Establishment of rat ankle post-traumatic osteoarthritis model induced by malleolus fracture

Osteoarthritis (OA) is the most common cause of disability in the elderly [1]. Disability stems from pain and limitations in mobility secondary to the degeneration of articular cartilage, a trademark of the disease. Unfortunately, current pharmacological therapy targeting the mechanism of OA is relatively ineffective, largely because the etiology and pathogenesis of OA remain poorly understood. The complex pathobiological changes that occur in human OA may be influenced by a multitude of genetic and environmental factors. The effort to clarify the molecular events that occur in OA during the onset and the progression of OA has necessitated the use of in vivo models [2]. Researchers tend to utilize knee OA models to investigate these factors, but have neglected the establishment of other types of OA models, such as ankle OA.

A recent study indicates that the biomarker and mechanism of ankle OA may not be the same as those of knee OA [3‐5]. Researchers have reported that aggrecan (Acan), bone morphogenetic protein (BMP)-2, BMP-7, and fibronectin-aggrecan complex (FAC) can be used as key markers of OA in the ankle, but not in the knee [6, 7]. In the knee and hip, primary OA accounts for 67% and 58% of all cases, respectively. Meanwhile, 78% of all cases of ankle OA are post-traumatic (PTOA) [8, 9]. In addition, while the incidence of knee OA in the adult population rises from 6% to 10% after 65 years of age, the incidence of ankle OA remains unchanged with age [10]. Malleolar fractures are the most frequent type of fracture in the ankle, presenting in 37–53% of patients with advanced or end-stage ankle OA [11, 12]. More than 50% of patients with fractures of the distal tibial articular surface develop OA [13]. After intra-articular fracture, the ankle joint sustains increased contact stress; in addition, the inflammatory response is a contributory factor to the progress of OA [14]. Chondrocyte necrosis and apoptosis are observed following trauma in human and porcine knees, and associated with cartilage damage and degeneration [15]. An advantage of PTOA models is that there is temporal control of disease induction (when compared with spontaneous animal OA and with human disease), while mimicking the molecular pathology and histopathology of human disease [2]. Despite the high incidence of ankle trauma and OA, ankle-specific OA research is sparse, with the majority of clinical and basic research pertaining to the knee and hip joints [16]. This will greatly limit the study of ankle OA. Thus, there is a need to develop novel ankle PTOA models to facilitate research of this type of OA. Clinically, some patients with malleolar fracture present only, while others present with fracture and dislocation. Therefore, we developed two rat ankle PTOA models in this study: 1) the malleolus fracture with dislocation and reduction; and 2) the malleolus fracture alone. The contralateral ankles were used as controls. To validate the success of our models, X-ray and Safranin-O were used to observe morphological changes in the subchondral bone, joint space and cartilage. FMT, ELISA and immunohistochemistry were used to detect protein levels of several OA-related biomarkers, and qPCR was used to obtain the mRNA levels of several OA related genes.

Approximately 1% of the world’s adult population is affected by joint pain and disability resulting from ankle OA [8]. Although knee OA has been surveyed thoroughly, the diagnosis and treatment of ankle OA may be different from those of knee OA due to the difference of metabolism, articular surface thickness and biomechanical properties [16, 22]. In healthy cadaver joints, ankle chondrocytes had increased proteoglycan (PG) and collagen synthetic rates when compared with knee chondrocytes [23]. When fibrillations and fissuring occur on the cartilage surface during early OA, markers of collagen synthesis and aggrecan turnover are increased in the ankle, but down-regulated in the knee; while markers of collagen degradation are higher in the knee than that in the ankle [24]. Ankle chondrocytes are more resistant to the effects of interleukin-1 or fibronectin fragments than those of knee cartilage, and are able to reverse their effects under the influence of BMP-7 [25]. Furthermore, ankle cartilage is significantly thinner (1–1.45 mm) than knee cartilage (3–6 mm). Joints with higher congruency appear to have thinner cartilage and have a lower incidence of osteoarthritis than noncongruent joints such as the knee [26]. In addition, ankle cartilage has a higher dynamic stiffness and compressive modulus than knee cartilage in compression, that is, ankle cartilage is more resistant to compressive loads [26].

