Standardized Instability-induced Animal Models for Preclinical Studies of Osteoarthritis

Osteoarthritis (OA) is a highly prevalent degenerative joint disorder and one of the leading causes of disability worldwide^1,2. Despite extensive research, the development of effective disease-modifying OA drugs (DMOADs) has been relatively slow³. It is now widely acknowledged that OA is a multifactorial disease involving complex pathological changes across multiple peri-articular tissues, including articular cartilage, subchondral bone, tendons, and synovium^4,5. Effective preclinical research on the diverse pathological processes of OA requires the use of well-defined and appropriate animal models. Although a variety of OA animal models have been established and reported^{6,7,8,9,10,11}, many studies continue to employ unsuitable animal models, limiting the translational relevance of their findings. To address this issue, we present standardized instability-induced OA models, including reproducible Destabilization of the Medial Meniscus (DMM) and Medial Meniscus Tear (MMT) procedures, in a video format to serve as a practical reference for preclinical OA research and model standardization.

As reported in previous studies, meniscal instability is a major factor contributing to knee joint degeneration in clinical cases. Therefore, surgically induced meniscal instability in mice and rats serves as an effective approach for the experimental induction of OA. In this context, we present two representative instability-induced OA models, including the DMM model in mice and the MMT model in rats. The DMM model in mice is particularly valuable for investigating molecular mechanisms of OA, as mice are amenable to transgenic manipulation, and this model exhibits a gradual and prolonged disease progression, closely mimicking human OA. However, it is important to note that the DMM model is less suitable for studies focusing on pathological alterations in the subchondral bone, where alternative models may provide better translational relevance.

Furthermore, we established the MMT model in rats to complement the mouse DMM model. The MMT procedure induces more extensive structural damage involving both the articular cartilage and subchondral bone, thereby better reflecting the advanced pathological features of OA observed in clinical settings. This makes the rat MMT model particularly suitable for bioengineering and translational research, including the evaluation of biomaterials, regenerative therapies, and drug delivery systems targeting joint repair. Although the MMT and DMM surgical techniques have been described in previous studies, there remains a lack of a comprehensive, step-by-step visual protocol that ensures standardization, reproducibility, and technical precision across research laboratories. Hence, to address this gap, we have developed a detailed video-based demonstration illustrating each critical step of model establishment, from animal preparation to postoperative care. In addition, we introduce two commonly used behavioral evaluation methods, one for assessing pain threshold and another for analyzing spatial gait stability to provide a consistent framework for functional assessment following OA induction. Thus, by integrating surgical and behavioral protocols, this manuscript aims to promote methodological standardization in preclinical OA research, thereby improving data comparability and translational relevance across studies.

All experimental procedures were conducted in accordance with the guidelines of the Animal Ethical and Welfare Committee of Zhejiang Chinese Medical University and were approved under the protocol number IACUC-20251020-11. For the DMM surgery, male C57BL/6 mice aged 10-12 weeks were used, while male Sprague-Dawley rats weighing 200-225 g were selected for the MMT procedure. All animals were housed under specific pathogen-free (SPF) conditions and allowed a one-week acclimatization period prior to experimentation. Standard husbandry conditions were maintained, with food, water, and bedding replaced at least twice weekly to ensure animal welfare and experimental consistency.

1. Animal preparation

1. Administer appropriate analgesia (e.g., buprenorphine) before surgery according to institutional guidelines. Anesthetize the animal using an inhalational anesthetic (e.g., 2% isoflurane at 0.4 L/min with a 4 L/min fresh gas flow in an induction chamber) and maintain anesthesia throughout the procedure via a nose cone. Confirm the depth of anesthesia by ensuring the absence of a pedal withdrawal reflex.
2. Apply a 2-3 mm strip of lubricating ophthalmic ointment into the conjunctival sac of the lower eyelid and gently massage the eyelid to ensure even distribution of the ointment over the ocular surface.
3. Shave the fur surrounding the right knee joint and disinfect the surgical site using a three-step cleansing procedure with povidone-iodine followed by 70% ethanol. Position the animal on a sterile operating table.

2. DMM surgery in mice

NOTE: This procedure induces joint destabilization by transecting the medial meniscotibial ligament (MMTL), highlighted in red in Figure 1, which normally anchors the medial meniscus to the tibial plateau.

