Circulatory pathways

The fetus receives oxygenated and nutrient‐rich blood from the placenta via the umbilical vein and ejects desaturated blood through the umbilical arteries to the placenta, thus, the placenta, not the lung, serves as the organ of respiration. Blood flow largely bypasses the lungs in utero, accounting for only about 7% of the fetal combined ventricular output [1]. Pulmonary vascular resistance (PVR) is high, and the lungs are collapsed and filled with amniotic fluid. This forms the basis for fetal circulation, which is a parallel circulation, rather than the series circulation seen postnatally. Three fetal circulatory shunts exist to carry better‐oxygenated blood from the umbilical vein to the systemic circulation: the ductus venosus, ductus arteriosus, and foramen ovale (Figure 9.1A) [2]. Approximately 50% of the umbilical venous blood, with an oxygen tension of about 30–35 mmHg, passes through the ductus venosus, and then goes into the right atrium. There it streams preferentially across the foramen ovale, guided by the valves of the sinus venosus and Chiari network into the left atrium. Thus, the brain and upper body preferentially receive this relatively well‐oxygenated blood, which accounts for 20–30% of the combined ventricular output. Blood returning in the inferior vena cava represents about 70% of the total venous return to the heart, and two thirds of this deoxygenated blood passes into the right atrium and ventricle. About 90% of the blood flows through the ductus arteriosus to supply the lower fetal body.

After birth, there is a dramatic fall in PVR and an increase in pulmonary blood flow, with inflation and oxygenation of the lungs (Figure 9.2) [3]. The placental circulation is removed, and all of these changes lead to the closure of the ductus venosus, constriction of the ductus arteriosus, and reversal of pressure gradients in the left and right atria, leading to the closure of the foramen ovale. This leads to a state called the transitional circulation (Figure 9.1B), which is characterized by high pulmonary artery pressures and resistance (much lower than in utero, however), and a small amount of left‐to‐right flow through the ductus arteriosus. This is a labile state, and failure to maintain lower PVR can rapidly lead to reversion to fetal circulatory pathways, and right‐to‐left shunting at the ductus arteriosus and foramen ovale. This maintenance of fetal circulatory pathways is necessary for survival in many CHDs, particularly those dependent on a patent ductus arteriosus (PDA) for all or a significant portion of systemic or pulmonary blood flow, or atresia of atrioventricular valves. Maintenance of ductal patency with prostaglandin E1 (PGE₁) is crucial in these lesions. In a two‐ventricle heart with large intracardiac shunts, maintenance of the fetal circulation leads to right‐to‐left shunting at the foramen and ductal levels, and thus hypoxia results. Conversion to the mature circulation (Figure 9.1C) in the normal heart occurs over a period of several weeks, as the PVR falls further, and the ductus arteriosus closes permanently by thrombosis, intimal proliferation, and fibrosis. Factors favoring the transition from fetal to mature circulation include normal oxygen tensions and physical expansion of the lungs, normal pH, nitric oxide, and prostacyclin. Factors favoring reversion to fetal circulation include low oxygen tension, acidotic pH, lung collapse, and inflammatory mediators (leukotrienes, thromboxane A2, platelet‐activating factor) as seen in sepsis and other related conditions, and endothelin A receptor activators [4].

Myocardial contractility

The fetal myocardium is characterized by poorly organized cellular arrangements, and fewer myofibrils with a random orientation, in contrast to the parallel, well‐organized myofibrillar arrangement of the adult myocardium [5] (see later in this chapter). Fetal hearts develop less tension per gram than adult hearts, because of increased water content and fewer contractile elements. Calcium cycling and excitation‐contraction coupling are also very different, with poorly organized T‐tubules and immature sarcoplasmic reticulum, leading to more dependence on free cytosolic ionized calcium for normal contractility. Despite this immature state, the fetal heart can increase its stroke volume in a limited fashion up to left atrial pressures of 10–12 mmHg according to the Frank‐Starling relationship, as long as afterload (i.e., arterial pressure) is kept low [6]. These features continue throughout the neonatal and early infancy periods.

Gene expression in cardiac development

Progress has been made recently in understanding the genetic aspects of human cardiac development, and in contrast to the physiologic studies which are almost exclusively performed in animal models, small amounts of human cardiac tissue obtained from biopsy or autopsy specimens can be used for these studies. Some aspects of these developmental changes will be reviewed.

