What neurohormonal markers and proinflammatory cytokines are involved in cardiac remodeling?

Chronic heart failure (CHF) is a complex clinical syndrome characterized by cardiac abnormalities and a typical pattern of hemodynamic, renal, and neurohormonal response. Many variants of the syndrome have been described: high- and low-output, right and left, forward and backward, acute and chronic, systolic and diastolic. The hemodynamic response may change with the etiology, but a common feature is the potential for fluid retention. The two main hemodynamic changes in CHF reduced cardiac output and atrial hypertension trigger a cascade of neurohormonal responses (Figure 1).

Many of these neurohormonal changes occur in response to the inadequate arterial volume characteristic of systolic heart failure. In conditions characterized by a reduction in cardiac output and/or an increase in wall stress, a number of neurohormonal systems are activated, in particular the adrenergic, renin-angiotensin-aldosterone, and hypothalamic-neurohypophyseal systems. The vascular bed also releases endothelin. The activation of these systems initially serves to maintain perfusion to vital organs and expand the inadequate arterial blood volume by the renal retention of sodium and water. However, despite helping to prevent hypovolemia, each mechanism may be thought as a double-edged sword. As heart failure becomes chronic, several of these compensatory mechanisms may have unwanted effects, such as excessive vasoconstriction, increased afterload, excess salt and water retention, electrolyte changes, and arrhythmia. In contrast, other responses, such as the release of atrial natriuretic peptide (ANP) in response to atrial distension, may alleviate these effects by causing vasodilation, increased salt and water excretion, and sympathetic inhibition.

The myocardium offers ample evidence of adrenergic overactivity. Originally, low or depleted levels of the adrenergic neurotransmitter norepinephrine were observed in failing human myocardium. It was later shown that these depleted stores were the result of a sustained increase in the release of the neurotransmitter and a sustained decrease in its reuptake, resulting in a constant exposure to levels of norepinephrine that are quite certainly cardiotoxic. Chronic (3-adrenergic stimulation induces expression of the proinflammatory cytokines tumor necrosis factor-a (TNF-a) and interleukins 1 and 6 (IL-l/IL-6). These can impair cardiac contraction and promote chamber hypertrophy, thus playing a significant role in the development of the dilated cardiomyopathy phenotype. Cardiac reaction to such inappropriate signaling is readily measured: explanted severely failing human hearts reveal decreases in (^ -adrenergic receptor density, G protein coupling by (3j- and p2-receptors, P-adrenergic stimulation of adenylyl cyclase activity, and, in some studies, in intracellular cAMP. Phosphorylation by P-adren-ergic receptor kinase 1, an enzyme that is increased in

Once established, left ventricular dysfunction progresses to symptomatic CHF. Progression also involves ventricular remodeling, a self-perpetuating process that remains poorly understood, but may largely be conditioned by the neurohormonal response. In fact, the only agents that slow the development of heart failure and reduce mortality are angiotensin-convert- heart failure, is a major mechanism of Pj-receptor desensitization. P,-Receptor activation by a cAMP-dependent kinase, protein kinase A (PKA), causes phosphorylation of phospholamban, a protein that in its unphosphorylated state inhibits the uptake (and release) of Ca2+ by the sarcoplasmic reticular Ca-ATPase, SERCA-2a. Phosphorylation of phospholam- ban enhances Ca2+ uptake from the cytoplasm. Loss of the P-adrenergic mechanism in heart failure leaves phos-pholamban in the unphosphorylated state, thereby impairing Ca2+ movement and interfering with cardiac contraction and relaxation. In addition, genetic variants of the P-adrenergic receptors may be associated with rapid progression of heart failure.

Augmented sympathetic activity in heart failure is initially beneficial. It increases cardiac output and redistributes blood flow from the splanchnic region to the heart and skeletal muscles. Renal vasoconstriction causes salt and water retention, which may help to improve the perfusion of vital organs. However, sustained sympathetic stimulation, as in CHF, activates the renin-angiotensin-aldosterone system (RAAS) and other neurohormones causing progressive salt and water retention, vasoconstriction, edema, and increased preload and afterload.

These developments, in turn, increase ventricular wall stress, resulting in higher myocardial oxygen demand and myocardial ischemia. Sympathetic hyperactivity may also predispose to ventricular arrhythmias. Finally, norepinephrine has many direct effects on cardiac myocytes, including fetal gene program induction, downregulation of calcium regulating genes, myocyte hypertrophy, apoptosis, and necrosis. Therefore, although the initial sympathetic nervous system response appears appropriate by maintaining blood pressure and cardiac output, prolonged sympathetic hyperactivation may have deleterious effects.

