Osteoarthritis (OA) is the leading cause of disability among the elderly population, affecting approximately 27 millions Americans and costing $60 billion in related-health care costs. Mouse models of OA have been developed to study the mechanisms of OA and therapeutic interventions. However, traditional animal models induce OA pathology through traumatic surgeries, which only represent 10% of human OA patients. Thus, in this thesis, a novel noninvasive OA mouse model was developed, characterized, and applied to transgenic mice. The changes in articular cartilage and subchondral bone were analyzed by histology, immunohistochemistry, and microcomputed tomography.

To develop a noninvasive OA mouse model, an in vivo tibial loading model was used to investigate the adaptive responses of cartilage and bone to mechanical loading and to assess the influence of load level and duration. Peak cyclic compression of 4.5 and 9.0N was applied to the left tibia via the knee joint of adult (26-week-old) male mice for 1, 2, and 6 weeks at 1200 cycles/day. In addition, 9.0N loading was utilized in young (10-week-old) mice. Loading promoted cartilage damage, cartilage thinning, and subchondral cortical bone thickening in both age groups. Both age groups developed periarticular osteophytes at the tibial plateau in response to the 9.0N load, but no osteophyte formation occurred in adult mice subjected to 4.5N load.

Development of a novel noninvasive loading model was followed by investigating the traumatic vs. nontraumatic nature of cyclic loading of the mouse knee joint. To differentiate traumatic tissue damage versus cell-mediated processes in the development of OA pathology, a single nondestructive 5-minute loading session was applied to the left tibia of adult (26-weekold) mice at a peak load of 9.0N. Knee joints were subsequently analyzed at 0, 1 and 2 weeks after loading. At T = 0, no change was evident in cartilage or subchondral bone. However, cartilage pathology demonstrated by localized thinning, proteoglycan loss, and inhibition of chondrocyte autophagy occurred at 1 and 2 weeks after the single session of loading. Transient cancellous bone loss was evident at 1 week, associated with increased osteoclast number, reversed at 2 weeks.

Finally, the in vivo tibial loading model was implemented to study the role of Dickkopf1, an inhibitor of the Wnt pathway, in the development of load-induced OA. To identify the role of Dickkopf-1 protein in OA, novel viable mice with Dickkopf-1 knockout and Wnt3 knockdown (Dkk1-/-;​Wnt3+/-) were used. The left tibia of 10-week-old Dkk1-/-;​Wnt3+/- and respective control groups, littermate control (Dkk1+/+;​Wnt3+/+) and Wnt3 knockdown (Dkk1+/+;​Wnt3+/-) mice, underwent cyclic compression at a peak load of 9.0N for 2 weeks. As a result of loading, both Dkk1-/-;​Wnt3+/- and Dkk1+/+;​Wnt3+/+ mice demonstrated cartilage erosion, subchondral cancellous bone loss, and osteophyte formation. However, Dkk1+/+;​Wnt3+/- mice did not undergo cartilage degeneration and showed limited osteophyte formation, indicating knockdown of Wnt3’s potential chondroprotection against an altered joint loading environment.

In summary, an altered joint loading environment caused by in vivo tibial loading repeatedly and robustly produced OA pathology in mouse joints. This loading modality was nontraumatic as evidenced by the absence of physical damage and presence of biological events that led to OA. In addition, the in vivo tibial loading model can be applied to investigate potential chondroprotection from genetic or pharmacological interventions. The novel in vivo tibial loading model presents tremendous opportunities to study the etiology of OA from patients without a history of traumatic joint injury and will be an excellent platform to develop therapeutic interventions.