Anatomical Localization of Cartilage Degradation Markers in a Surgically Induced Rat Osteoarthritis Model

Abstract

Osteoarthritis (OA) is a degenerative disease characterized by an irreversible loss of articular cartilage. Although surgically induced animal OA models are commonly used in drug efficacy assessment, degradation of type II collagen, an important component of articular cartilage is not routinely evaluated. Here, the medial meniscectomy surgical model (MMT) in Lewis rats was evaluated for proteoglycan loss with toluidine blue staining and collagen degradation with immunohistochemical staining for a collagen cleavage C-neoepitope, using a novel anti-type II collagen neoepitope antigen (TIINE) antibody. Femorotibial joints were collected for histology at 0 (no surgery), 3, 7, 14, 21, 28, 35, and 42 days postsurgery. Following MMT surgery, the medial tibial articular cartilage had proteoglycan matrix loss by day 3 that reached subchondral bone by days 28–42. Femoral cartilage damage occurred by day 14. TIINE staining was present at basal levels in growth plates and articular cartilage of all joints while all MMT-treated animals had increased intensity and area of staining in erosions that colocalized with proteoglycan loss. The MMT model produces a progressive pattern of cartilage damage resembling human OA lesions, making it useful, when evaluated with cartilage biomarkers, for assessing changes in cartilage degradation.

Introduction

Osteoarthritis (OA) is the most common type of joint disease, affecting 80–90% of men and women after age 65, and is characterized by progressive loss of articular cartilage integrity and other joint changes (Rosenberg, 1999). The poor regenerative capacity of the avascular and aneural articular cartilage prevent the recovery of full function after initial injury (Vigorita, 1999). The disease pathogenesis leading to the cartilage degradation remains an important research field (Janusz et al., 2002), as currently available treatments cannot halt damage such as erosions, surface fibrillation, cell loss, increased osteophyte size, and decreased cartilage thickness. Recent work focuses on catabolic enzymes such as collagenases (Otterness et al., 2000) and degraded cartilage components (Billinghurst et al., 1997; Otterness et al., 1999; Downs et al., 2001) as possible biomarkers to monitor disease progression.

Collagen damage is particularly troubling, as the network of fibrils providing the articular cartilage with tensile strength have a slow turnover, often taking years (Rosenberg, 1999). The collagenases (matrix metalloprotinases (MMP) 1, 8, 13, and MT1-MMP) are the enzymes capable of the initial cleavage in intact collagen necessary to denature the triple helix structure at physiological conditions (Whitham et al., 1986; Hasty et al., 1990; Freije et al., 1994; Ohuchi et al., 1997). A C- and N-neoepitope are formed by the cleavage, and antibodies against the C-terminal neoepitope have been evaluated in human and animal model OA tissues. Antibodies against these collagen fragments, also known as type II collagen neoepitopes (TIINE) have become an important tool in the study of OA (Billinghurst et al., 1997; Vankemmelbeke et al., 1998; Otterness et al., 1999; Stoop et al., 2000).

Animal models are an important tool for elucidating OA pathogenesis. Numerous surgically and chemically induced models in multiple species have been characterized for proteoglycan matrix, collagen, and bone changes (Bendele, 2001). Desirable models for drug development generate consistent, reproducible articular cartilage lesions in a rapid period of time. Rat surgical models often meet these criteria (Stoop et al., 2000; Karahan et al., 2001; Janusz et al., 2002; Roberts et al., 2003). MMT surgery is a model that has been partially characterized in the literature; however, this and other models have not been evaluated for the specific anatomical localization and progression of collagen degradation. This study characterizes cartilage damage in the rat MMT model from 0–42 days postsurgery by correlating morphologic changes in toluidine-blue-stained sections with immunohistochemistry for collagen cleavage under light microscopy.