In the severely failing heart, acute P-adrenergic receptor blockade may remove hemodynamically critical P-adrenergic support, thus intensifying heart failure. Gradual escalation of an oral p-adrenergic blocker, however, can be of substantial clinical benefit. Indeed, P-blocker therapy is now recommended in all cases of symptomatic chronic systolic CHF. In fact P-blockade has diametrically opposite effects on myocardial function depending on whether administration is chronic or acute, long-term blockade being associated with improved intrinsic systolic function and decreased ventricular volumes.

CHF is also characterized by elevated circulating and tissue concentrations of angiotensin II, a vasoconstrictor that increases ventricular afterload and causes myocyte hypertrophy, apoptosis, interstitial fibrosis, and cardiac and vascular remodeling. It achieves the latter effect, together with the related effects of fibroblast proliferation and collagen deposition, by stimulating aldosterone secretion. These changes increase passive stiffness in the ventricles and arterial bed, impair ventricular filling, and reduce arterial compliance. Elevated circulating aldosterone levels are predictive of adverse outcome in CHF, while RAAS inhibitors all have positive therapeutic effects. Arginine vasopressin (AVP) is synthesized in the hypothalamus and stored in the pituitary from which it is released in response to osmolar stimuli and rising levels of norepinephrine and angiotensin II. Increased AVP release in CHF causes vasoconstriction, water retention, and dilutional hyponatremia.

Endothelin (ET) is a potent vasoconstrictor peptide released by endothelial cells throughout the circulation. Three variants have been identified: ET-1, ET-2, ET-3, together with at least two receptor subtypes (A and B). Several vasoactive agents (norepinephrine, angiotensin II, thrombin) and cytokines [transforming growth factor (TGF)-P, IL-1(3] enhance ET release from endothelial cells in vitro. Circulating ET-1 levels are increased in CHF. Plasma ET correlates directly with pulmonary artery pressures and, in particular, with pulmonary vascular resistance and the pulmonary to systemic vascular resistance ratio, prompting the suggestion that ET plays a pathophysiologic role in mediating pulmonary hypertension in CHF. In normal subjects, plasma ET levels increase with orthostatic stress. However, in CHF, ET levels are already elevated and cannot increase further, much like the patterns of response to stress by various other vasoconstrictor substances, including norepinephrine and angiotensin.

ET receptor antagonists are now available and have been used to demonstrate the physiologic effects of ET. In rats with heart failure due to myocardial infarction, the combined A and B receptor blocker, bosen-tan, lowers blood pressure and has an effect additive to that of an ACE inhibitor. In cultured cardiac myocytes, ET induces cellular hypertrophy associated with fetal gene induction. In rats with pressure overload-induced hypertrophy caused by aortic banding, administration of an ETA receptor antagonist transiently inhibited myocyte hypertrophy and prevented fetal gene induction. In rats with myocardial infarction, chronic ET antagonist administration decreased hypertrophic remodeling of the left ventricular chamber, enhanced hemodynamic function, and prolonged survival. These observations suggest that ET receptor antagonists may be useful in the acute and chronic treatment of human CHE They have already been shown to enhance hemodynamic function, but their long-term effects on disease progression and survival are unknown.

Three natriuretic peptides, atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and C-type natriuretic peptide (CNP), have been identified in humans. ANP is stored mainly in the right atrium and released in response to an increase in atrial distension pressure; it causes vasodilation and natriuresis, thus countering the water-retaining effects of the adrenergic, renin-angiotensin, and AVP systems. BNP a misnomer in that it is stored mainly in ventricular myocardium, not brain may be responsive, albeit less so than ANP, to changes in ventricular filling pressures; it displays high structural homology with ANP and also causes natriuresis and vasodilation. CNP is located primarily in the vasculature; although its physiologic role is unknown, it may play an important role in counterregulating the RAAS. At least three natriuretic peptide receptors have now been identified. The A and B receptors mediate the vasodilatory and natriuretic effects, while the C receptor appears to act primarily in clearance, coregulating the levels of available peptides with neutral endopeptidase.

