|Home | About | Journals | Submit | Contact Us | Français|
The collagen network, part of the myocardial extracellular matrix (EM), and other EM proteins transmit mechanical forces generated by cardiomyocytes to cardiac cavities. Network rearrangement and enlargement – fibrosis – is an essential component of cardiac remodelling at various pathological stages. In particular, similarly abundant fibrosis occurs in dilated, hypertrophic and restrictive cardiomyopathy, and it is unclear how this relates to respective changes in ventricular cavities and size. Recent studies of hereditary forms have provided evidence that, in hypertrophic cardiomyopathy, the primary changes occur within cardiomyocytes that are subjected to either necrosis or mutations in genes that code for contractile proteins, while gene mutations of EM proteins have not been found. In contrast, in dilative cardiomyopathy, gene mutations of dystrophin, merosin and other proteins connecting the collagen matrix to cardiomyocyte membranes and actin filaments have been found. A distortion of the mechanical link between the contractile apparatus and the collagen matrix may disturb force transmission in both directions and lead to decreased developed pressure and increased end-systolic volume, provoking cardiac dilation. Profound alterations in the EM have also been induced acutely by alcohol, adriamycin or a high pacing rate, resulting in the development of dilative cardiomyopathy. Thus, EM alterations may be the primary factor in the pathogenesis of dilative cardiomyopathy.
The main components of myocardial structure are cardiomyocytes and the extracellular matrix (EM). Cardiomyocytes account for approximately 76% of myocardial volume and extracellular space 24% (1). The latter contains vessels and nerves (approximately 14%), strictly water space (7%), collagen network (1%), fibroblasts and macrophages (2%). Although its volume is minor, the collagen network is the essential component of the myocardium because it transmits mechanical forces generated by cardiomyocytes to cardiac cavities and back. Obviously, the composition and properties of the EM affect myocardial mechanical properties during both systole and diastole. Primary EM alterations resulting in myocardial fibrosis even in intact cardiomyocytes may lead to cardiac insufficiency (2), probably because of a restricted supply to cardiomyocytes of substrates and oxygen as well as a distortion of electrical contacts between them.
EM alterations may contribute considerably to the pathogenesis of cardiomyopathy. One may expect that the dilative form would be associated with increased EM strain, and the hypertrophic and particularly restrictive forms with increased EM stress. However, intriguingly all of these forms have similarly abundant fibrosis (3–6), and it is unclear how this relates to respective changes in ventricular cavities and size in dilated and hypertrophic cardiomyopathy (HCM). Also, the continuous formation of fibrotic tissue during the development of cardiomyopathy (7–9) is poorly understood. This pattern differs from other forms of cardiac pathology such as myocardial hypertrophy or infarction, in which EM growth stops after a certain degree of fibrosis development. These differences in the pathogenesis of cardiomyopathy seem to depend on the control mechanisms of EM structure.
The collagen network consists of threads and fibres (diameter 0.01 to 3 μm ) and is formed mainly by collagen types I and III. It may be imagined as multiple rows of parallel fibrils with cross-links between them. Other types also presenting in minor quantities are collagen IV, located in the vascular wall, and collagen VI, which is inherent in the newborn heart. Another abundant component of extracellular space is proteoglycans that have glycosaminoglycan residues. They possess significant osmotic activity because of the presence of multiple negative charges and thus may determine elastic properties of the myocardium (11). The collagen network is also tightly connected with the basement membranes of cardiomyocytes and fibroblasts; these connections are important because they transmit force in both directions. Various proteins have a transfer function; among these, fibronectin seems to be the most abundant extracellular protein associated with external plasmalemma of cardiomyocytes and fibroblasts. This link is realized through special membrane heterodimeric proteins – integrins (12,13) – which transmit mechanical forces into cardiomyocytes through a multiprotein system consisting of vinkulin, spectrin and alpha-actinin (14–16). The other extracellular protein, merosin, links the collagen network with actin filaments through dystrophin (17,18). In turn, the sarcolemma is connected to the nucleus through cytoskeletal proteins (19), and probably such a structure can transmit mechanical efforts directly from the collagen network to nuclear membranes (20).