Cardiomyocyte maturation begins after birth and is crucial for efficient contractility and metabolism. A hallmark of cardiomyocyte maturation is isoform switching of contractile genes from the fetal to the adult state [13]. Myosin is the major protein component of the thick filaments of the cardiac myofibril, and differences in the expression of this protein may play a significant role in myocardial contractility. Chromosome 14 has the genetic material responsible for producing the myosin heavy chain (MHC) which makes up the backbone of the thick filaments, and two major isoforms, namely α and ß, exist. The ß isoform predominates at birth but the α isoform becomes predominates through maturation [14]. The myosin light chain has multiple isoforms, and the relative proportions of these isoforms change with development, and also in response to pressure loading of the heart. The isoforms that predominate in the newborn myocardium appear to confer a greater sensitivity to Ca²⁺ than those seen in the mature heart [15] and may contribute to the increased sensitivity of the neonatal myocardium to Ca²⁺.

Troponin I, C, and T are critical proteins that bind Ca²⁺ and regulate the interaction between myosin and actin, directly affecting the force of contraction. Troponin C, the Ca²⁺‐binding portion of the troponin moiety, does not change with development. Troponin I, however, has two major isoforms, a slow skeletal muscle type that predominates in the heart in fetal and neonatal life, and the cardiac isoform, which is the only isoform expressed in the mature heart [16]. Only the cardiac (mature) isoform responds to ß‐adrenergic stimulation, producing a faster twitch development and greater twitch tension. However, contractility in the neonatal myofibrils containing the immature myosin light chain (MLC) isoform is more resistant to acidosis. Four isoforms of troponin T are expressed in the fetal and neonatal heart, but only one in the mature heart. These isoforms exhibit different levels of ATPase activity and Ca²⁺ sensitivity (see later), with greater ATPase activity and Ca²⁺ sensitivity seen in the immature forms [12]. Tropomyosin [17] has two isoforms and actin [18] has three isoforms, which are expressed in different proportions as developmental changes occur, but the functional significance of these changes has yet to be elucidated.

Some enzymes are affected by the loading conditions of the heart. Protein kinase C (PKC) is an enzyme with a major role in transmembrane signal transduction through phosphorylation of a number of downstream intracellular components (see the section on “Calcium cycling in the normal heart”) [19]. There are six isoforms of this enzyme, and it is not affected during development. However, in aortic stenosis producing left ventricular hypertrophy, all isoforms except PKC‐β are dramatically upregulated, and in dilated cardiomyopathy, there is a dramatic upregulation of PKC‐β. Phosphodiesterase (PDE) is an enzyme involved in the termination of the action of cyclic adenosine monophosphate (cAMP), which regulates the contractile state of the myocardium (see later). Expression of the isoform PDE‐5 is dramatically increased in the hypertrophied human right ventricle in patients with pulmonary hypertension, and inhibition of this enzyme improves ventricular contractility [20].

New information is available on the molecular and cellular basis for normal cardiac development and the causes of CHD [21]. A missense mutation in the myocardial protein actin has been discovered to be the cause of isolated secundum atrial septal defect in some patients [22]. Pluripotent cardiac progenitor cells reside in the human neonatal myocardium in relatively high numbers during the first month of life [23]. This knowledge has given rise to the exciting notion that these stem cells could potentially be used to facilitate recovery from cardiac morbidity or to enhance surgical repair.

Recent advances in genomics, such as RNA sequencing (RNA‐Seq) have significantly contributed to the further understanding of cardiac development in humans. RNA‐Seq has provided a comprehensive transcriptomic analysis of human heart development during early gestation [24]. Nebulin‐related anchoring protein (NRAP) was among the genes that showed the largest fold change between 9‐ and 16‐week gestational age. This protein is expressed at intercalated discs in cardiac muscle and is an actin‐binding cytoskeletal protein [25, 26]. NRAP is involved during development in the myofibrillar assembly in the embryonic murine heart [27]. A patient with dilated cardiomyopathy with biventricular failure was found to have a homozygous truncating mutation (rs201084642) which introduced a stop codon to all NRAP isoforms [28]. Another study utilizing microarray RNA Seq analysis found that 316 genes were specifically associated with the human fetal heart, and 20 of those genes are the known candidates for congenital heart failure [29]. Specifically, the genes identified from the study associated with CHD include PLN, NPPA, ANKRD1, MYH6, MYH7, ACTC1, CACNA1C, TBX20, HEY2, SLC8A1, RYR2, MYOCD, GJA1, ATP2A2, FBN2, SRPX, SCN5A, TBX5, HAND2, and KCNJ2. This network of specific genes in the human fetal heart could represent a candidate common pathway close related to the development of CHD. An example of the network is Notch‐signaling cascade‐related genes [30, 31]. Notch‐signaling cascade plays a crucial role in cardiac cell fate regulation and orchestrates the morphogenesis of cardiac chambers and valves [32–34]. A recent study has suggested that NOTCH1 haploinsufficiency alters specific gene networks affecting valve development and osteogenic factors, which in turn result in aortic valve disease [35]. NOTCH1 ligands JAG1 and DLL4 have been shown to cause Alagille syndrome and aortic valve disease in human and animal models [36–38]. In addition, Genome‐wide Copy Number Variant (CNV) analysis for congenital ventricular septal defects in the Chinese Han population revealed that VSD‐related candidate genes are enriched in chromatin binding and transcription regulation, which are the biological processes underlying heart development. Specifically, this study suggested that PAX3 and LBX1 (in duplications) and CRKL, GP1BB, PDLIM3, TBX1, TXNRD2 (in deletions) are associated with CHD. Enriched gene ontology (GO) revealed that these genes are involved in heart development, cardiovascular system development, and circulatory system development [39].

