The present protocol describes the controlled microblade scratches on the surface of the articular cartilage after destabilizing the mouse knee by cutting the medial miniscotibial ligament. This animal model presents an accelerated form of osteoarthritis (OA) suitable for studying osteophyte formation, osteosclerosis, and early-stage pain.
Osteoarthritis is the most prevalent musculoskeletal disease in people over 45, leading to an increasing economic and societal cost. Animal models are used to mimic many aspects of the disease. The present protocol describes the destabilization and cartilage scratch model (DCS) of post-traumatic osteoarthritis. Based on the widely used destabilization of the medial meniscus (DMM) model, DCS introduces three scratches on the cartilage surface. The current article highlights the steps to destabilize the knee by transecting the medial meniscotibial ligament followed by three intentional superficial scratches on the articular cartilage. The possible analysis methods by dynamic weight-bearing, microcomputed tomography, and histology are also demonstrated. While the DCS model is not recommended for studies that focus on the effect of osteoarthritis on the cartilage, it enables the study of osteoarthritis development in a shorter time window, with special focus on (1) osteophyte formation, (2) osteoarthritic and injury pain, and (3) the effect of cartilage damage in the whole joint.
Osteoarthritis (OA) is the most prevalent musculoskeletal disease in people over 45, with over 8.75 million seeking treatment in the UK1. The growing prevalence of the disease has led to an increased economic and societal cost, is a major contributor to disability, and reduces the quality of life for patients1. Without treatments available, there is an urgent need to accelerate research to understand the development and progression of the disease. The disease is complex and also multifactorial in its nature. The main clinical measurements of the disease are pain and joint mobility2, and OA affects all the tissues in the joint, not just the cartilage3. One of the main challenges in understanding OA is that it can take years, sometimes decades, from initial presentation/injury to symptomatic disease progression with pain and immobility.
Modeling osteoarthritis in rodents has enhanced our knowledge of OA pathophysiology by allowing us to understand the initiation and progression in a much shorter time frame and with a detailed examination of the tissues involved. There are numerous murine models of osteoarthritis, from genetically modified animals to surgical intervention models. The most widely used murine model of post-traumatic OA is the destabilization of the medial meniscus (DMM)4,5. A caveat of the model is the variability between different operators. Experienced surgeons can perform the procedure with minimal joint damage, while inexperienced operators expose the joint capsule for longer periods of time and inflict damage on the cartilage. This variability in the process influences the severity of the model, with more initial damage leading to increased cartilage damage scores and osteophyte formation. Intending to reduce the variability between operators and mimic cartilage damage from clinical intervention, a modified version of this model is developed, whereby controlled additional damage onto the cartilage surface in the form of three superficial scratches are inflicted6. This also allows modeling the OA progression resulting from cartilage damage caused by some clinical interventions. Compared to the standard DMM model, the directly induced cartilage damage results in consistently accelerated protruding osteophyte formation, increased cartilage damage and inflammation, and measurable surrogate pain in male mice.
This model is particularly suitable for the study of early-stage post-traumatic OA, focusing on osteophyte formation, pain presentation (in male mice), synovitis, and early changes to bone parameters. The consistency of osteophyte formation in this model makes it pertinent to study bone repair and endochondral ossification since osteophyte formation is a process of repair via endochondral ossification7. The model also mimics damage introduced directly to the cartilage during clinical interventions, such as arthroscopic surgical procedures, and thus it is also suitable for the study of the effect of cartilage damage on the whole joint.
All experimental procedures were approved by the Ethical Review Panel of the University of Glasgow and the University of the West of Scotland, and carried out following the Animals (Scientific Procedures) Act 1986 (UK) guidelines. 10-week-old C57Bl6/J male mice, weighing around 25 g, were used for the present study. The mice were obtained from commercial sources (see Table of Materials).
1. Animal preparation
NOTE: Consider mouse gender in regards to the purpose of the study as post-traumatic OA models display important differences depending on gender8,9,10.
