Large animal models of pressure overload-induced cardiac left ventricular hypertrophy to study remodelling of the human heart with aortic stenosis

2.1 Molecular mechanisms of pathological hypertrophy

Upon various triggers, conditions of increased workload can occur. Genetic predisposition, pressure or volume overload or ischaemia in the cardiac muscle are some of those. Such stimuli, usually result from diseases such as, hypertrophic cardiomyopathy, hypertension, valvular heart disease (including AS, the most common valvular disease in the developed world),16 coronary artery disease and others, and are responsible for exposing the heart to chronic stress. The main consequence of this chronic stress is the initiation of an adaptive, with tendency to maladaptive, or, directly, maladaptive mechanism, in order to normalize the pump activity and maintain the balance between oxygen demand and supply.10^,17^,18 Thus, the cardiac muscle grows in mass, causing an increase in the size of the organ, mimicking a physiological hypertrophic response. During the process, wall thickening or thinning may appear, accompanied with unusual fibrosis, scar tissue formation, and immune infiltration. At this stage, a pro-inflammatory response triggers various molecular cascades, initiating a non-reversible procedure. Furthermore, fibrosis and extracellular matrix (ECM) deposition cause systolic and diastolic dysfunction, impeding electric conduction.15 This transition to a pathologic, maladaptive condition is characterized by compromised heart function.

2.2 Types of pathological hypertrophy

Two main types of pathological hypertrophy have been observed and are correlated to the stimuli that causes them. In the first case, an eccentric response appears when increased blood volume enters the heart ventricle. Volume overload triggers cardiomyocytes to stretch longitudinally. This results in an elongation of the cardiac muscle. Wall thinning and dilation characterizes those hearts, in which the contractility and overall function progressively attenuates. On the other side, even though it is known that the heart has the ability to calibrate small increases in the ventricular pressure by stretching of the cardiac muscle and higher force of contraction, according to Frank Starling law, when constant pressure overload occurs, due to chronic hypertension, valvular disease and other pathologic conditions, different cardiomyocyte molecular mechanisms are activated.19 This results in increased cellular mass, at a cross-sectional point of view, which is responsible for an increase in heart mass, observed as extended wall thickening, representing an effort to produce more tension. Wall thickening, often in parallel to a decrease in ventricular chamber volume, is accompanied with a wall stiffening, decreasing the elasticity of the muscle, and the functionality of the organ. This is the case of a concentric hypertrophic response. In summary, small cumulative cellular reactions and alterations are causing a changing effect in the organ architecture, in a concentric or eccentric way, which will finally result in cardiac dysfunction.10^,12^,14 Other triggers, such as inherited diseases or myocardial infarction may also cause local or general hypertrophic responses, but without leading specifically to one of those two classifications.

2.3 Dynamic cell population state under pathologic conditions

A normal cardiac ventricular muscle consists of many cell types, such as mural cells, including pericytes and vascular smooth muscle cells, fibroblasts, endothelial cells (ECs), immune cells, surrounding and assisting the cardiomyocytes, the essential cells for cardiac contractile function.20^,21 Due to the pathology of a maladaptive hypertrophic heart, this analogy changes, with differentiated percentages of cell types, having an effect on the features of hypertrophic response. The cellular composition turns into a more fibrotic and inflammatory content, surrounded by altered ECM. The percentage of cardiomyocytes tends to decrease, due to increased non-cardiomyocyte migration, as well as their number, due to cell death. Cell proliferation from cardiac progenitors does not have an important effect in the adult heart, as their percentage is very low and the whole organ renewal capacity attenuated.22 Mechanical triggers not only initiate intracellular alterations in cardiomyocytes but they also stimulate cardiomyocytes to start a neurohumoral reaction, attracting multiple cell types, with a protective intention. However, chronic stimulation and subsequent drastic changes in cardiomyocytes and cellular processes such as inflammation, fibrosis, and angiogenesis will ultimately lead to a dilated and failing heart.

3.1 Ploidy and multinucleation

Cardiomyocytes are the main regulators of a hypertrophic response. They suffer alterations in their structure and architecture at multiple levels, which, will cause humoral reactions with, apart from autocrine, also paracrine and endocrine-stimulating properties. Initially, polyploidy and multinucleation are characteristics of adult cardiomyocytes, not common in most of the other cell types. Human adult cardiomyocytes, which are known for their low proliferative capacity, appear to be mononuclear but polyploid.23^,24 This happens in a postnatal period, during maturation, as in the fetal period, normal cell cycle and hyperplasia characterizes the cardiac cells. During this postnatal period, normal heart stiffening is observed with alterations in ECM, accompanied by cardiomyocyte maturation and cell cycle exit.13 Upon a hypertrophic stimulus, in disease conditions, ploidy seems to increase, as a mechanism to facilitate additional protein transcription due to high metabolic demands.25 Ploidy in hypertrophic cardiomyocytes, and especially right ventricular cardiomyocytes, was positively correlated with cellular hypertrophic response.26 Hypertrophy does not seem to have an effect in multinucleation in human, in contrast to observations in animal models, which appear to have binucleated or multinucleated cardiomyocytes.25 This may be a limitation in the translation of animal experimentation to humans.

