According to statistics from the American Heart Association, valvular heart disease directly leads to about 20,000 deaths a year and contributes to an additional 50,000. While significant advancements have been made in the treatment options available for valvular heart disease, complications still occur. For this reason, the future of valvular heart disease treatment lies in understanding the physiology of the heart valve, and subsequently bioengineering a valve from one's own tissue to mimic native valve processes. Valvular interstitial cells (VICs) are the major cell type populating the valve matrix. In the inactive fibroblast-like state, these cells are responsible for extracellular matrix deposition. Activated VICs display a myofibroblast morphology characterized by the expression of alpha smooth muscle actin and are responsible for valve maintenance and repair. The activation of VICs is hypothesized to be stimulated by mechanical tension, which, in the presence of TGF-â1 allows the complete differentiation of VICs from the inactive to the active form. However, little is known about the potential for reversal or dedifferentiation from the active to inactive state. The purpose of this study was to determine whether substrate stiffness, the mechanical tension hypothesized to initiate VIC activation, modulates alpha smooth muscle actin expression in the presence and absence of TGF-â1. To mimic conditions found in vivo, substrates were varied from physiologic to pathological stiffness levels. Results showed that when freshly isolated VICs are cultured in the presence of serum, alpha smooth muscle actin expression increased on all substrate stiffnesses. In TGF-â-free medium, there was an apparent increase on all stiffness levels as well, but a statistical significance between groups could not be demonstrated. Immunoblots used to detect TGF-â1 showed that intracellular TGF-â1 was upregulated in VICs cultured in the presence of serum compared to those cultured in TGF-â-free medium. Taken together, these results suggest that freshly isolated VICs become activated, as indicated by increased expression of alpha smooth muscle actin, on all substrate levels in the presence of serum. It also appears as though unknown factors which are present in serum are required to stimulate significant autocrine production of TGF-â1. To determine whether VICs which had transitioned to the myofibroblast phenotype had the ability to dedifferentiate, cells were cultured on polystyrene for a minimum of four days then replated on substrates of varying stiffness. Analysis of alpha smooth muscle actin expression showed that, in the presence of serum and when replated on all of substrates used, alpha smooth muscle actin expression decreased, suggesting that these cells indeed have the potential to dedifferentiate. A change in cell morphology to a more rounded phenotype as well as the loss of visible stress fibers further supported this possibility. These studies represent a unique approach to studying phenotypic differentiation of valvular interstitial cells. Using acrylamide substrates of varying stiffness, and growth factor free media, we have shown that by altering substrate stiffness, changes in alpha smooth muscle actin expression consistent with differentiation and dedifferentiation can be induced. This potential for dedifferentiation suggests that in engineering the next generation of bioartificial valves, it may be possible to use the patient's own cells to seed the manufactured scaffold. This would avoid complications associated with current treatments, including immune rejections.

The aortic valve maintains unidirectional blood flow between the left ventricle and the systemic circulation. When diseased, the valve is replaced either by a mechanical or a bioprosthetic heart valve, that carry issues such as thrombogenesis, long term structural failure, and calcification, necessitating the development of more structurally and biologically sufficient long-term replacements. Tissue engineering provides a possible avenue for development, combining cells, scaffolds, and biochemical factors to regenerate tissue. The overall goal of this dissertation was to create a foundation for the rational design of a tissue engineered aortic valve. The novel approach taken in this thesis research was to view each of the three leaflets as a laminate structure. The first three aims consider the leaflet as a laminate structure comprising of layers of collagen, elastin, and glycosaminoglycans (GAGs). In the first aim, the effect of GAGs on the tensile properties and stress relaxation in the leaflet was investigated, by removing GAGs through increasing amounts of hyaluronidase. A decrease in GAGs led to significantly higher elastic moduli, maximum stresses, and hysteresis in the leaflet. In the second aim, the 3D elastic fiber network of the leaflet was characterized using immunohistochemistry and scanning electron microscopy. This structure was found to have regionally varying thicknesses and patterns. In the third aim, a novel hydrogel-fiber composite design was proposed to match the anisotropy of the leaflet. This composite composed of aligned electrospun poly(ε-caprolactone) (PCL) within a poly(ethylene glycol) diacrylate (PEGDA) matrix. Surface modification and embedding of the PCL did not significantly alter the anisotropy or strength of the underlying PCL scaffold, providing the basis for an anisotropic, biocompatible scaffold. In the last aim, a novel co-culture model was designed using magnetic levitation as a layered structure of valvular endothelial cells and interstitial cells. This technique was used to create co-culture models within hours, while maintaining cell phenotype and function, and inducing extracellular matrix formation, as shown by immunohistochemical stains and their gene expression profiling. The overall result of this dissertation is a clearer understanding of the layered structure-function relationship of the aortic valve, and its application towards heart valve tissue engineering.