1. Position the anesthetized mouse in a supine orientation and make a small medial parapatellar skin incision (approximately 3-5 mm) over the right knee.
2. Carefully dissect the overlying soft tissues to expose the joint capsule. Make an incision medial to the patellar tendon and gently displace the patella laterally to visualize the knee joint. Confirm that the MMTL appears as a thin, ligamentous band connecting the anterior horn of the medial meniscus to the tibial plateau. Use a stereomicroscope to enhance visualization and minimize the risk of damaging the articular cartilage.
3. Carefully transect the tibial attachment of the MMTL using micro-scissors, a #11 scalpel blade, or a bent syringe needle. When using a bent syringe needle, the blunt edge helps protect the underlying articular cartilage. Confirm successful transection by observing the resulting destabilization of the medial meniscus.
4. Close the incised joint capsule using 8-0 absorbable sutures, followed by closure of the skin incision with wound clips or 6-0 sutures.
   NOTE: Take care to avoid piercing or damaging the articular cartilage during suturing and ensure that joint tightness is properly maintained while closing the incision.
5. Allow the animal to recover on a heating pad and monitor it closely until full mobility is regained.

3. MMT surgery in the rat

NOTE: This procedure induces OA by creating a complete tear of the medial meniscus following transection of the medial collateral ligament (MCL), thereby generating joint instability and progressive cartilage degeneration.

1. Position the anesthetized rat appropriately and make a medial parapatellar incision, approximately 6-8 mm in length, over the right knee.
2. Perform blunt dissection to separate the muscle layers and clearly expose the MCL (Figure 2).
3. Using surgical scissors, fully transect the MCL at its midpoint.
4. Gently retract the transected MCL to expose the underlying medial meniscus. Using fine forceps, grasp the meniscus and perform a full-thickness transection at its narrowest region to create a complete tear.
5. Reposition the soft tissues and close the incision in layers, using 5-0 absorbable sutures for the muscle and fascia, followed by wound clips or 5-0 non-absorbable sutures for skin closure.
6. Allow the animal to recover by following the same procedure described in step 2.5.

4. Behavioral assessments

1. Conduct baseline assessments prior to surgery and repeat evaluations at designated postoperative time points (e.g., 2, 4, and 8 weeks).
   NOTE: To minimize interoperator variability, behavioral tests were performed by the same trained operator who was blinded to the experimental grouping. Each measurement was repeated three times per animal, and the average value was used for analysis.
   1. Simplified mechanical withdrawal threshold test
      1. Place the mice in individual compartments on a raised wire mesh platform and allow them to acclimate for at least 30 minutes before testing. Mechanical pain sensitivity is then assessed using an electronic von Frey anesthesiometer.
      2. Using a Von Frey filament, apply the probe to the plantar surface of the mouse's right hind paw and gradually increase the force¹².
      3. Record the response, noting that a positive reaction is indicated by sudden paw withdrawal, flinching, or licking. Assess the presence or absence of a consistent response to evaluate changes in sensitivity over time.
   2. Gait analysis
      1. Utilize an automated gait analysis system for assessment.
      2. Allow the animal to voluntarily walk across the glass walkway of the apparatus. Define a valid trial as one in which the animal moves at a steady pace without pausing. Wait for the system software to automatically identify paw prints and compute multiple gait parameters. Analyze key metrics for the right (operated) hindlimb, including stance phase duration, swing speed, and paw pressure, and compare them with those of the contralateral limb and sham-operated controls.
         ​NOTE: All animals were acclimated to the testing apparatus with daily training sessions prior to baseline assessment. For each animal, three runs were recorded per time point, and only trials in which the animal maintained a consistent speed without pausing were considered valid. All measurements were performed by trained operators under a double-blind design to minimize inter-operator variability, and repeated runs were averaged to ensure data reproducibility.

Characterization and validation of the DMM OA mouse model
After a 1 week acclimatization period, the DMM-induced osteoarthritis (OA) model in mice was established following the procedure illustrated in Figure 1. Pathological and behavioral assessments were performed two weeks post-surgery. As shown in Figure 2A, the most pronounced degenerative changes were localized to the medial femorotibial joint. The extent of cartilage degeneration was further quantified using the OARSI scoring system13,14, which demonstrated significantly higher scores in the DMM model group compared to the control group (Figure 2B). All OARSI scoring was performed by a pathologist blinded to experimental groups. Coronal sections of the medial compartment were collected for each animal, and every third section was analyzed. The section with the highest score was used as the overall score for the knee joint. The OARSI scores followed a normal distribution, as verified by the Shapiro-Wilk test. Independent samples t-test analysis confirmed a statistically significant difference between the two groups (control: 2.33 ± 2.08 vs. DMM: 15.33 ± 2.08; n = 3, p. = 0.0016). Histological examination revealed no evidence of cartilage regeneration, patellar dislocation, free cells within the synovial cavity, or ectopic bone formation in any of the samples.