The extracellular matrix

The extracellular matrix (ECM) of the heart is important in translating the force generated from the shortening of sarcomere length to the cardiac chambers, resulting in stroke volume. The major components of the ECM are collagen types I and II, glycoproteins, and proteoglycans and the expression of these elements changes with development. The neonatal heart has a higher content of both total and type I collagen (which is stiffer and less compliant than type III collagen) when compared with the total protein content of the heart [40]. The collagen to total protein ratio reaches mature levels by about 5 months of life. This change, along with the greater water content of the immature myocardium, may partially explain the diminished diastolic function. Also, this relative lack of contractile elements reduces the ability of the neonatal myocardium to increase its inotropic state. A network of collagen‐based connections, called the weave network, develops rapidly after birth, connecting myocytes and capillaries and allowing greater functional integrity to develop in response to the greater afterload stress on the heart [41]. This development of the ECM appears to be complete by approximately 6 months of age and results in a much more efficient transfer of force generated by sarcomere shortening to the cardiac chambers (Figure 9.4) [12].

The connection of the cardiac myocyte to the ECM is maintained by two specialized complexes that together comprise over 20 proteins. The costamere represents a major muscle multiprotein complex and produces a physical connection between the sarcomeres at the Z disk and the ECM. The integrin complex and the dystrophin‐associated glycoprotein complexes are two major components of the costamere [42, 43] (Figure 9.5). The integrins are transmembrane glycoproteins consisting of two subunits, α and ß, which express several isoforms. Integrins such as integrin α5β1and α7β1 bind to collagen and fibronectin, which causes the attachment of the ECM to the myocytes, allowing force transduction to occur [44]. These integrins’ expression is very dynamic; For example, the α5 subunit is prevalent in fetal and neonatal cardiomyocytes, while α7 replaces α5 at the onset of postnatal development [45]. However, subunit expression can switch and return to the fetal form under stress‐producing myocyte hypertrophy and ischemia. Dystrophin‐associated glycoprotein complexes also contribute to a substantial mechanical linkage from the ECM to the cardiac cytoskeleton, and contribute to force transduction [42, 43]. The proteins dystrophin, sarcoglycans, dystroglycan, and dystrobrevins are included in these complexes. These complexes play an integral role in cardiac function, and mutations in these proteins can be associated with cardiomyopathies, especially muscular dystrophy‐associated cardiomyopathies. Reduction of dystrophin activity results in dilatation of all four cardiac chambers and reduced ventricular function. Mutations in dystrobrevins have been associated with left ventricular non‐compaction.

Cell‐to‐cell connectivity

The intercalated discs mediate the cell‐to‐cell interactions that coordinate cardiac myocyte activity resulting in synchronous contraction, and maintaining the structural integrity of cardiac tissue [46] (Figure 9.6). These discs are situated in between cardiac myocytes at the longitudinal ends of the cells and consist of three types of connections: desmosomes, fascia adherens junctions, and gap junctions. The desmosomes have both intracellular and intercellular components. These structures serve to integrate signals from both cell‐to‐matrix and cell‐to‐cell interactions, ensuring force transmission, cell membrane integrity, and biochemical signaling. The fascia adherens junctions are responsible for holding the cardiac myocytes tightly together, and they anchor myofibrils and ensure transmission of contractile forces from cell to cell. Finally, the gap junctions form the electrical coupling apparatus between individual myocytes, ensuring rapid propagation of the electrical impulse, forming an electrical syncytium, and thus triggering the coordinated contraction of cardiac myocytes. Mutations in intercalated disc proteins have recently been found to be associated with cardiac disorders. These include adherens junctions mutations associated with heart failure and dilated cardiomyopathy, and desmosome complex mutations associated with some forms of arrhythmogenic right ventricular cardiomyopathy.