All of these proteins can contribute to myocardial distensibility but collagens I and III are most important. Type I is stiffer and resembles steel; type III is more elastic. The ratio of type III to type I is relatively stable in normal cardiac muscle and may be considered as an index of myocardial distensibility. The collagen III:I ratio increases during postnatal development as well as in early stages of cardiomyopathy (21,22). However, stable myocardial fibrosis at late stages of cardiomyopathy is associated with predominant accumulation of collagen type I (22–26) and a decreased collagen III:I ratio.
The EM structure is maintained by the balance between protein synthesis and breakdown. Collagens I and III, as well as fibronectin and other minor extracellular proteins, are synthesized by fibroblasts and endotheliocytes, while collagen type IV is synthesized by pericytes (27). Collagen conformation is very resistant to the action of proteinases except for the specific enzyme intercellular collagenase, or matrix metalloproteinase-1 (MMP-1) (28). This enzyme splits the collagen molecule into two fragments that are subjected to further breakdown under the action of gelatinases MMP-2 and MMP-9 (29).
Myocardial metalloproteinases are secreted by fibroblasts, which are sensitive to some growth factors. They usually are kept in the latent form by a family of inhibitors called tissue inhibitors of MMPs. Collagenase activity accounts for only 3% of activity in normal atrial tissue after normalization for total protein concentration but can be increased to 80% to 90% by trypsin or plasmin (30). It has been found that MMPs and their inhibitors are colocalized and coexpressed throughout cardiac muscle (31). Thus, the myocardium as well as other tissues has an endogenous inhibitory system, suggesting that MMP activity is coordinated by their inhibitors at both the gene and the protein levels. Furthermore, MMPs and their inhibitors may be regulated both synergistically and separately to maintain the architecture of the interstitial tissue (31).
To induce collagen synthesis, fibroblasts first must be activated, and several factors serve as triggers of this activation. First, macrophages penetrating from capillaries into the extracellular space can produce two substances that activate fibroblasts, namely transforming growth factor (TGFβ1) and nitric oxide (29,32–34). Second, both circulating and tissue components of the renin-angiotensin-aldosterone system (RAAS) are potent activators of collagen synthesis.
It has been shown that angiotensin II activates collagen synthesis in cultured cardiac fibroblasts (14,35) through binding to angiotensin II type 1 (AT1) receptors in fibroblast membrane. This is followed by an increased concentration of cytoplasmic Ca2+ (36) and TGFβ1 (34). Prolonged angiotensin II injection into rats through a minipump results in the development of fibrosis in the heart and arteries (29). This is preceded by increased expression of mRNA of fibronectin, and then of collagen III and TGFβ1. The latter is believed to bind to its receptors in fibroblast plasmalemma and to induce the expression of EM protein genes (29,37). Fibroblasts play the main part in fibrogenesis in regions of myocardial damage – they express receptors for angiotensin II, TGF and endothelins (29). TGFβ1, being one of the regulatory peptides, or cytokines, is released by macrophages under the influence of increased concentrations of angiotensin II in the plasma as well as in response to myocyte necrosis.
Besides angiotensin II, angiotensinogen also participates in the renewal of EM structure, particularly in the endocardium, conductive system and right atrium (26). This is possible because of the conversion of angiotensinogen into angiotensin II by angiotensin-converting enzyme (ACE) in fibroblasts. ACE is found throughout the myocardium, although it is most abundant in regions with the highest rate of collagen synthesis: in valves and external layers of arteries (29). Moreover, ACE, being a kininase, promotes the breakdown of bradykinin, thus limiting its physiological action. Besides ACE, the human heart contains chimase, which also effectively transforms angiotensin I into angiotensin II. Chimase is found in the EM and endotheliocytes; its content is roughly two times higher in ventricles than in auricles (38), the opposite of ACE distribution.
An important function of fibroblasts is the expression of smooth muscle actin, an increased level of which facilitates shortening of connective tissue scar and may affect ventricular wall tension. This process can be affected by circulating angiotensin II. It has been shown in experiments in isolated hypertrophied heart that intracoronary angiotensin II administration significantly increases end-diastolic pressure in the left ventricle with fixed volume (39). This view is supported by observations in patients with left ventricular hypertrophy (40). Significantly reduced end-diastolic pressure combined with increased distensibility has been observed after just 15 min of injection of the ACE inhibitor enalaprilat.
