2.5. Cardiac extracellular matrix

The cardiac connective tissue is mainly composed of collagen, with smaller amounts of elastin, laminin and fibronectin. Approximately 85% of the total collagen in heart consists of collagen type I. In addition to type I, other fibril forming collagen types found in the heart are III and V. Type IV and VI collagens are located in basement membranes in heart as in other tissues. Collagen XIII, which has a transmembrane domain, is found also in the heart. Furthermore, type XV and XVIII collagens, two members of the heterogenous group of non-fibril-forming collagens, are found in the myocardium (Weber et al. 1994, Hägg et al. 1997a, Myllyharju & Kivirikko 2001, Sund et al. 2001). The obvious function of collagenous extracellular matrix is to serve as the structural network for translating the force generated by individual myocytes into organized ventricular contraction and to prevent myocyte slippage, but it also accounts for passive stiffness in diastole and prevents overstretch as well as interstitial edema (for review, see Weber et al. 1994). Interstitial connective tissue network may also have a role in mechanosensing process via integrins.

The alterations in cardiac collagen network occur in response to pressure or volume overload and after myocardial infarct. In hypertrophied and failing hearts, interstitial fibrosis is generally seen (for review, see Boluyt & Bing 2000). Reparative fibrosis occurs as a reaction to a loss of myocardial material after necrosis or apoptosis, due to myocardial ischemia or senescence, and it is mainly interstitial. Another type of fibrosis, reactive fibrosis, is observed in the absence of cell loss as a reaction to inflammation and is primarily perivascular. Usually during cardiac remodeling, both types of fibrosis exist (Swynghedauw 1999). Structural changes are one of the key features in cardiac failure. Type I and type III collagen mRNAs were not significantly elevated in ventricles of the non-failing SHR but were increased 4-fold in failing hearts (Boluyt et al. 1994). This suggests a role for collagen accumulation in transition from LVH to failure.

Initially, after myocardial injury, the remodeling process is characterized by collagen fiber degradation, edematous intermuscular spaces and increased formation of type III collagen (Weber 1989). Fibrosis may occupy as much as 30% of the myocardium. The high level of fibrosis alters the mechanical properties of myocardium significantly: stiffness increases, and impaired diastolic filling and cardiac function may result (Boluyt & Bing 2000, Lorell & Carabello 2000). However, fibrosis is not seen with all models of LVH, suggesting that it may be regulated by other factors besides load. For instance, infrarenal aortic banding induces blood pressure increase and LVH without fibrosis (Weber et al. 1994). LVH associated with exercise training is not associated with fibrosis. Accordingly, a major role for Ang II and also aldosterone in fibrosis and collagen I accumulation has been demonstrated (Weber et al. 1994). Also ETs can increase collagen synthesis and decrease collagen degradation in cultured cardiac fibroblasts (Guarda et al. 1993), and ET receptor blockers inhibited fibrosis in SHR independently of blood pressure changes (Karam et al. 1996). The effect may be partially mediated through ETB receptor mediated aldosterone release (Wada et al. 1997). Natriuretic peptides and AM have been suggested to exert an antifibrotic effect in the heart (Tamura et al. 2000, Shimosawa et al. 2002).

Type I collagen in the heart is mainly synthesized by cardiac fibroblasts, and it is subject to slow metabolism with a half-life of 100 days (Swynghedauw 1999). The degradation of collagens occurs via specific collagenases (matrix metalloproteinases; MMP). The MMPs are activated by extracellular serine proteases. Tissue inhibitors of MMP form a complex with MMP in extracellular space, inhibiting collagen degradation (Weber et al. 1994, Swynghedauw 1999). The finding that inhibition of MMPs attenuates left ventricular dilatation in mice with experimental myocardial infarct has led to the proposal that MMP inhibitors could be used as a therapy for patients at risk for the development of heart failure after myocardial infarction (for review, see Creemers et al. 2001).

Type XV collagen is a homotrimer consisting of three α1(XV) chains (Rehn & Pihlajaniemi 1994). The mouse type XV collagen gene is 110 kb in size and contains 40 exons. It is characterized by a central highly interrupted triple helical domain and large N and C-terminal domains (Myers et al. 1992). The mRNA shows wide tissue distribution, but the highest levels in the mouse can be detected in the heart (Hägg et al. 1997b). The protein is localized mainly to basement membrane zones, but it is also found in the fibrillar collagen matrix near the basement membranes in certain human tissues (Myers et al. 1996, Hägg et al. 1997a). Type XV collagen has been suggested to act as a link between basement membrane and the underlying collagen matrix. The NC1 domain of the protein binds strongly to extracellular matrix proteins (Fig. 5) (Sasaki et al. 2000).

Interestingly, type XV and XVIII collagen share homology in the C-terminal domain, which contains the 20-kDa endostatin peptide (Sasaki et al. 2000). Endostatin, similar to that derived from type XVIII collagen, inhibits endothelial proliferation and potently inhibits angiogenesis and tumor growth (O"Reilly et al. 1997). Tumor growth and metastasis are dependent of the formation of blood vessels, and consequently the inhibition of tumor angiogenesis has been suggested as a strategy for treating cancer (Folkman 1971, Saaristo et al. 2000). The homology between type XV and XVIII collagen endostatin fragments is 60%. As expected, also type XV-derived endostatin has antiangiogenic functions. Proteolytically released XV-endostatins are found in mouse tissues, but the physiological function remains unclear (Sasaki et al. 2000). So far the significance of collagen XV for the cardiovascular structure and function has remained unclear.

Figure 5. A schematic presentation of basement membrane structure surrounding cardiac cells. Modified from Eklund et al. 2000, Sasaki et al. 2000 and Towbin & Bowles 2001.

The sarcoglycan subcomplex contains five subunits that are laterally associated with β -dystroglycan: α-, β -, δ−, γ − and ε-sarcoglygans. The sarcoglycan complex is involved in coupling of cells to basement membrane and to the extracellular matrix (for review, see Towbin & Bowles 2001). A common pathogenic feature for many muscular diseases could be disruption of the link between the ECM and the cytoskeleton, which may occur in the subsarcolemmal part (e.g. dystrophin), at sarcolemmal level (e.g. sarcoglycans and integrin α7) or in the ECM (e.g. laminin α2 chain and type VI collagen). Mice lacking δ-sarcoglycan and thereby disruption of the muscle cytoskeleton and the sarcoglycan-sarcospan complex in vascular smooth muscle develop cardiomyopathy (Coral-Vazquez et al. 1999). The primary cause of the heart phenotype is thought to be a perturbation in vascular function, and the vasodilator verapamil can save the hearts (Cohn et al. 2001). Dystrophin deficiency in mdx mice leads to myopathy associated with impaired running performance: The adult mice of this strain ran less than half of the distance achieved by wild-type mice in voluntary tread wheel tests (Carter et al. 1995, Wineinger et al. 1998).