This study was approved by the Animal Ethics Committee of the Third Hospital of Hebei Medical University (approval number: Z2019-005-1). Female 4-week-old Sprague-Dawley rats (50–80 g) were provided by the Laboratory Animal Center of Hebei Medical University. Rats (n=120) were randomly divided into two groups: the control (n=60) and PI group (n=60). In the control group, the rats’ left knees did not undergo any surgery. In the PI group, the rats’ left knees underwent surgery to induce PI. Afterwards, left distal femur tissues were collected (n=20 knees/time point in each group) (Figure 1). All the rats were sacrificed by an overdose of anesthesia at 4, 8, and 12 weeks after surgery. According to animal euthanasia guidelines [26, 27], rats were intraperitoneally injected with an overdose of pentobarbital sodium (200 mg/kg). Rats were placed in a controlled environment (60±5% humidity, 12/12 h light/dark cycle, 22±3 °C) and were free to obtain food and water. The animals were accustomed to the laboratory environment with an adaptation period of 1 week before experimentation.
Before surgery, pentobarbital sodium (30 mg/kg, intraperitoneal injection) was used to induce anesthesia in all rats; the left knee was placed in an extended state. After the surgical area was shaved, the surgery was prepared aseptically. A midline skin incision was performed; after the skin and subcutaneous tissue were separated, the capsule was exposed through the medial approach of the knee. Similar methods of surgery-induced PI have been reported in our previous studies [28, 29]. In this study, the medial retinaculum of the knee joint was cut off, and PI could be observed during surgery. The incision was then closed without reconstruction of the medial patellofemoral ligament, retinaculum, or capsules. After surgery, dressing was applied (Figure 2). For post-operative care, animals were provided with 24 h quiet recovery time in a warm and dry environment. They were also provided with food to promote gastrointestinal peristalsis and to prevent stagnation after fully awakening. To control post-operative pain, rats were administered acetaminophen (30 mg/kg, daily) for 5 days.
Micro-computed tomography (micro-CT) assessment
In order to assess PI-induced changes in the subchondral bone microstructure, the bone tissues of distal femurs were carefully harvested and scanned via a micro-CT (SkyScan model 1076, Skyscan, Belgium) at 4, 8, and 12 weeks after surgery (n=5 knees/time point in each group). All scans were performed with a voxel size of 10 μm at 50 kV and 800 μA. The region of interest (ROI) was located transversely below the trochlea with a green rectangle (4 × 3 mm) (Figure 3). Based on previous studies, the following representative parameters were selected to assess changes in subchondral bone : bone volume to total volume fraction (BV/TV fraction), bone mineral density (BMD), trabecular (TB) thickness, TB separation (SP), structure model index (SMI), TB pattern factor (PF), TB number, and degree of anisotropy (DA).
The bone tissues of distal femurs were carefully harvested at each time point (n=5 knees/time point in each group), fixed with 4% polyformaldehyde, decalcified in 10% EDTA until completely demineralized, and embedded in paraffin. Sections (5 μm) were cut along the femoral axis to obtain the transverse images of the trochlea sulcus. Safranin O was used to evaluate cartilage glycosaminoglycans. Cartilage degradation was evaluated by subjecting protein to fast green counter staining and scored using a semi-quantitative histopathological grading system based on the Osteoarthritis Research Society International (OARSI) score [30, 31]. After observation with a microscope, a camera was used to acquire images of the representative sections. Chondrocyte numbers were quantified in the superficial and middle zone and expressed as numbers/mm2. Values were calculated from at least five nonconsecutive sections per region.
The slices were deparaffinized in xylene, rehydrated, and washed three times for 5 min with phosphate buffered saline at room temperature. Endogenous peroxidase activity was blocked by 3% hydrogen peroxide (H2O2) for 10 min. Antigen retrieval was performed by microwave treatment in 10 mm of sodium citrate (pH 6.0) for 10 min. This was followed by overnight incubation at 4 ℃ with anti-collagen X (BoAoSen, BeiJing, China), anti-MMP-13 (Servicebio, WuHan, China), anti-TNF- (ABclonal, WuHan, China), and anti-NFKP65 (Servicebio, Wu Han, China) at a dilution of 1:50. The primary antibody was omitted from the negative control reaction. Afterwards, images were acquired with a 20×100 objective lens from five randomly selected areas from each slice in each group (n=5 knees/time point in each group). The tissue filled the entire field of view, and the background light of each photo was the same. Image-Pro Plus 6.0 software (Media Cybernetics, Rockville, MD, USA) was used for image analysis and choose the same brown color as the uniform standard for judging all positive photos. Data regarding integrated optic density of positive cell and the pixel area of tissue were acquired for all images. The areal density was defined as the integrated optic density divided by the pixel area of tissue and was positively correlated with positive cell expression.
At 4, 8, and 12 weeks after surgery, the cartilage was peeled off from the trochlear with a blade and was analyzed by qRT-PCR for mRNA expression (n=5 knees/time point in each group). Trizol reagent (Servicebio, Wuhan, China) was used to extract RNA from the cartilage. mRNA was translated into cDNA by the RevertAid™ first strand cDNA synthesis kit (Thermo, #K1622) according to manufacturer’s instructions. The target genes were MMP-13, collagen X, TNF-, and NF-κB p65; all genes were analyzed by applying the sequence detection system, and the expression of related mRNA was standardized relative to the GAPDH gene and calculated with the 2-ΔΔCt (cycle threshold) method. The primers used are shown in Table 1. We performed each experiment three times and obtained the average values.
All data values were expressed as mean ± standard deviation. Data were analyzed with the statistical software GraphPad Prism (GraphPad software V7, La Jolla, CA, USA).The Shapiro-Wilk test was used to evaluate normality, the Levene’s test was used to evaluate the homogeneity of variance, and two-way analysis of variance was used to evaluate the parametric data between the two groups at each time point. p < 0.05 was considered statistically significant. According to the results of the preliminary experiments, each group and each time point required at least six rats, with a confidence of 90% and an efficacy of 80% (1−β) .