These data suggest that, besides the static component contributed by the content and properties of collagen, myocardial stiffness is also determined by a dynamic component determined by the concentration of circulating angiotensin II. Their interdependence is seen in that the hypertension-associated increase in systolic pressure requires a concomitant increase in myocardial stiffness. The synthesis of collagen and other EM proteins, besides angiotensin II, is also sensitive to aldosterone, deoxycorticoster-one and estrogen (28,41,42). Collagen synthesis induced by angiotensin II is concomitant with reduced MMP activity (35). On the contrary, prostaglandin E2 increases MMP activity (28) and simultaneously decreases collagen synthesis. Therefore, fibrosis develops when collagen is simultaneously synthesized at an increased rate and degraded at a decreased rate.
EM protein synthesis is also controlled by nitric oxide, although its mechanism of action remains unclear. Blockade of nitric oxide synthase (NOS) by NG-nitro-l-arginine methyl ester (L-NAME) administration in normal rats is accompanied by the development of hypertension and myocardial fibrosis by the end of the first week (34). This is preceded by a raised angiotensin II concentration, which, through AT1 receptors, activates TGFβ1 formation. Administration of L-NAME within two weeks was accompanied by increased arterial pressure; however, myocardial hypertrophy and fibrosis developed only after accompanying injections of angiotensin II (43).
These data suggest that, under natural conditions, NOS counteracts the action of angiotensin II not only on vascular tonus but also on collagen synthesis. Because the main source of NOS is endotheliocytes, their participation in the control of collagen synthesis and degradation appears obvious (44). Another source of nitric oxide is activated macrophages (29), which synthesize NOS molecules (45). Nitric oxide, as prostaglandin E2, probably limits the growth of the collagen network through MMP-1 activation. An even simpler suggestion is that nitric oxide stops EM protein synthesis in endotheliocytes themselves. Thus, nitric oxide along with prostaglandin E2 is a natural antagonist of the action of angiotensin II. Their formation is probably intensified by high concentrations of angiotensin II, which serves to limit excess fibrosis.
The collagen network promptly responds to various ontogenic situations. Among them, three principal factors should be considered: the load determined by diastolic or systolic myocardial tension; ischemic damage; and myocardial inflammation caused by infectious or toxic agents. The usual EM response to the prolonged action of various mechanical and chemical factors is thickening and multiplication of collagen fibrils, resulting in the development of myocardial fibrosis. The type of fibrosis depends on the nature of the inducing factor. First, fibrosis can be induced by increased cardiomyocyte force generation as, for example, in hypertensive disease; in this case the fibrosis has a reactive (additive) nature. Second, fibrosis may develop to substitute necrotized cardiomyocytes to preserve the integrity of collagen network as, for example, in myocardial ischemia; in this case fibrosis has a reparative (substitutive) nature.
Cardiac hypertrophy may be an example of myocardial remodelling associated with the development of increased intraventricular pressure. Most often hypertrophy is caused by hypertension. Fibrosis develops concomitantly with reduced MMP activity (45). An increased collagen content in spontaneously hypertensive rats either is proportional to the extent of myocardial hypertrophy, thus keeping collagen concentration constant up to 86 weeks (46), or exceeds the extent of hypertrophy after 80 weeks (47), which is associated with decreased distensibility of isolated papillary muscles. Fibrosis in hypertrophied heart develops under the action of both circulatory and local components of RAAS.
Ischemic myocardial damage affects not only cardiomyocytes but also EM structures. The sequence of events in the ischemic area has been extensively reviewed elsewhere (29). Activation of both latent MMPs (48) and NOS in macrophages (33) contributes to EM destruction. Later on, activated fibroblasts promote collagen synthesis, and fibrosis develops simultaneously with reduced MMP activity. It is important to note that AT1 receptors in fibroblasts are activated not only in the ischemic but also in the undamaged area (29), creating conditions for a corresponding increment in the collagen network in the heart. Moreover, increased ACE activation and AT1 receptor density are maintained several months after coronary artery ligation. This important observation suggests that cardiac muscle remains sensitized long after the action of damaging factors disappears.
Myocarditis resulting from viral infection is considered as a possible cause of cardiomyopathy development (49). Cox-sackie virus is found in the heart during the first days after infection, and its activity peaks by the end of the first week, while fibrosis develops two to four weeks after the virus has disappeared (50). Fibrosis is associated with fibroblast activation induced by cytokines and growth factors as well as by B and T cells (51).
