Non-invasive murine joint loading model: a novel model for osteoarthritic pain

Purpose: Osteoarthritis (OA) is a common degenerative joint disease associated with chronic and debilitating pain in the affected joints which significantly reduces mobility and quality of life in patients. The mechanisms for osteoarthritic pain are poorly understood and current therapies to reduce pain are often insufficient. The lack of murine models of osteoarthritic pain, that show both histological OA cartilage lesions and a robust reproducible pain phenotype, has deterred the understanding of the disease. In this study we used the recently established, adjustable and non-invasive murine joint loading model of OA to characterize the development of a pain phenotype. Methods: The right knee of male mice (12 week-old, C57BL/6) were loaded at 2N, 9N and 11N three times per week for two weeks under anaesthesia. Behavioural measurements were taken at baseline before loading, and monitored on a weekly basis for six weeks post loading. Mechanical hyperalgesia was measured using manual von Frey in both ipsilateral and contralateral hind paws; thermal hyperalgesia was measured using the hot and cold plates whilst the rotarod and weight bearing tests were used to measure disuse of affected limb. Development of pain was compared to control mice which did not undergo a loading regime but were subjected to isoflurane anaesthesia for the same duration as loaded mice. Severity of OA lesions in the articular cartilage was determined post-mortem on joint sections using the grading scheme from the OARSI histopathology initiative. Results: Following two weeks of loading 9N- and 11N-loaded mice developed a pain phenotype which was more severe than that of the 2N-loaded mice. Both the 11N- and 9N-loaded mice developed ipsilateral mechanical hypersensitivity two weeks post loading whilst contralateral hypersensitivity developed at a later stage. The mechanical hypersensitivity was accompanied by reduced motor activity and altered weight bearing. The non-loaded control group did not show changes over time in any of the pain measurements. There was no thermal hypersensitivity phenotype in any of the mice. Histological analysis revealed that loading at 11N induces OA-like lesions of significantly higher severity than in 9N- and 2N-loaded mice. Mice loaded at 2N showed signs of low OA in the ipsilateral knee which suggests that the static load of 2N is sufficient to induce mild damage in the knee. The contralateral knee of 2N-loaded mice showed no lesions. This correlates with the weak ipsilateral mechanical hypersensitivity but no contralateral pain seen in the 2N-loaded mice. The contralateral knee of both 9N and 11N-loaded mice showed mild OA lesions suggesting that in these mice the ipsilateral knee is damaged to such an extent that they compensate with the contralateral leg. This is in agreement with the contralateral development of mechanical hypersensitivity seen in these mice. Conclusions: We have characterized for the first time the pain phenotype in a novel non-invasive murine joint loading model. Our results show that this model of OA produces a robust and reproducible pain phenotype as well as showing OA lesions, both of which vary in severity depending on the load magnitude. This establishes the use of the non-invasive murine joint loading model as an appropriate model to measure osteoarthritic pain.