Molecular Basis of Heart Failure

Types of Heart Failure

Heart failure (HF) is a multisystemic disorder characterized by profound disturbances in circulatory physiology and a plethora of myocardial structural and functional changes that adversely affect the systolic pumping capacity and diastolic filling of the heart. A discrete inciting event, such as myocardial infarction (MI) or administration of a chemotherapeutic agent, may be identifiable as a proximate trigger in some cases. However, in the vast majority of instances, contributory risk factors (e.g., hypertension, obesity, ischemic heart disease, valvular disease, or diabetes) or genetic and environmental cues are uncovered during the diagnostic workup. These processes adversely affect myocardial biology and trigger cardiomyocyte hypertrophy, dysfunction, and cell death. They also provoke alterations in the extracellular matrix and vasculature, promoting neurohormonal signaling as an adaptive response that paradoxically worsens the pathophysiology. At the cellular level, loss of cardiomyocytes occurs focally with an acute MI or diffusely with some chemotherapeutic agents and with viral myocarditis. This leads to sustained hemodynamic stress, which results in increased hemodynamic load on the surviving myocardium. Simultaneously, molecular changes are triggered in various cardiac cell types, either in response to the inciting stress or as a secondary consequence of increased hemodynamic load, culminating in contractile dysfunction, altered relaxation and stiffness, fibrosis, and vascular rarefaction.

Evolution of the disease process entails inexorable progression of these cellular and molecular changes in the face of unremitting stress, often despite state-of-the-art antiremodeling therapies. When the process reaches its end stage, mechanical circulatory support or heart transplantation is required. Elucidation of the molecular and cellular bases of these changes during the course of HF pathogenesis is therefore paramount in developing the next generation of therapeutic approaches to address the growing epidemic of HF. For the convenience of the reader, a glossary of abbreviations used in this discussion is presented at the end of this chapter.

Investigative Techniques and Molecular Modeling

Contemporary investigation into the molecular pathogenesis of heart disease has been driven by parallel advances in preclinical modeling, genetic manipulation, and imaging technologies coupled with rapid refinements in high-throughput sequencing technology. Together, these developments have permitted the integration of unbiased approaches with candidate gene–based reductionist strategies to interrogate cellular pathways in animal models of HF and in specimens from patients with HF. Simultaneously, the framework for understanding normal cardiac growth and development as well as physiologic myocardial function has been refined. With these advances, insights gained from genome-wide analyses of human disease and small animal preclinical studies can be tested in large animal models. As a consequence, a pipeline-based approach for the development and evaluation of therapeutic strategies has emerged.

The existing paradigm for deciphering the molecular basis of HF is based on a reductionist strategy to define events in myocyte and nonmyocyte cell types triggered by disease-related injury (e.g., ischemia and reperfusion, viral infection, chemotherapeutic agents) or biomechanically transduced due to changes in hemodynamic load (either pressure or volume overload). These stimuli elicit specific changes in gene expression, resulting in perturbations in proteins and signaling pathways that affect the structure and function of the heart. Preclinical model systems ranging from in vitro experimentation in isolated cardiomyocytes to in vivo studies in large animal models have been employed to dissect the molecular and cellular pathways involved. A clear advantage of large animal models is the close resemblance of the cardiac structure and function and coronary vasculature in these animals to those of the human heart. On the other hand, small mammals, such as mice, zebrafish, and invertebrates (e.g., Drosophila ), allow for genetic manipulation with progressive ease as one moves down the evolutionary tree. Investigative approaches have evolved from an early focus on the pharmacologic manipulation of specific pathways in large animals to experimentation involving gain of function and loss of function of candidate genes and/or proteins in small animals in order to recapitulate human pathology.

In vitro techniques in isolated cardiomyocytes have evolved from the development of isolated neonatal rat cardiomyocytes to studies in isolated adult cardiomyocytes and in reprogrammed induced pluripotent stem (iPS) cells differentiated into beating cardiomyocytes and cardiac microtissues. Neonatal rat and mouse cardiomyocytes continue to be widely used, as these cells are easily isolated and cultured. Also, they respond to hypertrophic stimuli with an increase in cell size associated with increased protein synthesis and changes in gene expression, mimicking the cardiomyocyte hypertrophic response in vivo. This model system allows the study of cellular changes occurring in hemodynamic overload–induced hypertrophy. A major shortcoming of neonatal cardiomyocytes, however, is the incompletely developed sarcomeric architecture and sarcoplasmic reticulum network. To overcome these limitations, techniques to isolate calcium-tolerant adult cardiomyocytes have been developed to allow for the measurement of contraction, relaxation, and calcium transients. These cells are also amenable to gene transfer with viral vectors. Given that the mouse is the predominant mammalian model for genetic manipulations, isolated field-paced cultured adult myocytes are an attractive model system for assessing the effects of genetic manipulations on cardiomyocyte function.

A major breakthrough in defining patient- and disease-specific alterations in cardiomyocytes was achieved with the observation that isolated somatic cells, such as fibroblasts obtained from a skin biopsy, can be reprogrammed with a cocktail of transcription factors to acquire the characteristics of stem cells. These induced pluripotent stem cells (iPS cells) can subsequently be transdifferentiated into beating cardiomyocytes in vitro, with the structural and functional characteristics of adult cardiomyocytes. When they are transdifferentiated from iPSCs derived from a human, these cardiomyocytes are uniquely suited to deciphering the biology of human cardiomyocytes and to dissecting the unique effects of human disease-causing mutations therein. Indeed, as this technology has become technically more accessible, there has been an explosion of studies that have utilized the iPSC platform to begin to elucidate the basis of human cardiac disease.

