Bone is the most commonly replaced organ, with nearly 1 million grafting procedures performed annually in the United States. Inherent limitations associated with bone grafts, such as graft availability and donor site morbidity, leave room for alternative grafting solutions. Current mineralized tissue engineering approaches include the use of synthetic hydroxyapatite as cement or as nano- or microparticles pre-incorporated into a tissue engineering scaffold prior to cell seeding or implantation. While promising results have been reported with such methods, these constructs are not biomimetic as they fail to replicate neither the size, distribution, nor density of mineral inherent in the native bone, leading to inferior mechanical properties and supra-physiologic levels of calcium phosphate that can disrupt healing, alter cell response and inhibit normal tissue homeostasis. To address these issues, inspiration is taken from the native biomineralization process which is often facilitated by matrix vesicles, a lipid-based nanocarrier within which calcium and phosphate ions are combined to form calcium phosphate mineral in hard tissues such as bone. Synthetic matrix vesicles (SMV) formulated from self-assembling liposomes have emerged as a promising model both for studying the biomineralization process as it relates to matrix vesicles and for use in regenerative medicine. The ideal SMV system is defined as follows:​ the mineral formed should match the native calcium phosphate in both structure and chemistry, the mineral must be stable in the physiological environment and can continue to grow in size when necessary and the matrix vesicles should also be able to work in conjunction with a scaffold tailored for bone tissue engineering. It is hypothesized that the formation of native bone-like calcium phosphate can be achieved with the controlled optimization of matrix vesicles in terms of fabrication parameters, ion transport, cell response and interactions with a gelatinous matrix.

To this end, a liposome-based, biomimetic matrix vesicle system was designed to facilitate vesicle-mediated biomineralization for regeneration of calcified tissues. Synthetic matrix vesicles were fabricated from two different phospholipids, DPPC (1,2-dipalmitoyl-sn-glycero-3-phosphocholine) and DMPC (1,2-bis(myristoyl)-sn-glycero-3-phosphocholine) and optimized in terms of membrane composition, alkaline phosphatase bioconjugation, and ion encapsulation. Calcium (Ca²⁺) and phosphate (Pi) ions were successfully encapsulated within the liposomes. Ion permeability across the bi-layer membrane, which is necessary for Ca²⁺ and Pi to combine within the SMV for mineralization, was found to increase with increasing DMPC composition, validated through ion release studies and diffusion modeling through Fick’s 2nd Law. In addition, alkaline phosphatase (ALP), an enzyme which cleaves Pi from organic phosphate molecules for mineral formation with Ca²⁺, was successfully conjugated to the SMV membrane through the use of biotin-functionalized phospholipids and streptavidin-ALP. Human osteoblast-like cells were dosed with the optimized SMV and the effects of SMV type and dosage on mineralization response was evaluated. Mineralization potential of human osteoblast-like cells was found to decrease through exposure to Pi-encapsulated SMV similar to the response found for human osteoblast-like cells supplemented with β-glycerophosphate (β-GP), an organic phosphate source typically used in mineralization in vitro studies. Human osteoblast-like cells were also dosed with two different configurations of ALP SMV liposomes with ALP bound within (ALP-inside SMV) and liposomes with ALP bound to the membrane on the outside (ALP-outside SMV). ALP-outside SMV were ultimately selected for further study since the location of the ALP in the outside configuration more closely mimics the structure of native matrix vesicles. While mineral-like structures were observed in several types of SMV under cryo-electron microscopy, no bulk mineralization was observed by human osteoblast-like cells from SMV supplementation alone. This motivated a dosage study conducted with the Pi SMV which optimized the cell-to-liposome ratio and the concentration of Pi encapsulated. The optimized ALP-outside SMV and Pi SMV were individually combined with an electrospun gelatin nanofiber scaffold to further promote cell mineral deposition by acting as a biomimetic substrate for calcium phosphate nucleation. It was demonstrated that in the absence of growth factor stiumulation, culture of human osteoblast-like cells with SMV+βGP and Pi SMV resulted in mineral deposition on the gelatin nanofiber scaffold. Human mesenchymal stem cells (hMSC), a more clinically relevant cell type, were also cultured on the SMVgelatin scaffold system. Mineralization potential was found to increase for hMSC cultured with ALP SMV, and the osteogenic marker osteocalcin was upregulated for cultures with Pi SMV. Dosage of hMSC with SMV+β-GP and Pi SMV alone resulted in the formation of a mineralized matrix.

In summary, this thesis focuses on the design of a biomimetic, liposome-based synthetic matrix vesicle system and elucidates the compositional and dosage parameters for the formation of calcified tissue by human osteoblast-like cells and MSCs. The synthetic matrix vesicle system developed in this thesis can be utilized for further investigation into the mechanisms of biomineralization, in addition to its potential for use in promoting cell-mediated regeneration of a variety of calcified tissues, including bone, teeth and mineralized soft-to-hard tissue interfaces.