3.2 Metabolic modifications and mitochondria

Energy metabolism of cardiomyocytes differs from most of other cell types due to their increased demand of energy, required for contractility. Fetal cardiomyocytes are known to mainly use glucose, as their main substrate for energy production, which switches after birth to fatty acid β-oxidation.27 During hypertrophy, cardiomyocyte metabolic demands increase, as stated above, corresponding to an excessive need for anabolic products, like nucleic acids, proteins and lipids, as well as energy.17 Hypertrophied cardiomyocytes are subjected to a metabolic switch, similar to the embryonic condition, from fatty acids to glucose.2^,28^,29 Interestingly, it was suggested that this switch exists independently of the energy demand, probably correlating to biosynthetic/anabolic demands of the cell.30 Furthermore, fructolysis, fructose metabolism, provides has been implicated in the pathology of hypertrophy.31^,32 Mitochondria, the responsible organelles for energy production, undergo changes during the transition to pathologic hypertrophy and energy metabolism switch. Normally, they are characterized by high density, occupying in approximately 30% of the cardiomyocyte mass and located in the perinuclear area. During hypertrophy, they are subjected to morphological and functional alterations, resulting in an impaired function.33 Their dimensions increase, in contrary to their volume density, while their morphology transforms from ovate to elongated-filamentous, with ruptured membranes and structures.34^,35 Quality control mechanisms such as mitochondrial biogenesis, mitophagy, fusion and fission appear to be imbalanced, while mitochondrial Ca²⁺ uptake, important for ATP production and cardiac contractility, is abnormal.33 Finally, increased release of Reactive Oxygen Species (ROS) from the damaged mitochondria, exceeding the antioxidant capacity of the cell, leads to oxidative stress, which may further stimulate apoptotic cascades and maladaptive extracellular remodelling.36^,37

3.3 Subcellular homeostasis

Excessive ROS generation in cells can trigger mechanisms like autophagy. Autophagy is considered to be the housekeeping mechanism of the cell, recycling damaged macromolecules and organelles inside it, or occasionally driving the cell to death.27 This homeostatic mechanism can provide biosynthetic components to the cell and has been found to play a controversial role in cardiac hypertrophy.38 Initially, autophagy, triggered by metabolic demands, provides a compensatory mechanism to haemodynamic stressors facilitating the heart function and morphology alterations.39 During the passage to the pathologic state of a failing heart, autophagy has been found to be importantly overactivated in human, fact that may be associated with increased mitochondrial dysfunction and energy crisis, worsening progressively the prognosis of the disease.3^,27 In contrast, some studies in mice strongly support that autophagy induction has a reversing effect in hypertroph, while autophagy deficiency is associated with hypertrophic reactions and dysfunction of mitochondria, calcium handling, and contractility mechanisms.40–43 This may be attributed to an early stage of the disease or organism-specific implicated mechanisms. Proteostasis, the mechanism responsible for protein balance inside the cells, plays an important role in cardiac pathology.44 A dysfunctional ubiquitin–proteasome system, the system that subcellularly invigilates the protein function, is suspected to be a hallmark of cardiomyopathies, especially during stress conditions, nevertheless, further investigation on the topic is required.5^,45

3.4 Oxidative stress

It has been demonstrated that different cardiovascular abnormalities such as functional hypoxia, metabolic derangements, uncoupling of mitochondrial electron transport, and inflammation produce oxidative stress in the hypertrophied failing heart. A small amount of oxyradicals is generated at the onset of the hypertrophic process but these radicals are readily removed by the endogenous scavengers to maintain cardiac redox homeostasis and proper function. However, prolonged exposure of hypertrophied hearts to pathological stimuli promotes the development of oxidative stress as a consequence of functional hypoxia due to constriction of the coronary arteries, reduction in the capillary density, inflammation, metabolic alterations, and mitochondrial dysfunction. Oxidative stress has been suggested to cause Ca²⁺-handling abnormalities in association with subcellular remodelling, defect in energy production, inflammation, apoptosis, fibrosis, and loss of cardiomyocytes. These abnormalities are considered to result in cardiac dysfunction and heart failure (reviewed in46). Regarding AS, oxidative stress is mainly associated with calcific AS by correlating with the degree of inflammation and calcification of human aortic valves. In fact, ROS mediate the increase in BMP2 expression and signalling, favouring osteogenesis (reviewed in46).

3.5 Cell homeostasis and death

Cell death is another reparative process, removing damaged or dysfunctional cells, usually triggered by mechanical forces, oxidative stress, hypoxia, and others, common characteristics of a hypertrophic response.47 It is known that cardiac homeostasis is preserved with the balance of CM regeneration and cell death, including apoptosis, necrosis, and autophagy associated with the ubiquitin-proteasomal pathway. Necrosis, a non-regulated and non-physiological form of cell death, does not seem to play an important role in hypertrophic hearts.48 On the other side, apoptosis has a controversial role in the progress of a hypertrophic response. In the initial stages, hypertrophy promotes cell survival and evasion of apoptotic mechanisms.49 Large, aged cardiomyocytes are more susceptible to apoptosis, especially when they suffer of mitochondrial dysfunction, as a result of chronic growth signalling exposure.48 They are progressively led to atrophy and death, facilitating muscular remodelling.48^,50 At late stages of the disease, cellular apoptosis increases due to the increasing number of damaged cells. Organ failure is then irreversible, with the causality relationship of failure and apoptosis being still unclear. Finally, in a study in AS patients, autophagic cell death appeared to be more significant than the programmed cell death via apoptosis.3 Further investigation of the homeostatic mechanisms of the heart could elucidate important time points indicative of the transition from compensatory to pathologic condition and failure.

