, such that in failing myocardial cells L-type Ca²⁺ channels have a reduced ability to activate the adjacent RyRs. In addition, hyperphosphorylation of RyRs has been implicated in the pathogenesis of heart failure.⁴⁸ PKA phosphorylation modulates RyR function by changing the sensitivity of RyR to Ca²⁺, resulting also in “leaky” channels that may cause diastolic Ca²⁺ release and generate delayed after depolarizations.^35,49

In both failing human hearts and in experimental models of heart failure, there is a reduction in the expression and enzymatic activity of SERCA2a.^39,48 Reduced SERCA2a protein level and activity result in impaired intracellular Ca²⁺ uptake into the SR (see Fig. 6–2). Enhanced expression of the NCX has been also reported as a compensatory mechanism for the reduction of SERCA2a because the exchanger can operate to extrude Ca²⁺ out from the cell.³⁸ The elevated intracellular diastolic Ca²⁺ is central to the heart failure pathophysiology; therefore, therapies to enhance SERCA2a activity have emerged as a rational approach. Gene transfer of SERCA2a in human failing cardiomyocytes rescued the depressed contractility and improved myocardial energetic.⁵⁰ The beneficial effects of restoring SERCA2a levels were also confirmed in vivo using experimental animal models of heart failure.^38,48,51 The evidence for key regulatory role of SERCA2a in Ca²⁺ handling has been recently fortified by two gene transfer studies. Post-translational modification of SERCA2a by small ubiquitin-like modifier 1 (SUMO-1) called SUMOylation has been shown to be critical for maintaining SERCA2a stability and activity.⁵² In heart failure, SERCA2a SUMOylation was significantly decreased, and gene transfer of SUMO-1 or small molecule activators of SUMO-1 restored the SERCA2a level, which led to improved cardiac function.⁵³ S100A1 is a Ca²⁺-binding protein abundantly expressed in the normal heart, and it promotes cardiac contraction and relaxation via SERCA2a and RyR activation. The S100A1 level is decreased in heart failure, whereas overexpression of S100A1 restored contractility by increasing SERCA2a activity and modulating RyR.⁵⁴

In contrast to decreased SERCA2a expression and activity in heart failure, phospholamban appears to maintain its protein levels, resulting in relative increase in SERCA2a inhibition.³⁸ In addition, phospholamban phosphorylation is reduced, further augmenting the inhibitory effect. PP1 protein expression is unchanged whereas global PP1 activity is increased in failing hearts.³⁹ This discrepancy can be explained by reduction in I-1 mRNA, protein, and phosphorylation levels, which have been found to be reduced by 50%, 60%, and 80%, respectively, in failing human hearts.⁴⁴ Expression of a constitutively active I-1 rescued function of isolated cardiomyocytes as well as failing hearts.^45,55,56

A functional imbalance between signaling elements that increase diastolic Ca²⁺ (Ca²⁺ channel, RyR, partly also NCX) and those that reduce it (SERCA, possibly also NCX) would elevate cytosolic free Ca²⁺ as described above. It appears possible that this is a final pathway of many alterations that lead to heart failure. What might be the downstream mechanisms exerting the detrimental action of cytosolic Ca²⁺? A multitude of experimental studies has refined our picture how enhanced Ca²⁺ levels might chronically harm cardiomyocytes. Specifically, activation of calcineurin⁵⁷ and the calmodulin/CaMK⁴⁷ pathway appear to be critically involved in Ca²⁺-mediated cardiomyocyte hypertrophy as described later in this chapter.

Contractile Proteins

A number of abnormalities also occur at the level of the contractile proteins in failing myocardium. There is a decrease in the fast isoform of the myosin heavy chain (α-MHC) and an increase in the slower isoform β-MHC.⁵⁸ This isoform switch correlates with systolic dysfunction, and its reversal by β-blockers was associated with functional restoration.⁵⁹ An increase in β-MHC results in a decrease in cross-bridge cycling rate and an increase in energy conservation because fewer ATP molecules are split. It also leads to slower contraction and relaxation phases which are characteristic of failing myocardium. Omecamtiv mecarbil, an activator of cardiac myosin, increases the transition rate of myosin into the actin-bound state and prolongs systolic ejection time without increasing Ca²⁺ transient.⁶⁰ Intravenous administration of omecamtiv was recently shown to increase stroke volume in a phase II clinical trial in patients with stable heart failure.⁶¹

Hypertrophy is the predominant method for the heart to increase its size and mass, because cardiomyocytes are terminally differentiated and have very limited capacity for hyperplastic growth. It is usually categorized as physiologic hypertrophy and pathologic hypertrophy.^62,63 Physiologic hypertrophy occurs in response to growth, exercise, and pregnancy without apparent abnormalities in both systolic and diastolic function. In contrast, pathologic hypertrophy takes place in settings of chronically increased cardiac load represented by hypertension and valvular diseases. Pathologic hypertrophy is described as concentric (characterized by addition of sarcomeres in parallel, leading to increased thickness of the myocyte) and eccentric (characterized by the addition of sarcomeres in series, leading to increased length).⁶² If the load placed on the heart is not normalized, the heart may continue to hypertrophy, eventually leading to elevated filling pressures caused by increased ventricular stiffness. The hypertrophied heart may also begin to decompensate in the long term, leading to progressive dilatation and systolic dysfunction. Not surprisingly, cardiac hypertrophy is a significant risk factor for the development of heart failure and, in addition, for sudden death.⁶⁴

Irrespective of the type of hypertrophy, the processes of hypertrophy require a dramatic reprogramming of gene expression to upregulate gene products necessary for growth of the cardiomyocytes. These include genes encoding contractile elements and proteins of the basic transcriptional and translational machinery that allow new protein production, and the genes encoding proteins that remodel the extracellular matrix, allowing growth to proceed. This process of reprogramming occurs in response to respective growth signals that are generated from a multitude of sources. In order to reprogram gene expression, these signals must be sensed at the cardiomyocyte membrane and transmitted into the interior of the cell, eventually into the nucleus, by a process called signal transduction. A large number of growth factors as well as signaling molecules have been reported to induce growth. Here, we will focus on the major arms for which strong evidence exists implicating components of the pathway in the hypertrophic response.

