Tendons, essential tissues that connect muscles to bones, are susceptible to rupture/degeneration due to their continuous use for enabling movement. Often surgical intervention is required to repair the tendon; relieving the pain and fixing the limited mobility that occurs from the damage. Unfortunately, post-surgery immobilization techniques required to restore tendon properties frequently lead to scar formation and reduced tendon range of motion. Our ultimate goal is to create an optimal tendon prosthetic that can stabilize the damaged muscle-bone connection and then be remodeled by resident cells from the surrounding tissues over time to ensure long-term function. To achieve this, we must first understand how cells respond to and interact with candidate replacement materials. The most abundant extracellular matrix (ECM) protein found in the body, collagen, is chosen as the replacement material because it makes up the majority of tendon dry mass and it can be remodeled by cell-based homeostatic processes. Previous studies found that Di-catechol nordihydroguaiaretic acid (NDGA) cross-linked fibers have greater mechanical strength than native tendons; and for this reason this biomaterial could be used for tendon replacement. This work focuses on investigating the behavior of fibroblasts on NDGA cross-linked and uncross-linked collagen samples to determine if cross-linking disrupts the cell binding sites affecting cell spreading, attachment, and migration. The in-vitro platform was designed by plasma treating 25 mm diameter cover slips that were exposed to 3-aminopropyl-trimetoxysilane/toluene and glutaraldehyde/ethanol solutions. The collagen solution was then dispensed onto the glutaraldehyde-coated cover slip and incubated for fibrillar collagen matrix formation. The collagen matrices were submerged in NDGA cross-linking solution for 24 hours to ensure the surface was completely cross-linked. Collagen films were made by allowing the uncross-linked gels to dry overnight before and after NDGA treatment, resulting in a more compacted structure. A spinning disk device was employed to quantify the ability of cells to remain attached to the collagen samples when exposed to hydrodynamic forces. To avoid any cell-cell interaction and focus on cell-surface interactions, 50-100 cells/mm2 were seeded carefully on each sample. Temporal studies demonstrated that cell adhesion strength and spreading area reached steady-state by 4 hr. Adhesion and spreading studies along with migration experiments demonstrated that NDGA treatment affects cellular behavior on films, partially reducing adhesion strength, migration, and spreading area. However, on the cross-linked gels which are less dense, the only change in cell behavior observed was in migration speed. We hypothesize that these differences are due to the collapsing of the collagen films. This compaction suggests a less open organization and could be allowing the collagen fibers to form more inter-chain bonds as well as bonds with the small NDGA cross-linker; while NDGA treatment of the fully hydrated gels may rely more on NDGA polymerization to span the greater distance between collagen fibrils. From these results, we can determine that the chemical/physical masking of the adhesion sites by NDGA on collagen films affects cellular behavior more than the masking that occurs in the cross-linked gels. Although this study shows an effect in cell behavior on the cross-linked films, it also demonstrates that cells can adhere and migrate to this NDGA biomaterial supporting the idea that this biomaterial can be utilized for tendon replacement.

The mechanisms of interaction between inorganic matter and biomolecules, as well as properties of resulting hybrids, are receiving growing interest due to the rapidly developing field of bionanotechnology. The majority of potential applications for metal-biohybrid structures require stability of these systems under vacuum conditions, where their chemistry is elusive, and may differ dramatically from the interaction between biomolecules and metal ions in vivo. Here we report for the first time a photoemission and X-ray absorption study of the formation of a hybrid metal-protein system, tracing step-by-step the chemical interactions between the protein and metals (Cu and Fe) in vacuo. Our experiments reveal stabilization of the enol form of peptide bonds as the result of protein-metal interactions for both metals. The resulting complex with copper appears to be rather stable. In contrast, the system with iron decomposes to form inorganic species like oxide, carbide, nitride, and cyanide.