3.6 Contractile activity and sarcomeric reorganization

Mechanotransduction is a machinery enabling cardiomyocytes to sense and respond to mechanical load.11 The main sensors of this system are the three cellular compartments: the sarcomere, the intercalated discs and the sarcolemma. During increased chronic workload, those sensors are triggered to promote changes that will facilitate a hypertrophic response. Sarcomeres are structures responsible for force generation in cardiomyocytes, that by actin and myosin filament interactions promote contraction reactions inside cardiomyocytes.51^,52 In case of a workload stressor, a cellular response is initiated leading to an increase in the amount of connected sarcomeres.25 This increase can happen longitudinally or in parallel, resulting in longer or wider cardiomyocytes, respectively. Longer cardiomyocytes are observed in hearts with eccentric hypertrophy, while cardiomyocytes with larger cross-sectional area (CSA) are characteristic of a concentric response.53 Intercalated discs, the cardiac specific junctions, are responsible for intercellular communication related to mechanical and electrical coupling.52 Those terminal structures consist of adherence junctions, where the myofibrils anchor, desmosomes, for mechanical stabilization and gap junctions, for ion transfer.54 It is shown that ultrastructural changes affect the assembly of sarcomeres, while, in hypertrophic conditions, intercalated discs appear to have a denser architecture, fact that requires further research.53^,55 Finally, the sarcolemma consists of many proteins and complexes, associated with mechanotransduction, and in hypertrophic responses appears to be altered indicated by an irregular cell surface, t-tubule dilation, and enhanced Ca²⁺ efflux, through stretch-sensitive ion channels.17^,52^,56^,57

Despite the various pathological stimuli, there are many common features in the hypertrophic response in different cardiac diseases. In addition to increased cardiomyocyte mass, sarcomere rearrangement, and ECM deposition, other common features have recently been appreciated including inflammatory signalling, immune cell activation, and decreased angiogenesis. Numerous cell types are involved in orchestrating this complex pathological response. The heart consists of a heterogeneous population of cells including cardiomyocytes and non-cardiomyocytes, and it is now clear that intercellular signalling and communication between these cell types is critical in the pathophysiology of ventricular hypertrophy and remodelling. Although we will not elaborate on the different intercellular communication pathways contributing to the cardiac response to stress, here below we discuss the main different cells and processes that occur during the development of the disease and that involve other cell types rather than CMs alone.

4.1 Inflammation

Upon a hypertrophic stimulus, a strong inflammatory response is also observed, which is triggered in the first stages of pathologic hypertrophy. Due to haemodynamic overload and subsequent mechanical stress, mostly cardiomyocytes, but other cardiac cell types also, release autocrine and paracrine factors such as hormones, cytokines, growth factors (GFs), and chemokines. These factors, that may be meant to contribute to the adaptation to stress, when sustained, become maladaptive mediators contributing to decompensation to heart failure (reviewed in58^,59).

Initially, the Sympathetic Nervous System (SNS) is activated, as a compensatory mechanism to the hypertrophic trigger, followed by activation of Renin-Angiotensin-Aldosterone-System. Neurohumoral systems are found to interact with the immune system and trigger fibrosis, while chronic activation of those systems is correlated with the maladaptive responses and characteristics of the failing heart.60^,61 Inflammatory cytokines target many cell types in the hypertrophic heart, such as cardiomyocytes, immune cells or fibroblasts, in a complex process that remain poorly studied. However, they positively correlate to the disease progress and severity.62 Some studies show evidence that cardiac inflammation is directly initiated by the cardiomyocyte as being a potent source of inflammatory cytokines. Early after transverse aortic constriction (TAC), cardiomyocytes are known to release cytokines such as monocyte chemoattractant protein 1, chemokine (C-X-C-motif) ligand1, interleukin-18, -6, and -32 (IL-18, IL-6, and IL-32), which may all contribute to the pro-inflammatory environment of the stressed heart.62^,63 Other researchers reported cytokine immunoreactivity also in the vasculature in a TAC model, implicating ECs in the initiation of immune cell infiltration in the heart.64^,65 Additionally, the natriuretic peptide system appears to affect the inflammatory modulators in a double and contradicting way, having an impact in the following cardiac events. It includes atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP) and C-type natriuretic peptide, as major molecular regulators, each of which acts through different mechanisms during remodelling processes in the failing heart.61

4.2 Immune activation and infiltration

It is known that cardiac tissue hosts an important amount of resident immune cells. Macrophages, B cells, T cells, Dendritic Cells (DCs), Natural Killer cells (NK), neutrophils, and mast cells are the immune cellular subpopulations, present in the healthy, as well as the diseased heart.66–68 During a hypertrophic reaction, they become active, serving different aspects of a strong inflammatory response and attracting additional immune subpopulations with significant roles in the progress of the disease. In pressure overload induced hypertrophy, they infiltrate the heart in a time-dependent manner, determining the cardioprotective, or cardiodestructive effects.68 Neutrophils are the first responders of the innate immune system during inflammation. In hypertrophic animal models of pressure overload, their population remained increased during the whole study period, and it was positively correlated to the mass index.69 They have a pivotal role in recruiting macrophages and DCs, by the secretion of immunomodulating factors. Mast cells are responsible for vascular homeostasis and angiogenesis in heart, but also play a modulator’s role during hypertrophy. With the process of degranulation, they release inflammation mediators and their accumulation has been positively correlated with adverse remodelling of pressure overload models.70 DCs are antigen-presenting cells, serving as a bridge between innate and adaptive immune system. Evidence show that they play a pathogenic role. Although they are found to increase during the progress human heart failure, no statistical significance was proven.71 Those subcellular populations are proposed to have detrimental role in cardiac hypertrophy due to pressure overload, while others appear to have time-dependent and controversial roles.68