The ability to conduct genome editing with the CRISPR-Cas9 system (see further on), zinc-finger nucleases, and transcription activator–like effector nucleases (TALENs) offers tremendous promise for manipulating molecular pathways in patient-specific iPS cells (reviewed in Hockemeyer and Jaenisch )s. Also, given their extensive potential for self-renewal and differentiation as well as for reduced immunogenicity in an autologous setting, these cells have been investigated to determine their potential for developing cellular therapy for cardiac disease. Another exciting development has been the direct reprogramming of resident cardiac fibroblasts into cardiomyocytes with delivery of a cocktail of transcription factors in vivo and the discovery of RNA processing and splicing factors that regulate this process ( see also Chapter 2 ). This has set the stage for novel strategies for the in vivo manipulation of cardiac regeneration with contemporary genetic targeting techniques ( see also Chapter 3 ).

Although in vitro systems are well suited to the study of myocyte cell biology, in vivo modeling is required to determine the effect of disease processes on organ structure and function. The prerequisites for an ideal model system are as follows: (1) a high degree of similarity to human cardiac structure and function, (2) ease of surgical manipulation with development of structural and functional changes that mimic human pathology, (3) superior fidelity to the implementation of targeted genetic interventions to perturb molecular pathways and mimic human genetic alterations, and (4) suitability for application of analytic assays in the live organism to permit serial evaluation in a high-throughput fashion. None of the currently available model systems offers all these advantages, thus necessitating the use of combinations to interrogate the wide range of pathophysiologic, molecular, and cellular changes observed in cardiovascular disease.

Large animal models are well suited to studies involving disease-related stresses, such as valvular stenosis or insufficiency, ischemia/reperfusion, pressure overload, and cardiomyopathy (e.g., pacing-induced HF or coronary microembolization). These allow for evaluation of hemodynamic and neurohormonal events in disease progression; however, such animal models do not allow for genetic manipulation. Small mammals, particularly mice, have served as an almost ideal system for experimental in vivo studies. Techniques for genetic perturbations, surgical intervention, and the assessment of cardiac structure and function with noninvasive and invasive approaches have been developed over the last three decades. Despite persistent concerns regarding the translation of findings from the mouse to the human, many observations regarding disease pathophysiology mimic those noted in human disease. Indeed, data obtained in murine models are the backbone of our contemporary understanding of the molecular basis of HF. At the other extreme, model systems such as zebrafish and Drosophila are ideally suited for rapid and high-throughput modeling to unveil the effects of genetic perturbations; these models, however, can be less informative with respect to alterations in myocardial structure and function or circulatory pathophysiology.

A transgenic gain-of-function approach is typically employed to evaluate whether a gene or its product, by virtue of its structure or its functional involvement in a particular signaling pathway, is sufficient to stimulate myocardial pathophysiology. Forced expression in cardiomyocytes of proteins is conventionally achieved by driving their expression with cardiomyocyte-specific promoters, such as Mlc2v and αMHC. This strategy achieves a high level of gene expression in the early embryonic heart or starting at birth, respectively. For proteins that may have lethal effects following forced expression or when temporal evaluation of the effects of forced expression are to be studied, conditional bitransgenic systems are employed, whereby the expression of the protein of interest can be switched on or off with drug administration, typically tetracycline derivatives or mifepristone.

A loss-of-function approach is designed to determine whether a certain gene (or its product) is necessary for a specific phenotype. One approach is to forcibly overexpress a modified protein that has dominant-negative effects by virtue of its structure or function. The potential limitations of this methodology stem from the unpredictable effects of a modified protein, which may be difficult to discern experimentally, and the possibility of noncanonical effects of high-level protein expression, whereby even otherwise inert proteins may induce pathology. A more scientifically robust strategy is gene ablation, which has been achieved using homologous recombination in embryonic stem cells. This strategy has permitted the evaluation of nonredundant functional roles of mammalian proteins in normal development and homeostasis and in pathophysiologic processes relevant to human disease states. To overcome limitations pertaining to the systemic effects of germline gene ablation, tissue-specific ablation has been achieved in the heart using cre-lox (or flp-FRT: flippase, flippase recognition target) technology. Cardiomyocyte-specific gene deletion may be achieved in the embryo with Cre expression driven by the Nkx2.5 promoter (knocked into the Nkx locus) or conditionally at any age with Cre expression induced by tamoxifen treatment in Cre-ER (mutant estrogen receptor)–expressing (αMHC promoter driven) Mer-Cre-Mer transgenic mice. Simultaneously, a “tool box” has been developed to target diverse cell types at various developmental stages, driving an explosion of knowledge in cardiac development and regeneration.

In parallel, there have been exciting developments in tools to target resident cardiac fibroblasts, which form 10% of all cells in the heart and can be activated to transform into myofibroblasts to drive the fibrotic response. Specifically, tamoxifen-inducible systems have been generated driven by promoters for genes expressed in fibroblasts of epicardial origin with persistent expression in mature fibroblasts, namely encoding transcription factor 21 (Tcf21) and platelet-derived growth factor receptor-α (Pdgfr-α), as well as with periostin (Postn gene), which is expressed in activated cardiac fibroblasts (reviewed by Tallquist and Molkentin ). Coordinated international efforts have created repositories of targeted genes with an ever-expanding list of available targets. This is likely to facilitate further expansion of experimentation with these technologies in the future.