Several OA models are available for knee OA study such as anterior cruciate ligament transection (ACLT) and destabilization of the medial meniscus (DMM), but only two ankle PTOA models have been established recently in mouse (by medial and lateral ligament resection) and mini-pig (by fracture) [27‐29]. The ankle joint of mouse ligament resection model is too small to collect SF and cartilage for biomarker studies using ELISA and gene arrays using qPCR. The mini-pig model is expensive for most research groups and is not suitable for drug screen. However, the rat models have the advantages of being low cost and relative large size of joints to collect SF lavage and cartilage, as well as genetically similar within a specific breed strain, and amenable to genetic manipulation [16]. Furthermore, the anatomical and histological features of human and rodent ankle joints are comparable [28]. Therefore, it is necessary to create an innovative ankle fracture OA model for ankle OA research in the field.

Valderrabano and his colleagues showed that ankle PTOA was seen in 78% of 406 ankle OA cases; among these, malleolar ankle fractures accounted for 39% [8]. Moreover, ankle PTOA induced by fracture with dislocation is also common in clinic. Therefore, it is necessary to compare whether there is a difference between the medial malleolus fracture with dislocation injury and the medial malleolus fracture alone. Our results indicated that both the malleolar fracture models with and without dislocation/reduction were healed at 8 weeks post fracture. Based on our pilot study, we found that the 37 degree of the angle fixator created a stable fracture model located at medial 1/3 tibiotalar joint. Our results indicated that these fracture models were stable and no fixation was required. All animal fractures were completely healed without ankle joint deformity at 8 weeks after the fracture. Histology data determined by Safranin-O staining demonstrated that OA changes were similar to changes in human ankle and knee OA, including cartilage fibrillation, rough articular surface, decrease of Acan, matrix and chondrocyte numbers [17, 24, 28, 30, 31]. Increased MMP-13 and type X collagen as well as decreased type II and Acan were further detected by FMT, immunohistochemistry, qPCR and ELISA respectively. Noticeably, the fracture of medial malleolus resulted in mild to moderate OA cartilage lesions. The lesions were primarily located on the central weight-bearing region of the tibiotalar joint with a rare subchondral sclerosis. Compared with our model, the loss of cartilage and subchondral sclerosis are severe and common in the mouse ligament transection ankle OA model [28]. Our model resembles the slowly-progressive human ankle OA and should allow for evaluation of target drugs studies.

There are a few potential limitations to our study. Firstly, we used the contralateral ankle joints as controls instead of sham injured joints and unoperated joints control. The surgically-induced gait alteration may occur in the contralateral side as a result of altered loading. However, our surgical results are significantly more evident than the contralateral limbs. Nevertheless, future studies should add additional sham control groups as an ideal control. Secondly, the growth plates close at skeletal maturity and longitudinal growth ceases in adult human while the rats maintain a growth plate into old age [32]. Although this provides the potential for continued longitudinal growth, in reality bone growth ceases after a certain time (the rate of growth increases before first 5 weeks, then declines at 11.5–13 weeks, ceases until 26 weeks) [32]. Despite of all this, the difference between rodents and human should not affect our results.

The results of this study suggest that the two models can successfully induce OA, but the differences between them are not significant. Compared with the mouse ankle OA models created by transecting several ankle ligaments, our rat fracture OA model is more relevant to human ankle OA and allows for the collection of enough cartilage tissue and SF lavage for gene and inflammation biomarker analyses. Furthermore, compared with the mini-pig model, our model will be beneficial for rapid screening of targets drug with low cost.

We have successfully established two rat ankle PTOA models induced by malleolus fracture. Although the OA changes observed in the two ankle OA models are similar to the changes that occur in human ankle OA cartilage, we recommend the fracture alone model as it is simpler and there is no significant difference between the two models. Thus, our ankle PTOA models will accelerate ankle OA research in the future, especially for ankle OA induced by the fracture injury.