2. Pre-operative care (carried out by a surgical assistant)
3. Destabilization of the medial meniscus procedure followed by cartilage scratch
4. Post-operative care
5. Evaluation of osteoarthritic disease
The percentage load per total body weight of the rear operated/OA leg was compared to the contralateral/control leg. Although other parameters may also give significant differences, like the increase in front paw load after surgical intervention, a consistent change in rear paw load indicates a preference to use one leg over the other and is a more direct indicator of significant discomfort for the mouse due to OA development. There were no significant changes in rear leg load in the DMM model within 8 weeks post-induction, while DCS mice favor the contralateral/control leg significantly 2 weeks after intervention (Figure 3A).
Subchondral bone was analyzed by focusing on the volume under the medial loaded region of the tibial condyle. Here we assessed the bone density of this area by determining the percentage of mineralized bone within the region of interest and calculated the ratio between the contralateral and the ipsilateral leg. The ratio indicates that both models have increased bone density in the affected limb (ratio above 1) 4 weeks after induction (Figure 3B). The emergence of osteophytes is more prominent in the DCS model, where there is a significant increase in the number and volume compared to the DMM model 2 weeks after intervention (Figure 3C,D). DCS presents elevated cartilage damage in the medial tibial and femoral compartments and synovitis (Figure 3E,F) 4 weeks after induction.
Figure 1: Surgical intervention to induce post-traumatic OA in the mouse. Sequential images represent the different stages of the procedure. (A) Exposure of joint capsule cutting the superficial membrane around the knee by inserting a number 11 scalpel blade on the medial side of the patellar ligament and away from the ligament. This will expose the infrapatellar fat pad. (B) Identification and transection of the medial meniscotibial ligament. To identify the ligament, move the patellar ligament toward the lateral side and then push the fat pad upward. This allows for the visualization of the ligament as a small horizontal white line just above the tibial condyle (indicated here with a black arrow). To cut the ligament, the lower blade of the spring scissors is placed under the ligament, taking care not to damage the cartilage. Move the meniscus towards the medial side to visualize the tibial condyle. (C) Scratching the surface of the exposed cartilage and closure of the wound. To scratch the cartilage, the microblade is inserted toward the posterior side, where it contacts the cartilage and then moves forward toward the anterior part of the joint. Once the scratches are done, pull the skin over the knee and close the wound either by subdermal suturing or with wound clips. Please click here to view a larger version of this figure.
Figure 2: Evaluation of osteoarthritis in the mouse. (A) Dynamic weight-bearing consists of matching the load on a pressure mat to the corresponding paw. The load is then expressed as a percentage of the total weight. (B) Subchondral bone is measured by selecting a volume of interest in the loading region of the medial tibial condyle and selecting the subchondral plate or trabecular bone. These images are at a resolution of 4.5 µm. (C) Osteophytes are identified and quantified in a three-dimensional view of the acquired µCT images. The volume of osteophytes is measured by selecting an ROI delineating the edge of the osteophyte. The bone density is calculated as the bone volume per osteophyte volume. The images presented here were taken at a resolution of 2 µm, but quantification is usually done with a 4.5 µm resolution. (D) Cartilage and synovitis scores are taken from 6 µm sections stained with Safranin-O and Fast green. A coronal section of the mouse knee where all quadrants, marked with a black box, are visible for scoring and a magnification of the medial side is shown. The synovitis surrounding the medial side of the knee joint is also visible, especially above and below the displaced meniscus. Please click here to view a larger version of this figure.
Figure 3: Representative evaluation of OA in the DMM and the DCS models. (A) DWB measured up to 8 weeks post-induction on experiments carried out by the same expert operator. Load is expressed as a ratio between the operated/OA load versus the contralateral/control load. Paired t-tests of both legs are also shown in the Sham (grey), DMM (blue), and DCS (pink) models. µCT analysis 4 weeks after surgical intervention. (B) Subchondral bone was analyzed 4 weeks after surgical intervention and expressed as the ratio of the ipsilateral over the contralateral %BV/TV. (C) Osteophyte number and (D) osteophyte volume was analyzed 2 weeks after induction. Histological evaluation 4 weeks after induction of (E) cartilage damage of the medial tibial and femoral articular cartilage and (F) synovitis were scored with standardized methods19,20. Data are expressed as mean ± standard deviation, n ≥ 5. Data were compared by repeated-measures ANOVA with a Šídák test correction (A), paired t-test (A), or standard Student's t-test (B–F). *P < 0.05, **P < 0.01, ***P < 0.001, ns = not significant. Please click here to view a larger version of this figure.