The discovery of the CRISPR-Cas9 system, a bacterial system for adaptive immunity, has spurred major breakthroughs in gene editing technologies, enabling precise targeting of the mammalian genome. Specificity is achieved in this system by small RNAs termed guide RNAs (gRNAs) that direct Cas-9 to a specific site in genomic DNA for inducing cleavage, which is subsequently repaired by nonhomologous end-joining and homology-directed repair mechanisms to result in deletions or specific targeted mutations directed by a homology repair template that is coadministered, respectively. This technique has been rapidly applied to facilitate gene targeting in iPSCs to discern the mechanistic basis of human cardiac disease and to correct mutations in engineered human tissue as a proof of principle of its therapeutic potential in a future human application. This technology has essentially become the preferred approach to rapidly and efficiently target genes for generating knockouts, knockins, and transgenic mice.

With the development of this technology, genetic manipulation offered a powerful experimental strategy to clarify the role of specific genes and proteins in cardiovascular homeostasis. Interactions among genetic changes and various stressors could be examined with the advent of microsurgical techniques to mimic human cardiovascular disease. Examples of such approaches include the induction of ventricular pressure overload or MI. Pressure overload can be induced by thoracic aortic constriction or pulmonary artery banding. Induction of MI can be performed either by reversible ligation or permanent occlusion of murine coronary arteries to simulate ischemia-reperfusion injury or permanent infarction, respectively. This can also be performed in a closed-chest model to minimize inflammatory changes related to open surgery, thereby more closely mimicking human disease. Miniaturization of invasive hemodynamic monitoring has been achieved with the development of micromanometer-tipped catheters for pressure measurement and conductance catheters for pressure-volume loop assessment of load-independent indices. Noninvasive assessment by echocardiography and magnetic resonance imaging has also advanced significantly for myocardial function and tissue characterization, and high-resolution positron emission tomography with computed tomography imaging has been miniaturized to evaluate various facets of cardiac metabolism in the mouse heart. Finally, telemetry-based cardiac rhythm monitoring has permitted rapid throughput evaluation of arrhythmic phenotypes in mutant mouse models.

The isolated perfused working heart preparation and the ejecting heart model are experimental approaches well suited to investigating cardiac function and metabolism in the setting of disease (e.g., ischemia-reperfusion injury) coupled with pharmacologic perturbations in genetically manipulated mice. Together, these approaches comprise a comprehensive “tool kit” that allows for the detailed evaluation of molecular pathways in the context of pathologic stress.

Molecular Determinants of Physiologic Cardiac Growth, Hypertrophy, and Atrophy

Based on early studies in rodents, cardiomyocytes have traditionally been regarded as terminally differentiated cells that rapidly exit the cell cycle early in the postnatal period. Cardiomyocytes manifest increases in cell size and nuclear division, leading to bi- and even multinucleated mature cells, but true cell division appears to occur at a low frequency. As a consequence, cardiomyocyte hypertrophy has been understood as the dominant response of adult cardiomyocytes to injury, as opposed to the hyperplasia observed in tissues with robust regenerative capabilities. Indeed, cardiomyocyte loss due to cell death has been considered largely irreplaceable. Recent work has challenged these notions, and accumulating evidence indicates that adult cardiomyocytes retain a modest capability of reentering the cell cycle for replication, occurring at a rate much lower than that observed prior to the early postnatal period ( see also Chapter 3 ).

Cardiac hypertrophy has been conceptualized as “physiologic” to indicate normal postnatal growth and the cardiac enlargement observed with the increased workload demands of pregnancy or exercise conditioning; conversely, “pathologic” hypertrophy is observed in response to disease-related stress, such as hemodynamic overload or myocardial injury. Hypertrophy serves to normalize wall stress occurring with increased hemodynamic load, thereby diminishing oxygen consumption, and is traditionally viewed as an “adaptive” response. In pathologic states, however, hypertrophy may be considered maladaptive, as it often progresses to a decompensated state, with the development of cardiomyopathy and HF.

Although these descriptive terms reflect the nature of the inciting stimulus and the probable outcome, it is the specific intracellular signaling events that are closely correlated with outcome. Indeed, the hypertrophic response may match the stimulus but not track its pathologic characteristics. In a study where intermittent pressure overload was induced with reversible transverse aortic constriction, quantitatively less severe hypertrophy was observed as compared with persistent pressure overload. However, the key pathologic characteristics of maladaptive pressure overload hypertrophy were comparable and resulted in functional decompensation with both intermittent and persistent pressure overload, suggesting that it is the nature of the inciting stress, not its frequency or intermittency, that is most relevant.

Normal embryonic and postnatal cardiac growth, termed cardiac eutrophy , and physiologic hypertrophy of the adult heart share important traits that distinguish physiologic from pathologic hypertrophy. Physiologic hypertrophy is associated with normal contractile function and normal relaxation. Myocardial collagen deposition is not observed, and capillary density is increased in proportion to the increase in myocardial mass. Additionally, favorable bioenergetic alterations are observed with enhanced fatty acid metabolism and mitochondrial biogenesis. Also, the characteristic expression of a “fetal gene program” seen in pathologic hypertrophy is not observed with physiologic hypertrophy. Physiologic hypertrophy is typically mild (∼10%–20% increase in mass over baseline) and regresses without permanent sequelae upon termination of increased hemodynamic demand. Indeed, induction of physiologic hypertrophy by exercise and molecular manipulation of cardiac growth signaling pathways induced primarily in physiologic hypertrophy (see further on) have been reported to prevent or ameliorate the effects of pathologic hypertrophy and HF.

Eutrophy occurs via activation of signaling pathways similar to those observed in exercise-induced hypertrophy ( Fig. 1.1 ). At birth, a dramatic increase in circulating thyroid hormone levels transcriptionally upregulates the synthesis of contractile and calcium handling proteins in the heart and induces a myosin heavy chain isoform shift. Concomitantly, the peptide growth factor IGF-1 is secreted primarily from the liver in response to growth hormone released from the pituitary gland, stimulating physiologic growth. An essential role for IGF-1 in normal growth is evidenced by growth retardation and perinatal lethality in IGF-1 and IGF receptor (IGFR)-1 null mice.