The Signal at the Cell Membrane

Two types of stimuli are believed to trigger the hypertrophic response: mechanical deformation of the membrane (cell stretch) and growth-promoting ligands binding to their cognate receptors in the myocyte cell membrane. Stretch of cardiomyocytes in culture leads to gene transcription and protein synthesis in a pattern that closely resembles the load-induced hypertrophic response in vivo.⁶⁵ Stretch triggers responses both via the direct activation of signaling molecules and by inducing the release of humoral factors. Direct activation involves recruitment of integrin signaling and stretch-activated ion channels.⁶⁶ For example, melusin, an integrin bound protein, is activated by mechanical stretch and promote adaptive hypertrophy via PI3-kinase/Akt and GRK3.⁶⁷

The release of growth factors acts in an autocrine or paracrine fashion to amplify hypertrophic responses. Angiotensin II activates a number of prohypertrophic signaling pathways.⁶⁸ The interleukin-6 family of cytokines, including cardiotrophin-1, is another group of factors released by mechanical stretch, which act via receptors specific to the cytokine and via the common receptor gp130.⁶⁹ A wealth of data indicates that insulin-like growth factor-1 (IGF-1) signaling is also activated by a variety of stimuli that induce hypertrophy. Many of the prohypertrophic peptides bind to receptors that are linked to heterotrimeric G proteins of the Gq family. These G proteins convert receptor activation into mobilization of intracellular signaling pathways, and are the initial trigger for downstream events.⁷⁰ Cardiac-specific overexpression of the α subunit of Gq led to cardiac hypertrophy.⁷¹ Later, it was reported that expressing a peptide that inhibits Gq-dependent signaling significantly limits the hypertrophic response to pressure overload in vivo.⁷² These studies, taken together, confirm a critical role of Gq in hypertrophic signaling and in the hypertrophic response to pressure overload. Activation of these proximal mediators of the stretch response results in increased cytosolic Ca²⁺, which plays a role in activating several signaling pathways, including calcineurin.

Signaling Within the Cell

There are two essential features of hypertrophic growth: reprogramming of gene expression and protein synthesis. Each is regulated by a series of intracellular pathways that are activated by events occurring at the membrane as described above. Signal transmission from the cell membrane to the nucleus is essential for gene reprogramming.⁴⁶ This is generally accomplished by linear cascades of proteins, most commonly protein kinases and also by the protein phosphatase calcineurin, activating one another in sequence, culminating in the phosphorylation and activation of one or more transcription factors. The transcription factors then bind to promoter elements, promoting the gene transcription. Each transcription factor will usually target several genes, and the net result of the activation of the entire set of genes is the hypertrophic response. Characteristic of the response is a reestablishment of a gene program often described as “embryonic” or “fetal,” because some of the gene expression pattern resembles that normally expressed in utero. The genes induced by the hypertrophic response are often divided into three groups based on their time of expression: immediate early, intermediate, and late. Immediate early genes include the neurohormonal mediator, brain natriuretic peptide, and several stress-induced genes or genes involved in growth control. These include c-Fos, c-Jun, c-Myc, Egr-1, and heat shock protein 70 (HSP70). Intermediate response genes include atrial natriuretic peptide and angiotensinogen as well as several sarcomeric components, β-MHC (and corresponding downregulation of α-MHC), myosin light chain-2, and skeletal α-actin (replacing cardiac α-actin). Late response genes include ACE and the NCX.

In-Depth Evaluation of Calcium-Dependent Hypertrophic Signaling

The calcineurin pathway is the most consistently reported pathway that has been shown to regulate hypertrophic growth in a number of different models. Calcineurin, also known as protein phosphatase 2B (PP2B), is a Ca²⁺-calmodulin–activated protein phosphatase, and it is uniquely activated by sustained elevations in intracellular Ca²⁺³⁵ (Fig. 6–4). Transgenic mice overexpressing an activated form of calcineurin in the hearts induced a profound hypertrophic response (two- to three-fold increase in heart size) that rapidly progressed to dilated heart failure within 2 to 3 months.⁵⁷ Such data implicate calcineurin as a sufficient inducer of the hypertrophic response and as a potential causative factor associated with the transition to decompensation and heart failure. Calcineurin catalytic activity can be inhibited by the immunosuppressive drugs cyclosporine A (CsA) and FK506. Inhibition of calcineurin with either CsA or FK506 can antagonize cardiac hypertrophy and/or disease progression in pleiotropic rodent models.⁷³ However, their potential usefulness in humans remains uncertain because both CsA and FK506 have a number of side effects in humans, including nephrogenic and neurogenic toxicity and immunosuppression.

CaMKII is another target of Ca²⁺–calmodulin activation. CaMKII is involved in the regulation of intracellular Ca²⁺ handling, affecting contractility and relaxation of the cardiomyocytes.⁴⁷ This molecule also modulates gene transcription by phosphorylating several transcription factors, including class II histone deacetylases, thereby enhancing the myocyte enhancer factor 2 activity. Increased CaMKII activity was found in hypertrophied and failing myocardium.⁴⁷ CaMKII overexpression induced cardiac hypertrophy,⁷⁴ whereas loss of CaMKII was associated with attenuated cardiac hypertrophy on hypertrophic stimulation.⁷⁵

In summary, these pathways illustrate the paradigm of how a signal, increased cytosolic Ca²⁺, generated in response to either deformation of the membrane or to hypertrophic agonist binding to its receptor, activates key signaling factors (calcineurin/CaMKII), which then activate transcription factors that, in turn, reprogramming gene expression.

Regulation of Protein Synthesis

The second critical component of hypertrophic growth is the ability to dramatically upregulate protein synthesis capabilities. This is regulated by two very complex, interacting pathways. One is the PI3-kinase pathway that, in addition to its role in regulating protein synthesis, also plays a major role in reprogramming gene expression. The other is the mammalian target of rapamycin (mTOR) pathway. These pathways are essential in determining cell, organ, and body size (ie, normal growth) in species as diverse as Drosophila and human, but they are also recruited in, and regulate the response to, pathologic stress-induced hypertrophic growth. Both the PI3-K and mTOR pathways regulate protein synthesis by modulating the activity of various translation factors.

