Home SVCC                                                  Area: English - Español - Português

Genesis of Extracellular Matrix Variations
in Regurgitant Valvular Diseases and Relation to Development of Congestive Heart Failure

Jeffrey S. Borer, MD

Weill Medical College of Cornell University, New York, NY, USA

   Fibroblasts comprise the most numerous cell line in the myocardium. Their functions include secretion of proteins that form the extracellular matrix, or ECM, which forms the scaffold that determines the relation of cellular and vascular components of the myocardium. Despite the prominence of fibroblasts and their products, the physiologic role of the ECM and the pathophysiologic importance of specific ECM variations are incompletely understood. During the past decade, some important new knowledge has come from studies of regurgitant valvular diseases and, most particularly, from studies of aortic regurgitation (AR).

   The potential pathophysiological and clinical importance of the cardiac ECM in AR must be related to the predominant functional abnormality associated with CHF in this disease, i.e., loss of intrinsic contractility. Thus, the sequelae of valvular regurgitation are both caused and predicted by alterations in contractility, which is generated by intramyocyte contractile proteins, as indicated by our recent publication based on a prospective 15 year follow-up of more than 100 patients with severe AR, normal LV ejection fraction at rest and no symptomatic debility (1). Outcome events, including clinical heart failure, subnormal LVEF at rest, or sudden death, all were related to contractility at baseline, with the third of our patients manifesting the best contractility at baseline having the slowest rate of endpoint development, about 1.5%/year, and the third with the worst contractility at baseline, despite normal resting LVEF, having the fastest progression to endpoints, about 10 times faster than the "low-risk" tercile. Consequently, to justify focus on the ECM, perhaps we should be able to relate it to contractility.

   Our first suggestion of such a relation in humans came from our prospective study of the effects of aortic valve replacement for AR in patients followed for up to 10 years after operation2. LV performance eventually normalized after operation, but it took three years before improvement was maximal. Since contractile protein turnover requires only a few days, recovery of contractility takes one or two orders of magnitude longer than could be accounted for by complete replacement of the intracellular contractile proteins. In three years, 100 new hearts could be generated if only contractile proteins were considered. Thus, other processes are likely to be involved in post-op recovery of contractility and, equally likely, in its loss. Several candidate processes may modify the effects of contractile protein alterations. Because of differences in the inciting mechanical stresses these processes may differ to some extent in AR versus mitral regurgitation. The bulk of available data from my own work has resulted from studies in AR, as described below.

   The apparent disjunction between known contractile protein metabolic kinetics and recovery after valve replacement might be explained by variations in neurovascular modifications that occur after operation, by changes in non-myocyte myocardial proteins or by other processes. We chose to study non-myocyte ECM protein changes because, at the time of the index clinical study, we made a collateral observation in an animal model we had created with severe chronic AR. In summary, prior to the development of CHF, exuberant fibrosis often surrounded and enmeshed hypertrophied myocytes (3). The involved myocytes sometimes manifested myocytolysis either due to necrosis or apoptosis. However, inflammatory cell infiltrate invariably was absent. In the past, we thought that fibrosis was secondary to myocyte injury and resulting inflammation. These new observations made us begin to wonder whether we had the process backwards - in other words, that the fibrosis was primary and cell destruction was, at least in part, secondary. Of course, myocyte destruction, itself, could cause loss of contractility. However, ECM alterations may affect contractility by another, more complex mechanism.

   Data from studies in other models or diseases states indicate that the cardiomyocyte cytoskeleton is directly bound to the ECM by specific proteins that pierce the sarcolemma (4). In certain disease states, like some muscular dystrophies, genetically determined absence of one or more of these connectors is associated with loss of cardiac contractility. Even if the connecting proteins are synthesized, alteration in the nature of the connections, by a quantitative or qualitative change in the ECM, plausibly could cause a similar contractility deficit since the normal magnitude and vector of contractile force, that is, of systolic function, requires normal connection between myocytes and matrix - myocytes alone are not enough. Other experimental evidence suggests that the myocyte-matrix interaction is mediated at least in part by fibronectin, a highly expressed ECM protein that acts as an integrin (5).

   These latter possibilities led us to culture cardiac fibroblasts from rabbits with chronic experimental AR and to compare them with fibroblasts from normal rabbit hearts. Using both differential display PCR and selective subtractive hybridization, we that found several genes were abnormally expressed by AR fibroblasts (6). Some were upregulated and some were downregulated, but only three coded for ECM elements, these, here. All were upregulated, most prominent among them was fibronectin, and none was collagen.

   Substrate incorporation into the ECM was consistent with this finding (6). ECM incorporation of tritiated proline, a collagen component, by the cultured AR fibroblasts, was no different from incorporation by normal fibroblasts. In contrast, incorporation of tritiated glucosamine, a component of non-collagen ECM glycoproteins like fibronectin, was twice as great by AR fibroblasts as by normal fibroblasts.

   To further demonstrate that the alterations in gene expression were translated into protein synthesis changes, we performed Western blots for two common collagen ECM isoforms as well as for fibronectin. Collagen synthesis by AR fibroblasts was no different than normal (6). However, fibronectin synthesis clearly was upregulated in AR, by a ratio of about 2 to 1. Preliminary data from a parallel study indicates that activity of MMP-2, a metalloprotease that degrades myocardial collagen, is increased by AR, while fibronectinase activity is unaffected (7).

   These data describe a variety of interesting phenomena, but do not prove that the processes are primary results of AR, as I suggested above, rather than secondary results of myocyte destruction. To do this, we used an experimental design allowing us to grow normal cardiac fibroblasts in culture while applying mechanical strain at 60 cycles per second (8). We divided our cells so that some were grown during cyclical strain mimicking that impacting on AR fibroblasts at LV midwall. Others were grown during cyclical strain mimicking that found in the normal LV, and others were grown without strain.