Macrophages and their different subpopulations play key roles in the progress of the inflammatory reaction, upon hypertrophic stimulation. Cardiac-resident macrophages have cardiac pro-repair and angiogenesis regulatory capacity, while a smaller subpopulation of them is capable of activating T cells by antigen presentation, serving as immunomodulators.58^,68 Furthermore, blood-circulating monocytes are reported to reach the heart, during health and in higher degree in disease conditions, and differentiate in macrophages with a pro-inflammatory character.66^,72 The recruitment of non-resident macrophages [specifically the C-C chemokine receptor 2⁺ (CCR2⁺) population] in the heart is known to have a pivotal role in the adverse cardiac remodelling induced by pressure overload.73^,74 NK cells are known by their cytolytic and tissue homeostatic function, as well as their immunomodulatory properties of macrophage and T cell activation, in the cardiac tissue. Decrease in number or dysfunction of NK cells was positively correlated with chronic heart failure in human.75 The adaptive immune system also participates in the inflammatory reaction of a pathologically hypertrophic heart. T cells have the ability to activate or suppress inflammation, according to their subtype, and are known to participate in the pathophysiology of a diseased heart.58 Activated T cells migrate from lymphoid organs to the heart, where they expand in response to pressure overload.67 B cells, on the other side, are well known for antibody production. In the diseased heart, they release pro-inflammatory molecules, by which LV function is attenuated, and activate the complement, acting by damaging the cardiomyocytes.66^,76 Finally, eosinophils, able to control cardiac death and fibroblast activity, as well as a specific subset of T cells, NK T cells, provide evidence of cardioprotective actions.67^,68

4.3 Fibrosis

Cardiac fibrosis, where activated myofibroblasts are the main effector cells, is primarily driven through activation of resident interstitial cell populations. Although, considered a protective mechanism upon tissue or ECM damage occur, aiming at wound healing and repair, complete restoring of the initial tissue functionality is seldomly achieved.77 Several other cell types, including cardiomyocytes, ECs, pericytes, macrophages, lymphocytes, and mast cells may also contribute to the formation of fibrotic lesions by secreting fibrogenic mediators and matricellular proteins, or just by modifying specific communication axes that will influence the fibroblast’s phenotype.78 It was suggested that the first stage of a hypertrophic response is accompanied by loss of cell–cell interactions and protective factors secreted by fibroblasts.79 The mechanisms of induction of fibrosis and the contribution of the various cell types to the process are dependent on the underlying cause of myocardial injury.80 Activation of neurohumoral pathways stimulates fibroblasts both directly and by affecting certain immune cell populations. Immune cell infiltration, a strong inflammatory response, and increased oxidative stress due to ROS accumulation, are major regulators of the fibrotic response that characterizes the maladaptive hypertrophic heart.36 During chronic hypertrophic stimuli, the small number of cardiac-resident fibroblasts become activated and acquire the ability to proliferate37 and differentiate into myofibroblasts.81 Myofibroblasts, apart from the ability to produce excessive amounts of collagen and other ECM elements, are recognized as cellular types with both fibrotic and contractile properties, playing an important contribution to muscle repair.78^,82^,83 Furthermore, they present high resistance to cell death, as a survival mechanism in the abundant cell death signals microenvironment of a wound, fact that may have negative consequences, as persistent cardiac fibrosis can be detrimental. The extensive fibrosis and ECM deposition are main factors that directly stiffen and impair contractility of the cardiac muscle, both characteristics of the failing heart.81 Attenuation or inhibition of the fibrotic response, through pharmacological (personalized) interventions, is believed to have a positive effect in heart failure.78^,81^,84

4.4 Angiogenesis and hypoxia

Vasculature density and angiogenesis in the heart have been positively correlated with physiological hypertrophy that is observed in a growing organism towards adulthood or in cases of athletic or pregnancy hypertrophy, attributed to the increased CM size and consequent oxygen and metabolic product demands.85 In healthy conditions, CMs have the ability to secrete angiogenic factors and promote angiogenesis, through an important crosstalk with ECs, the most abundant cell type in the cardiac muscle.86^,87 Although it is also proven that angiogenesis alone is able to stimulate hypertrophic responses in cardiomyocytes, this review mainly focuses on cases of an already existing hypertrophic stimulus, apart from angiogenesis. Miscommunication and disproportional CM growth, compared to the capillary net/vasculature bed leads to hypoxic conditions within the cardiac muscle, contributing to the transition into a pathologic hypertrophied heart.15^,85 During this stage, capillary rarefaction and endothelial dysfunction cause underperfusion of the heart, leading, apart from hypoxia, to fibrosis, CM death, and triggering of inflammation.88 Moreover, hypoxia is known to contribute to the metabolic switch in CMs and facilitate further pathological alterations. Tube-like structure formation as well as EC number appears to decrease with the progress of heart failure. The latter may be attributed to molecular signals that either inhibit endothelial proliferation, promote cell death or the Endothelial to Mesenchymal transition.87^,88 Recently, our own study revealed a new process by which, stressed cardiomyocytes can influence the surrounding endothelium through release of exosomes that are enriched in miR-200c-3p. Once stressed, cardiomyocyte-derived exosomes are taken up by cardiac ECs, these will lose their angiogenic capacity and become dysfunctional contributing to the severity of the cardiac phenotype caused by prolonged increased pressure overload of the LV.89 All these, indicate that restoring the angiogenetic ability of the failing heart may have a positive impact in delaying or even preventing the progress of the disease.