The Phosphoinositide 3-Kinase Pathway

This highly conserved pathway (Fig. 6–5) is remarkable in the fact that virtually every component of the pathway has been shown in animal models to regulate cell and organ growth, including growth of the heart. The pathway is activated by most (if not all) of the agonists implicated in inducing cardiac hypertrophy, including pressure overload.⁶³ When activated, PI3-K phosphorylates the integral membrane phospholipid, phosphatidylinositol, leading to the recruitment of the protein kinase Akt (also known as protein kinase B, PKB) to the cell membrane. This brings PKB/Akt into proximity to its activator, the 3-phosphoinositide-dependent protein kinase-1 (PDK1), and results in activation of PKB/Akt.⁷⁶ PKB/Akt then plays a role in activating mTOR and, consequently, p70S6K and the protein translation machinery.⁷⁷ Consistent with this, all PI3-K, PDK1, and PKB/Akt have been shown to regulate cell and organ size, including size of the heart. Recently, inhibition of PI3-K p110γ was shown to normalize β-AR density and improve cardiac contractility in mouse heart failure,⁷⁸ and its efficacy in humans is being explored in clinical trials.⁴

Evaluation of mTOR

The importance of the protein kinase mTOR is illustrated by its conservation throughout evolution, from yeast to human. The critical importance of mTOR in regulating hypertrophic growth in vivo was demonstrated when its inhibitor, rapamycin, was found to attenuate the hypertrophic response to pressure overload in mice.⁷⁹ Distinct functions of mTOR are reflected in its assembly as structurally distinct multiprotein complexes, mTORC1 and mTORC2 (see Fig. 6–5). Growth factors—including those leading to cardiac growth, such as insulin and IGF-1—activate mTORC1. Activation of mTORC1 leads to disinhibition of eukaryotic translation initiation factor 4E, resulting in enhanced protein synthesis.⁷⁷ Because the mTORC pathway is also required for physiological hypertrophic response, complete inhibition of this pathway results in rapid ventricular dilatation and dysfunction upon pressure overload.⁸⁰ In contrast, partial inhibition of mTORC1 seems to be able to reduce only pathological hypertrophy.⁷⁷

Other Pathways Involved in Growth Regulation

MAPKs have been reported to be involved in a repertoire of biological processes including proliferation, differentiation, metabolism, motility, survival, and apoptosis.¹³ Stress-activated MAPKs, the c-Jun N-terminal kinases (JNKs), and p38 have been exhaustively studied (see Fig. 6–5); however, their exact role in hypertrophy remains obscure.¹³ Given the central role of these kinases in very basic responses of all cells to a wide range of cellular stressors (eg, oxidant stress, ionizing radiation, cytokine stimulation, osmolar stress, heat shock), compensatory adaptations are likely to be profound in various circumstances. Assessment of the present data suggest that the JNKs and p38 may be more involved in the progression of heart failure, including remodeling of the matrix, and may play much less of a role in hypertrophic growth.⁸¹ The extracellular signal-regulated kinase (ERK) family of MAPKs is also potently activated by hypertrophic stimuli. Interestingly, cardiac-specific overexpression of MEK1 (one of the immediate upstream activators of the ERKs) produced concentric cardiac hypertrophy, but unlike most other models, the MEK1 transgenic animals did not progress to contractile failure.⁸² These data raised the concept of “beneficial” hypertrophy versus “detrimental” hypertrophy by clearly demonstrating that hypertrophy per se need not inexorably progress to heart failure. This may have to do with the fact that the ERKs (in contrast to many other prohypertrophic signaling molecules, including calcineurin, the JNKs, and the p38) are, in many circumstances, cytoprotective.¹³

Glycogen synthase kinase-3 (GSK-3) is another target of PKB/Akt involved in hypertrophic signaling. GSK-3β is a negative regulator of cardiac growth. Transgenic animals expressing GSK-3β have dramatic reductions in normal cardiac growth and also have a markedly reduced hypertrophic response to pressure overload.⁸³ Furthermore, inhibition of GSK-3β is necessary for the hypertrophic response to a number of agonists, likely mediated via several mechanisms. The GSK-3β targets include the nuclear factor of activated T cells (NFATs), and thus GSK-3β acts in opposition to the calcineurin pathway.⁸⁴ Other known growth regulators negatively regulated by GSK-3β include c-Myc, GATA-4, β-catenin, and c-Jun.⁸⁵

Alterations in Signaling in the Diseased Human Heart

Despite the continuous efforts to characterize molecular pathways in hypertrophy using transgenic and pressure overload animal models, analysis in human heart tissues exhibits somewhat inconsistent results. For example, in one study, various signaling factors (calcineurin, ERK1/2, JNK, p38, PKB/Akt, and GSK-3) were examined in the hypertrophic heart without heart failure, and calcineurin was the only factor that was consistently found to be activated.⁸⁶ In contrast to the paucity of data on hypertrophied hearts, several studies have examined the signaling profile of hearts explanted from patients with advanced heart failure undergoing either transplant or left ventri­cular assist device placement. Where examined, calcineurin expression and activity were increased significantly.^87,88 Activation of p38 was generally reported, although there seems to be a substantial variability.⁸¹ Meanwhile, no consistent results have been reported for JNKs and ERKs. PKB/Akt was found to be activated, irrespective of the etiology of the heart failure, and accordingly, its downstream target, GSK-3β, was inhibited.⁸⁹ Variabilities among the studies could be accounted for by a number of factors, including different medical therapies, device therapies, status of the patient, and methods of preserving the tissue.

To summarize, current data are insufficient to conclude whether any signaling alterations are causal or simply a consequence of the hypertrophy and heart failure. The complexity of the heart failure signaling abnormalities and the changing activities of various pathways at various times in the progression of disease creates a “moving target” and leads to significant challenges to determine the key pathways. The patients with compensated hypertrophy versus advanced heart failure are obviously at different ends of the pathophysiologic spectrum of heart failure, and in between these points, including the transition to and early progression of heart failure, we have very little data on what pathways might be reasonable targets.