Myocardial remodelling during inflammation has early, adaptive and late, disadaptive phases, and involves all myocardial cells and structures (2). In the initial stage of septic myocarditis, the expression of inducible NOS has been noted (52). Subsequent remodelling includes fibroblast proliferation and EM growth. An ACE inhibitor, captopril, supresses fibrosis, hypertrophy and myocardial remodelling during the inflammatory, but not the fibrotic, phase of viral myocarditis (53). This suggests that angiotensin II participates in the early phases of the inflammatory process.
Studies of myocardial samples taken from patients with HCM during surgery or at autopsy, as well as investigations of hereditary HCM of hamsters, have shown the presence of abundant fibrosis at any term of the disease. The collagen concentration was found to be 72% higher in the hearts of patients with HCM than in hearts in which hypertrophy had another origin (54). A significant increase in both scar and matrix fibrotic tissue was noted, meaning that the nature of fibrosis may have both reparative and reactive. Even more abundant fibrosis in patients with HCM exceeding that in hypertensive hypertrophy by approximately four times has been reported (55). Another feature of HCM is that fibrosis is more abundant in the interventricular septum, while in hypertrophy the degree of fibrosis is roughly alike in all the walls of the heart. In addition, a correspondence between the degree of hypertrophy and of fibrosis in HCM has not been found, suggesting a weak link between hypertrophy and fibrosis in HCM.
Studies in hamster HCM have shown that the collagen matrix in the early stage exhibits features of the growing heart, with a prevalence of collagen III, but in the late stage type I dominates, and myocardial distensibility is decreased (22–26). Collagen growth is observed after a lag period in which the concentrations of mRNA collagens I and III rise and remain elevated in the left ventricle at all stages (56). Except for collagen, such abundant extracellular proteins as fibronectin linking collagen matrix and external plasmalemma of cardiomyocytes have been observed to grow throughout the lifetime of hamsters with HCM (9), indicating a permanent progression of fibrosis.
As would be expected, total ACE activity is raised in the heart of hamsters with hereditary HCM (57). The use of ACE or AT1 receptor inhibitors in the course of development of HCM in hamsters affected the degree of fibrosis. The ACE inhibitors ramipril and captopril, when given before the necrotic stage, did not prevent myocardial necrosis and EM protein synthesis (58,59); thus, fibrosis still may be reparative. On the other hand, enalapril or captopril administration to hamsters after the end of the necrotic stage prevented increases in the collagen III:I ratio at 20 weeks of age (26,60). Similar, but even more significant and statistically reliable changes, have been observed under the action of K+-canrenoate, an aldosterone antagonist (60). An inhibitor of AT1 receptors, losartan, did not influence the growth of the collagen network, but caused regression of hypertrophy, and reduced an increased level of MMP-1 and MMP-2 activity through 60 to 200 days (56). These results suggest that fibrosis in HCM after the necrotic stage is reactive and is controlled separately from the process of hypertrophy.
Genetic mutations affecting protein synthesis have been described in HCM. The first gene discovered to be responsible for the development of HCM is that encoding the beta-myosin heavy chain (61), a main myofibrillar protein that forms approximately 35% of total proteins. An inducible point mutation in the myosin heavy chain that changes one amino acid, arginine, for another, glutamine, did not change myosin ATPase activity, but reduced actomyosin activity by 3.5 times and motility by five times in vitro (62). In addition, the direct introduction of mutant beta-myosin heavy chain into isolated cat cardiomyocytes caused sarcomere disorganization within five days (63).
Besides beta-myosin heavy chain, mutations in genes encoding other myofibrillar proteins, namely troponin T and alpha-tropomyosin, have been found (61,64). In contrast, an extensive search for other genes that encode collagens of different types, as well as fibronectin, laminin, fibrillin, desmin, titin, alpha-actinin, vinculin, ankirin and spectrin, was unsuccessful, and they were consequently excluded from being responsible for the development of familial HCM (65). However, a deficiency in membrane glycoproteins has been reported. One of them is CD36 protein, which functions as a transporter of long chain fatty acids into cardiomyocytes. Approximately a third of patients with asymmetrical septal hypertrophy were found to be deficient in CD36 (66). Another membrane defect is associated with a deficiency of all four subunits of sarcoglycane in both cardiac and skeletal muscles of hamsters with HCM (67), although only the delta subunit mRNA has been found to be absent.