6.1 Ferret

Among the animals used to study LVH induced by pressure overload is the ferret (Mustello purtorius furo). In a study conducted in 1980,99 adult male animals were submitted to Banding of Ascending Aorta (AAB), by a circular wire clip, covered with polyethylene. The constriction occluded 70% of the aorta, achieving an overall LV mass increase of 45% of LV, already after two weeks and determined by the Left Ventricle/Body Weight (LV/BW) ratio. In parallel, hypertrophic growth regression was studied in a different group of animals, that were allowed to recover from the banding period of two weeks, for six more weeks after debanding. The results showed that LV mass was able to regress and adjust to similar values in comparison to the control animals. Haemodynamic data indicated an immediate increase in Left Ventricular Systolic pressure (LVSP), upon banding, but no significant differences in the Left Ventricular End Diastolic pressure (LVEDP). Heart rates, immediately upon banding, were slightly increased in both banding and regression groups, but their mean rates indicated no significant difference compared to controls. In a tissue/cellular level, it was shown that both the banded and regression animals had greater fractional cell volume of myofibrils in endocardial regions compared to the anterior papillary muscle, a difference that was not observed between different heart regions of control animals. Subcellularly, the morphology and cell fraction volume were examined, showing no significant differences rather than the mean relative mass of mitochondria to be greater in banded animals compared to controls, similarly to the mean relative mass of the myofibrils. This fact was attributed to the increased need of energy by the producing and consuming structures, i.e. mitochondria and myofibrils, as a response to the hypertrophic stimulus of pressure overload.

6.2 Lapine

New Zealand White Rabbit (Oryctolagus cunigulus) has, lately, become very common as an experimental animal model to study CVDs. It can combine the benefits of smaller animal models, while rabbits are phylogenetically closer to human, with more physiological and genetic resemblance.100 Experiments conducted in adult as well as neonatal/infant animals achieved a pathologic hypertrophy formation and were able to correlate with the human pathologic hypertrophic condition, caused by AS, by presenting comparable results. Nevertheless, the size and life span differences to human, deprive lapine models of high representativeness necessary to model cardiac diseases.

Haemodynamically, it has been shown that AAB, resulting in 75% of reduction of the aortic CSA, in adult rabbits, can cause significantly elevated LVSP, while preserving the animal viability.102 Pressure overload for the whole period of the constriction (up to 24 weeks) was confirmed by weekly echocardiography, while decreased to normal upon removal of the constriction. Until a period of two weeks after the banding, Fraction Shortening (FS) increased, starting to diminish afterwards. In contrast, Left Ventricle Systolic and Diastolic Dimensions appeared to decrease until week 2, but both showed an extravagant increase between weeks 4 and 24. During the first four weeks after banding, LV and Interventricular Septum (IVS) thickness were progressively increasing, while right after appeared to attenuate, in accordance with the diastolic wall thickness to LV cavity radius (h/r), marking the passage from a concentric compensated hypertrophic response to an eccentric decompensation state. No statistical significance was shown by the slight increase of heart rate in the banding animals and the total LV Weight to Body Weight ratio (LVW/BW), was found to be at the peak of 92,3% at week 8. In a different, ‘non-surgical’, attempt to induce LV hypertrophy in adult rabbits, by the placement of a partially deployed stent in descending aorta by catheterization, after three months an increase of 10% in Heart weight to Body weight ratio (HW/BW) was achieved, indicating the possibility of a more chronic model of the disease.107 Experiments in neonatal (10 days old) rabbits indicated a compensated state, four weeks after the banding, by a preserved FS and highly increased LV mass to volume ratio, which both appear to be reduced at week 6, representing the decompensated phase; during those time points the HW/BW appeared to be equal and higher than the control animals.103 Recently, it was proven that severe TAC (75%) can establish heart failure in a period of 6 weeks in neonatal rabbits, while a moderate constriction (50%) failed to do so.104 While the pressure after ligation was increased 26 times in six weeks the LV wall thickness also appeared to increase with time, being two times higher than in the control animals. Moreover, the progress from a concentric compensated to an eccentric decompensated heart, in these animals, was found to happen between week 4 and 6, indicated by alterations in LV dimension, EF, and FS.103^,104

Histology results indicate that AB in rabbits can cause myocyte and myofibrillar disarray, accompanied by interstitial fibrosis. Collagen Volume Fraction indicates two to three times increase in collagen, which does not normalize to control levels upon removal of the banding, in contrast to other parameters determined.102^,104 Myocytes showed increased thickness, with an increased CSA by 38.4% after three months in moderate model of chronic hypertrophy and almost 200% in a severe neonatal model, after six weeks.104^,107 An old study, focusing on the mitochondria of CMs, indicated an increase of their function following pressure overload which, gradually decreases in time.105 Mitochondrial alterations can already be observed 2 days after banding and include larger size, decreased number of cristae or occasionally emptiness. At day 7, completely destroyed or swollen mitochondria are present with significantly augmented mass. At day 45, a relative normalization in size and structure of the organelles is observed, in contrast to the observations at eight months where the cell’s mitochondrial fraction appears to be smaller, with many elongated incompletely divided complexes. Moreover, experimental models of rabbits have evidenced a decrease in expression of Vascular Endothelial Growth Factor (VEGF) and CD34, and increased Pleiotrophin, suggesting impaired angiogenesis and an imbalance between the microcapillary network and the cellular needs to cope with hypertrophic growth and remodelling.107 Finally, increased pro-inflammatory factors such as cytokine IL-18, IL-18 Receptor, Matrix metallopeptidase 9 (MMP9) and ANP, and decreased antagonist IL-18 Binding Protein, during all the stages of remodelling induced by pressure overload, are believed to play a role in endothelial cell death and other pathological mechanisms of heart failure.103

6.3 Feline

Cats (Felis catus) have been also used as large animal models in variable conditions. In CVDs, and especially cardiac hypertrophy due to pressure overload, young cats were used to reproduce a mild state of the disease that increases in severity as the animal grows. The slowly evolving constriction mimics the progressive nature of the chronic human compensated hypertrophy towards a decompensated state. The majority of the experiments were towards generating a model of Heart Failure with preserved Ejection Fraction (HFpEF). Feline appear to be suitable animal models, as their physiological properties can be comparable to human with translational power, even though their small size may be a limitation for the use of human equipment and measurements. Nevertheless, metabolic differences, such as concentration of high density lipoproteins, may affect responses to drugs or the development of a disease in a way that is not common in humans, confounding experimental results.106