Myocardial cell loss by any mechanism is an important component in the genesis of heart failure. Although necrosis has been regarded as the predominant mode of myocardial cell death, there is increasing evidence that cells die through a variety of programmed and nonprogrammed mechanisms. In addition to necrosis, cardiac cells may be lost through apoptosis as well as by misfolded protein accumulation associated with protein degradation mechanisms. The proportion of cells dying through each mechanism may differ at various stages in the natural history of heart failure, and therapeutic interventions may radically alter the fate of cells dying via each mechanism. Apart from mechanistic insights, understanding of such mechanisms of cell survival and death may offer clue for novel therapeutic approaches.

Cell Death by Necrosis in Heart Failure

The causes of necrotic cell death in heart failure are multiple.^90,91 Ongoing ischemia and the injury following reperfusion are the most common causes and infringe on aerobic oxidative respiration. Infective agents, postinfective immune processes, hypersensitivity phenomenon, and autoimmune diseases result in predominant inflammatory injury. In addition, chemical insults, including alcohol and doxorubicin toxicity, are infrequent causes of myocardial damage. Regardless of the inciting agent, several common biochemical pathways mediate cell necrosis, the most important of which is ATP depletion, which occurs commonly during hypoxia and ischemia. Partially reduced reactive oxygen species, produced during oxidative phosphorylation, are exaggerated during reperfusion injury and lead to cellular damage.⁹² In addition to ATP depletion and oxidative stress, an increase in intracellular Ca²⁺ activates potentially deleterious enzymes, such as phospholipases, proteases, ATPase, and endonucleases. These biochemical events contribute to mitochondrial and cell membrane damage. Formation of a permeability transition pore in the mitochondrial membrane leads to loss of mitochondrial membrane potential, which is critical for oxidative phosphorylation. On the other hand, the loss of integrity of the cell membrane, which is the hallmark of necrotic cell death, leads to cell swelling, extrusion of intracellular content and induction of inflammation.⁹³ Morphologically, there is severe disruption of cell membrane in necrosis that is associated with loss of cell contents.⁹⁴ Most necrotic cells and their debris are removed by a combined process of digestion and fragmentation, with phagocytosis of particulate debris. If not destroyed and removed, they attract Ca²⁺ salts and other minerals and develop dystrophic calcification. Recently, a programed necrosis, namely necroptosis, was focused as another form of necrosis,⁹⁵ and its role in the progression to heart failure is actively explored.⁹⁶

Cell Death by Apoptosis in the Failing Myocardium

Apoptosis is a genetically programmed and energy-requiring series of events that permits the cell to die without inducing an inflammatory response. The classical changes of apoptosis include cell shrinkage and formation of apoptotic bodies.⁹⁷ The activation of DNAses results in cleavage of chromatin into multimers of oligonucleosomal-length DNA fragments, evident on agarose gel electrophoresis. Sarcolemmal integrity is preserved but progressive convolution of membrane breaks up the cell into clusters of membrane-bound subcellular organelles referred to as apoptotic bodies. Unlike necrosis, because cell swelling and sarcolemmal disintegration do not occur, cytoplasmic contents are not released and inflammation not provoked. Apoptotic bodies are phagocytosed and removed.

External and internal stimuli can trigger death receptor-mediated (also known as extrinsic) or mitochondrial-mediated (intrinsic) apoptosis⁹⁴ (Fig. 6–6). Both involve activation of highly specific proteolytic enzymes referred to as caspases.⁹⁸ Apoptosis in human heart failure was initially demonstrated by Narula et al⁹⁹ based on the histochemical demonstration of DNA fragmentation; apoptosis occurred both in ischemic and dilated cardiomyopathy. It was proposed that slow ongoing apoptotic loss of myocytes (rate of 0.12%–0.70%)⁹⁴ may contribute to inexorable progression of chronic heart failure. The ultrastructural studies in explanted hearts demonstrated cytochrome c release from mitochondria to cytoplasm in failing hearts.¹⁰⁰ Downstream to cytochrome c release, caspase 3 was also found to be activated in the cardiomyopathic hearts. Although the majority of myocytes demonstrate ultrastructural evidence of cytochrome c release and there is variable loss of cytoplasmic proteins, the nuclei in all these cells remain essentially normal.¹⁰¹ Intactness of nuclei suggests the viability of cardiomyocytes. Lack of clear damage to nucleus despite the induction of apoptotic process in cardiomyocytes appears to be different from that seen in the classic apoptosis. A myocyte that maintains its nuclear integrity, but has compromised its respiratory chain and oxidative phosphorylation and allows variable destruction of its contractile proteins should contribute to systolic dysfunction of the cell.¹⁰¹ Such cells with intact nucleus may be amenable to recovery and can be a therapeutic target.

Role of Ubiquitin–Proteasome Pathway and Autophagy in Myocyte Death in Heart Failure

There are multiple mechanisms that the body uses to remove inappropriately folded or damaged proteins. In the failing heart, these defective proteins increase as a result of the endoplasmic reticulum (ER) stress and play important roles in heart failure pathophysiology.¹⁰² Failing to remove abnormal proteins in cardiomyocytes can form toxic aggregates, which leads to development of cardiomyopathies. Two major protein degradation systems are the ubiquitin–proteasome system and autophagy. The evolutionarily highly conserved ubiquitin system labels substrate proteins with a ubiquitin molecule on lysine residues en route to degradation. The ubiquitination-protease system is responsible for 80% to 90% of protein degradation in vivo with a high specificity.¹⁰³ Numbers of experimental animal and human heart tissue studies suggest impaired function of the ubiquitin-proteasome system,¹⁰⁴ rendering it as an attractive therapeutic target. Autophagy is another protein quality control system that is involved in a bulk degradation of long-lived proteins and aggregated proteins. Damaged proteins are isolated in double-membrane vesicles, known as autophagosomes, and degraded by lysosome. Autophagy is activated in heart failure; however, this is likely a secondary activation, and whether inhibiting autophagy is beneficial currently remains controversial.¹⁰⁵

Unfolded protein response (UPR) is a mechanism in which the cells try to protect themselves from increased ER stress stimuli such as ischemia, altered redox status, or impaired calcium homeostasis.¹⁰⁶ UPR serves to relieve the ER stress by downregulating the protein translation, generating chaperone proteins, and degrading proteins through ER-associated degradation as well as autophagy.^105,107 Such mechanisms help the cells survive by reducing the entry of new proteins into the ER, and increasing protein translocation out of it. Activation of UPR is found in heart failure; however, prolonged ER stress can lead to accumulation of protein aggregates, which eventually triggers cell death signals.