Development of physiologic cardiac hypertrophy in response to exercise is also triggered by IGF-1, levels of which are increased in trained athletes and in cardiomyocytes in response to hemodynamic stress. Indeed, IGF-1 signaling is required for exercise-induced hypertrophy; the hypertrophic response to swimming was completely suppressed in mice with cardiomyocyte-targeted ablation of the IGF-1 receptor. Interestingly, induction of IGF-1 production and secretion by cardiac fibroblasts is observed in pressure overload–induced hypertrophy mediated via the activation of the Kruppel-like transcription factor KLF5. This has been implicated in provoking cardiomyocyte hypertrophy by paracrine signaling to preserve cardiac function in the short term, possibly by maintaining an adequate “adaptive” hypertrophic response. Insulin signaling also transduces physiologic cardiac growth, in addition to governing metabolism, as mice with ablation of the insulin receptor manifest reduced cardiomyocyte size with depressed myocardial contractile function. Insulin receptor ablation as well as ablation of insulin receptor substrates IRS1 and IRS2, individually in cardiomyocytes, attenuate development of exercise-induced hypertrophy. Additionally, ablation of the insulin receptor exacerbates pathologic hypertrophy, suggesting an increased propensity for decompensation in the absence of protective physiologic hypertrophic signaling.

IGF-1 and insulin signaling converge on heterodimeric lipid kinases, termed PI3Ks ( see Fig. 1.1 ), and class I PI3Ks catalyze the formation of phosphatidylinositol-3,4,5-trisphosphate. PI3P recruits downstream effectors such as Akt via a PH-3 domain. Phosphoinositide phosphatases, namely phosphatase and tensin homolog (PTEN) and Inpp5f, extinguish PI3P signaling. Class IA PI3 kinases (PI3Kα, β, and δ) mediate signaling downstream of receptor tyrosine kinases, namely IGFR, the insulin receptor, and integrins (see further on). Class IA PI3Ks are heterodimers composed of a regulatory subunit (p85α or p85β, or their truncated splice variants p50α or p55α) and a catalytic subunit (p110α, p110β or p110δ). PI3Kγ, a class IB PI3 kinase, is composed of the p110γ catalytic subunit bound to either p84/87 or p101 regulatory adaptor and primarily signals downstream of G protein–coupled receptors (such as the β-adrenergic receptor) in cardiomyocytes. Although genomic ablation of PI3K (p110α) is embryonically lethal in a murine model at day 9.5 of gestation, expression of a dominant-negative mutant of p110α in the postnatal heart reduces adult heart size and blunts development of swimming-induced cardiac hypertrophy. Additionally, a gain-of-function approach with forced cardiomyocyte expression of p110α results in cardiac growth with characteristics of physiologic hypertrophy; ablation of PTEN kinase promotes cardiac growth, confirming a role for this pathway in physiologic cardiac hypertrophy. PI3K (p110α) is also essential for maintaining ventricular function via membrane recruitment of protein kinase B/Akt ( see Fig. 1.1 ). Ablation of Akt1 and/or Akt2 downstream of PI3K activation, as well as its activator PDK1, reduces cardiac mass, further emphasizing the essential role of this signaling pathway in normal cardiac growth.

C/EBPβ, a transcription factor repressed by Akt activation, was discovered in a genetic screen for transcriptional determinants of swimming-induced hypertrophy, and its conditional ablation resulted in increased cardiomyocyte proliferation and mild physiologic hypertrophy. This study suggested the exciting possibility that exercise may promote cardiomyocyte proliferation. C/EBPβ inhibition also ameliorated pressure overload induced hypertrophy, bolstering the paradigm that stimulation of physiologic hypertrophy pathways may offer benefits by hitherto undiscovered mechanisms.

Metabolic reprogramming and angiogenesis are other essential components of the physiologic hypertrophic response. Indeed, targeted ablation of LKB1 (an activator of AMPK, which is activated in response to energy deficit) and of vascular endothelial growth factor (VEGF), a proangiogenic signaling protein (in vegfb- null mice), provokes reduced postnatal cardiac size and vascular rarefaction. Remarkably, cardiomyocyte-specific ablation of the glucose transporter (GLUT4) triggers cardiomyocyte apoptosis and interstitial fibrosis in the setting of swimming exercise. These events are associated with Akt dephosphorylation by increased levels of phosphatase protein phosphatase 2A (PP2A), pointing to a role for glucose uptake in maintaining mitochondrial metabolism and cardiomyocyte survival during physiologic heart growth. Intriguingly, activation of AMPK (which senses the ratio of AMP/ATP and ADP/ATP, i.e., the energetic state of the cell) is observed with physiologic hypertrophy and has been demonstrated to inhibit pathologic hypertrophy. The mechanism for this observation appears to involve inhibition of O-GlcNAcylation of proteins by AMPK-mediated inhibition of glutamine:fructose-6-phosphate aminotransferase (GFAT), which reduces the supply of N-acetylglucosamine.