It is believed that enhanced sympathetic action on the heart has an important role in the pathogenesis of HCM (68,69). In fact, catecholamines in high doses induce alterations of cardiomyocytes and EM, resembling that in HCM. Isoproterenol (150 mg/kg) caused cardiomyocyte necrosis followed by abundant fibrosis and hypertrophy, as well as by increased concentrations of renin, aldosterone and ACE (59). Ramipril, an ACE inhibitor, prevented the development of hypertrophy and raised the left ventricular end-diastolic pressure, but weakly influenced myocardial necrosis and synthesis of EM proteins. Constant infusion of isoproterenol in rats (0.5 mg/kg/day) by a minipump within seven days caused hypertrophy without hypertension, and this was accompanied by raised concentrations of mRNA of collagen I by 1.9 times, collagen II by 2.7 times and fibronectin by 3.2 times (70). However, concentrations of mRNA of TGFβ1, collagen IV, and laminin B1 and B2 were not changed. Manidipine, a calcium antagonist, prevented the development of hypertrophy and increased the mRNA concentrations of the aforementioned proteins. This finding, suggesting that an increased concentration of cytoplasmic Ca2+ is a necessary component of catecholamine action, does not exclude the participation of angiotensin II in causing similar changes (36).
Interestingly, the direct effect of a catecholamine overdose on the collagen network has been reported. Distortions and the partial disappearance of the myocardial collagen network were observed right after noradrenaline administration (0.3 mg/kg for 90 min) (71). This effect may probably be attributed to increased activity of MMPs and peptidases.
Cardiac dilation inevitably involves corresponding EM changes. At least the initial EM alterations should be associated with an increased strain accompanied by reduced total collagen content (2). However, in the late stage of the disease, the collagen concentration is raised by two (24) or even five times (3) in hearts of patients with dilated cardiomyopathy (DCM). The degree of fibrosis was more prominent in endocardium than in epicardium. Although cardiomyocyte diameter was substantially higher in these hearts than in hearts without DCM, a correlation between this parameter and the degree of fibrosis has not been found.
The collagen network in DCM has been shown to consist predominantly of collagens I and III (4,5) with a 13% decreased III:I ratio (24). Also, moderately increased type VI collagen content has been reported (4,5) while type IV content was either decreased (4) or increased (5). This was combined with the appearance of gigantic spring-like perimysial fibres with a diameter of 20 to 30 μm. The meaning of increased type VI collagen content is unclear (72). This preliminary form of collagen probably always appears during the continuing fibrotic process. This suggestion is in agreement with the finding of another precursor of collagen, tenascin, an EM glycoprotein, at the margins of fibrotic tissue (73).
The content of fibrotic tissue in the hearts of patients with DCM is characterized not only by collagens but also by fibronectin, laminin, multiple macrophages and fibroblasts (7). This obviously is associated with an enlarged extracellular space. In addition, a varied combination of hypertrophied, atrophied and normal cardiomyocytes with the presence of some degenerative alterations, namely a loss of myofibrils and disorganized cytoskeleton, has prompted some authors to hypothesize the existence of a vicious circle (7). According to this hypothesis, altered myocytes can release some substances that stimulate matrix growth but, in turn, its enlargement can overload myocytes. If this is the case, the continuous growth of the collagen network in DCM may be explained by progressive cardiomyocyte degeneration.
Studies of EM alterations in the most popular experimental model of DCM, hereditary cardiomyopathy in hamsters, have shown a basic similarity with clinical findings but have added some important information concerning the time course of the alterations. It has been found at the early stage of five to 10 weeks of age either that the collagen network is not changed or that only collagen III content increases predominantly (21,26), which is inherent in the growing heart possessing increased distensibility. The collagen III:I ratio then either returned to normal at 20 weeks (74) or remained increased during weeks 20 to 40 (21). The latter observation is in accordance with the steadily increasing collagen concentration observed from 20 to 40 weeks, when it is approximately three times the normal value (21).
