Elevated cyclic stretch induces aortic valve calcification in a bone morphogenic protein-dependent manner

Abstract

Calcified aortic valve (AV) cusps have increased expression of bone morphogenic proteins (BMPs) and transforming growth factor-beta1 (TGF-beta1). Elevated stretch loading on the AV is known to increase expression of matrix remodeling enzymes and pro-inflammatory proteins. Little, however, is known about the mechanism by which elevated stretch might induce AV calcification. We investigated the hypothesis that elevated stretch may cause valve calcification via a BMP-dependent mechanism. Porcine AV cusps were cultured in a stretch bioreactor, at 10% (physiological) or 15% (pathological) stretch and 70 beats per minute for 3, 7, and 14 days, in osteogenic media supplemented with or without high phosphate (3.8 mmol/L), TGF-beta1 (1 ng/ml), as well as the BMP inhibitor noggin (1, 10, and 100 ng/ml). Fresh cusps served as controls. Alizarin red and von Kossa staining demonstrated that 15% stretch elicited a stronger calcification response compared with 10% stretch in a fully osteogenic medium containing high phosphate and TGF-beta1. BMP-2, -4, and Runx2 expression was observed after 3 days on the fibrosa surface of the valve cusp and was stretch magnitude-dependent. Cellular apoptosis was highest at 15% stretch. Tissue calcium content and alkaline phosphatase activity were similarly stretch-dependent and were significantly reduced by noggin in a dose dependent manner. These results underline the potential role of BMPs in valve calcification due to altered stretch.

Figure 1

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 (A). An ex vivo tensile stretch bioreactor was used in this study (B). A magnified image of the tissue chamber is shown on the right. A loading curve was used in this study. The gradients of the extension and relaxation approximated those experienced in vivo (C).

Figure 2

Normal tri-layered morphology of aortic valve cusps was maintained in stretch experiments. Tissue morphology and structure was comparable between fresh controls and samples stretched for 3, 7, and 14 days (A). There was no significant difference in tissue thickness between fresh controls and samples stretched for 3, 7, and 14 days (B) [n = 10]. Samples cultured to 3, 7, and 14 days showed some evidence of DNA deterioration as demonstrated by arrows (C); however, the number of necrotic cells was less than 3.5% of the total number of cells (D) [n = 8] F indicates fibrosa. *P < 0.05.

Figure 3

Alizarin red staining of porcine aortic valve cusps that were cultured for seven days under 10% (A–D) or 15% (E–H) stretch in normal DMEM, osteogenic DMEM, osteogenic DMEM with high phosphate (3.8 mmol/L), and osteogenic DMEM with 1 ng/ml TGF-β1, as indicated. Small calcific nodules were found only in G and H as marked with arrows. F indicates fibrosa; V, ventricularis.

Figure 4

Cyclic stretch induced calcification of porcine aortic valve cusps. Aortic valve cusps were exposed to 10% (A and B) or 15% (C and D) cyclic stretch for seven days in fully osteogenic medium containing both high phosphate and TGF-β1. Cusps were then stained with Alizarin Red (A and C) or Von Kossa (B and D). Calcific nodules are marked with arrows. F indicates fibrosa; V, ventricularis.

Figure 5

Elevated stretch-induced apoptosis of porcine aortic valve cusp cells. Fresh valve cusps and cusps exposed to 10% or 15% cyclic stretching in fully osteogenic medium were assayed for apoptosis by terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling staining (green) with 4′,6-diamidino-2-phenylindole nuclei counterstain (blue), with apoptotic cells detected marked with arrows (A). The normalized number of apoptotic cells per image field was then quantified for each treatment condition (B). F indicates fibrosa; V, ventricularis. *P < 0.05; **P < 0.05 vs. fresh control.

Figure 6

Representative micrographs of BMP-2 and BMP-4 immunohistochemical staining (A) and semiquantification (B) after three days of culture (10% or 15% stretch) in fully osteogenic DMEM compared with fresh controls. Expression of BMP was primarily observed on the fibrosa surface of the valve cusp. Arrows represent immunopositive cells. F indicates fibrosa. *P < 0.05; **P < 0.05 vs. fresh control.

Figure 7

Representative micrographs of BMP-2 (green) and von Willebrand Factor (red) double immunohistochemical staining for samples stretched at 10% (A and B) and 15% (C and D) in fully osteogenic media. Cell nuclei are stained with 4′,6-diamidino-2-phenylindole (blue) and presented in a separate image for clarity of viewing. Arrows represent immunopositive cells. F indicates fibrosa.

Figure 8

Calcium content in stretched (10% and 15%) valves cultured in fully osteogenic DMEM in 0, 1, 10, and 100 ng/ml noggin for 14 days (A) [n = 8]. ALP activity in stretched (10% and 15%) valves cultured in fully osteogenic DMEM in 0, 1, 10, and 100 ng/ml noggin for seven days (B) [n = 8]. *P < 0.05; **P < 0.05 versus fresh control.

Figure 9

Runx2 expression in stretched (10% and 15%) valves cultured in fully osteogenic DMEM in 0 and 100 ng/ml noggin for three days (A) [n = 6]. Osteocalcin (OST) expression in stretched (10% and 15%) valves cultured in fully osteogenic DMEM in 0 and 100 ng/ml noggin for seven days (B) [n = 6]. *P < 0.05; **P < 0.01.