   The results for substrate incorporation, protein synthesis and gene expression comparing unstrained, normally strained or AR-strained cells closely paralleled the results of comparisons of cultured AR versus normal fibroblasts. Thus, ECM remodeling in AR is likely to be a primary response to the mechanical stresses of volume loading (8).

   Additional studies are needed before novel therapy can result from these observations. However, we have already shown that drugs can affect the process. Using the quinalone derivative, vesnarinone, for suppression of cardiac fibroblast survival, we reported that, at doses an order of magnitude lower than those used clinically, normal cardiac fibroblast survival was significantly suppressed. AR cells were affected to a greater extent and at a lower dose, providing a potential therapeutic advantage (9).

   While the vesnarinone work proves that the process can be modulated with drugs, the best pharmacological targets still need to be defined. For this, it will be helpful to define the range of possible targets, beginning with the pathway responsible for activation of the ECM genes by volume loading. Transduction of mechanical stress may begin with alteration of membrane receptors or other integrins, by functional distortion of the cytoskeleton, or by some other mechanism.

   The initial mechanically-induced alteration stimulates a cascade of intracellular reactions that culminate in activation of a gene, for example, the fibronectin gene. Several pathways impacting on fibronectin expression have been identified in other cell systems, and some are known to be activated by specific mechanical stresses. The relevant pathway in AR fibroblasts is not known. We have demonstrated that AR upregulates a reaction that normally is the last step prior to activation of a fibronectin promoter (10). This step involves activation of Jun kinase leading to phosphorylative activation of the promoter. We are now studying the effect of AR on the reaction that precedes Jun kinase activation in other systems, that is, activation of a mitogen-activated kinase, MEK 4/7. We plan to move backward through the known sequence pathway until we find a reaction that doesn't respond to AR. We'll then seek a relevant alternative, until we reach the mechanical initiator of the pathway. In this way, we hope to define the appropriate targets for novel therapy and, at the same time, to define descriptors for refinement of prognostication, as well.

   In summary, it seems likely that the mechanical strain of volume loading in regurgitant valvular diseases and, specifically, in AR, directly stimulates fibroblasts to cause ECM remodeling with fibrosis. Fibronectin and, perhaps, other non-collagen proteins may affect myocyte-matrix interaction, potentially modulating contractility. To some extent, this process may be adaptive and compensatory. Ultimately, however, it may be destructive, compromising contractility and leading to heart failure. Therefore, suppression of this process may be therapeutic. Additional study of the phenomena related to myocardial fibrosis in regurgitant valvular diseases is likely to result in knowledge that can abet efforts to define new and novel strategies for preventing or reversing heart failure in these and related conditions.


1. Borer JS, Hochreiter C, Herrold EM, Supino P, Aschermann M, Wencker D, Devereux RB, Roman MJ, Szulc M, Kligfield P, Isom OW. Prediction of indications for valve replacement among asymptomatic or minimally symptomatic patients with chronic aortic regurgitation and normal left ventricular performance. Circulation 1998; 97:525-534.

2. Borer JS, Herrold EM, Hochreiter C, Roman MJ, Supino P, Devereux RB, Kligfield P, Nawaz H. Natural history of left ventricular performance at rest and during exercise after aortic valve replacement for aortic regurgitation. Circulation 1991;84(Suppl. III):III-133-III-139.

3. Liu S-K, Magid NM, Fox PR, Goldfine SM, Borer JS. Fibrosis, myocyte degeneration and heart failure in chronic experimental aortic regurgitation. Cardiology, 1998; 90:101-109.

4. Leiden J. The genetics of dilated cardiomyopathy - emerging clues to the puzzle. New England Journal of Medicine 1997;337:1080-1081.

5. Ahumada G, Saffitz J. Fibronectin in the rat heart: a link between cardiac myocytes and collagen. J. Histochem. Cytochem. 1984;32:383-388.

6. Borer JS, Pena M, Falcone DJ, Truter SL, Kolesar JA, Dumlao TF. Chronic aortic regurgitation; pre-heart failure fibrosis characterized by fibronectin excess. Circulation, 1999;100(Suppl I):I-519.

7. Truter SL, Lee JA, Dumlao TF,Young K, Borer JS. Collagenase activity selectively increases in aortic regurgitation to suppress collagen content in fibrosis. Journal of the American College of Cardiology, 2001;37:2:1047-107:473A.

8. Herrold EM, Borer JS, Truter SL, Carter JN, Liu F, Dumlao TF. Myocardial fibrosis in aortic regurgitation: fibroblast response to in vitro strain is magnitude dependent. Circulation: 2000;102(Suppl II):II-530.

9. Ross JS, Goldfine SM, Herrold EM, Borer JS. Differential response to vesnarinone by cardiac fibroblasts isolated from normal and aortic regurgitant hearts. American Journal of Therapeutics 1998; 5:369-375.

10. Truter SL, Lee JA, Dumlao TF, Borer JS. Increased fibronectin expression in aortic regurgitation: role of stress activated protein kinase/c-Jun NH2-terminal kinase pathway. Journal of the American College of Cardiology, 2001;37:2:875-3:487A.



Your questions, contributions and commentaries will be answered
by the lecturer or experts on the subject in the Clinical Cardiology list.
Please fill in the form (in Spanish, Portuguese or English) and press the "Send" button.

or commentary
Name and Surname:
E-Mail address:


2nd Virtual Congress of Cardiology

Dr. Florencio Garófalo
Steering Committee
Dr. Raúl Bretal
Scientific Committee
Dr. Armando Pacher
Technical Committee - CETIFAC

Copyright© 1999-2001 Argentine Federation of Cardiology
All rights reserved


This company contributed to the Congress