Here, we report that proteoglycan loss occurs as early as day 3 postsurgery and lesions progress into fibrillation, erosions, and chondrocyte degeneration similar to the human form of the disease. TIINE staining colocalized in both control and surgically treated animals in the growth plate and in the calcified layer of the articular cartilage. Rats that underwent surgery showed increased levels in and surrounding cartilage lesions that had no comparable staining in control animals.

Materials and Methods

Animal Model

The procedures used in this study were consistent with the guidelines of the Pfizer Institutional Animal Care and Use Committee. Animals were provided with living conditions, food, and housing consistent with the approved animal care operating procedures. In order to minimize any potential discomfort, animals were housed in solid bottom cages as opposed to wire bottom cages. Male Lewis rats (81–90 days of age) were received and acclimated to their environment 3–5 days before surgery or joint collection. The medial meniscal tear surgery was performed under isoflurane anesthesia (Abbott Laboratories, North Chicago, IL, USA) as previously described (Janusz et al., 2002). Briefly, animals were anesthetized with isoflurane and the skin over the medial aspect of the right femorotibial joint was clipped to remove the hair and surgically prepped with Nolvasan Surgical Scrub (Fort Dodge Animal Health, Overland Park, KS) followed by 70% alcohol. The medial collateral ligament was exposed by blunt dissection and transected to reflect the meniscus toward the femur. The joint space was visualized and the meniscus was cut through the full thickness at its narrowest point to simulate a complete tear. The skin was closed with 4.0 silk suture. Sham surgery (medial collateral ligament transected without direct meniscal damage) was performed on 1 rat to assess the effects of joint destabilization and sacrificed 42 days postsurgery.

Histology

Four rats per time point at 0 (no surgery), 3, 7, 14, 21, 28, 35, and 42 days postsurgery were used in the study. The entire femorotibial joint was removed, fixed in 10% phosphate buffered formalin, decalcified in formic acid (Immunocal, Decal Corporation, Tallman, NY), bisected in the frontal plane, and routinely processed into paraffin-embedded blocks. Three step sections from each MMT and sham surgery animal and 2 sections from each control were stained with toluidine blue. Damage to the articular cartilage and osteophyte size were scored as previously described (Janusz et al., 2002), with the modification that osteophyte size was scored as 0–4 (no osteophyte, <200, <300, <400, >400 μm) and the ratio of depth of cartilage damage to depth of total un-calcified cartilage was determined at four equidistant points across the medial tibial articular cartilage using an ocular micrometer.

Immunohistochemistry

Deparaffinization and probing with the anti-type II collagen monoclonal antibody (Oncogene Research Products, San Diego, CA) was performed at 37°C (Discovery Instrument, Ventana Medical Systems, Tucson, AZ). Antigen retrieval was a 16-minute enzyme digest with Protease, followed by Avidin and Biotin blockers (Ventana Medical Systems, Tucson, AZ). A 1:200 dilution of primary antibody was followed by a 1:1000 dilution of linker antibody (Biotinylated Anti-Mouse IgG, Vector Laboratories, Burlingame, CA) for 30 minutes each before counterstaining with hematoxylin and Bluing reagent (Ventana Medical Systems, Tucson, AZ).

The mouse monoclonal antibody used for TIINE immuno-histochemistry was developed in house (clone 1B12) and is specific for a collagen degradation neoepitope sequence similar to that detected by the monoclonal 9A4, described elsewhere (Otterness et al., 1999). The specificity of this antibody was validated by western blotting as shown in Figure 1. Routine deparaffinization for TIINE staining was performed manually, followed by antigen retrieval with a thiocyanic acid and guanidine solution 10 minutes at room temperature (LAB, Polysciences Inc, Warrington, PA). Probing was performed at room temperature (DAKO Autostain Plus, DAKO Cytomation, Carpinteria, CA). Then, 3% hydrogen peroxide (Pfizer Consumer Health, Morris Plains, NJ) suppressed endogenous peroxidase activity and DAKO Protein Block (DAKO Cytomation, Carpinteria, CA) decreased nonspecific background staining. A 1:800 dilution of 1B12 for 1 hour was followed by secondary antibody for 30 minutes (DAKO Envision (+) System murine DAB kit, DAKO Cytomation, Carpinteria, CA) before counterstaining (DAKO Automation Hematoxylin, DAKO Cytomation, Carpinteria, CA).