The preceding short review is meant to give the reader an idea of some of the aspects of the cellular biology of the developing circulation. The explosion of new information in this area, especially new data from human tissue, will lead to a more thorough understanding of the pathophysiology of disease states and will suggest avenues for future treatment. For a more complete treatment of this area, the reader is referred to several excellent reviews [47–49].

Innervation of the heart

Clinical observations in newborn infants have led to the hypothesis that the sympathetic innervation and control of the cardiovascular system are incomplete in the newborn infant compared with older children and adults and that the parasympathetic innervation is intact [5]. Examples of this include the frequency of bradycardia in the newborn in response to a number of stimuli, including vagal and vagotonic agents, and the relative lack of sensitivity in the newborn to sympathomimetic agents. Histologic studies in animal models have demonstrated incomplete sympathetic innervation in the neonatal heart when compared with the adult, but no differences in the number or density of parasympathetic nerves [50, 51].

Autonomic cardiovascular control of cardiac activity can be evaluated by measuring heart rate variability in response to both respiration and beat‐to‐beat variability in systolic blood pressure [52]. The sympathetic and parasympathetic inputs into sinoatrial node activity contribute to heart rate variability changes, with greater heart rate variability resulting from greater parasympathetic input into sinoatrial node activity [53]. Studies using these methodologies for normal infants during sleep suggest that the parasympathetic predominance gradually diminishes until approximately 6 months of age, coinciding with greater sympathetic innervation of the heart similar to adult levels [54].

The adrenergic receptor

The adrenergic receptors (ARs) are a part of a large superfamily of receptors that mediate their biological responses through the coupling of a specific guanine nucleotide regulatory protein, or G protein [68]. This superfamily of receptors shares a common structural motif, characterized by seven hydrophobic domains spanning the lipid bilayer. The seven domains are attached by three internal loops and three external loops between the amine terminus and the cytoplasmic carboxy terminus. The function of this receptor family is dependent on a specific agonist (or ligand) binding to the receptor, which causes a conformational change in the receptor. This structural change permits the interaction between the intracellular portion of the receptor and the G protein. This interaction, also referred to as “coupling,” inevitably links the activated receptor to a specific biological response. The regulation of the biological response is initiated by the specificity of the receptor for a particular extracellular agonist and the coupling of a specific G protein to that activated receptor.

Once an extracellular ligand (or agonist) is specifically recognized by a cell surface receptor, the receptor goes through a conformational change that exposes a specific region of the receptor complex to the intracellular side of the plasma membrane [69] (Figure 9.9). This conformation change triggers the interaction of the G protein with the amino acids of the third intracellular loop of the receptor and hence leads to G protein activation. There are three different G proteins: stimulatory G protein (G_s), inhibitory G protein (G_i), and the G_q protein. Under normal conditions, all of the ß receptors interact with the G_s, the α1 interacts with G_q, and the α2 interacts with G_i. Each G protein is a heterotrimer made up of three subunits: α, ß, and γ. The activation of the G‐protein‐coupled receptor causes an exchange of bound guanosine diphosphate (GDP) for guanosine triphosphate (GTP) within the α subunit and initiates the disassociation of the ßγ subunit from the α subunit. The GTP‐activated α subunit modulates the activity of a specific effector enzyme within a specific signaling pathway by catalyzing the hydrolysis of GTP to GDP and inorganic phosphate. This causes the transference of a high‐energy phosphate group to an enzyme and, in turn, causes the deactivation of the α subunit. This process will eventually lead to the reassociation of the α subunit with the ßγ complex. This cycle is continuously repeated until the agonist becomes unbound from the receptor. Downstream of enzyme activation, the production of a second messenger regulates the biological response.

Adrenergic receptors have been subdivided into two groups of receptors based on the results of binding studies using a series of selective agonists and antagonists. In 1948 Ahlquist used the difference in rank orders of the potency of a series of agonists to separate the ARs into two principal receptor groups, the α and ß receptor groups [70]. These findings have been confirmed repeatedly with the development of drugs that function to selectively antagonize the α receptor with no effect on the ß receptor. Soon after the distinction between the α and ß receptor types was known, it became more evident that the separation of α and ß receptors was not sufficient to explain pharmacological studies using rank order of potency for an antagonist, differing from an agonist because it blocks the biological response. With the advent of radioligand‐labeled antagonists and new molecular cloning techniques examining receptor gene expression, it became clear that the two principal receptor groups could be further subdivided into additional subtypes.