Collagen growth in the early stage of hamster hereditary DCM is combined with increased expression of the angiotensinogen mRNA gene (26), suggesting the participation of RAAS in the development of fibrosis in DCM. In the end stage of human DCM, a concomitant increase in MMP activity has been reported (75,76) coinciding with either a sharp fall (76) or rise (75) in tissue inhibitors of MMP activity. Meanwhile, MMP-1 activity that is most important for the initial collagen degradation falls sixfold (75), which seems to facilitate network preservation.
Genetic forms constitute 20% to 30% of all cardiomyopathic events (77). In human DCM, mutations of genes of two interconnected proteins, dystrophin and merosin, which link the collagen network and actin filaments within cardiomyocytes, are found (17,18,78,79). A deficiency in these proteins in a group of children was combined with muscular and myocardial weakness, especially in cardiac muscle in which the dystrophin deficiency was more profound (17).
Dystrophin deficiency is believed to be responsible for the development of Duchenne muscular dystrophy combined with DCM (80). Some membrane dystrophin-associated glycoproteins are deficient in hamster (67,79,81) and human hereditary DCM (82). In addition, defects in actin molecules at their point of attachment to Z-lines and intercalating discs (83), and distortion of intercellular contacts (84) have been observed. Another genetic defect in DCM is a shifted coronary endotheliocyte phenotype (85), which may substantially change the control of EM synthesis in view of the intermediary role of these cells in the action of circulating components of RAAS.
From the presented data, it can be concluded that hereditary DCM is associated with profound EM alterations resulting in a weakening of links that transmit force between myofibrils and collagen network. This disrupts the interdependence between sarcomere length and intraventricular pressure, which may lead to decreased developed pressure and increased left ventricular end-systolic volume, promoting dilation of the heart.
Besides idiopathic DCM, cardiac dilation is a common feature of some other diseases caused, for example, by overdoses of alcohol or adriamycin, an effective antitumour drug. The complex of structural, metabolic and functional alterations under their action is very similar to that observed in idiopathic DCM (86–88) and thus they are often considered as toxic DCMs. This similarity is probably explained by the ability of alcohol to destroy the immune system, promoting the development of viral or idiopathic myocarditis. As for adriamycin, its destructive action on the collagen network was noted a decade ago (10). Significant reduction and disorganization of endo- and perimyseum in rat hearts have been observed one week after adriamycin injection (6 mg/kg) (89). These changes were preserved after four and even eight weeks, and might cause an insufficiency of the valvular apparatus and congestive heart failure (10,89).
DCM may also result from induced continuous pacing of the heart at a high rate. After three weeks of supraventricular tachycardia in pigs, left ventricular fractional shortening fell by 60% and left ventricular end-diastolic dimension increased by 47% compared with controls (90). Cardiac failure was accompanied by profound alterations of EM protein properties. Particularly, the authors found significantly decreased collagen concentration and cross-linking. Moreover, myocyte adhesion for all of the basement membrane components was found to be halved. These alterations suggest a weakening of major links inside the collagen network and its connections with cardiomyocytes, favouring cardiac dilation. Recovery from tachycardia-induced failure was followed by restoration of collagen concentration and normalization of myocyte adhesion.
Thus, DCM caused by alcohol, adriamycin or a high pacing rate also exhibits alterations of the EM similar to those observed in idiopathic DCM. Destruction of the collagen network associated with decreased collagen content has been observed in the acute, damaging stage of the disease; then fibrosis develops as a part of myocardial remodelling.
Fibrosis in cardiomyopathy, as in other cardiac diseases, is an essential component of myocardial remodelling but has some distinctive features. Several findings in the growth of the collagen network that are common to DCM and HCM may be pointed out.
First, the degree of fibrosis in both forms of cardiomyopathy is greater than the extent of accompanying hypertrophy and is not related to the latter (3,54). This may probably be explained by a substantial contribution of the extensive fibrosis that follows focal cardiomyocyte necrosis and probably is reparative, while hypertension-induced hypertrophy is associated with reactive fibrosis. Careful morphological study of the hearts of patients with DCM leads to an estimate of the value of fibrosis in total myocardial volume of 20%, and the authors suggest that reparative and reactive fibrosis are distributed equally (91). In HCM, the degree of fibrosis is increased by four times (55) and collagen concentration by 72% above that associated with hypertrophy of the heart (54). Fibrosis in hamster DCM is higher than in HCM within the period 20 to 40 weeks (21).