Mouse normal serum IgG₁ was used in place of antibody for the negative control while the growth plate served as an internal positive control for all staining. Primary, negative control, and the Vector linker antibody dilutions were made in phosphate-buffered saline solution (ChemMate, Ventana Medical Systems, Tucson, AZ).

Statistical Analysis

OA scoring was analyzed using a Trend test (SAS).

Results

Histology

MMT-induced lesions increased in severity over time (Figure 2A). The trend test found all damage score parameters to be significant (p < .015) at all time points except femoral damage at time points 3 and 7 days (no damage present). Within surgical time points, lesion severity was consistent (Figure 2A). Osteophytes first developed at day 3 and rapidly increased in size between days 7 and 14 (Figure 2B), indicating a proliferative cartilage reaction to the joint destabilization. Occasional osteophyte fissuring and cracking was seen by day 35 and consistent ossification at days 35–42. The average depth of cartilage damage increased over time, reaching 50% of total cartilage thickness by day 42 (Figure 2C).

The most severe lesions consistently occurred in the outer region of the medial tibia (Figure 3). An initial loss of proteoglycan matrix with mild irregularity of the articular cartilage surface was visible at day 3 that progressed into fibrillation, complete matrix loss, and chondrocyte degeneration by day 7 (Figures 3B, 3C). Cartilage degeneration reached the tide-mark by day 21 and extended into the deep calcified cartilage and subchondral bone by days 28–42 (Figure 3D). Chondrocytes adjacent to erosions initially underwent cloning with eventual loss of cells as lesions deepened. In the corresponding femoral condyle, proteoglycan loss was evident by day 14 and progressed to significant cartilage degeneration with fibrillation and erosions by day 42 (Figure 3D). Significant sclerosis and occasional cysts were evident in subchondral bone. The sham surgery animal showed mild osteophyte formation, but otherwise no overt cartilage damage.

Immunohistochemistry

In normal articular cartilage, weak-to-moderate diffuse staining for TIINE was typically observed in the calcified region, below the tidemark (Figure 4A). In addition, the growth plate of all animals had strong staining in cartilage between the zone of hypertrophy and adjacent bone spicules (Figure 4B). Staining was present in the extracellular matrix and partially surrounded cells in the last row of hypertrophied chondrocytes. Staining was abrogated when the sections were incubated with an irrelevant mouse antibody.

Mild superficial erosions by day 3 showed an increase of stain intensity in a narrow band deep to the surface damage. Deeper erosions at days 7–42 showed increased stain in the area deep to and immediately surrounding the erosion, often extending to the tidemark (Figure 4C). There was an eventual loss of stain in the most superficial fibrillated cartilage, while increased stain remained in deeper portions of erosions. The sham and control animals lacked the increased staining of articular cartilage seen in MMT animals. Moderate diffuse staining for type II collagen was present in the growth plates and in the articular cartilage except for the most superficial layers. Increased intensity of the staining was observed in areas of proteoglycan loss and fibrillation, which colocalized with TIINE staining (Figure 4D).

Discussion

We characterized the morphology and progression of osteoarthritis-like lesions in the rat MMT OA model up to 42 days. The model induced consistent, reproducible lesions that closely resemble the human disease. More importantly, we demonstrated by immunohistochemistry that areas of cartilage damage contain significantly increased levels of a type II collagen degradation product, an important biomarker that could be used to monitor disease progression (Downs et al., 2001) as well as response to therapeutic intervention in animal efficacy studies and in human clinical trials.