Despite these observations, it is critical to proceed cautiously with strategies to stimulate physiologic hypertrophy in hopes of ameliorating pathologic hypertrophy. This is underscored by the observations that forced cardiac expression of IGF-1 initially produces functionally compensated ventricular hypertrophy that evolves, over time, into pathologic hypertrophy with fibrosis and systolic dysfunction. Also, forced cardiomyocyte expression of Akt, its pivotal downstream signaling effector, results in compensated hypertrophy, which transitions to cardiac failure due to inadequate angiogenesis. Mechanistically, it is plausible that exuberant cardiomyocyte hypertrophy, whether initially physiologic or pathologic, may outstrip concordant angiogenesis and exceed the individual’s capacity for oxygen and nutrient delivery sufficient to meet the demands of the hypertrophied myocyte. This may explain the rare clinical observations of irreversible ventricular hypertrophy and dilation observed in athletes after long-term participation in endurance sports with a strength component, such as rowing and cycling.

Cardiomyocyte size is remarkably plastic ; the heart undergoes atrophy with a reduction in hemodynamic load or metabolic demand, as may occur in conditions of weightlessness and extended bed rest, as with spinal cord injury. This may involve the inhibition of growth pathways, as seen with suppression of insulin signaling in cancer-induced cardiac cachexia in mice, which was reversible with exogenous insulin treatment. A parallel induction of proteolytic and catabolic pathways, such as activation of the ubiquitin-proteasome system, facilitates the atrophic response. Indeed, activation of muscle ring finger1 (Murf 1) ligase has been demonstrated to be essential for the regression of pressure-overload hypertrophy, and loss of function of Murf1 and Murf2 in the heart lead to spontaneous cardiac hypertrophy and HF via the downregulation of E2F transcription factor signaling. Cardiac growth and atrophy are also observed as physiologic responses to metabolic demand, as demonstrated in fascinating studies in Burmese pythons, wherein a large meal induces a 40% increase in cardiac mass rapidly over 48 hours and a 50% increase in stroke volume. These dramatic changes revert to baseline values as the meal is digested over a period of days. The increased cardiac mass is due to cardiomyocyte hypertrophy (and not hyperplasia); it is associated with transcriptional induction of synthesis of contractile elements with activation of PI3K-Akt and mTOR pathways and is mechanistically driven by increased circulating free fatty acids and stimulation of fatty acid uptake and oxidation in the myocardium.

Molecular Determinants of Pathologic Hypertrophy

Cardiac hypertrophy occurring in response to injury, hemodynamic overload, or myocardial insufficiency has been conceptualized as a compensatory response to normalize wall stress as defined by LaPlace’s relationship (S = PR/2H, where S is wall stress [force per unit area], P is the intraventricular pressure, R is the radius of the ventricular chamber, and H is the wall thickness). Hypertrophy is measured at the organ level using electrocardiographic, echocardiographic, and/or magnetic resonance imaging (MRI) indices of myocardial mass and cardiac size. In pressure overload, cardiomyocytes enlarge in the short axis by adding sarcomeres in parallel. In volume overload, sarcomeres are added in series, lengthening the cell. Hypertrophic remodeling is then characterized as concentric (increased wall thickness without dilation) or eccentric (chamber dilation with a mild increase in wall thickness), respectively, with pressure and volume overload stress.

Transcriptional Regulation of Pathologic Cardiac Hypertrophy

A hallmark of pathologic hypertrophy in the adult heart is reexpression of embryonic cardiac genes, a process often referred to as the “fetal gene program,” as this aspect of the cardiac response to stress or injury recapitulates aspects of cardiac development. The earliest detectable change (within hours of increasing afterload or stimulating cultured cardiomyocytes to hypertrophy with norepinephrine) is induction of regulatory transcription factors c-fos, c-jun, jun-B, c-myc, and Egr-1/nur77, and heat shock protein (hsp) 70, thereby mimicking changes observed with cell cycle entry. Induction of these “early response genes” drives expression of other genes in the fetal program. The prototypical gene, atrial natriuretic factor (ANF), is expressed early during heart development through the coordinated interactions of Nkx2.5, GATA-4, and PTX transcription factors, but only in the atria of normal adult hearts. A robust induction of ventricular ANF expression (and related brain natriuretic peptide [BNP]) is observed in pathologic hypertrophy and HF. In fact, increased BNP secretion from the stressed heart is used widely as a biomarker of HF. Other elements of the fetal gene program induced in pathologic hypertrophy and HF encode sarcomeric genes, such as βMHC, MLC2v, α-skeletal actin, and β-tropomyosin, proteins that are prominent in embryonic, but not adult, ventricle.

Cardiac gene expression in pathologic hypertrophy is driven by the reactivation of many developmentally regulated transcription factors ( Fig. 1.2 ). GATA4 and GATA6 are two such zinc finger DNA-binding transcription factors that are individually essential for heart tube development and myocyte proliferation during embryogenesis. Both of these transcription factors are also essential for homeostatic expression of various cardiomyocyte genes in the adult heart, including ANF, BNP, ET-1, α-skeletal actin, αMHC, βMHC, cardiac troponin c, and AT1Ra; their individual or combinatorial ablation results in progressive cardiomyopathy. In pressure-overload hypertrophy, cardiomyocyte-specific deletion of either GATA4 or GATA6 markedly attenuates development of pressure-overload hypertrophy, resulting in accelerated decompensation. Importantly, GATA4 signaling may be essential for sustaining angiogenic responses in pressure-overload hypertrophy via VEGF activation, underscoring its critical role in this setting.