Today there is an increasing clinical demand and expectation of patients for biomaterials, which underscores the importance of discovering the correlations between biomaterials and biological systems, especially blood. When an artificial material makes contact with blood, the first event is a rapid adsorption of plasma protein on the material surface, on top of which the innate immune system is triggered, with potentially detrimental consequences. The work presented in this thesis, reported in four papers, was designed to investigate complications associated with (a) biomaterial-induced immune systems, including activation mechanisms and crosstalk between cascades on the biomaterial surface, and with (b) clinical investigations. In Paper I and Paper II, a series of studies led to the development of a direct prediction of the subsequent biological events based on the pattern of initially bound proteins. A reciprocal relationship was demonstrated between activation of the contact system and the complement system when they were induced on artificial material surfaces. Based on these studies, a robust and simple method for biocompatibility testing was proposed and validated, yielding high specificity and sensitivity when compared to today’s gold standard. Paper III investigated biomaterial-induced activation of complement and leukocytes in dialysis treatment-related conditions. The results suggested that citrate is more biocompatible than the conventionally used acetate. This reduction in activation could be further enhanced with higher citrate concentrations, suggesting that dialysis fluid containing citrate is a promising alternative to acetate dialysis fluid. Paper IV investigated complement initiation mechanisms with clinical implications. An experimental system was set up to revisit the initiation of the complement alternative pathway, and correlations were found between chaotropic or nucleophilic agents and iC3 generation under physiologically relevant conditions. A clinical study of hepatic encephalopathy patients indicated a direct correlation between elevated plasma ammonia and iC3 formation, as well as with complement activation in vivo.  Taken together, these studies have provided a model for a robust biomaterial test and have investigated biomaterial-induced complications in the fluid phase in clinically related conditions; furthermore, the basic mechanisms of complement activation have been dissected in relation to disease symptoms. Keywords: Complement system, contact system, blood, biomaterials, biocompatibility, in vitro screening, iC3, dialysis

Over the past several decades, our understanding of the immune system has advanced considerably. With it, an appreciation for its role in a number of diseases, such as cancer and infection has significantly grown. While our increased understanding of the immunological mechanisms underlying these diseases has improved treatment, considerable morbidity and mortality from these illnesses still exists signifying the need for more effective and innovative therapies. Dendritic cell (DC) therapy has been shown to be a promising approach to induce strong immune responses for immunotherapy, and biomaterial-based strategies have been developed to target DCs in vivo to facilitate this purpose. Given the importance of calcium in DC function and activation, we hypothesized that we could develop a biomaterial-based approach to locally and specifically control calcium signaling in DCs in vivo as a novel strategy for immunotherapy. Our first sub-hypothesis was that the calcium used to crosslink alginate gels, a commonly used biomaterial, could activate DCs in vitro; our second sub-hypothesis was that calcium ionophore A23187 could be delivered from biomaterials to activate DCs in vitro; and our third sub-hypothesis was that calcium used to crosslink alginate gels and/or controlled delivery of A23187 could increase local inflammation in vivo. We found that both the calcium released from calcium alginate gels and A23187 matured DCs and enhanced TLR-induced inflammatory cytokine secretion in vitro. Although we were unable to effectively deliver A23187 in vivo, calcium alginate gels injected subcutaneously were able to upregulate a number of inflammatory cytokines and chemokines relative to barium alginate gels. Likewise, when LPS was delivered from calcium alginate gels, the inflammatory effects of LPS on surrounding tissue were enhanced compared to when it was delivered from barium alginate gels. Thus, we confirmed that the calcium crosslinker in alginate gels could activate DCs, and provided a proof-of-principle that calcium signaling could be harnessed in vivo to enhance the immune response. Not only does this work impact the future of biomaterial design, but it may also enhance our understanding of DC biology. This thesis lays the groundwork for a novel and potentially effective strategy for enhancing DC activation in vivo, and suggests that ion signaling pathways in other cell types (both immune and non-immune) could also be targeted using biomaterials. / Engineering and Applied Sciences

The field of immunology has advanced tremendously over the last 40 years, with seminal findings that have guided the development of powerful new therapies. However, the ability to induce safe and long-lasting antigen-specific tolerance has remained elusive. A therapy that could prevent the immune system from aberrantly destroying self-tissues, without impairing its capacity to eliminate dangerous pathogens, would be transformative for the treatment of autoimmune diseases. In addition, such a therapy could also greatly advance the field of organ transplantation by inducing antigen-specific tolerance to prevent graft rejection. / Engineering and Applied Sciences