Supplementary File 1: Treatment and incubation condition for paraffin embedding. Please click here to view a larger version of this figure.
To perform surgical induction of post-traumatic osteoarthritis (PTOA), support from an assistant is strongly recommended (e.g., to prepare the mice while the operator focuses on the surgery). This facilitates aseptic surgery, thereby reducing the risks of infections and making the intervention more efficient in large experiments. It is easy to lose the plane of focus during the surgery, so a microscope that includes pedals for focusing is a valuable feature in helping to maintain sterility throughout the surgery. The position of the mouse and the knee is crucial. The knee must be facing upward and sufficiently bent to maximize the opening of the knee joint space, facilitating easier access to the ligament for introducing the microblade to scratch the condyle surface. Identifying the MMTL can be challenging, especially when the fat pad is larger than usual or there is a small bleed. To avoid bleeds, push the fat pad upward to prevent tears and subsequent bleeding. If the fat pad is large, this might take a little longer, but patiently continue to push it upwards.
The MMTL is quite close to the tibial condyle, so one must take care not to injure the cartilage when positioning the lower blade of the curved spring scissors under the MMTL. The curved blades should point toward the medial side and slightly upward, parallel to the condyle. For best sectioning of the MMTL, ensure the scissors are sharp. Check that the meniscus can move medially after cutting the ligament, as sometimes a small attachment remains that needs further cutting. When introducing the microblade to scratch the condyle, it must be perpendicular to the condyle. Make the first scratch closer to the middle of the joint but take care not to damage the anterior cruciate ligament. Then move toward the medial side and then behind the meniscus. The scratches might be visible as faint white lines on the cartilage. Because we usually use clips, the initial incision is performed on the lateral side, so the clips are positioned on the side of the leg after closing the wound. This avoids the clips rubbing the knee as the mouse regains movement. When using sutures, the use of subdermal stitches is strongly recommended. If using external stitches, the mice are likely to gnaw at the stitches and open their wound, which will increase the chances of infection. When done right, this surgery must not take more than 5-10 min, from incision to wound closure, thus minimizing the exposure of the cartilage and any additional uncontrolled damage that may occur. After the surgery, the mice recover very quickly and almost immediately can climb into the cage and move around normally. If the mice are not active, the appropriate expert in the unit should be consulted.
For the behavioral evaluation of pain, dynamic weight-bearing was assessed. However, this method may be considered less sensitive than other evoked pain tests, such as von Frey testing15. It is recommended that more than one method is used to monitor and assess pain. The changes observed 2 weeks after intervention in DCS, even though transient, indicate a generally decreased loading of the OA leg compared to the healthy leg. Therefore, 2 weeks after DCS intervention may be used to evaluate early osteoarthritic or injury pain in mouse models. Visualization of mineralized osteophytes by µCT allows for three-dimensional quantification, which can also be matched to the histological sections12, adding another dimension to the study of osteophyte emergence and evolution. In our group, osteophyte presence was variable in the DMM model between and within operators (2.3 ± 1 vs. 1.2 ± 1, n > 7, P = 0.0183), whereas induction of DCS robustly led to osteophyte generation in all cases irrespective of the operator (2.6 ± 0.7 vs. 2.4 ± 0.5, n > 7, P = 0.711). Also, there are significantly more and larger osteophytes in the DCS model compared to DMM. Thus, DCS is an ideal model for the study of osteophyte formation. Quantification of osteosclerosis limited to the loading area of the subchondral bone is also an improvement in detecting small changes. Comparing the medial compartment of the operated leg to the contralateral leg also offers a way to normalize against the intrinsic bone phenotype of that particular mouse12. The addition of the cartilage scratches in the DCS model is a controlled means of inducing focused cartilage damage during surgery that accelerates many of the aspects of the disease. One of the consequences of the experimental procedure involving intentional damage to the cartilage itself is that this artefactual damage needs to be excluded or adjusted for in the cartilage grading system. Because of this limitation, we do not recommend this model if the study's main aim is to understand the effect of osteoarthritis on the cartilage itself. Finally, it is also strongly recommended to have at least two blinded scorers grade the cartilage damage and synovitis scores. This validates and enhances the standardization of the scoring systems.