Cross-Talk in Death Pathways

It is becoming clearer that various forms of death may be a part of the same spectrum, and multiple morphologies of cell death may coexist in a very close proximity.¹⁰⁸ It has been demonstrated that inhibiting one pathway of death may simply switch it to die by another pathway.^109,110 One could speculate that cross-talk in death pathways might be dependent on a number of associated factors, such as the energy state of cell and intensity of damage. Preserved energy states and lower intensity stimuli can drive a cell toward “energy hungry” death pathways such as apoptosis. On the other hand, exhaustion of energy during these (energy-requiring processes) and exaggeration of severity of inducing stimuli may upset Ca²⁺ homeostasis and mediate necrosis. For instance, a low-intensity stress, such as a short period of ischemia or interruption of ischemia by reperfusion, may lead to apoptotic death, whereas a longer period of ischemia may bring on necrotic death. Repletion of the cytosolic ATP pool before irreversible damage may redirect the death program toward energy-dependent death pathways.

A number of pathways and regulators discussed above play crucial roles in cardiac biology, mainly in contractile cardiomyocytes. Another cell type involved is the fibroblast and, indeed, cardiac fibrosis is a hallmark of cardiac remodeling processes usually initiated in response to stress. Myocardial fibrosis refers to a variety of both quantitative and qualitative changes in the interstitial myocardial collagen network that usually is the consequence of cardiac insults or other forms of cardiac stress, such as pressure overload of the left ventricle. Cardiac fibrosis describes a multidimensional process of exaggerated activity of cardiac fibroblasts, as well as inflammatory reactions leading often to excessive extracellular matrix production within the myocardium with detrimental consequences. Thus, cardiac fibrosis leads to significant changes in the myocardial architecture facilitating the development of cardiac dysfunction caused by cardiac stiffness, but also the occurrence of cardiac arrhythmias and ischemia at the cellular level. Next to a focus on the cardiomyocyte, it is thus clear that alterations of the cardiac extracellular matrix leading to myocardial fibrosis play also a major role in the development and evolution of cardiac disease.⁶² Fibrosis is generally part of a maladaptive cardiac remodeling process. Remodeling is a term that is used to describe a cardiac pathological condition in response to ischemia or pressure and/or volume overload of the heart.^62,111 There are also various genetic forms of cardiac diseases showing enhanced fibrosis, such as hypertrophic cardiomyopathies.¹¹²

Biology of Cardiac Fibroblasts and Collagen Turnover

Myocardial fibrosis occurs when the net effects of collagen turnover become unbalanced (ie, increased synthesis predominates over unchanged or decreased degradation).^113,114 Excess collagen may accumulate in the interstitial and perivascular space.¹¹⁵ At the cellular level, cardiac cells such as myocytes, endothelial cells, and interstitial cells such as cardiac fibroblasts are reactive in response to pathological stress. Although cardiomyocytes represent the majority of the cardiac mass, in terms of numbers they only represent about 30% to 50% of the cell types that build up the heart. A considerable number of cells are “cardiac fibroblasts,” which is a heterogeneous population of important and as yet not well characterized cells contributing to myocardial homeostasis. The origin of cardiac fibroblasts is debated and may be multifarious: that is, there are resident cardiac fibroblasts but also those that may transdifferentiate from other cell types, such as inflammatorycells, mesenchymal, epicardial or endothelial cells.^116,117. However, a recent finding suggests that the majority of fibroblasts responsible for the cardiac fibrosis response originate from endogenous cardiac tissue–derived fibroblasts.¹¹⁸

Like the fibroblasts residing in other organs, the main function of cardiac-derived fibroblasts is the production and homeostatic maintenance of the extracellular matrix. This extracellular matrix surrounds the cardiomyocytes and provides a solid, flexible scaffold to a tissue that is constantly subject to great mechanical workloads. Thus, alterations of the composition of the extracellular matrix may have strong impact on the cardiac function.¹¹⁴ In the normal heart, the extracellular matrix is composed of a network of macromolecules, including collagens, elastin, glycoproteins and proteoglycans, with type I collagen being the main component.¹¹⁹ In healthy conditions, the net deposition of collagen in the cardiac interstitium is rather low, with a good balance between production and degradation. This homeostasis is severely impaired during diseased states of the heart; high collagen deposition is a consequence of either reparative or reactive fibrosis.¹²⁰ Here, cardiac fibroblasts may differentiate into more specialized myofibroblasts, which are characterized by increased expression of contractile proteins such as alpha smooth muscle actin. Myofibroblasts are critically involved in the normal scar formation and contraction processes occurring after tissue injuries such as myocardial infarction.¹²¹ Importantly, the extracellular matrix is capable of electrically isolating regions of the myocardium.¹²² This has both a physiological relevance in the sinoatrial node, where thicker and thinner sheets of connective tissue determine the expansion of excitation signals in a correct direction, as well as a pathological relevance. Excessive deposition of extracellular matrix can favor the development of detrimental arrhythmias in the remodeling myocardium.^123,124

Fibroblast Characterization

So far, cardiac fibroblasts are defined on the basis of some markers and phenotypical characteristics such as a flat, spindle-shaped morphology, with multiple processes originating from the cell body and lack of a basement membrane.¹²⁵ Characterization of fibroblasts of several tissue origins is difficult because of the lack of specific molecular markers, a feature that represents a limiting factor for in vivo investigations on fibroblast biology. However, several substitute markers have been used over time, especially in vitro, for the evaluation of purity in primary cardiac fibroblast cultures, but also in vivo, in order to map the origin of proliferating and collagen-producing fibroblasts during cardiac injury. Examples of such markers are vimentin, FSP1, DDR2, periostin, TCF21 and prolyl-4-hydroxylase. However, it remains to be shown which of those can serve as rather specific markers for identifying cardiac fibroblasts. Furthermore, common in vitro cell culture models do not reliably reproduce the in vivo properties of cardiac fibroblasts because of a spontaneous differentiation of cells toward a myofibroblast phenotype¹²⁶ and because of the lack of a three-dimensional structural organization and interrelation with myocytes. For these reasons, characterization of cardiac fibroblasts from a molecular point of view is still ill defined. Treating cardiac fibrosis may prevent the onset of arrhythmias and improve the functional performance of the heart in several disease conditions. It thus will be critical to identify new profibrotic mechanisms not targeted by currently available therapies and to translate these mechanisms into individualized diagnostic tools and specific therapeutic targets. There is a tremendous need for a systematic and collaborative interaction between clinical investigators and basic scientists, as well as with the industry—to introduce more advanced approaches and methodologies into the field of myocardial fibrosis.