The toluidine blue staining demonstrated rapid, reproducible damage to the medial tibial articular cartilage that was consistent with previous studies (Janusz et al., 2002). Damage to medial tibial articular cartilage was expected because it is destabilized by the surgery and rats primarily bear weight on the medial tibial cartilage (Bendele, 2001). Although the lesions were severe, they were highly localized; the inner regions of the medial and all of the lateral tibial cartilage were unaffected or only mildly damaged as late as day 42 and medial femoral articular cartilage lesions were absent before day 14. Even though the surgery rapidly induced lesions, the slower and consistent progression as compared to other models, such as the mono-iodoacetate (MIA) model (Guzman et al., 2003), may make assessment of therapeutic response easier during treatment trials.

Collagen type II is known to be the most common type of collagen in hyaline articular cartilage (Bullough, 1997). In our study, there was moderate staining of normal articular cartilage for type II collagen except in the most superficial layers. This is consistent with previous report indicating that these superficial layers contain primarily type I and III collagens (Teshima et al., 2004). TIINE staining colocalized with strong collagen type II staining in areas of cartilage degeneration and fibrillation, confirming that these neoepitope fragments are derived from type II collagen. The increased intensity of type II collagen in cartilage lesions is likely due to increased exposure of antigenic epitopes due to loss of the surrounding proteoglycan matrix. The loss of TIINE staining in the most superficial fibrillated areas of erosions at late time points may indicate a complete loss of collagen, while the deeper portions of those lesions continued to show increased active collagen cleavage. Consistent neoepitope detection over all time points in the present study differs from a cranial cruciate ligament (CCL) rat model that lacks collagen degradation fragments after 4 weeks (Stoop et al., 2000). That study uses an antibody against the N-terminal collagen neoepitope, which is smaller and more rapidly degraded. (Vankemmelbeke et al., 1998), which could hinder immuno-histochemical detection. Staining for denatured collagen is detected up to 10 weeks postsurgery in the CCL model, which indicates continued damage to the collagen network (Stoop et al., 2000).

Collagen type II and C-neoepitope were also observed the deep calcified cartilage. As neoepitope was present in both control and surgically treated animals, this likely indicates the basal turnover of collagen in those tissues. This observation of neoepitope in calcified cartilage is similar to descriptions in nontransgenic control and collagen-deficient transgenic mice (Salminen et al., 2002), as well as unoperated rats from a CCL-transection study (Stoop et al., 2000). In the present study, C-neoeptiope was also observed in the reserve and hypertrophic zones of the tibial growth plates. The observation in the hypertrophic zone is in agreement with recent reports of collagen cleavage in Wistar rats (Bae et al., 2003) as well as bovine fetal physis (Mwale et al., 2002). MMP-13 mRNA is detected in the hypertrophic zone of Sprague-Dawley rat tibial growth plates of a similar age to the Lewis rats used in this study (Alvarez et al., 2000), although collagen cleavage was not investigated in that study.

There is interest in using antibodies against TIINE in clinical biomarker assays to assess the efficacy of osteoarthritic therapeutics during a study (Downs et al., 2001). In the MMT rat model, differences in urinary C-neoepitope levels may be difficult to detect because of the relatively large amount of endogenous C-neoepitope produced in the growth plate and articular cartilage compared to the localized effects in the lesioned femorotibial joint. Rabbits could also serve as a model utilizing TIINE biomarkers, as the rabbit growth plate closes with age and the multiple rabbit collagenase genes share greater homology with human collagenases than the single rodent collagenase shares with human forms (Vincenti et al., 1998).

In conclusion, the proteoglycan and collagen changes in articular cartilage in the MMT model demonstrate rapid and consistent disease progression after the initial meniscal injury. This study allows for increased understanding of the localization and time progression of collagen degradation in this commonly used model. The use of antibodies against TIINE show promise for monitoring joint changes during human clinical and animal treatment trials, adding important data to the normal histological end points. By more accurately characterizing the osteoarthritic lesions in this and other animal models with histology and biomarkers, an enhanced ability to evaluate the in vivo effects of candidate molecules on cartilage degradation will be possible.