In adult mice, banding of the aortic arch with a nylon thread, causing an average pressure gradient of 64 mmHg, after six weeks of banding, resulted in a significant decrease of Left Ventricular Inner Dimension in systole, accompanied by an increase in LV wall thickness, EF and FS, compared to the baseline measurements. Left Ventricular Inner Dimension in diastole, Cardiac Output and Stroke Volume remained unchanged. Altogether, these data suggest a concentric hypertrophic response, that can be classified as level 1 of the New York Heart Association.108 Young cats subjected to AAB with a clip of a standard diameter of 2,8–3 mm, presented signs of hypertrophy after one week of banding, including increased IVS wall thickening and Left Atria (LA) size, while LV diameter and FS remained comparable to sham animals, in a period of two to three months of banding. Animals were categorized, one week after the banding, according to the constriction degree with a cut-off of 85%; animals with a constriction greater than 85% presented a steadily increased pressure gradient, around the banding, which was almost doubled compared to sham after two to three months. Animals with a constriction of less than 85%, also presented increased LA size and pressure gradient in the banding site, compared to sham, but still at a lower degree than the ones with more severe constriction.109 Another study aiming at comparing sex-related differences in HFpEF and where loose constrictions were performed in the aorta of male and female young cats, resulted in similar phenotype in both sexes without any disturbances in the physiological growth of the animals. In detail, four months post-surgery the pressure gradient remained the same in all the banded animals and HW/BW and LV wall thickness were increased, compared to sham animals. EF indicated a small but significant decrease only in male animals. LVEDP was also higher in males but this increase did not reach the significance cut-off.111 In these studies, while there was a sex-dependent increase in CM CSA compared to sham, in female a decrease in CMs was observed, accompanied by increased endothelial cell number. Moreover, sex-independent significantly high endocardial fibrosis developed, while increased epicardial fibrosis was only significant in male animals with many of the ECM proteins, such as periostin and collagen VI, to be upregulated.111 Regarding the contractility properties, a depression in magnitude contraction of the myofibrils was observed, with a tendency to increase proportionally to the constriction degree.109 In females, the isovolumic relaxation was found to be prolonged compared to sham and males, while resting tension was significantly increased only in males. Furthermore, the credibility of hypertrophic response in this animal model was proven by RNA sequencing for hypertrophic markers including Myosin Heavy Chain 7, Solute Carrier Family 2 Member 2, Actin Alpha 1, and Solute Carrier Family 2 Member 4, which appeared to be elevated in the banded animals compared to sham. Finally, there are indications of fetal gene program activation, and many altered CM gene sets, which seem mainly to occur in male. Similarly, macrophage-associated activation of heart growth, tissue regeneration, and leucocyte differentiation signalling were found to be upregulated in male, but depressed in female. In contrast, fibroblasts appeared to have a common response mechanism, in both sexes, in terms of gene activation.111 The sex-related activation of specific gene sets may elucidate the reasons of the different prevalence of disease observed in human male and female.

6.4 Canine

Dogs (Canis lupus familiaris) used to be one of the most common animal models in the field of biomedical sciences during the past years. Due to their abundancy in the human environment, they are easy to access and approach, but also facilitate the understanding of pain or discomfort expressions, necessary for animal welfare. Furthermore, canine organ physiology, and in particular the cardiovascular system, presents many similarities to human physiology, allowing the utilization of same medical instruments, such as human catheter and echocardiography instruments.112 Being phylogenetically closer to human when compared to smaller animals, they are more likely to share common cellular and molecular mechanisms both in healthy and disease conditions. Lately, due to the increase of ethical concerns and restrictions, this model tends to be less used and is being substituted by other methods. Importantly and similarly to feline models, dogs do not develop common comorbidities that are observed in human CVDs, such as atherosclerosis, and therefore, may not be suitable for the study of particular cardiac conditions.106

Nevertheless, large studies in dogs have been essential to achieve a better disease representation, while avoiding mortality factors, such as aortic rupture, congestive heart failure and others. A 3-year-long large study, apart from establishing a progressive aortic constriction model in adult animals, concluded that animals presenting high baseline wall stress are more prone to heart dysfunction.113 Comparisons between chronic hypertrophy, during normal growth of young animals, and hypertrophic stimuli during adulthood, indicate a similar hypertrophic degree in all banded animals, increase in coronary perfusion pressure proximal to the banding site and attenuated minimum coronary resistance. The myocardial blood flow was, however, only significantly increased in the young models compared to their respective controls.114 Furthermore, regression experiments in banded puppies with normal FS indicated that a significant increase of more than 100% in LV mass and three times in pressure gradient, after 10 months, could be reversed with tendency to normalize, upon removal of the band, presenting an important decrease in both pressure gradient and LV mass.101 When a fixed-diameter band was placed in young animals for 12 months, two different types of cardiac hypertrophy developed, varying between 40 and 100% of increase in LV mass. One presented compensated heart failure with an LV/BW ratio 7.7 g/kg, LVEDP of 8 mmHg and mean systolic pressure of 202 g/cm², while the other, decompensated, presented values of 9.8 g/kg, 25 mmHg, and 313 g/cm², respectively, in addition to a decrease in subendocardial blood flow.112

