Articular cartilage is a load-bearing tissue covering the ends of long bones. It has a zonal structure, with variations in cellular and extracellular matrix properties, that helps it to function as a low-friction bearing surface through millions of cycles of loading. Unfortunately, cartilage has only a minimal capacity for healing after damage. Therefore, even small cartilage defects commonly progress to widespread degeneration in a condition called osteoarthritis (OA), which is one of the primary causes of pain and disability in the elderly. OA increases the financial burden on both the individual and society. Many surgical strategies have been used for cartilage defect treatments, but they typically result in mechanically and biologically inferior repair tissue and lack the zonal structure of healthy articular cartilage. Cartilage tissue engineering is an alternative approach for cartilage defect treatment in which cells, biomaterials, and environmental factors are combined to grow new cartilage tissue. While this approach is promising, the lack of optimal cell sources and optimal culture conditions have limited the success of cartilage tissue engineering for clinical application. Articular cartilage is a highly organized tissue with cellular and matrix properties that vary with depth zones. Recapitulation of the zonal organization remains an elusive goal in cartilage tissue engineering. We are interested in developing zonal cartilage tissue engineering approaches to improve the properties of the tissue-engineered cartilage. In this thesis, we aim to evaluate the effects of culture environments that mimic aspects of the native cartilage environment on chondrocyte subpopulations for zonal cartilage tissue engineering.

Chondrocytes are the most commonly used cells for clinical cartilage tissue engineering, but have limitations based on changes in cell phenotype, donor-site morbidity, difficulties accessing the tissue and number of cells that can be isolated. Therefore, alternative cell sources are being investigated. Human umbilical cord perivascular cells (HUCPVCs), which are non-invasively isolated from discarded umbilical cords, have been shown to have multilineage differentiation capacity and therefore may be a useful source for cartilage tissue engineering. In the first study of this thesis, we investigated the potential of HUCPVCs as a novel cell source for cartilage tissue engineering. HUCPVCs maintained their capacity to proliferate in culture at least 56 days. However, the capacity for multilineage differentiation was not demonstrated by HUCPVCs from passage 5 and passage 10 using several protocols for adipogenic, osteogenic and chondrogenic differentiation. Due to the limitations of these cells, further experiments focused on optimizing the culture environment for zonal cartilage tissue engineering using human chondrocytes.

Mimicking the natural cellular environment is one approach to find the optimal culture environment. This approach has been successful, for example, in promoting chondrocyte bioactivity in low oxygen and dynamic compression environments. Applying this approach for zonal cartilage tissue engineering is more complex, as chondrocytes from different zones of cartilage are subjected to different stimuli and also display a zone-specific phenotype. For example, superficial zone chondrocytes, which synthesize less glycosaminoglycans (GAG) and more lubricating molecules such as proteoglycan 4 (PRG4), are subjected to higher shear stress than chondrocytes from deeper zones. The aim of the second study was to evaluate the effect of fluid induced shear stress created by rocker plate bioreactor on chondrocyte subpopulations. Fluid induced shear stresses did not have a significant effect on the chondrogenic gene expression of superficial zone chondrocytes or middle deep zone chondrocytes high density monolayer. The high density monolayer culture system lacks the 3-dimensional extracellular matrix surrounding the chondrocytes, and may not be the optimal environment for in vitro chondrogenesis of chondrocytes.

A wide variety of scaffolds have been used to provide a 3-dimensional environment and structural support in cartilage tissue engineering. However, there are no existing scaffolds that truly mimic the cartilage ECM environment. We hypothesized that zone-specific decellularized cartilage ECM would provide the optimal environment to the sub-population chondrocytes of their zone and promote their activity. Superficial chondrocyte mixed with decellularized ECM of its own zone showed up-regulation of chondrogenesis markers (ACAN, PRG4) compared to the gold-standard chondrocyte pellet control, suggesting the potential use of decellularized ECM to improve in vitro zonal cartilage tissue engineering, especially in the superficial zone.

While the in vitro results with decellularized ECM were promising, cartilage is not an isolated tissue in the joint. Rather, it is in close contact with bone and synovial fluid. To further evaluate the effects of decellularized ECM on zonal chondrocytes in a more realistic environment, constructs were tested in an in vitro full-thickness defect model. Cartilage constructs composed of decellularized ECM in alginate with and without zonal structure were made and cultured in free-swelling and inside a full-thickness defect model. The results showed the increased GAG production and accumulation in the full-thickness defect model compared to free-swelling culture. However, the results cannot rule out the effect of GAG absorbed from degraded bovine cartilage construct.

We continued to evaluate the effect of full-thickness defect model in a subcutaneous SCID mouse model. We hypothesized that the full-thickness defect model provides the optimal cartilage environment to the cartilage construct in the SCID mice. The chondrocytes were embedded in alginate with and without the zonal structure and implanted in the subcutaneous of SCID mice for 42 days. The chondrogenic gene expression and protein production were evaluated from the SCID mice constructs compared to the in vitro control. The results were not significantly different for ACAN and PRG4 gene expression between in vitro and in vivo. However, there was high COL1A1 and low COL2A1 gene expression in the mouse model. The results suggest that the full-thickness defect model may be used to provide the cartilage environment in SCID mice. However, modification is needed.

In the previous studies, we evaluated the effect of the healthy joint environment from bovine cartilage ECM or bovine explants on the chondrocytes biosynthesis of cartilage construct to improve cartilage tissue engineering culture conditionsin vitro and in an animal model. The results showed the positive effect of the healthy cartilage or cartilage and bone environment on the chondrocytes biosynthesis in cartilage construct. However, the cartilage tissue constructs need to survive in the OA joint environment for clinical application in OA. It is important to know how the OA environment affects the chondrocytes in tissue engineered constructs and how cyclic compression affects the chondrocyte biosynthesis in the OA condition. The final experiment aimed to evaluate the effect of soluble factors from sclerotic OA osteoblasts and normal osteoblasts on the chondrocytes biosynthesis in cartilage constructs and the effects of cyclic compressive stimuli, which have been proven to stimulate OA chondrocyte biosynthesis, on the chondrocytes after co-cultures in both conditions. Our data suggest that there is a clear zonal difference in how chondrocytes respond to OA osteoblast-conditioned medium and compressive stimulation. OA osteoblast condition media decreased the regulatory cytokine (IL-6), PRG4 gene expression and IL-6 secreted to the medium only in the S chondrocytes. Compression led to an increasing PRG4 and IL-6 gene expression in the MD chondrocytes and reduced the MMP13 and IL-6 gene expression in the S chondrocytes. The results highlight the importance of mechanical stimulation, and their role in decreasing the gene expression levels of catabolic enzymes such as MMP-13, and how soluble factors from the OA osteoblasts can mediate the decrease in gene expression levels of lubricating molecules such as PRG4, which is necessary in preventing mechanically-induced surface fibrillation.