Prohypertrophic signaling pathways, such as activation of MAP kinase cascades downstream of Gαq-coupled α-adrenergic (α1A) receptors, trigger activation of GATA4. GATA4 also complexes with other transcription factors such as Nkx2.5, Mef2, a coactivator, p300, SRF, and nuclear factor of activated T-cells (NFAT) to effect cardiac gene expression ( see Fig. 1.2 ). Serum response factor (SRF) is another cardiac-enriched transcription factor that coordinately induces sarcomerogenesis with other transcription factors, including SMAD1/3, Nkx2–5, and GATA4. Cardiomyocyte-specific ablation of SRF in the adult heart results in progressive development of cardiomyopathy with disorganization of the sarcomeres and HF. SRF also interacts with myocardin and HOP transcription factors. HOP antagonizes SRF signaling, and conditional ablation of HOP results in aberrant cardiac growth with evidence of both lack of myocyte formation and excess cardiomyocyte proliferation. Myocardin acts as a cardiac and smooth muscle–specific coactivator of SRF and is essential for embryonic cardiomyocyte proliferation and the maintenance of normal sarcomeric organization in the adult heart.

There has been much speculation about the functional impact of reexpression of the fetal gene program in pathologic hypertrophy. It has been postulated that an increased ratio of β-MHC/α-MHC isoforms impairs myocardial contractility due to relative inefficiency of the β-isoform, culminating in reduced sarcomeric shortening, prolonged relaxation, and adverse remodeling. Although this may have a major impact in the adult mouse ventricle, which predominantly expresses the faster α-isoform, its relevance in the adult human heart, in which 90% of the myosin heavy chain is the β-isoform, is less clear. In contrast, downregulation of the gene encoding the SERCA affects the activity of this important Ca ²⁺ pump, which is responsible for the rapid diastolic reuptake of calcium into the sarcoplasmic reticulum. This has been established as an important mechanism for the contractile dysfunction observed in human HF.

Multiple studies focusing on transcriptome profiling of cardiac pathology have identified a panoply of gene expression changes in both human HF and animal models of pathologic hypertrophy ( http://www.cardiogenomics.org ). These myocardial mRNA signatures and the different patterns of gene expression in normal, early-failing, late-failing, and recovering hearts might be useful as prognostic biomarkers and could help guide therapeutics. Also, in the past decade, there have been dramatic advances in sequencing technologies. A rapidly accumulating list of individual variations in genetic sequence (termed single-nucleotide polymorphisms ) and their combinations, through genome-wide association studies (GWAS), holds promise to uncover novel targets for further mechanistic exploration in HF.

Another layer of complexity in gene regulation has been revealed by studies of microRNAs (miRNAs). These are short, noncoding, naturally occurring single-stranded RNAs that regulate gene expression negatively by promoting the degradation of mRNAs and/or inhibiting mRNA translation, thereby suppressing protein synthesis ( see Fig. 1.2 ). miRNAs are abundantly expressed in the myocardium and differentially regulated in animal models and human HF. miRNAs are essential for homeostatic gene regulation, as targeted cardiomyocyte-specific ablation of Dicer, an enzyme essential for miRNA processing, causes HF with profound transcriptional dysregulation of cardiac contractile proteins. Upstream signaling pathways alter miRNA expression in response to developmental cues and hypertrophic stimuli, such as the regulation of miR-1 and miR-133 from a common precursor by SRF and MEF2. These transcription factors control cardiomyocyte proliferation during development and downregulate miR-1 and miR-133, facilitating prohypertrophic pathways in swimming-induced and pressure overload hypertrophy. Another group of miRNAs is localized to the myosin heavy chain genes, miR-208a, miR-208b, and miR-499 (termed MyomiRs). These have been shown to regulate transcriptional repressors and thyroid hormone signaling to transduce the changes in myosin heavy chain gene expression observed in pathologic hypertrophy. Importantly, targeted deletion of miR-208a prevented reexpression of the fetal gene program and attenuated pathologic remodeling with pressure overload. Upregulation of the miR-15 family at birth plays an important role in suppressing cardiomyocyte proliferation in the immediate postnatal period. Indeed, miR-15 inhibition with locked nucleic acids stimulates continued cardiomyocyte proliferation after birth and induces the regeneration of myocardium when administered post-MI in adult mice. This restores the regenerative capacity otherwise observed only in 1-day-old pups. Therefore targeting the regulation of gene expression with miRNA-targeted strategies holds promise in developing novel therapeutics to treat HF.

Cellular Mechanisms of Impaired Cardiomyocyte Viability

Hypertrophy of the ventricular myocardium is an independent risk factor for cardiac death and is observed with near-universal prevalence in patients with HF. LV hypertrophy may, in part, underlie the diastolic dysfunction observed in HFpEF. In addition, in patients with HFrEF, pathologic LV hypertrophy progresses inexorably from a compensated or nonfailing state to dilated cardiomyopathy and overt failure. It is important to recognize that although the essential feature of cardiac hypertrophy is increased cardiomyocyte size/volume, other myocardial alterations—such as fibroblast hyperplasia, deposition of extracellular matrix proteins, and a relative decrease in vascular smooth muscle and capillary density —also contribute to the progression from functionally compensated pathologic hypertrophy to overt HF.

Activation of Cell Death Pathways ( see also Chapter 2 )

Evidence for cardiomyocyte “dropout” due to death or degeneration is observed in failing hearts and in pathologic hypertrophy before the development of cardiomyopathy. The extant literature indicates that hypertrophied cardiomyocytes are likely to die from a number of different processes, and cardiomyocyte death can be a causal factor in cardiomyopathic decompensation, although the relative contribution of specific pathways appears to vary with pathologic context. Cardiomyocyte death may be programmed (i.e., cell suicide) by apoptosis, necrosis, or autophagy or it may be accidental (as in conventional necrosis due to interruption of vascular supply). Histologic evidence for all forms of death is seen in end-stage human cardiomyopathy.