Second, in contrast to myocardial ischemia and hypertrophy, in which EM growth stops after reaching a certain degree, fibrosis formation in cardiomyopathy is continuous (7–91,73,92). This may be stimulated both by sustained cardiomyocyte necrosis and by prolonged activation of fibroblasts. The latter is suggested by the finding that increased ACE activation and AT1 receptor density in fibroblasts were maintained several months after coronary artery ligation (29). Progressive fibrosis may become an important factor in the pathogenesis of cardiomyopathy because the more relative volume of the myocardium that is occupied by the EM, the higher the probability that adequate blood supply to cardiomyocytes is restricted and the stiffer the myocardium becomes (23). Thus, EM growth is transformed from being a compensatory factor in the initial stages of disease into an important pathogenic factor in the gradually progressive chronic insufficiency of the heart.
Third, in both forms of cardiomyopathy, EM alterations are accompanied by increased MMP activity (56), indicating an increased turnover rate of collagen molecules. Although corresponding data in cardiomyopathy induced by adriamycin, high pacing rate or isoproterenol have not been reported, the observed destruction of the collagen network and reduction of collagen content make concomitant MMP activation highly probable. Similar MMP activation has been observed in restrictive cardiomyopathy (6). Subsequent growth of the collagen network is associated with an initial dominance of type III collagen synthesis over type I in both DCM and HCM (21,22,26,56,93). This feature seems to be pertinent to the growing heart (26) and to restrictive cardiomyopathy (6). Thus, activation of MMPs and a raised III:I collagen ratio are the initial steps in remodelling of the collagen network when it becomes more distensible and susceptible to other alterations.
Despite these similarities, some differences in EM alterations between DCM and HCM may be noted, namely in fibrosis location and nature, as well as in genetic defects:
Development of fibrosis in HCM and DCM has more commonalities than differences. EM growth is continuous and is not correlated with the extent of accompanying hypertrophy. The extent of fibrosis may exceed that in normal myocardium by several times, and fibrosis may be both reparative and reactive in nature. The differences between them are related to fibrosis location and the character of myocardial alterations.
In HCM, profoundly disorganized myofibrils are a peculiar feature (61–65) while in DCM, at least in its familial form, predominant alterations of several EM proteins and the collagen network are observed (17,18,78–82). In this regard, the pathogenesis of HCM and of DCM may be quite different. In HCM, the primary defect seems to be within cardiomyocytes that are subjected either to necrosis, as in hereditary cardiomyopathy of hamsters of line Bio 14.6, or to mutations of genes of contractile proteins, including myosin, in patients. This may upload work to other myofibrils, whose hyperfunction then stimulates the development of myocardial hypertrophy, which may be nonuniform because of focal myofibrillar lesions. Progressive fibrosis may then be caused both by fibrous replacement of focal cardiomyocyte lesions and by associated prolonged fibroblast activation. Thus, fibrosis in this disease may be secondary, although it contributes to the development of cardiac failure.
On the other hand, in DCM, the primary lesion seems to be associated with a deficiency in EM proteins as a result of either mutations of the encoding genes or their increased breakdown under the action of toxic factors (10,86–90). These situations would be followed by a disturbance of force transmission between myofibrils and the collagen network, breaking the interdependence between sarcomere length and intraventricular pressure. This transmission may be disrupted even more by defects in actin attachments within cardiomyocytes. The consequent reduced developed pressure and increased end-systolic volume provoke cardiac dilation, which, unlike that in the normal heart, cannot be followed by appropriate sarcomere distension and improved contractile function. The concomitant expansion of the collagen network may differentially affect neighbouring cardiomyocytes, distending some much more than others and provoking their hypertrophy or atrophy. Progressive fibrosis increases the probability of capillary clamping, resulting in focal myocardial ischemia with further continuous development of fibrosis. Thus, fibrosis in this form may be primary and contribute directly to the development of cardiac failure, as suggested (25).
This hypothesis is based on the possibility of a primary weakening or breaking of the mechanical links between EM and cardiomyocytes. The ineffectiveness of multiple attempts to treat DCM with inotropic drugs, ACE inhibitors, nitrates, beta-blockers and other drugs that are effective for the treatment of other heart diseases indirectly confirms the particular pathological nature of a given disease. Substantial progress in this field may probably be expected from genetic therapy.