From a historical perspective, proteins such as transcription factors were thought to be the main players in gene expression regulation, but with the development of new, high-resolution technologies for the analysis of the transcriptome, such as next-generation RNA deep sequencing, this paradigm has fallen. In fact, it has now been determined that although about three-fourths of the mammalian genome is transcribed, less than 2% is ultimately translated into proteins¹²⁷ (Fig. 6–7). These non–protein-coding transcripts form a relatively unexplored non–protein-coding RNA (ncRNA) dimension that we have only recently started to study. The fundamental importance of this new ncRNA world becomes more clear by noticing that the degree of complexity of a species correlates better with the amount of DNA that is transcribed into ncRNA than with the number of protein-coding genes.¹²⁸ As of today, various ncRNAs with a gene regulatory function have been identified and were shown to act at various steps along the protein biosynthetic process, including transcription, RNA maturation, translation, and protein degradation. These discoveries add to a new level of understanding also in cardiovascular pathophysiology.

Based on their size, ncRNAs can be subdivided into two major groups: (1) small ncRNAs (< 200 nucleotides long) include microRNAs (miRNAs), PIWI-interacting RNAs, and endogenous short interfering RNAs, and (2) long ncRNAs (lncRNAs), which have a length between 0.2 and 2 Kb.^{129,130,131,132} Indeed, ncRNAs emerged as crucial players in animal and human pathology, including that of the cardiovascular system.^133,134,135 Around 1500 to 2000 human miRNAs have been catalogued and are believed to regulate about half of all messenger RNAs (mRNAs)—maybe more. Many of the miRNAs show developmental stage– and cell type/tissue–specific patterns of expression. MiRNAs regulate gene expression at the level of mRNA translation, via binding to mRNAs leading to repression and/or degradation.¹³⁵ In the following subsections, a number of examples are given where miRNAs have been shown to be directly involved in cardiac pathological processes.

MicroRNAs Involved in Cardiac Hypertrophy

One of the first studies that mechanistically investigated the importance of a miRNA in cardiac hypertrophy showed a miRNA termed miR-208 to be crucially involved in hypertrophic signaling.¹³⁶ Another early report demonstrated that left ventricular pressure–overloaded mice showed low cardiac miR-133 levels and that inhibition of miR-133 with an anti–miR-133 oligonucleotide-generated cardiac hypertrophy in vivo.¹³⁷ A first pioneering study showed then a therapeutic effect of miRNA modulation in a murine cardiovascular disease model.¹³⁸

Some miRNAs become strongly activated during pathological stress, making them then functionally important during disease. An example is the miR-212/132 family, which becomes activated during heart failure in humans¹³⁹ and animal models.¹⁴⁰ This miRNA family regulates cardiac hypertrophy and autophagy, a crucial process to maintain cardiac homeostasis,¹⁴¹ in cardiomyocytes by targeting the antihypertrophic and proautophagic transcription factor forkhead box O3, leading to hyperactivation of the prohypertrophic calcineurin/NFAT signaling pathway. Both genetic and pharmacologic inhibition of one of these miRNAs, miR-132, rescued cardiac hypertrophy and heart failure development in mice, offering a possible therapeutic approach for cardiac remodeling.¹⁴⁰ In addition, certain miRNAs regulate the intracellular Ca²⁺ homeostasis, directly contributing to cardiac contractility. For instance, impaired myocardial SERCA2a activity is a hallmark of failing hearts. In studies with rats, it was shown that SERCA2 therapy after myocardial infarction restored miR-1 levels, leading to normalized expression levels of the NCX1.¹⁴² MiR-214 also has a cardioprotective role during ischemic/reperfusion injury by repression of NCX1, a key regulator of Ca²⁺ influx.¹⁴³ Several miRNAs, such as miR-25 or miR-132, directly regulate SERCA2; thus their inhibition might be beneficial for general Ca²⁺ handling in the cardiomyocyte.¹⁴⁴

MicroRNAs in Cardiac Fibrosis

As described on p. 11 under “Cardiac Fibrosis,” cardiac fibrosis is the exaggerated activation of (mostly) resident cardiac fibroblasts, leading to increased secretion of extracellular matrix proteins, growth factors, cytokines, as well as genetic material. There are several miRNAs enriched in cardiac fibroblasts that have been shown to serve as powerful regulators of various fibrotic processes, including fibroblast proliferation and growth factor secretion. For instance, miR-133 and miR-30, which control expression of connective tissue growth factor—a protein tightly linked to fibrosis development—are also usually downregulated under cardiac stress.¹⁴⁵ MiR-21 was shown to silence the ERK–MAP kinase inhibitor sprouty-1, which is mechanistically involved in fibrosis development.¹³⁸ Because miRNAs usually have dozens of targets, it is interesting that miR-21 has also been found to promote cardiac fibrosis by regulating other targets, such as transforming growth factor beta 1 receptor III¹⁴⁶ and matrix metalloprotease-2.¹⁴⁷ Surprisingly, miR-21 knockout mice still developed cardiac fibrosis under cardiac stress,¹⁴⁸ whereas pharmacologic inhibition of miR-21 inhibited fibrosis in the heart^138,149,150 and other organs, such as lung¹⁵¹ and kidney.¹⁵² In contrast to miR-21, which is activated during cardiac stress, miR-29 is downregulated and leads to a derepression of many fibrosis-related genes, such as collagens, elastin, and fibrillin.¹⁵³ Both miR-21 and miR-29 have been shown to serve as potential therapeutic targets for cardiac fibrosis. Interestingly, circulating levels of miR-29 were found to positively correlate with hypertrophy and fibrosis in patients with hypertrophic cardiomyopathy.¹⁵⁴ Thus, other functions of this miRNA (eg, in cardiomyocyte hypertrophy) might be expected, and monitoring the level of this or other miRNAs could be a useful for prognostic/therapeutic purposes in this clinical setting.