The term apoptosis is derived from the Greek expression for “the deciduous autumnal falling of leaves” ( apo means “away from” and ptosis means “falling”); it is an orderly and highly regulated energy-requiring process that mediates targeted removal of individual cells during development without provoking an immune response. The rates of apoptosis, measured as apoptotic indices (e.g., the number of TUNEL-positive nuclei/total nuclei), parallel the rates of cell division and are highest in the outflow tract (∼50%); intermediate in the endocardial cushions, which are sites of valve formation, and in the LV myocardium (10%–20%); and lowest in the right ventricular myocardium (∼0.1% at birth). Both cardiomyocyte apoptosis and mitosis in the LV myocardium virtually cease soon after birth and within the first 2 weeks of life in the right ventricle. Abnormal persistence of apoptosis in right ventricular myocardium contributes to the pathogenesis of arrhythmogenic right ventricular dysplasia, a disorder caused by mutations provoking abnormal localization of desmosomal proteins leading to the suppression of Wnt signaling. This stimulates de novo adipogenesis from resident cardiac stem cells to cause right ventricle–specific cardiomyocyte apoptosis and fibrofatty replacement associated with arrhythmias and sudden death. Apoptotic cardiomyocytes are extremely rare in normal adult myocardium (1 apoptotic cell per 10,000–100,000 cardiomyocytes). Together with reactivation of the “fetal gene program” in hypertrophied and failing hearts, the prevalence of cardiomyocyte apoptosis is markedly increased in chronic cardiomyopathies. Apoptotic cardiomyocyte death may also play a role in the transition of pressure overload hypertrophy to dilated cardiomyopathy. Emerging evidence suggests that necrosis, a form of cell death associated with rupture of the plasma membrane and inflammatory infiltration, may also be programmed and controlled by the cell. The death machinery that orchestrates these processes exhibits cross talk at multiple levels, whereby features of either or both forms may be dominant in a specific pathophysiologic setting.

Cell death may be initiated by ligand-dependent signaling from the cell exterior through the extrinsic or receptor-mediated pathways; conversely, it may occur by induction of the death machinery within the cell through mitochondrial pathways ( Fig. 1.3 ). Sustained experimental pressure overload is sufficient to induce expression of the prototypical death-promoting cytokine, TNF. This molecule signals via the type 1 TNF receptor (TNFR1) to stimulate cardiomyocyte hypertrophy and apoptosis and provoke contractile dysfunction. A potentially causal role for elevated levels of this cytokine is suggested by elevated TNFα plasma levels that are correlated with the degree of cardiac cachexia in end-stage HF. Death receptor signaling downstream of TNFR1 is triggered by TNF binding to a receptor homodimer, resulting in formation of the death-inducing signaling complex (DISC) with recruitment of adaptor protein FADD and caspase 8 (an upstream member of a family of executioner cysteine proteases). Activated caspase 8 then cleaves caspase 3 and Bid, a proapoptotic Bcl2 family member. Activated caspase 3, the effector caspase, activates a nuclear DNAase (CAD-caspase activated DNAse), resulting in internucleosomal cleavage of DNA and chromatin condensation. The generation of truncated tBid links the extrinsic pathway to activation of the intrinsic pathway. This leads to their simultaneous activation in TNF-induced cardiomyocyte apoptosis in the setting of TNF-induced depletion of antiapoptotic signaling proteins in the mitochondria. Although elevated TNF levels signal to provoke myocardial hypertrophy with increased cardiomyocyte apoptosis, adverse ventricular remodeling, and systolic dysfunction in rodent models, endogenous TNF signaling is cytoprotective in ischemia-reperfusion injury. This indicates that precise context-dependent modulation of TNF signaling may be required to attenuate cell death in pathologic hypertrophy.

The intrinsic mitochondrial pathway of programmed cell death is triggered by stress-induced upregulation or activation of BH3 domain-only prodeath proteins ( see Fig. 1.3 ), such as with Gαq/PKC/SP-1–mediated transcriptional induction of BNIP3L/Nix in pathologic hypertrophy. Nix targets and permeabilizes mitochondria to induce the release of prodeath mediators such as cytochrome c. Nix-induced mitochondrial permeabilization may be direct via outer membrane permeabilization (MOMP) or may occur via Nix targeting to the ER, triggering ER-mitochondrial cross talk to provoke calcium overload and opening of the mitochondrial permeability transition pore (MPTP). In the cytosol, cytochrome c binds to the adaptor protein Apaf-1 (apoptotic protease activating factor-1), resulting in sequential recruitment and cleavage-mediated activation of caspase 9 and caspase 3. Together with the release of AIF (apoptosis inducing factor) and endoG from the mitochondrial intermembranous space, this results in activation of PARP and DNA cleavage in the nucleus ( see Fig. 1.3 ) and cell death. Stress-induced cardiomyocyte death is an important determinant of pathologic hypertrophy and decompensation, as cardiomyocyte-specific ablation of Nix attenuates pressure overload–induced ventricular remodeling and programmed cell death.

Calcium overload–induced opening of MPTP is also implicated in programmed necrosis. Inhibition of MPTP formation with mitochondrial ablation of ppif (the gene encoding cyclophilin D), a critical mitochondrial matrix component of the MPTP, reduces cell death provoked by ischemia-reperfusion injury. However, a subsequent study reported a critical physiologic role for cyclophilin D-mediated mitochondrial Ca ²⁺ efflux in maintaining adequate mitochondrial function to match metabolic demand. In this study, mice with cyclophilin D deficiency developed exaggerated hypertrophy and HF with exercise or pressure overload, which was corrected by transgenic restoration of cyclophilin D levels. This emphasizes a paradigm similar to that observed with TNF signaling, in which a precise modulation of the MPTP will be required to therapeutically address programmed necrosis in HF. Indeed, mitochondrial calcium overload can be countered by the mitochondrial sodium-calcium exchanger (encoded by Slcb81); stimulating its activity reduced the propensity for mitochondrial permeability transition and prevented necrotic cell death in models of HF.