Healing large bone defects remains a clinical challenge. While autografts are the gold standard treatment for large bone defects, they are limited by availability and donor site pain. Growth factor treatments such as BMP therapy provide a promising alternative but are expensive and present clinical safety concerns, primarily due to delivery of BMPs at supraphysiological doses. Integrins are ECM receptors which mediate crucial cell functions such as adhesion and differentiation. Therefore, understanding the role of integrins in bone formation and directing desired interactions may enable modulation of host cell functions for therapeutic applications. In this work, beta 1 integrins were deleted in osteolineage cells of transgenic mice at three different stages of differentiation to elucidate their role in bone development. We also engineered bioartificial PEG-based matrices which target the pro-osteogenic alpha 2 beta 1 integrin to promote bone healing. Conditional deletion of beta 1 integrins in osteochondroprogenitor cells under the Twist 2 promoter resulted in severe pre-natal skeletal mineralization defects and embryonic lethality. Targeted deletion of beta 1 integrins in osterix-expressing osteoprogenitors resulted in growth abnormalities, reduced calvarial mineralization, impaired femur development, and tooth defects. However, mice lacking beta 1 integrins in osteocalcin-expressing osteoblasts and osteocytes displayed only a mild skeletal phenotype, indicating that beta 1 integrins play an important role in early skeletal development, but are not required for mature osteoblast function. PEG hydrogels functionalized with the integrin-specific GFOGER ligand enhanced bone regeneration, induced defect bridging in combination with low doses of rhBMP-2 and stimulated improved bone healing compared collagen sponges, which are the clinical standard delivery vector for BMP-2 therapy. These results suggest that treatment with bioartificial integrin-specific PEG hydrogels may be a promising clinical strategy for bone regeneration in large bone defects.

Despite the increasing clinical demand in reconstructive, cosmetic and correctional surgery there remains no optimal strategy for the regeneration or replacement of adipose tissue. Previous approaches to adipose tissue engineering have failed to create an adipose tissue depot that maintains implant volume in vivo long-term (>3 months). This is due to inadequate mechanical properties of the biomaterial and insufficient vascularization upon implantation. Modular tissue engineering is a means to produce large volume functional tissues from small sub-mm sized tissues with an intrinsic vascularization. We first explored the potential of a semi-synthetic collagen/poloxamine hydrogel with improved mechanical properties to be used as the module biomaterial. We found this biomaterial to not be suitable for adipose tissue engineering because it did not support embedded adipose-derived stem cell (ASC) viability, differentiation and human microvascular endothelial cell (HMEC) attachment. ASC-embedded collagen gel modules coated with HMEC were then implanted subcutaneously in SCID mice to study its revascularization potential. ASC cotransplantation was shown to drive HMEC vascularization in vivo: HMEC were seen to detach from the surface of the modules to form vessels containing erythrocytes as early as day 3; vessels decreased in number but increased in size over 14 days; and persisted for up to 3 months. Early vascularization promoted fat development. Only in the case of ASC-HMEC cotransplantation was progressive fat accumulation observed in the module implants. Although implant volume was not maintained, likely due rapid collagen degradation, the key result here is that ASC-HMEC cotransplantation in the modular approach was successful in creating vascularized adipose tissue in vivo that persisted for 3 months. The modular system was then studied in vitro to further understand ASC-EC interaction. Coculture with ASC was shown to promote an angiogenic phenotype (e.g. sprouting, migration) from HUVEC on modules. RT-PCR analysis revealed that VEGF, PAI-1 and TNFα was involved in ASC-EC paracrine signalling. In summary, ASC-HMEC cotransplantation in modules was effective in rapidly forming a vascular network that supported fat development. Future work should focus on further elucidating ASC-EC interactions and developing a suitable biomaterial to improve adipose tissue development and volume maintenance of engineered constructs.