MicroRNAs Regulating Cardiac Vessel Density

Each cardiomyocyte in the heart is usually surrounded by several capillaries to provide sufficient energy and oxygen transport. This balance is derailed under cardiac stress, such as with myocardial infarction, cardiac fibrosis, or heart failure. There is a growing interest in strategies to improve vessel density leading to smaller scar formation after myocardial infarction.¹⁵⁵ Endothelial knockout of Dicer—a key protein involved in miRNA maturation—led to endothelial dysfunction, demonstrating the key role of miRNAs in endothelial biology.¹⁵⁶ The endothelial cell-enriched miRNA miR-126 was found to mediate developmental angiogenesis in vivo.¹⁵⁷ Indeed, deletion of miR-126 led to loss of vascular integrity and defects in endothelial cell proliferation, migration, and angiogenesis. Thus, approaches to deliver miR-126 are likely to be beneficial to endothelial cells and the improvement of vascular integrity and repair.¹⁵⁸ Also, an miRNA involved in cardiac angiogenesisis miR-92a.^159,160 This miRNA is expressed in human endothelial cells, but also elsewhere in many other cells, and it controls the growth of cancer cells,¹⁶¹ blood vessel growth, and many other cellular functions throughout the body. By using mouse models of limb ischemia and myocardial infarction, it was shown that systemic blockade of miR-92a led to enhanced blood vessel growth and functional recovery of damaged tissue.¹⁵⁹

Another miRNA involved in the regulation of cardiac vascularization is miR-24. This miRNA is one of the most highly expressed in cardiac endothelial cells and is further upregulated after ischemia in various organs.^162,163 High levels of miR-24–induced endothelial cell apoptosis and abolished endothelial capillary network formation, partly through targeting the endothelial transcription factor GATA2 and the p21-activated kinase PAK4. Importantly, blockade of endothelial miR-24 limited myocardial infarction size in mice, via enhancement of vascularity, eventually leading to preserved cardiac function and survival. In addition, miR-24 regulates smooth muscle cell functions, and miR-24 overexpression inhibited development of the vasculature in a model of engineered heart tissue.¹⁶⁴

MicroRNAs Regulating Cardiac Metabolism

Another important function of miRNAs in the heart is the control of energy homeostasis and metabolism. The idea that a failing heart is comparable with an engine out of fuel is relatively old.¹⁶⁵ A cardiac system to create energy is the creatine kinase system, which is controlled by several nuclear-receptor transcription factors, such as nuclear receptors of the peroxisome proliferator–activated receptor (PPAR) family, which also are responsible for substrate utilization from fatty acids to glucose.¹⁶⁶ Interestingly, the miR-199a/214 cluster shares PPARδ as a common target. Silencing of both miR-199a and miR-214 in mice undergoing pressure overload improved cardiac function and restored mitochondrial fatty acid oxidation.¹⁶⁷ Thus, miRNA manipulation has been shown to modulate the metabolic shift from initial reliance on fatty acid utilization in the healthy myocardium toward increased reliance on glucose metabolism at the onset of heart failure. This may open up new miRNA-based therapeutic concepts to improve energy utility in the failing heart. A further regulator of energy homeostasis is the peroxisome proliferator-activated receptor γ coactivator 1β (PGC-1β), which is a transcriptional coactivator regulating metabolism and mitochondrial biogenesis through stimulation of nuclear hormone receptors. The PGC-1β gene harbors the genes for both miR-378 and its counterpart miR-378*, which counterbalance the metabolic actions of PGC-1β: indeed, mice genetically lacking miR-378 and miR-378* were resistant to high-fat diet-induced obesity.¹⁶⁸ Other targets of miR-378 and miR-378* include the carnitine O-acetyltransferase, which is a mitochondrial enzyme involved in fatty acid metabolism, as well as MED13, a component of the mediator complex. A heart-specific miRNA, miR-208a, directly regulates MED13, and pharmacologic inhibition of miR-208a in mice led to resistance to high-fat diet-induced obesity, demonstrating that a cardiac-specific miRNA may be of importance in regulating whole body energy hemostasis.¹⁶⁹

MicroRNAs as Potential Biomarkers and Paracrine Mediators

MiRNAs can be released into the extracellular space and circulation either within small vesicles called exosomes or by binding to specific proteins.^170,171 The secretion of miRNAs can also occur via other vesicular bodies arising from the plasma membrane, by simple extrusion through membrane shedding, or by their presence within apoptotic bodies.¹⁷⁰ A number of reviews have covered the role of these circulating miRNAs as biomarkers of cardiovascular diseases.^135,172 Indeed, different kinds of vesicles are produced by cardiovascular cells, including apoptotic bodies, microvesicles, and exosomes. They all contain specific cargoes of protein, miRNA, and mRNA.¹⁷³ It was recently shown that are particularly enriched in small RNAs, such as miRNAs.¹⁷⁴ Exosomes with a mean diameter of 30 to 100 nm are membrane vesicles and are formed by fusion of multivesicular bodies with the plasma membrane, leading to a release of exosomes into the extracellular space. The RNA content of exosomes does not simply reflect the composition in the cytoplasm, suggesting a specific transport and shuttling system that very specifically brings only selected RNAs to the exosomes.

Of potential interest for the clinician, secreted exosomes can be isolated from biological fluids, and their miRNA or other RNA profiles assessed, enabling them to serve as potential biomarkers for certain diseases.¹⁷⁰ This has recently been also suggested for extracellular lncRNAs, which can be detected in the plasma of patients with myocardial infarction.^175,176 Of considerable biological importance are recent findings that demonstrate that secreted ncRNAs have a paracrine potential (eg, through uptake by recipient cells and subsequent binding to and regulation of the recipient cells’ mRNAs¹⁷⁷).

