Fibrosis in hypertensive heart disease

Hypertension and heart failure are major health problems of worldwide proportions. Prevention is of paramount importance with efforts focused on the identification and modification of risk, and the presumption of risk reduction favourably influences cardiovascular-related morbidity and mortality. Left ventricular hypertrophy (LVH) is recognized as an important risk factor for adverse cardiovascular events [1–3]. Why should the hypertrophic growth of cardiomyocytes prove detrimental? Athletes have the same left ventricular mass as individuals with mild, moderate and severe hypertension [4]. If it is not the quantity of myocardium that confers increased risk, could it be its quality? The growth of parenchyma (or cardiomyocytes) in hypertensive heart disease (HHD) need not be faulty. Instead, abnormalities of stroma could account for ventricular dysfunction and a predisposition to re-entrant dysrhythmias. An adverse accumulation of fibrous tissue, or fibrosis, is an established morphological feature of HHD whether ventricular tissue is examined postmortem [5–7] or antemortem by endomyocardial biopsy [8]. Morphologically, it presents as a reactive perivascular fibrosis of intramural coronary arteries and arterioles with extensions into the contiguous interstitial space to create an interstitial fibrosis. In response to cardiomyocyte necrosis, a replacement fibrosis, or microscopic scarring, is also seen. What pathophysiological mechanism(s) account for the appearance of fibrous tissue in HHD? Is it simply the longstanding pressure overload? The haemodynamic burden placed on the left ventricle? Numerous studies conducted over the past 30 years have shown that haemodynamic factors govern cardiomyocyte growth while reactive fibrosis is blood pressure independent [9]. Instead, chronic inappropriate (relative to dietary salt) elevations in such circulating hormones as angiotensin II, aldosterone, deoxycorticosterone and endothelin-1 (ET-1) are operative. These same substances are produced de novo within cardiac tissue. Together with fibrogenic cytokines such as transforming growth factor (TGF)-β1 and connective tissue growth factor (CTGF) acting as intermediaries, these signals contribute to stromal growth. How humoral factors and cytokines lead to a fibrogenic cardiac phenotype is an area of ongoing investigation. In the case of primary and secondary aldosteronism, one working paradigm implicates neuroendocrine and immune systems (a neuroendocrine–immune interface) predisposing to an activation of circulating peripheral blood mononuclear cells (PBMC) that subsequently target the intramural vasculature of the heart and systemic organs [10–12]. This immunostimulatory state is induced by iterations in PBMC divalent cation composition (Mg2+ and Ca2+) and is transduced by oxi/nitrosative stress [10, 12–14]. What cells are responsible for the adverse accumulation of connective tissue in HHD? Cardiomyocytes constitute one-third of all cells found in myocardial tissue; the remainder include fibroblasts, endothelial and vascular smooth muscle cells. Fibrous tissue is composed of fibrillar collagens: predominantly type I and to a lesser extent type III collagens [15]. Only fibroblasts have the transcriptome and proteome needed to, respectively, express these fibrillar collagen genes and process the structural proteins. What is known about cardiac fibroblasts? Usual fibroblasts of the normal cardiac interstitium account for the gradual turnover of collagen, a stable structural protein with a half-life of 80–120 days [15]. Studies performed in vivo have addressed the behaviour of fibroblasts involved in a more rapid turnover of collagen that accompanies cardiac remodelling. Various experimental models have been used, including myocardial infarction, pericardial injury, reaction to foreign body (i.e. silk ligature placement), elevations in endogenous effector hormones of the circulating renin–angiotensin– aldosterone system (RAAS) induced by suprarenal aortic or unilateral renal artery banding, and treatment with exogenous angiotensin II, aldosterone or deoxycorticosterone acetate. Fibroblast-like cells responsible for the deposition of fibrillar collagens at sites of repair are phenotypically modified [16]. They are termed myofibroblasts (myoFb) because they express α-smooth muscle actin (α-SMA) microfilaments and therefore are contractile. The source of myoFb is uncertain and may include a pool of circulating fibroblasts, termed fibrocytes, tissue-based fibroblasts, or both [16–18]. Unlike injured skin, where myoFb promote healing and wound closure, and thereafter disappear via apoptosis, myoFb are persistent and active at sites of cardiac fibrosis [19,20]. Whether their prolonged survival is related to suppressed apoptosis remains uncertain. MyoFbs do not operate in isolation. Cell–cell signalling is an integral feature of their behaviour and function. For example, their interplay with invading inflammatory cells leads to fibrous tissue formation. Without an inflammatory cell response, fibrous tissue does not appear. Such is the case following cardiomyocyte apoptosis. Cardiac fibroblasts found at sites of high collagen turnover communicate with one another by gap junctions [16,21]. These communications allow a transmission of signals. Together with the attachment (via adhesion molecules) of myoFb to fibrillar collagen, the myoFb-connective tissue assembly has a