Circulating ANP and BNP are both elevated in CHF, and correlate directly with functional class. In normal human hearts, ANP predominates in the atria, where there is also a low level of expression of BNP and CNP. In CHF, the atrial content of ANP is unchanged, while that of BNP and CNP increases 10-fold and 2.3-fold, respectively. ANP and BNP secretion appears to be regulated mainly by wall tension. Free N-terminal ANP has a longer half-life and greater stability than ANP and is a potent independent predictor of cardiovascular mortality and progression to CHF. ANP levels normalize after cardiac transplantation. The hemodynamic and natriuretic responses to ANP infusion are attenuated in human and experimental CHF. However, studies with an ANP receptor antagonist in canine heart failure showed that, despite attenuated hemodynamic and renal effects, the peptide remained a potent suppressant of norepinephrine and the RAAS. One approach that attempts to capitalize on the beneficial effects of natriuretic peptides is to inhibit their degradation through the use of neutral endopeptidase inhibitors. Infusion of the endopeptidase inhibitor candoxatril into CHF patients mimics the action of infused ANP; it lowers left and right heart filling pressures, suppresses plasma norepinephrine levels, and transiently lowers plasma vasopressin, aldosterone, and renin activity. In additional to their beneficial effects on neurohormones, renal function and hemodynamics, there is evidence that the natriuretic peptides may directly inhibit myocyte and vascular smooth muscle hypertrophy and interstitial fibrosis.

A number of intrarenal hormonal systems may be activated in CHF. The most important are the arachi-donic acid cascade and the kallikrein-kinin system. The renal arterioles, glomeruli, and portions of the tubules and collecting ducts synthesize the vasodilator prostaglandins PGI2 and PGE2. Their predominant effect is to protect the glomerular microcirculation during states of renal vasoconstriction by causing vasodilation, mainly in the afferent arterioles, and also by promoting sodium excretion by directly inhibiting sodium transport in the distal tubules. Prostaglandin synthesis is increased by activation of the RAAS and renal sympathetic nervous system, and in clinical and experimental heart failure. Prostaglandins probably do not modulate renal hemodynamics or sodium excretion in normal subjects but may play a major role in situations such as CHF with elevated sympathetic and RAAS activity. Inhibiting prostaglandins using cyclooxygenase inhibitors may thus markedly reduce cardiac output and renal blood flow, while increasing peripheral vascular resistance and sodium retention. The distal renal tubule synthesizes kallikrein, a protease that cleaves kininogen to bradykinin and kallidin. Both bradykinin and kallidin cause vasodilation and natriuresis, and the former also stimulates prostaglandin production. Although the exact role of this system in CHF is unknown, there is evidence that at least some of the reduction in ventricular remodeling achieved by ACE inhibitors derives from an increase in bradykinin.

Growth hormone is secreted by the anterior pituitary and mediates its effects via the activation of insulin like growth factor-I (IGF-I). Its role in CHF is unknown. High levels are seen in severe untreated low- and high-output heart failure and in cardiac cachexia. Treating CHF with human growth hormone has proved beneficial in some but not all studies. Cortisol, a hormone produced by the adrenal under anterior pituitary control, is also elevated in various CHF syndromes, possibly as part of a general stress response.

Calcitonin gene-related peptide (CGRP) is a potent vasodilator released during heart failure. It is colocalized with substance P and vasoactive intestinal peptide (VIP) in parasympathetic nerve endings in the heart, blood vessels, and nervous system. Short-term CGRP infusion has beneficial effects in patients with CHF. CHF can induce endogenous overproduction of not only neurohormones but also cytokines. Inflammatory cytokines, including TNF-a, IL-ip and IL-6, may play an important role in the pathogenesis of CHF. In vitro, with other inflammatory cytokines, they regulate growth and gene expression in cardiac myocytes and other cells present in the myocardium. Circulating TNF-a and IL-6 levels are increased in CHF. The failing myocardium itself may produce inflammatory cytokines, which will thus be present in high local concentrations, with protean effects. TNF-a can induce immediate myocardial dysfunction and has been shown to attenuate intracellular calcium transport in vitro. In cultured cardiac myocytes, TNF-a, IL-ip, and IL-6 can stimulate hypertrophy, fetal gene program reexpression, and apoptosis, possibly mediated in part by nitric oxide. Chronic systemic infusion of TNF-a in rats caused left ventricular failure, while myocardial overexpression of TNF-a in mice caused dilated cardiomyopathy associated with increased myocyte apoptosis. Pilot clinical trials with soluble TNF-a receptors that reduce the level of TNF-a available to the tissues have suggested that this maybe a feasible form of therapy for patients.

Figure 2 is a simplified illustration of current ideas on the interplay between cardiac function and the neuro-hormonal-cytokine systems activated by myocardial injury, many of which confer benefit in acute heart failure. However, their chronic activation aggravates the myocardial injury and further depresses cardiac function. By causing myocyte hypertrophy and apoptosis, as well as ventricular remodeling and fibrosis, they set up a series of vicious circles. Fortunately, many of these maladaptive processes can now be blocked, thereby pre-

Keywords

pathophysiology; cytokine; neurohumoral activation; ventricular remodeling; neurohormone; apoptosis