Evaluating pharmacological inhibition of PPARdelta in a rat model of post-traumatic osteoarthritis

Purpose: Osteoarthritis (OA) is a painful chronic disease affecting all joint tissues. Previous studies in our lab have demonstrated that nuclear receptors are involved in regulation of cartilage homeostasis. In particular, activation of nuclear receptor PPARdelta in-vitro results in the upregulation of genes coding for fatty-acids and cartilage-matrix degrading proteases. We have also shown that genetic inactivation of PPARdelta specifically in cartilage can be protective in the DMM model of post-traumatic osteoarthritis in mice (Ratneswaran et al, Arth Rheum 2015), thus illustrating it could be a potential therapeutic target. In order to establish the viability of PPARdelta inhibition as a pharmacologic therapy for OA, we conducted a pilot study (N=5) using two pharmacological PPARdelta inhibitors in a rat model of post-traumatic OA. Our pilot study indicated that 4 weeks post-surgery rats treated PPARdelta inhibitors did not experience the same pain-like behaviour (spontaneous activity, weight-bearing) as vehicle-control treated rats. Structural protection to the cartilage and underlying bone had trended towards protection but was not statistically significant. Our current study is an extension of the original pilot, examining pharmacological PPARdelta inhibition in rat post-traumatic OA. We hypothesize that inhibition of PPARdelta will delay the progression of post-traumatic OA in animal models, and that rats treated with PPARdelta inhibitors will present with less pain like behaviour. Methods: In our first cohort, 300-350g male Sprague-Dawley rats (n=5 /group) were subjected to sham control surgery or an anterior cruciate ligament transection (ACLT)/partial medial meniscectomy (PMMx) surgery and were systemically treated with two different inhibitors of PPARdelta (GSK0660, 3787) or vehicle control DMSO for six days per week, starting one day post-surgery, for four weeks. Weekly testing for spontaneous exploratory behavior (Open Field Testing) and load-bearing asymmetry (Incapacitance Testing) was conducted to measure pain associated behaviours. End point blood glucose and liver weight were measured. Rats were compared through measures of OA progression including Safranin-O staining with OARSI scoring, immunohistochemistry for cartilage matrix breakdown products, and picrosirius red staining for collagen structure. Our second cohort of rats (n=6/group), 300-350g male Sprague-Dawley rats were subjected to sham control surgery or ACLT/PMMx surgery and systemically treated with GSK0660, GSK3787, vehicle control DMSO, or saline control for six days per week, starting one day post-surgery, for two or four weeks (early, and moderate OA). End-point weight, blood glucose and liver weight were measured for all animals. Gait and MicroCT analysis were performed in animals 4 weeks post surgery, while classical OA histopathology was scored for subchondral bone and cartilage 2, and 4 weeks post-surgery. Results: Rats treated with either PPARdelta inhibitor were not different in physiological profiles (blood glucose, liver weight, body weight) from vehicle treated or SHAM operated animals in either cohort, at 2 or 4 weeks post-surgery. In our first cohort, vehicle treated rats demonstrate changes in behaviour after ACLT/PMMx surgery, such as decreased vertical activity and increased rest time. In contrast, ACLT/PMMx animals treated with PPARdelta inhibitors or SHAM operated animals do not demonstrate these changes in behaviour post-surgery. However, preliminary OARSI scoring indicate minimal differences in subchondral bone and cartilage damage between inhibitor treated and vehicle treated rats post-surgery. In our second cohort, 2 weeks post-surgery our preliminary results indicate very early OA development in all groups except sham animals in the medial and lateral compartment cartilage, with changes such as cell death and proteoglycan loss. Subchondral bone changes, such as increased basophilia, and marrow changes are evident in these groups as well. 4 weeks post-surgery, DMSO-vehicle controls, saline-treated controls, and GSK0660 treated animals present with greater cartilage damage to the medial tibial plateau compared to sham operated animals, while GSK3787 treated animals do not. Similarly, DMSO-vehicle treated controls, saline-treated controls and GSK0660 treated animals present with greater damage to the medial femoral condyle compared to sham treated animals, unlike GSK3787 treated animals. Conclusions: This study suggests that PPARdelta inhibition may have a role in chondroprotection at intermediate stages of OA progression. Similarly, PPARdelta may have a role in modulating pain-like behaviour in OA. However, additional studies are required, in particular with regard to the differential effects of the two PPARdelta inhibitors at the 4-week time point.