A limitation of this study is that the extent of variability across all the parameters comparing the DCS and DMM models was not fully evaluated. This will be addressed in the future with more extensive studies, which could also include an assessment of the variability between operators from different institutions.
In conclusion, the accelerated OA pathogenesis in the current DCS model allows representation of post-traumatic OA and provides a powerful and robust research tool to investigate and elucidate underlying OA pathophysiological mechanisms driving this chronic debilitating joint disease. Additionally, it enables OA to be explored in a shorter time window, focusing on osteophytogenesis, OA pain, and the effect of cartilage damage on the whole joint.
The authors have nothing to disclose.
We would like to acknowledge the work of Gemma Charlesworth and Mandie Prior at the University of Liverpool, who acquired the µCT images used in this publication. Work was funded by Versus Arthritis (grants 20199 and 22483). Lynette Dunning was funded by Versus Arthritis (grant 20199). Kendal McCulloch was funded by a UWS PhD scholarship. Carmen Huesa was funded by Versus Arthritis (grants 20199 and 22483).
#11 scalpel blade (and scalpel handle). | World precision instruments | 500240 | access the joint capsule |
15° Cutting Angle microsurgical stab knife | MSP | REF7503 | scratch the cartilage |
6-0 vicryl rapide | Any medical supplies provider | – | alternative method to close wound |
Anaesthetic rig | Generic (many different suppliers) | – | |
Antibacterial skin clenser (Hibiscrub) | Amazon | – | To sterilise surgical skin area |
Applicator for 7 mm clips | World precision instruments | 500343 | close the wound |
Balance | Generic (many different suppliers) | – | To weigh mouse |
Blunt curved forceps | Fine science tools | 500232 | move the patellar ligament to the side |
Buprenorphine (Vetergesic) | Supplied by unit as it is a prescription drug | – | Analgesia |
CT analyser | Bruker | 3D.SUITE software | Software |
Ctvol | Bruker | 3D.SUITE software | Software |
Data viewer | Bruker | 3D.SUITE software | Software |
Dynamic weight bearing equipment | Bioseb | BIO-DWB-DUAL | Measure limb loading and has cage, pressure matt and software for analysis |
EDTA | Merck | E9884 | 10% solution in PBS (or water) to decalcify bone pH 7.4 |
Ethanol | Generic (many different suppliers) | – | for embedding decalcified bones |
Fast Green FCF | Merck | F7252 | For staining sections |
Glacial acetic acid | Merck | 1005706 | For stianing sections |
Haematoxylin solution | Merck | GHS132 | Nuclear staining in paraffin sections. |
Hoskins #21 micro-tweezers. | Cameron surgical limited | PHF1085 | move the fat pad |
Isofluorane | Supplied by unit as it is a prescription drug | – | |
Mice | Charles river | – | C57Bl6/J male 8 weeks old (to allow acclimatisation in the unit) |
Microcomputed tomography scanner | Bruker | SKYSCAN 1272 CMOS | µCT |
Micropore surgical paper tape | FisherScientific | 12787597 | hold leg in position |
Paraffin wax | Generic (many different suppliers) | – | for embedding decalcified bones |
Reflex 7 mm stainless steel wound clips or | Fine science tools | 12032-07 | close the wound |
Remover for 7 mm clips | World precision instruments | 500347 | remove wound clips |
Rotary Microtome | Generic (many different suppliers) | – | To cut section of Paraffin embedded tissue. |
Safranin-O | Merck | S2255 | For staining sections |
Serrated curved forceps | Fine science tools | 15915 | hold the skin |
Sterile Drape | Generic (many different suppliers) | – | To ensure sterility of surgical area |
Sterile Drape with key hole | Generic (many different suppliers) | – | To cover mouse and expose leg |
Sterile saline | Generic (many different suppliers) | ||
Sterile surgical drape | Generic (many different suppliers) | – | maintain sterile environment for surgical tools |
Sterile surgical drape with key hole | Generic (many different suppliers) | – | cover the mouse and keep leg through key hole |
Straight Scissors | World precision instruments | 14393 | open the wound |
Surgical microscope. | Generic (many different suppliers) | – | Adjustable focus. |
Vannas spring scissors with 2 mm blades. | Fine science tools | 15000-04 | cut the MMTL |
Xylene | Generic (many different suppliers) | – | for embedding decalcified bones |