To further confirm the successful establishment of the model, a series of behavioral assessments was conducted. During forward locomotion, control mice exhibited normal, well-aligned fore- and hind-paw prints, whereas mice in the DMM group showed irregular and unstable gait patterns characterized by altered stride length (Figure 2C,D). Quantitative analysis demonstrated that the stride length parameter was significantly greater in the DMM group (0.097 ± 0.021) compared to the control group (0.047 ± 0.015) (mean ± SD, n = 3; p = 0.0285). Consistently, the Von Frey test indicated a marked reduction in pain threshold among DMM mice, reflecting enhanced mechanical hypersensitivity (Figure 2E). The mean withdrawal threshold was significantly higher in the control group (0.650 ± 0.105) relative to the DMM group (0.400 ± 0.089) (mean ± SD, n = 6; p. = 0.0012). Collectively, the gait abnormalities and heightened pain sensitivity confirm the successful induction of OA in the DMM model, supporting its robustness and translational relevance for OA research.

Establishment and validation of the MMT OA rat model
Following the procedure illustrated in Figure 3, the MMT-induced OA model was established in rats, and subsequent pathological and behavioral evaluations were conducted two weeks post-surgery. As shown in Figure 4A, distinct structural disruption of the meniscus was evident, accompanied by cartilage erosion and characteristic osteoarthritic alterations. The extent of cartilage degeneration was further quantified using the OARSI scoring system, which demonstrated markedly higher scores in the MMT group compared to the control group (Figure 4B), signifying advanced joint damage and disease progression. Quantitative analysis confirmed a significant elevation in OARSI scores in the MMT group (16.33 ± 1.53) relative to controls (2.00 ± 1.00) (mean ± SD, n = 3; p. = 0.0002).

Furthermore, behavioral assessments revealed that rats in the MMT group displayed irregular and unstable gait patterns during forward locomotion, characterized by a pronounced alteration in stride length (Figure 4C,D). Quantitative analysis showed a significant increase in stride length variation in the MMT group compared with controls (MMT: 0.093 ± 0.015 vs. Control: 0.037 ± 0.015; mean ± SD, n = 3; p = 0.0105). Consistently, the Von Frey test demonstrated a significant reduction in pain threshold in the MMT group, indicative of heightened mechanical sensitivity (Figure 4E). The mean withdrawal threshold was significantly lower in the MMT rats (8.17 ± 1.47) compared to controls (12.67 ± 2.07; mean ± SD, n = 6; p. = 0.0015). Together, the gait and pain threshold analyses validate the successful establishment of the MMT rat model, confirming its reliability in recapitulating osteoarthritic pathology.

Figure 1: Establishment of DMM mouse model. (A) Schematic illustration showing the procedure for establishing the DMM model in mice.(B) Representative intraoperative images demonstrating key surgical steps in the creation of the DMM model. Abbreviation: DMM = Destabilization of the Medial Meniscus. Please click here to view a larger version of this figure.

Figure 2: Evaluation of the DMM mouse model at 2 weeks post surgery. (A) Representative Safranin O/Fast Green-stained sections of knee joints from control and DMM groups (scale bar = 500 µm).(B) Osteoarthritis Research Society International scoring results. (C) Representative gait pattern images of control and DMM mice. (D) Statistical analysis of spatial symmetry in gait. (E) Statistical analysis of heel pain threshold. Data are expressed as mean ± SD (n = 3). Data normality was assessed using the Shapiro-Wilk test, and group comparisons were conducted using a two-tailed independent samples t.-test (*P < 0.05, *P < 0.01, ns = not significant, P > 0.05). Abbreviations: DMM = Destabilization of the Medial Meniscus; OARSI = Osteoarthritis Research Society International. Please click here to view a larger version of this figure.

Figure 3: Establishment of the MMT rat model. (A) Schematic diagram depicting the procedure for establishing the MMT model in rats.(B) Representative intraoperative photographs showing the surgical steps of MMT model induction. Abbreviation: MMT = Medial Meniscus Tear. Please click here to view a larger version of this figure.

Figure 4: Assessment of the MMT rat model at 2 weeks. (A) Representative Safranin O/Fast Green-stained sections of knee joints from control and MMT groups (scale bar = 500 µm). (B) OARSI scoring results. (C) Representative gait images of control and MMT rats.(D) Quantitative analysis of spatial gait symmetry. (E) Statistical analysis of heel pain threshold. Data are presented as mean ± SD (n = 3). Normality was verified using the Shapiro-Wilk test, and statistical comparisons were made using a two-tailed independent samples t-test (*P < 0.05, *P < 0.01, ns = not significant, P > 0.05). Statistical analysis was performed using a two-tailed t.-test. *P < 0.05, **P < 0.01, ns = not significant (P > 0.05). Abbreviations: MMT = Medial Meniscus Tear; OARSI = Osteoarthritis Research Society International. Please click here to view a larger version of this figure.

