The role of bone morphogenic proteins in human aortic valvular endothelial cells

Thumbnail Image
Ankeny, Randall Francis
Jo, Hanjoong
Nerem, Robert M.
Associated Organization(s)
Supplementary to
In the United States alone, there are nearly 49,000 aortic valvular repairs or replacements each year, and this number is expected to rise. Unlike atherosclerosis, the molecular mechanisms contributing to this side-dependent disease development are limited, which contributes to the lack of therapeutic treatments. Once clinically manifested, options for treatment are limited to valvular replacement or repair. Therefore understanding the mechanobiology and cellular responses in aortic valves may provide important information for disease development and possible biomarkers or therapeutic treatments. Aortic valve disease occurs on one side of the valvular leaflet. The fibrosa side, which faces the aorta, is prone to disease development, while the ventricularis remains relatively unaffected. The hemodynamics is hypothesized to play a role in side dependent disease formation. The fibrosa endothelium is exposed to oscillatory flow while the ventricularis endothelium is exposed to a pulsatile unidirectional flow. Previous work by our group has shown that bone morphogenic protein-4 is a mechanosensitve inflammatory cytokine in the vasculature. In the following study, we proposed that mechanosensitive bone morphogenic proteins play a role in side specific aortic valve disease. Recently, the bone morphogenic proteins (BMPs) have been found in calcified human aortic valves. Furthermore, BMP-4 in vascular endothelial cells is increased by oscillatory shear stress. However, the role of the BMPs in aortic valve endothelial cells and their contribution to aortic valve calcification remains unstudied. Therefore, the overall objective of this dissertation was to investigate how disease and hemodynamics affects the BMP pathway and inflammation in human aortic valvular endothelial cells. By understanding how the bone morphogenic proteins are regulated and what roles they play in aortic valve disease, we will have better insight into endothelial cell regulation and contribution in aortic valve pathology. The central hypothesis of this project was that oscillatory flow conditions on the fibrosa side of the aortic valve stimulate endothelial cells to produce BMP-4, which then activates an inflammatory response leading to accumulation of inflammatory cells, calcification, and ultimately valve impairment. This hypothesis was tested through three specific aims using calcified human aortic valves, non-calcified human aortic valves, and side-specific human aortic valve endothelial cells. We first worked to establish the importance of the BMPs in the aortic valvular endothelium by looking at two populations of aortic valves: 1) calcified human aortic valves were obtained from patients undergoing valve replacement, and 2) non-calcified valves were obtained from recipient hearts of patients undergoing heart transplantation. Using immunohistochemical techniques, we examined the BMPs, BMP antagonists, and SMADs. Surprisingly, we identified that the ventricularis endothelium had higher BMP expression in both calcified and non-calcified human aortic valves. Furthermore, no disease-dependent BMP expression was detected. Next, we examined the BMP antagonists and found that there was robust BMP antagonist expression in the ventricularis endothelium and very low expression in the fibrosa endothelium. Finally, to determine if the BMP pathway was activated, we stained for the canonical BMP signaling molecule phosphorylated-SMAD 1/5/8 and found increased staining in the endothelium of calcified human aortic valves. Furthermore, a significant increase in SMAD 1/5/8 phosphorylation was seen in the endothelium of calcified fibrosa when compared to the non-calcified fibrosa. Finally, inhibitory SMAD 6 was significantly increased in the ventricularis endothelium of non-calcified human aortic valves. These findings suggest that preferential activation of BMP pathways, controlled by the balance between the BMPs and their inhibitors, play an important role in side-dependent calcification of human AVs. We next wanted to examine the role of shear stress in BMP regulation, but before doing so, we needed to examine the endothelial response to fluid shear stress to validate the phenotype of our isolated human aortic valve endothelial cells. KLF2 and eNOS expression in vascular endothelial cells has been shown to be increased by laminar flow and to have anti-inflammatory effects by decreasing VCAM-1 levels. Conversely, oscillatory shear stress has been shown to increase NF-kappa B translocation and increase ICAM-1 and E-selectin. We found laminar shear stress causes human aortic valve endothelial cells align parallel to flow and have robust increases of KLF2 and eNOS and decreases in VCAM-1 levels; however, laminar shear-treated cells had similar levels of NF-kappa B activation as oscillatory treated cells while ICAM-1 and E-selectin was not affected by shear stress. In contrast, oscillatory shear had higher levels of monocytes bound which may be due to eNOS's protective effects under laminar shear and robust VCAM-1 expression in oscillatory shear. These studies suggest differential regulation of human aortic valvular endothelial cells than published reports on human aortic endothelial cells which adds to the growing evidence that valvular endothelial cells are phenotypically different than vascular endothelial cells. After verifying the shear response of our endothelial cells, we next determined the shear response of the BMPs and BMP antagonists and described BMPs' effect on inflammation. Previously, BMP-4 has been shown in vitro and in vivo to be increased in endothelial cells exposed to oscillatory flow, while the closely-related BMP-2 has not been shown to be shear sensitive. In this study we have found that BMPs -2 and -4 are shear sensitive while BMP-6 is not. Furthermore, we have found that follistatin is decreased by laminar flow only in the ventricularis, while MGP1 is decreased in the fibrosa valvular endothelial cells under both oscillatory and laminar flow. Finally, incubation with noggin did not affect monocyte adhesion after shear, suggesting differential regulation of inflammation in human aortic valvular endothelial cells. By addressing the specific aims of this project, we have investigate disease- and side-dependent valvular endothelial BMP expression in vivo, shear regulation of valvular endothelial inflammation in vitro, and shear regulation of valvular endothelial BMP expression in vitro. Our results suggest that the BMP pathway is playing a role in side specific aortic valve disease development; however, regulation of the BMPs does not appear to be shear regulated in vivo. Other factors that may be affecting BMP production include including pulsatile pressures, bending stresses, cyclic stretch, and humeral stimuli present in the blood of the patients. However, in vitro we have found BMPs -2 and -4 to be shear-regulated in human aortic valvular endothelial cells. Shear-induced inflammation in human aortic valve endothelial cells seems to be VCAM-1-dependent, and BMP-independent. Finally, by identifying factors that are modulated in calcific- and shear-dependent processes, new targets for the early detection of aortic valve disease can be determined and new therapeutics to slow or stop the progression of aortic valve disease may be discovered.
Date Issued
Resource Type
Resource Subtype
Rights Statement
Rights URI