Cardiomyocytes of banded animals under the same conditions, that presented compensated or decompensated hypertrophic response after 12 months, appeared to reach a plateau in their CSA between six and nine months, with the CMs from failing hearts to grow first in size, followed by those from compensated. In the same study, a decrease in circumferential fibre shortening was observed in failing hearts and an increase in compensated hearts, compared to control, after 12 months.112 Regarding the contraction abilities of CMs from puppies subjected to banding for 15 months, it appears to be impaired at baseline but also upon Ca²⁺ trigger, in accordance to a decreased ability of L-type Ca²⁺ channels to release Ca²⁺. The number of receptors of SERCA2a and RyR2 was also assessed, showing a significant decrease of 68 and 35%, respectively, and proteins such as RyR2 and phospholamban appeared hyperphosphorylated.110 Studies involving a chronic model of fully compensated hypertrophic hearts, revealed β-adrenergic receptors with higher density of binding sites but with lower affinity and decreased myocyte norepinephrine, indicating the implication of SNS.124 Later, in another study, a decrease of 59% in guanil nucleotide binding protein (G_s) was demonstrated, justifying the observed deficiency in G_s coupling. The decreased adrenergic responsiveness and defects in the adenylate cyclase system were attributed to the G_s coupling deficiency, in the sarcolemma of a hypertrophied heart.125

6.5 Ovine

Lambs and sheep (Ovis aries) have also been subjected in aortic banding procedures in order to reproduce models of LVH related to AS. These animals are manageable and easily operated, with nutritional and housing tolerance, go through wound healing without complications and display decreased risk of arrhythmogenic incidences during open-heart surgeries. Interestingly, it was observed that intraspecies interaction acts in a motivational way regarding animal recovery after surgery, limiting the factor of space availability.126 The large size and anatomy relatively similar, to human allow for chronic and translational instrumentation of the animals.127 Also, compared to other experimental animals, such as dogs and pigs, and despite similar degree of anatomical differences, sheep seem to develop more consistent degrees of LVH126 and have a moderate body mass increase wile ageing, which facilitates surgical procedures.98

A large experiment was performed to evaluate AAB in juvenile sheep as a model of LVH induced by AS. The model resulted in a condition of progressive AS with low probability of complications during the surgical procedures.126 When subjecting both lambs to six weeks of banding, both groups developed a similar increase in LV/BW, as wells as similar pressure gradients between the LV and aorta, similar LVEDP and heart rates. Despite the similar conditions, the sheep group showed impaired cardiac function, while the lamb group remained similar to the control (sham) group, revealing age-related differences in response to cardiac stress.108 These observations could be attributed to the hyperplastic capability of the juvenile heart, to the different degree of neovascularization or to altered calcium handling properties between the two groups. Nonetheless, those results indicate the existence of a possible confounder, related to maturity of the myocardium, which in many, if not most, studies were not taken into account. A series of studies in adult animals, where an inflating device was used to partially and gradually constrict the aorta, demonstrated a reproducible model in which the LVH, evaluated by Left Ventricular mass index, is established three to four weeks after the constriction, with stable cardiac function and progressive LV dilation up to seven to eight weeks, were the failure stage initiates.128–130 The failure stage was reflected by decreased FS, the animals’ inability to move, unwillingness to eat and tachypnea. Ultimately, a more chronic model, using the same device, was achieved, with the peak of LVH at approximately 32 weeks after the constriction, LV dilation starting at 103 and decreased EF at 117 weeks. Notably, removal of the hypertrophic stimulus by deflation of the device resulted in restoration of EF to control levels.131

Assessment of the vasculature bed capacity, during hypertrophic stimulus, in young sheep, revealed an increase of 32% in the number of capillaries in LV, while capillary density, coronary flow reserve, and minimal coronary resistance remained stable, in comparison with control animals. In the same study, when protamine, a protein that correlates to neovasculature impairment during hypertrophy, was administered, the subendocardial capillary density was reduced and the minimal coronary resistance was increased, with the LVH degree remaining the same compared to the control LVH animals.121 As the experiment involved young animals, with increased capacity of neovascularization, further investigation of the mechanisms involved in the process may unveil new therapeutic targets and approaches. Adult animals, in turn, revealed a significant increased expression of the pro-apoptotic genes Bax and Fas, upon the development of hypertrophy, while the anti-apoptotic gene Bcl-2 were not altered.128 This increased Bax to Bcl-2 ratio observed during pathologic hypertrophy, positively correlates with the expression of p53, a direct regulator of the anti-apoptotic family of Bcl-2. Furthermore, Bcl-xL, another member of the Bcl-2 family, was found to progressively increase during the process and was associated with LV dilation and the failing stage, suggesting its role in the transition to failure.129 Apoptosis was assessed, by investigating the Expression levels of caspases 3, 8, and 9, supported further implication of mitochondrial- and death receptor-mediated pathways. Caspase 3 was progressively increased to almost eight times in all stages of the disease and, in particular, in the failing stage; caspase 8 was found almost two times overexpressed in the failing stage and caspase 9 progressively increased its expression up to approximately six times through all the stages of the disease.130 The role of matrixins, GFs and inflammatory mediators, in the ECM changes during the progress of the disease was also investigated. MMP-1 and MMP-3 were significantly increased during cardiac dilation and MMP-2 was increased in the failing stage, but were decrease again upon removal of the hypertrophic stimulus. In turn, VEGF and Fibroblast growth factors-2 were found to significantly decrease towards failing, while VEGF increased again at the recovery stage. Regarding inflammation, Interferon alpha-2 was decreased along with cardiac hypertrophic growth, in contrast to the soluble ligand sCD40l that increased during the dilation stage.131

Sheep and lamb models are currently being used and constantly being improved to study different aspects of cardiac disease, and as they exhibit important cardiac pathology differences, similar to other animal species, the choice of model should always be based on the specific scientific question.