In addition, apoptotic bodies enriched in miR-126 have been reported to limit atherosclerosis development.¹⁷⁸ Further evidence for the crucial role of miR-126 as an important protector of the endothelial system was recently obtained with miR-126 KO mice, which have impaired endothelial proliferation.¹⁷⁹ A further example of atheroprotective actions based on miRNA trafficking was shown for shear stress–stimulated human endothelial cells that produced vesicles enriched in miR-143/145 that led to reduced atherosclerotic lesion formation in aortas of ApoE knockout mice.¹⁸⁰ Other cell types with evidence of an ability to secrete miRNAs are cardiac fibroblasts and inflammatory cells, both involved in the cardiac fibrosis process.¹⁸¹

Thus, there is cumulative evidence that cardiovascular cells communicate with each other via the transfer of ncRNAs. Their use as diagnostic markers is intriguing and may open a new avenue for cardiovascular biomarker research. Finally, circulating vesicles containing ncRNAs may serve as new treatment targets, through manipulation of derailed intercellular communication systems.

The lncRNAs are transcripts more than 200 nucleotides long that have no known protein-coding function.¹⁸² However, this definition may be preliminary; in fact, many lncRNAs have been recently reported to encode short peptides in human tissues.¹⁸³ The lncRNAs can be classified as (1) sense lncRNAs, when they overlap one or more exons of another transcript on the same strand; (2) antisense lncRNAs, when they overlap one or more exons of another transcript on the opposite strand; (3) bidirectional lncRNAs, when their expression and that of a neighboring coding transcript on the opposite strand are initiated in close genomic proximity, (4) intronic lncRNAs, when they are derived from an intron of a second transcript; and/or (5) intergenic lncRNAs, when found as an independent unit within the genomic interval between two genes¹³⁴ (Fig. 6–8).

The detailed mechanisms of the various functions of lncRNAs are currently being intensely investigated; lncRNAs can interact with many macromolecules within the cell, including other RNA species, proteins, and DNA. Complementary sites on the lncRNA allow them to recognize and bind to mRNAs, miRNAs, or even other lncRNAs and act as highly specific sensors for their regulation. On the other hand, protein-binding sites allow them to interact with proteins, generating specific ribonucleoprotein particles with different functions. Their ability for conformational switching is remarkable, whereby the activity of the lncRNA is activated or suppressed by an external signal, thus making them exquisite regulatory devices.¹⁸⁴ LncRNAs have a high folding energy, which distinguishes them from other mRNAs.¹⁸⁵

In general, lncRNAs can be divided into nuclear lncRNAs and cytoplasmic lncRNAs, although many lncRNAs also may be present in both compartments or even shuttle between those compartments. Thousands of eukaryotic lncRNAs have been identified, with many found to be species specific, but less conserved than protein-coding genes.¹⁸⁶ However, the expression profile of lncRNAs is more cell-type specific than that of protein-coding genes, and the profile changes with differentiation and the developmental stage of an organism.¹⁸⁷ The lncRNAs are being implicated in many biological processes, including cardiovascular ones.^134,188,189

Indeed, the role of lncRNAs in the heart has just begun to be uncovered. In the vascular system, MALAT1, LINC00323 or MIR503HG are enriched in endothelial cells and regulate the angiogenic features of vascular cells.^190,191

An lncRNA Braveheart was shown to be crucial for cardiac development in mice but is not present in humans, limiting translational relevance.¹⁹² Another lncRNA, Fendrr, was able to control chromatin modifications and, thus developmental signalling also, in the rodent heart.¹⁹³ In mice, the lncRNA Chrf was found to act as a ceRNA sequestering miR-489, a miRNA targeting the mRNA of myeloid differentiation primary response gene 88 (Myd88), the upregulation of which is known to induce hypertrophy.¹⁹⁴ Another lncRNA that has been very recently linked to cardiac pathology is cardiac apoptosis–related lncRNA (Carl).¹⁹⁵ In mouse cardiomyocytes, Carl bound to and sequestered miR-539, a miRNA found to target the mRNA of the PHB2 subunit of prohibitin, a protein localized to the inner mitochondrial membrane, where it has a role in mitochondrial homeostasis. Carl acted as the endogenous sponge for this miRNA, regulating mitochondrial morphology and cell death under normal conditions. One recent study identified the lncRNA Mhrt, which originates from MYH7 locus, as cardioprotective, and restoration of Mhrt levels protects the heart from hypertrophy and failure.¹⁹⁶ The mode of action of Mhrt is described in Fig. 6–9. An lncRNA that is conserved between rodents and humans is the cardiac hypertrophy–associated transcript (Chast) lncRNA. Chast was identified in a wide screen of lncRNA profilling during pressure overload–induced cardiac remodelling in mice.¹⁹⁷ The human CHAST homolog is significantly upregulated in hypertrophic heart tissue from aortic stenosis patients and in human embryonic stem cell–derived cardiomyocytes on hypertrophic stimuli. Viral-based overexpression of Chast was sufficient to induce cardiomyocyte hypertrophy, whereas pharmacological silencing of Chast both prevented and attenuated TAC-induced pathological cardiac remodeling with no early signs on toxicological side effects. These results indicate that Chast can be a potential target to prevent cardiac remodeling and highlight a general role of lncRNAs in heart diseases.

In conclusion, the biology of lncRNAs is remarkable, and many new aspects in cardiovascular physiology and pathology are emerging. For instance, lncRNAs also may form circular structures, so-called circular RNAs, for which also first examples for biological functions in the cardiovascular systems exist.^198,199 Also, lncRNAs may sometimes code for micropeptides with biological function; a putative muscle-specific lncRNA that encodes a peptide of 34 amino acids was named dwarf open reading frame (DWORF). DWORF is the only endogenous peptide known to activate the SERCA pump by physical interaction and provides a means for enhancing muscle contractility.²⁰⁰

We have reviewed representative classical pathways and newer mechanisms involved in cardiac pathophysiology. There remain numbers of important pathways/mechanisms that are not covered here and the readers are referred to recent reviews.^{201,202,203,204,205,206} Successful clincial application of β-blockers and ACE-inhibitors was supported by not only the clinical trials but also by intense research in the molecular and cellular biology related to these drug pathways. Significant advances in our understanding of cell death, fibrosis, and microRNA have been achieved in the past deacade. Many new exciting discoveries about other noncoding RNAs, such as lncRNAs and circular RNAs, involved in the control of cardiovascular biology, including chances for new therapeutic concepts and diagnostic strategies, can be expected.