To date, within the ß‐adrenergic group, four different subtypes have been identified: ß₁, ß₂, ß₃, and ß₄. Pharmacologically, ß₁ and ß₂ are differentiated by their affinities to different catecholamines: epinephrine, norepinephrine, and isoproterenol. ß₁ has a similar affinity to epinephrine and norepinephrine, while ß₂ has a higher affinity to epinephrine than to norepinephrine. Both ß₁ and ß₂ have the same affinity to isoproterenol. The ß₃ and ß₄ receptors have minor roles in cardiovascular function and will not be discussed further.

The expression and distribution of each subtype are highly dependent on the organ, which adds another level of specificity. The distribution of a particular receptor in two different tissue types may result in two different functions. When examining the cardiovascular response to adrenergic stimulation, the ß₁ receptor is predominantly expressed in heart tissue. The stimulation of the receptor subtype leads to both inotropic and chronotropic effects on cardiac function, resulting in an increase in the myocardial contractile force and a shortening of contractile timing, respectively. While ß₂ can also be found in the heart, it is mostly expressed in vascular smooth muscle tissue. The distribution and function relevance of this receptor subtype in the heart are controversial and may change with alterations in cardiac function. The percentage of ß₂ receptors in the non‐failing heart averages about 20% in the ventricle [71] and 30% in the atrium. The ratio of ß₁ to ß₂ receptors is approximately 75: 25% in the ventricles of younger hearts [72, 73].

Each signaling pathway is specific to each adrenergic receptor. Once the agonist binds to the ß₁ receptor causing the coupling of the G protein, the G‐protein α subunit becomes activated followed by an increase in adenylate cyclase activity, which induces the conversion of ATP to cAMP. The second messenger, cAMP, phosphorylates protein kinase A (PKA). The function of a kinase is to phosphorylate other target proteins, which initiates a biological response. PKA phosphorylates many intracellular targets, including calcium channels, troponin I, and ryanodine receptors. PKA also plays a central role in regulating calcium sensitivity [74].

The ß₂ receptor has also been shown to function through the cAMP signaling pathway, causing the activation of PKA, but not nearly to the extent of ß₁ in cardiomyocytes [74]. The response of this stimulation appears to have a larger effect on smooth muscle; for example, the vascular smooth muscle. In this tissue type, the stimulation of ß₂ and the subsequent increase in cAMP promote the vasodilation of vascular smooth muscle and may lead to alterations in blood pressure. In these tissues, the effect of ß₁ stimulation appears to be minimal, due to the lack of ß₁ receptors in the smooth muscle.

Similar to the ß receptors, the α receptors can be pharmacologically subdivided into α₁ and α₂. The α receptor is distributed in most vascular smooth muscle and, to a lesser extent, in the heart. The α₂ receptor has been found in some vascular smooth muscle; however, its major functional importance is as a presynaptic receptor in the central and peripheral nervous systems. The use of molecular techniques has identified three additional subtypes of the α₁ receptor (α_1A, α_1B, and α_1D) and three additional subtypes of the α₂ receptor [68]. Binding of an agonist to an α₁ receptor in the heart or vascular smooth muscle results in activation of the G_q subunit of the G protein, which activates phospholipase C (see later), producing diacylglycerol and inositol‐1,4,5‐triphosphate, which releases Ca²⁺ from the sarcoplasmic reticulum and increases vascular smooth muscle tone or cardiac contractility. A schematic classification of adrenergic receptors incorporating recent knowledge of molecular pharmacology and signal transduction is presented in Figure 9.10 [68].

The adrenergic receptor concentration in cardiac tissue is very small and measured as fentomoles/mg protein. However, the response to stimulation of the receptor is significantly amplified by the signal that occurs downstream of the receptor. In rat ventricular myocytes, the ratio between the ß receptors and the next two downstream signaling components (ß receptor: G protein: adenylate cyclase) is 1 : 200 : 3 [76]. This demonstrates how a large response can be initiated by the activation of a small number of receptors. In addition, it also shows that the rate‐limiting component that ultimately regulates intracellular production of cAMP is receptor density and the enzyme concentration of adenylate cyclase.