In pathologic hypertrophy, cardiomyocyte necrosis may also be triggered by ischemia due to mismatch between the degree of hypertrophy and vascular supply. An adequate blood supply for growing myocardium is critical to normal cardiac function, and capillary density is closely coupled to myocardial growth during development. As discussed earlier, pathologic hypertrophy is associated with relative vascular rarefaction (as compared with normal capillary density observed with physiologic hypertrophy), with decreased capillary density, decreased coronary flow reserve, and increased diffusion distance to myocytes. There is a temporal correlation between decreased capillary density and cardiomyocyte “dropout” during decompensation in both human disease and experimental animal models. Indeed, in pressure overload–induced hypertrophy, impairment of angiogenesis with cardiomyocyte-specific GATA4 ablation (and resultant deficiency of angiopoietin and Vegf) and sustained increases in p53 (with suppression of HIF1α signaling) accelerate the progression to decompensated HF. Conversely, restoration of angiogenesis with p53 inhibition or exogenous administration of proangiogenic agents markedly attenuates cardiomyocyte death in this setting, confirming a central role for this paradigm in decompensation. Also, studies have shown that enhanced generation of placental growth factor (PGF) by various cell types in response to hypertrophic stress in the myocardium stimulates angiogenesis and cardiomyocyte hypertrophy via IL6 generation, and this angiogenic response can be stimulated further to prevent decompensation of pressure-overload hypertrophy. Additionally, during the physiologic hypertrophy of pregnancy, proangiogenic signaling via peroxisome proliferator activating factor γ coactivator α (PGC1α) counters the antiangiogenic effects of a VEGF inhibitor secreted by the placenta, termed sFLT-1. Deficiency of the angiogenic response provokes peripartum cardiomyopathy, pointing to a critical need for coordinated increases in angiogenesis and myocyte size to maintain physiologic hypertrophy.

Cell Survival Pathways

Countervailing pathways that promote cell survival play critical roles in regulating cell death during pathologic cardiac remodeling. One such pathway is elicited by the IL-6 family of cytokines, comprising IL-6, cardiotrophin, and LIF and signaling via a shared membrane receptor, namely the gp130 glycoprotein, which has intrinsic tyrosine kinase activity ( Fig. 1.4 ). Binding of ligand induces gp130 homodimerization or oligomerization with β-subunits of other cytokine receptors, stimulating autophosphorylation on receptor cytoplasmic tails and activating intrinsic tyrosine kinase activity. This permits the binding of adaptor proteins Grb2 and Shc to SH2 binding domains to activate Janus kinases (JAKs), which phosphorylate STAT transcription factors. Activated STAT dimers then migrate to the nucleus to (1) regulate gene expression; (2) activate SH2 domain-containing cytoplasmic protein tyrosine phosphatase (SHP2), which subsequently activates the MEK/ERK pathway; and (3) activate the Ras/mitogen-activated protein kinase, leading to MAPK activation and extracellular signal-regulated kinase (ERK) signaling ( see Fig. 1.4 ). A simultaneous transcriptional upregulation of SOCS family proteins via STAT signaling provides for feedback inhibition of these pathways ( see Fig. 1.4 ). Gp130 activation is transiently observed early after the onset of pressure overload, and this survival pathway is deactivated during the transition to failure, possibly related to interruption of gp130-JAK-STAT signaling by stress-induced SOCS3 and the resulting suppression of STAT3 signaling. Importantly, ablation of gp130 provokes exaggerated cardiomyocyte apoptosis with fulminant HF and pressure overload hypertrophy.

Autophagy (in Greek, meaning to eat oneself) is a lysosomal degradative pathway essential for breaking down intracellular constituents to recycle defective proteins and mitochondria and to replenish nutrients during periods of deprivation. Autophagy of intracellular constituents is critical for cardiomyocyte survival during the early perinatal period of starvation, as mice deficient of key autophagy proteins, such as ATG5 and ATG7 (both essential for the initial step of autophagosome formation), develop fatal myocardial deterioration. Autophagy is also essential for cardiomyocyte homeostasis in adult mice, as adult-onset cardiomyocyte-specific ablation of ATG5 results in fulminant HF, and postnatal onset of ATG7 deficiency leads to the insidious development of cardiomyopathy with aging. Stimulation of autophagy enhances cardiomyocyte survival during myocardial ischemia, as mice expressing a dominant-negative mutant of AMPK are unable to mount an effective autophagic response to ischemia and manifest larger infarct sizes as compared with controls. Interestingly, perinatal onset of ATG5 deficiency is well tolerated yet induces rapid decompensation in the setting of pressure overload. Foci of degenerated cardiomyocytes with autophagic vacuoles are observed in human dilated cardiomyopathy and aortic stenosis, suggesting that this pathway plays an important role in human pathophysiology.

Interestingly, other studies have suggested that the induction of autophagy may be potentially deleterious in certain forms of cardiac injury. In particular, mice with haplo-insufficiency of Beclin-1 demonstrate decreased infarct size with ischemia-reperfusion injury and attenuation of pressure overload–induced adverse ventricular remodeling. Whether these effects are due to a role for Beclin-1 in autophagy or involve Beclin-1–induced signaling via other mechanisms remains to be fully delineated. Therefore, whereas evidence suggests a predominantly pro-survival role for autophagy, the therapeutic targeting of autophagy in HF will require careful experimental validation in large animal models.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here