Elevated cyclic stretch alters matrix remodeling in aortic valve cusps: implications for degenerative aortic valve disease

Abstract

Matrix metalloproteinases (MMPs) and cathepsins are proteolytic enzymes that are upregulated in diseased aortic valve cusps. The objective of this study was to investigate whether elevated cyclic stretch causes an increased expression and activity of these proteolytic enzymes in the valve cusp. Circumferentially oriented fresh porcine aortic valve cusp sections were stretched to 10% (physiological), 15% (pathological), and 20% (hyperpathological) in a tensile stretch bioreactor for 24 and 48 h. The expression and activity of MMP-1, MMP-2, MMP-9, tissue inhibitor of MMP-1, and cathepsin L, S, and K were quantified and compared with fresh controls. Cell proliferation and apoptosis were also analyzed. As a result, at 10% physiological stretch, the expression and activity of remodeling enzymes were comparable with fresh controls. At 15% stretch, the expression of MMP-1, -2, -9 and cathepsin S and K were upregulated, whereas the expression of cathepsin L was downregulated compared with controls. A similar trend was observed at 20% stretch, but the magnitudes of upregulation and downregulation of the expression were less than those observed at 15%. In addition, there were significantly higher cell proliferation and apoptosis at 20% stretch compared with those of other treatment groups. In conclusion, elevated mechanical stretch on aortic valve cusps may detrimentally alter the proteolytic enzyme expression and activity in valve cells. This may trigger a cascade of events leading to an accelerated valve degeneration and disease progression.

Fig. 1.

A: preparation of aortic valve cusps for the experiment. A circumferentially oriented 15 × 10-mm section was excised from the central region of porcine aortic valve cusps. B: ex vivo tensile stretch bioreactor used in this study (Ref. 8). A magnified image of the tissue chamber is shown (right), showing the 8 tissue wells with aortic valve cusps. C: loading curve used in this study. The temporal strain waveform imposed (Fig. 2A) in the circumferential direction was derived from the variations of the cusp length in the circumferential direction over one cardiac cycle (35). The extension of the cusp corresponded to diastole (2/3 of the cardiac cycle), whereas the unstretched state of the cusp corresponded to systole (1/3 of the cardiac cycle). The gradients of the extension and relaxation approximated those experienced in vivo.

Fig. 2.

A: cell proliferation was increased by cyclic stretch in a magnitude-dependent manner. Proliferation as measured by bromodeoxyuridine (BrdU) staining (arrows) is highest at 20% stretch after 48 h, showing numerous proliferating cells across the thickness of the cusp. V, ventricularis; n , number of valve cusps. B: cell apoptosis was increased by cyclic stretch in a magnitude-dependent manner. Apoptotic cell number as determined by terminal deoxyribonucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) staining (arrows) was significantly higher at 20% (70 immunopositive cells/unit area of tissue) compared with 10% and 15% stretch (10 and 27 immunopositive cells/unit area of tissue respectively) after 48 h. These immunopositive cells at 20% were distributed throughout the thickness of the tissue.

Fig. 3.

Cyclic stretch increased collagen in aortic valve cusps. Collagen content was significantly reduced (P < 0.05) at 10% stretch but significantly increased (P < 0.05) at 15% and 20% stretch after 48 h compared with fresh controls (n = number of valve cusps). amt, Amount.

Fig. 4.

A: effect of cyclic stretch on cathepsins L, S and K. There was a clear stretch magnitude dependence in the regulation of cathepsins. Among the three cathepsins analyzed, cathepsin L was the dominant cathepsin in the fresh valve, whereas 15% stretch significantly increased expression of cathepsins S and K. F, fibrosa; n, number of valve cusps. B: effect of stretch magnitude on cathepsin expression.

Fig. 5.

A: effect of cyclic stretch on matrix metalloproteinase (MMP)-1, -2, and -9 and tissue inhibitor of MMP-1 (TIMP-1). Immunoblotting revealed that MMP expression increased in a magnitude-dependent manner, peaking at 15% cyclic stretch before falling at 20%. Conversely, TIMP-1 expression was reduced under stretch. B: correlation between MMP-1 expression and TIMP-1 expression. There was a strong (R² = 0.755, P < 0.05) negative correlation between TIMP-1 and MMP-1 expression. C: relation between MMP-to-TIMP ratio (MMP/TIMP) and stretch. There was a significant increase in MMP/TIMP under 15% stretch compared with 10% or 20% stretch.

Fig. 6.

A: effect of cyclic stretch on gelatinases. Gelatinase activity was highest at 15% and 20% cyclic stretch after 48 h. Pro-MMP-9 was observed at 15% stretch (48 h only). B: effect of cyclic stretch on metalloproteinase inhibitors. TIMP inhibitory activity was reduced by cyclic stretch. The maximum reduction was seen at 10% stretch. C: effect of cyclic stretch on collagenases. Collagenase activity progressively increased with increasing cyclic stretch. Activity was significantly higher at 15% and 20% cyclic stretch compared with that of fresh controls. D: correlation between enzyme activity and expression. There was good (R² = 0.597, P < 0.05) correlation between MMP/TIMP activity and expression for all the treatment groups (stretch level or time). OD, optimal density.