contractile tonus. Contraction is induced by angiotensin II, ET-1 and catecholamines, while relaxation occurs in response to papaverine [22,23]. In infarcted hearts, gap junctions are also found between myoFbs and cardiomyocytes where impulse propagation can occur over extended distances, leading to conduction delays, and this could account for the arrhythmogenic potential of such tissue [24]. What regulates myoFb collagen turnover? Insights to this question are provided by heart valve leaflets. Leaflets are a normal site of high collagen turnover. By immunohistochemical labelling and transmission electron microscopy, myoFbs (also known as valvular interstitial cells) are found to be normal residents of leaflet tissue [21]. The active phenotype (vimentin and α-SMA positive) of leaflets and myoFb found at sites of tissue repair has an impressive portfolio of activities (e.g. angiotensin and ET-1 generation). By in situ hybridization, myoFb mRNA expression of angiotensinogen, renin and cathepsin D has been found in myoFb located at the site of infarction (not in normal myocardium); and these myoFb are a site of high-density angiotensin-converting enzyme (ACE) and angiotensin II receptor binding (predominantly AT1 in rats) [25–27]. Angiotensin II regulates myoFb collagen synthesis via AT1 receptor-ligand binding. A comparable expression of angiotensin peptide requisites is found at other sites of repair involving the heart (pericardium, endocardial fibrosis, foreign-body fibrosis), kidney and skin irrespective of the nature of injury [20,27–31]. MyoFb of the infarct scar also express the ET-1 precursor (prepro-ET-1), endothelin-converting enzyme-1 (ECE-1), a metalloproteinase required to convert Big ET-1 to ET1 and ETA and ETB receptors [32]. Auto/paracrine properties of locally produced ET-1, together with TGF-β1, similarly contribute to myoFb collagen synthesis [32]. Angiotensin II and ET-1 have other auto/paracrine properties. MyoFb expression of vascular endothelial growth factor and its receptors, which may contribute to neovascularization at sites of repair, are upregulated by angiotensin II [33]. Adhesion molecule expression (e.g. integrins) and de novo generation of aldosterone are each regulated by angiotensin II [34,35]. AT1 receptor-ligand binding induces the generation of oxi/nitrosative stress from NADPH oxidase and also where oxygen metabolites participate in intracellular signalling [36]. ET-1 activated mitogen-activated protein kinases and activator protein-1 are each inhibited by antioxidants [36–38]. Petrov et al. [39–41] have utilized cultured cardiac fibroblasts in an attempt to elucidate their behaviour. They have forged several new frontiers. For example, they have shown TGF-β1 induces ACE and aminopeptidases in cardiac fibroblasts that differentiate into myoFb, and that this fibrogenic cytokine regulates myoFb collagen production [39–41]. In this issue of the journal, Lijnen et al. [42] demonstrate the presence of ACE (lisinopril-dependent) and non-ACE (lisinopril-independent) activities in cultured adult rat cardiac fibroblasts and a role for aminopeptidases in regulating TGF-β1-stimulated collagen production. Despite the inherent elegance and simplicity of such studies performed in vitro, their clinical relevance must be interpreted with caution given that the fibrous tissue response in HHD appears gradually and may involve a complex interplay with other cell populations. This caveat notwithstanding, the work of Lijnen et al. [42], together with that of others, suggests an emerging paradigm of tissue repair, where myoFb-derived angiotensin II and ET-1 regulate their turnover of collagen, likely via an intermediary TGF-β1 [43]. In this connection, and when circulating levels of angiotensin II are normal or are suppressed by mineralocorticoid-salt treatment, pharmacological intervention with either an AT1 or a combined ETA and ETB receptor antagonist has demonstrated the importance of local angiotensin II and ET-1 in governing the expression of TGF-β1 which, in turn, regulates the synthesis of type I and III collagens and the appearance of fibrosis [20,44–46]. In summary, an accumulation of fibrous tissue is an integral feature of the structural remodelling of myocardial tissue that appears in HHD. Cardiac fibrosis confers an increased risk for such adverse cardiovascular events as ventricular dysfunction and dysrhythmias. Fibrous tissue formation in the heart is the result of a population of cardiac fibroblasts whose diverse activities include angiotensin II and ET-1 production which, in turn, regulate their turnover of collagen in an auto/paracrine manner. A role for haemodynamic factors in regulating cardiomyocyte growth is firmly established while their contribution to fibrogenesis has been seriously challenged. Given the body of knowledge that has emerged over the past 30 years with regard to such myocardial remodelling in HHD, it would no longer seem tenable to assume: (i) that a reduction in cardiovascular risk can be accomplished by simply reducing blood pressure; and (ii) that the nature of the pharmacological intervention used to reduce blood pressure does not matter. Moreover, current management strategies for the reduction in adverse cardiovascular risk that accompanies existing abnormalities in myocardial structure, such as fibrosis, should focus on pharmacological intervention(s) with proven cardioreparative properties that regress fibrous tissue [47–51].