This protocol provides a standardized framework for the induction and evaluation of two widely used surgical OA models. These models are particularly valuable because they closely replicate the pathophysiological features of human post-traumatic OA, thereby offering strong translational relevance. The DMM model involves transection of the medial meniscotibial ligament, whereas the MMT model entails a direct tear of the medial meniscus. In our laboratory, the DMM procedure is routinely performed in mice, as the medial meniscus is difficult to accurately identify in this species, and a complete meniscal tear can lead to excessive cartilage damage. Additionally, we have observed that subchondral bone damage is considerably milder in the DMM model compared with the MMT model. This suggests that while the DMM model is ideal for studying cartilage degeneration and mechanistic pathways, it is less suitable for investigating subchondral bone remodeling in OA research.

A critical determinant of success in both the DMM and MMT procedures is the precise transection of the target ligament, the MMTL in the DMM model and the MCL in the MMT model, without inducing iatrogenic injury to the articular cartilage. Such unintended cartilage damage can alter joint biomechanics and confound histopathological outcomes, particularly the OARSI scores used to evaluate disease severity. In the DMM model, achieving a clear visualization of the MMTL under a stereomicroscope is essential to ensure accurate transection and to prevent inadvertent disruption of adjacent structures. In contrast, for the MMT model, complete transection of both the MCL and the medial meniscus is required to induce consistent and reproducible joint pathology that mirrors the progression of post-traumatic OA. Despite careful surgical execution, procedure-related complications may still arise. The most common include iatrogenic cartilage injury, which can artificially elevate OARSI scores but can usually be identified and excluded by an experienced pathologist during histological assessment. Another important variable is behavioral testing variability; factors such as environmental stress or inadequate acclimatization can significantly influence pain threshold and gait measurements. Therefore, it is strongly recommended to provide sufficient adaptation time for animals prior to behavioral assessments to ensure reliability and reproducibility. Postoperative infection represents an additional concern, particularly in rats, where surgical wounds are larger and more susceptible to contamination. To minimize this risk, prophylactic antibiotics are routinely administered. Animals exhibiting joint swelling, erythema, or signs of infection following surgery are excluded from subsequent analyses to maintain data integrity and ensure the validity of experimental outcomes.

A key limitation of the instability-induced OA models is the anatomical differences between species, which affect the technical feasibility and reproducibility of surgical procedures. The MMT model, although robust in larger rodents such as rats, is less suitable for mice due to the extremely small size and delicate structure of the murine meniscus. Performing a clean and consistent meniscal tear in mice is technically challenging and highly dependent on the operator's skill, often resulting in variable outcomes and unintended damage to cartilage. In contrast, the DMM model, which involves transection of the MMTL rather than the meniscus itself, offers a more accessible and reproducible approach in mice. This model reliably induces joint instability and gradual cartilage degeneration, making it the preferred method for OA induction in small animal studies aimed at mechanistic or genetic investigations. Another practical consideration concerns the simplified mechanical sensitivity test used to assess postoperative pain. While this test is highly convenient, cost-effective, and suitable for high-throughput screening, it primarily provides a qualitative assessment of mechanical allodynia (i.e., sensitivity to normally non-painful stimuli) rather than an exact quantitative force threshold. Despite this limitation, it remains a valuable behavioral tool for evaluating pain-related changes and monitoring functional outcomes across experimental groups, particularly when complemented by more quantitative gait or nociception assays in advanced studies.

Compared to chemically-induced models, surgically induced instability models such as DMM and MMT offer a slower and more physiologically relevant disease progression, closely resembling the natural course of human post-traumatic OA^15,16. For instance, agents such as sodium iodoacetate or tert-butyl hydroperoxide can induce rapid inflammatory and degenerative changes within the joint cavity, making them useful for studying antioxidant or anti-inflammatory therapies. However, these models are limited in translational value, as chemical injury is rarely the underlying cause of OA in patients, thereby restricting their applicability to a narrower subset of research contexts. In contrast, instability-based models such as DMM in mice and MMT in rats better replicate the mechanical and structural etiologies of human OA, particularly cases arising from meniscal instability or ligamentous injury, which are common clinical scenarios. Consequently, these models offer a more clinically relevant platform for preclinical investigations, eliminating the conceptual limitations associated with chemically induced damage.

Thus, by providing this comprehensive, step-by-step video protocol, our goal is to minimize inter-laboratory variability, enhance technical reproducibility, and standardize OA research methodologies across laboratories. The potential applications of these validated models are extensive as they serve as powerful tools for evaluating the efficacy of emerging DMOADs and analgesics, for dissecting the genetic and molecular mechanisms of OA using knockout or transgenic animals (particularly with the DMM model in mice), and for advancing diagnostic imaging and biomarker development. Overall, these standardized models will significantly strengthen the translational bridge between experimental OA research and clinical application.