Despite the increasing clinical demand in reconstructive, cosmetic and correctional surgery there remains no optimal strategy for the regeneration or replacement of adipose tissue. Previous approaches to adipose tissue engineering have failed to create an adipose tissue depot that maintains implant volume in vivo long-term (>3 months). This is due to inadequate mechanical properties of the biomaterial and insufficient vascularization upon implantation. Modular tissue engineering is a means to produce large volume functional tissues from small sub-mm sized tissues with an intrinsic vascularization. We first explored the potential of a semi-synthetic collagen/poloxamine hydrogel with improved mechanical properties to be used as the module biomaterial. We found this biomaterial to not be suitable for adipose tissue engineering because it did not support embedded adipose-derived stem cell (ASC) viability, differentiation and human microvascular endothelial cell (HMEC) attachment. ASC-embedded collagen gel modules coated with HMEC were then implanted subcutaneously in SCID mice to study its revascularization potential. ASC cotransplantation was shown to drive HMEC vascularization in vivo: HMEC were seen to detach from the surface of the modules to form vessels containing erythrocytes as early as day 3; vessels decreased in number but increased in size over 14 days; and persisted for up to 3 months. Early vascularization promoted fat development. Only in the case of ASC-HMEC cotransplantation was progressive fat accumulation observed in the module implants. Although implant volume was not maintained, likely due rapid collagen degradation, the key result here is that ASC-HMEC cotransplantation in the modular approach was successful in creating vascularized adipose tissue in vivo that persisted for 3 months. The modular system was then studied in vitro to further understand ASC-EC interaction. Coculture with ASC was shown to promote an angiogenic phenotype (e.g. sprouting, migration) from HUVEC on modules. RT-PCR analysis revealed that VEGF, PAI-1 and TNFα was involved in ASC-EC paracrine signalling. In summary, ASC-HMEC cotransplantation in modules was effective in rapidly forming a vascular network that supported fat development. Future work should focus on further elucidating ASC-EC interactions and developing a suitable biomaterial to improve adipose tissue development and volume maintenance of engineered constructs.

Mucoadhesive drug delivery systems which are being used from 1980’s to avoid first pass metabolism of drugs, commercially exist for only systemic drug delivery with fast erosion times (15-60 min), that may not be appropriate for local oral disorders. The goal of this research was to develop and characterize mucoadhesive films with flexibility of carrying different drugs and proteins and provide sustained release for local treatment of oral disorders. Mucoadhesive films composed of polyvinylpyrrolidone (PVP) and carboxymethlycellulose (CMC) were formulated with imiquimod, an immune response modifier. Problems such as solubilization of imiquimod to increase drug loading, uniformity in films and total amount of drug released into supernatants were addressed by use of acetate buffer after investigating multiple methods. Subsequently, other relevant properties of mucoadhesive systems, such as adhesion (shear, pull-off), tensile properties, swelling profiles, transport kinetics, and subsequent changes in release profiles as a function of film composition were characterized. The potential of the system for local retention of imiquimod, determined in oral mucosa of hamsters showed time dependent decrease in imiquimod amount through 12 hours, with no traces of drug in blood. Further testing in humans revealed that the residence time of the mucoadhesive films depended on the application site, increasing in the order of tongue < cheek < gingiva. In parallel, mucoadhesive films loaded with epidermal growth factor (EGF) were developed to promote treatment of oral mucosal wounds. Bioactivity was tested in vitro on buccal tissues by creating a wound followed by application of films. Although EGF-loaded films did not accelerate wound healing, but rather elicited a hyperparakeratotic response. In vitro buccal tissues may not be appropriate for testing the effects of EGF in wound healing without incorporation of other biochemical factors. Overall, a mucoadhesive system capable of delivering bioactive small molecules and proteins in sustained manner was developed in this work. A thorough understanding of the system properties was achieved to further tune for future applications. In vitro studies and in vivo studies in hamsters and humans clearly showed the potential and usefulness of the system to translate in to clinic for treatment of oral precancerous lesions.

Hollow fiber membranes are widely used in blood oxygenators to remove carbon dioxide and add oxygen during cardiopulmonary bypass operations. These devices are now widely used off-label by physicians to perform extracorporeal blood oxygenation for patients with lung failure. Unfortunately, the hollow fiber membranes used in these devices fail prematurely due to blood plasma leakage and gas emboli formation. This project formed ultrathin (~100nm) polymer coatings on polymer hollow fiber membranes. The coatings were intended to “block” existing pores on the exterior surfaces while permitting high gas fluxes. This coating is synthesized using surface imitated control radical polymerization. The coating was durable and did not peel or degrade. Fibers modified using this coating technique did not substantially degrade the mechanical properties of the membrane. This coating technique prevented blood plasma leakage and gas emboli formation. The coating permitted blood oxygenation and carbon dioxide removal from in a mock circulation module. Coating formation on polymeric hollow fiber membranes using surface initiated controlled radical polymerization allows for the formation of membranes that have the potential for long term blood oxygenation. This coating technique would allow these long term blood membranes to be produced more inexpensively than currently existing membranes used for long term use.