Osteoarthritis (OA) is a complex disease that affects many people [9, 10], and leads to severe disability. The main risk factors for OA include age, obesity, and joint injury, and OA is more common in women [11]. The major symptom of OA is pain. Other nervous system changes have been documented including poor proprioception and chronic anxiety [12-14]. No efficient treatment or drugs are available, largely because the pathological mechanisms of OA joint damage and the associated pain are not fully understood. Therefore, it is urgent to elucidate the pathomechanisms of OA.

Few pain measurements assays have been established for experimental OA models. One goal of this dissertation was to establish an assay to measure knee hyperalgesia in a surgical mouse model, where experimental OA is induced by destabilization of the medial meniscus (DMM). We did this using the Pressure Application Measurement (PAM) device. This study showed how knee hyperalgesia developed and recovered using wild type (WT) mice after DMM and sham surgery. The results showed that DMM mice had more severe knee hyperalgesia than sham mice and that in the sham group hyperalgesia recovered faster than in the DMM group. To confirm that the hyperalgesia is pain-related, this study injected morphine (systemic analgesia) and lidocaine (local analgesia) to WT mice after surgery and confirmed that this reversed hyperalgesia, indicating it is a pain-related behavior. This study also used chemogenetic mice which artificially turns off activation of nociceptors. The result showed that transient chemogenetic blocking of nociceptors temporarily reversed knee hyperalgesia in WT mice, 4 weeks after DMM surgery.

In the next part of my work, we aimed to develop an assay to assess proprioception in mice. According to clinical studies, OA patients tend to develop poor proprioception, which causes a higher risk of falls and also limits activity [15]. Furthermore, a recent study showed that accuracy of proprioception function helps to determine onset and radiographic knee OA [16, 17]. To assess proprioception in mice, we used two tests, the beam test and the ladder test, which are one of the known methods used to assess balance in rodents [18, 19]. To determine whether the beam or ladder test is appropriate to measure proprioception function, this study used chemogenetics to inhibit parvalbumin (PV)-expressing neurons, which are proprioceptors. We determined whether the beam or the ladder test can assess proprioception function using PV/Pdi 10-week-old mice. The result shows that blocking activity in PV-neurons using CNO in PV/Pdi mice significantly altered (1) tail use duration;​ (2) frequency of hind paw slip;​ and (3) time to travel. The beam and ladder test were also conducted in WT mice after DMM or sham surgery. However, the parameters could not find any difference in proprioception ability between naïve, sham, and DMM groups up to 16 weeks after surgery. From these results, the DMM model might not be suitable for proprioception study. This result led us to pursue using the partial meniscectomy (PMX) model, a more severe OA model. Here, we found that the frequency of hind paw slips differentiated between sham and PMX, 12 weeks after surgery. These results show that the PMX model using the ladder test is suitable for proprioception assessment and that proprioception deficits occur at later time points.

The third focus for my thesis is anxiety development caused by OA pain, because chronic pain is commonly associated with development of chronic anxiety, as has been described in several diseases such as diabetes, coronary heart disease, and OA [14, 20-22]. However, anxiety in OA has not been well studied. Therefore, we used the DMM model to develop a method to assess anxiety-like behavior by the elevated plus maze (EPM) test. According to the results, mice showed increased duration of grooming 12-16 weeks after DMM surgery, comparing to age-matched WT naïve mice. This result suggests that this animal model develops anxiety-like behavior and increasing severity of anxiety gradually up to 16 weeks after surgery.

Finally, I examined how CCL2-CCR2 signaling may contribute to the pain-related behaviors described above. Ccr2 null mice were protected from knee hyperalgesia up to 16 weeks after DMM. To understand how CCR2 affects knee hyperalgesia systemically, we used systemic injection of CCR2RA, and found it reversed knee hyperalgesia 4 or 8 weeks after surgery. Local injection of CCR2RA into the OA knee also reversed knee hyperalgesia 7 weeks after surgery. To assess the direct contribution of neuronal CCR2 to development of knee hyperalgesia, we assessed the effect of intra-articular (i.a.) injection of CCL2 into the knee joint space. The result showed that this resulted in transient knee hyperalgesia WT but not in Ccr2 null mice. These results suggest that blocking CCR2 responses at peripheral terminals is sufficient to produce analgesic effects in OA. Recent studies have shown that CCR2 affects anxiety development in anxiety and neuropathic models [7, 23]. The results from this study showed that Ccr2 null mice did not develop anxiety behavior up to 16 weeks after DMM surgery, unlike WT mice. This study also showed that anxiety behavior was reversed by systemic injection of CCR2RA 16 weeks after DMM.

Overall, these studies indicate that primary knee hyperalgesia assessment by PAM device is a useful method in experimental OA induced by DMM, and that CCL2-CCR2 signaling locally in the joint contributes to knee hyperalgesia in experimental OA. Additionally, we developed a proprioception assay using a beam and a ladder test. We could not detect proprioceptive deficits after DMM surgery, but we did find proprioceptive deficits in the PMX model. Finally, we developed the EPM as a method to detect anxiety associated with OA in the DMM model, and found that CCL2-CCR2 signaling affects anxiety development in this model.