6.6 Porcine

Porcine (Sus scrofa) is a very common and highly efficient model that can predict and resemble human conditions and responses.115 Pigs display similar heart physiology, vasculature and function, disease progress, gene sequence, and chromosome analogy to human, while availability of tissue samples never being an issue.7^,118^,123 Comorbidities related to CVDs in human, are also observed in pigs, indicating similar reactions and underlying mechanisms, as well as a highly resembling immune system.122^,132 Furthermore, due to their size, the animals can be subjected to surgical practices with human equipment and conditions, mimicking procedures that are performed in patients. The downside is that research in pigs takes longer, requires expensive surgical equipment, specialized facilities, and expertise, with a very limited number of institutions being able to accommodate such procedures.117^,120

Studies describing surgical protocols of aortic constriction have demonstrated hypertrophic and fibrotic responses in animals 12 weeks after the intervention, with transition to a decompensated heart happening between weeks 4 and 6 and cardiac dilation establishing at the 8th week.116^,119 Furthermore, disappearance of muscle layer in the mitral valve may contribute to the onset of cardiac dysfunction.119 Immediately after banding, increased LVESD and myocardial strain were detected but again normalized to baseline 12 h later, similar to the increased LVEDD that was normalized 3 h later. Systolic wall thickness of the LV, declined during the first 6 h, but was significantly increased after 72 h, suggesting an acute cardiac adaptation to pressure overload.133 High adaptability of the cardiac muscle was also indicated by the active compensating remodelling that could initially prevent the increase in systolic pressure, accompanied by elevated active tension; leading though to a consequent decrease in diastolic stress but increase in diastolic pressure, during a period of approximately one month. Diastolic stress of the LV was attributed to cardiac stiffness and not fibrosis.134 Placement of an inflating cuff in the aorta of young animals, for 30 days, causing a 45% increase in LV/BW, indicated a concurrent increase of 16% weight of the right ventricle (RV) and decreased myocardial blood flow of the LV, during exercise, in comparison to control animals. Furthermore, minimal coronary vascular resistance, specifically in the endomyocardium, was significantly increased in the hypertrophic hearts compared to their respective controls.135 Transplantation of mesenchymal stem cells (MSCs) overexpressing VEGF, through the venous coronary system, in banded animals resulted in a significantly higher myocardial blood flow compared to banded, with normal MSCs or without transplantation, and sham animals, indicating the important role of VEGF in regulating myocardial blood flow during pathologic hypertrophy.136 A more chronic model of five months, suggested that a progressive stenosis is responsible for mean circumferential wall thickness and isovolumic relaxation constant (tau) increase, LV and RV mass increase and fibrosis, as well as a secondary pulmonary artery pressure increase.137

Investigation of the metabolic energetics of hypertrophied pig CMs, by magnetic resonance imaging, of animals with concentric hypertrophy and HFpEF, demonstrated the characteristic metabolic switch from fatty acids to glucose substrate.138 Furthermore, severe cardiac dysfunction was correlated with reduced ATP hydrolysis, decreased myocardial phosphocreatine to ATP ratio (phosphocreatine/ATP) and Creatine Kinase-mediated ATP production. Data indicated that ATP hydrolysis rates, which were proven to increase linearly with cardiac workload, dynamically change during the different stages of the disease.139 Immediate responses of CMs implicated a decreased contractility during the first 6 h, that was later normalized and progressively increased 4 and 7 days later, and a 13-fold angiotensinogen, 112-fold ppET-1 and 37-fold Insulin-like Growth Factor Receptor peak mRNA overexpression, reported at 2, 3, and 24 h, respectively. Taken the findings together, it was suggested that GFs provide an initial support in the CM contractility as acute response to the hypertrophic stimulus.133 In later stages of chronic models, high expression of ANP, BNP, and miR-29a and decreased miR-1 and SERCA2a were demonstrated, accompanied by increased circulating neutrophil gelatinase-associated lipocalin levels, indicative of fibrosis, and NT-pro-BNP levels, five months after the banding.137 Endothelial dysfunction in this model has been correlated with G_i and G_q protein-mediated relaxation of the coronary arteries, and a decrease in endothelial relaxation in response to serotonin and bradykinin (Bk) was also observed. In addition, no significant decreases were detected in the endothelium-derived hyperpolarizing factors pathway, in contrary with the plasma NO levels that appeared declined, indicating a contribution to the progression of heart failure, as NO is known for its antihypertrophic effect.140 Concerning the ECM, a study reported unchanged collagen concentrations in banded animals with a degree of 98% increase in HW/BW, four to six weeks later, compared to controls.141

Despite all the scientific advantages of using pig models to study cardiac disease, logistics remains and will remain an issue for many institutions. Furthermore, experiments in these animals are very costly not only due to the costs of breeding and maintaining the animals but any pharmacological study will require higher amounts of drugs, oligonucleotides, and reagents which also makes it often impossible to perform well-designed studies with significant statistical power.

Acknowledgements

L.D.W. acknowledges support from ERA-CVD JCT2016 EXPERT, the Dutch Heart Foundation, Dutch Federation of University Medical Centers, ZonMW and the Royal Netherlands Academy of Sciences. L.D.W. was further supported by grant 311549 from the European Research Council (ERC), the VICI award 918-156-47 from The Netherlands Organization for Scientific Research (NWO) and the Marie Sklodowska-Curie grant agreement no. 813716 (TRAIN-HEART). P.C.M. is supported by a Dutch CardioVascular Alliance (DCVA) awarded to the Phaedra consortium as well as the Impulse Grant 2018 awarded to the Phaedra IMPACT consortium (CVON-2018-29) and is an Established Investigator of the Dutch Heart Foundation (NHS2015T066). P.C.M. was further supported by the H2020 Twinning project RESETageing (GA 952266) and the TKI-LSH project MEDIATOR (LSHM21016) from Health-Holland.