Accelerated development of aging-associated and instability-induced osteoarthritis in osteopontin-deficient mice

Authors


Abstract

Objective

To investigate the role of osteopontin (OPN) in the development of osteoarthritis (OA) under in vivo and in vitro conditions.

Methods

Both instability-induced and aging-associated OA models were generated using OPN-deficient (OPN−/−) and control wild-type (WT) mice. An in vitro cartilage degradation model was also used, to evaluate the effect of OPN on proteoglycan loss from joint cartilage.

Results

OPN deficiency exacerbated both aging-associated and instability-induced OA. Both structural changes and an increased loss of proteoglycan from cartilage tissue were augmented in the absence of OPN. OPN deficiency also led to the induction of matrix metalloproteinase 13 (MMP-13), which degrades a major component of the cartilage matrix protein type II collagen. Both the loss of proteoglycan and the induction of the collagen-degrading enzyme MMP-13 facilitated the development of OA.

Conclusion

OPN plays a pivotal role in the progression of both instability-induced and aging-associated spontaneous OA. OPN is a critical intrinsic regulator of cartilage degradation via its effects on MMP-13 expression and proteoglycan loss.

Osteoarthritis (OA) is a noninflammatory degenerative joint disease that is mainly characterized by cartilage degradation, subchondral bone sclerosis, and osteophyte formation. This disease is the most common form of arthritis and a leading cause of physical disability. Approximately 80% of patients with OA have some degree of limitation of movement, and 25% of patients with OA cannot perform the major activities of daily living (1). The main causes of OA are considered to be aging and excessive mechanical stress to the joints, due to joint instability or obesity (2). A number of in vivo studies using various joint instability models have clarified some aspects of the molecular mechanisms comprising the response to excessive mechanical stress to the joints (3–8). However, the mechanisms of aging-associated development of OA have not been clearly elucidated. Effective prevention and therapeutic treatment of OA will require the identification of common mediators of joint destruction.

Articular cartilage is composed of abundant extracellular matrix (ECM) proteins, including collagens, proteoglycans, and noncollagenous proteins. Regulation of the synthesis and degradation of cartilage ECM is necessary for the maintenance of normal cartilage, and deregulation of cartilage metabolism undoubtedly plays a crucial role in the pathogenesis of cartilage degradation. This process is regulated by mechanisms that depend on the interaction of chondrocytes with ECM proteins (9). This interaction is mediated through specific cell surface receptors belonging to the integrin family. Integrins are transmembrane glycoproteins consisting of an α and a β subunit. Ligands for integrins on chondrocytes are often large ECM proteins such as collagen, laminin, fibronectin, or soluble proteins, including vitronectin and osteopontin (OPN) (10). Thus, it is increasingly recognized that these ECM proteins may contribute to the maintenance of articular cartilage homeostasis and are also involved in the pathogenesis of OA.

OPN was originally identified as a noncollagenous ECM protein in the bone (11, 12). This protein is secreted by various cells, including osteoclasts, osteoblasts, activated T cells, macrophages, smooth muscle cells, and tendon fibroblasts and exists in several tissues, such as the bone, kidney, placenta, smooth muscle, and secretory epithelia (13, 14). OPN contains an RGD sequence that interacts with integrin receptors to promote cell adhesion, chemotaxis, and signal transduction (15). In addition, we have demonstrated that the function of OPN in inflammatory reactions is regulated by syndecan 4, which is a major sulfated glycosaminoglycan (sGAG) on the cell surface (16). Investigators in our group (17) and other investigators (18) have generated mice with a targeted disruption of the OPN gene. OPN-deficient (OPN−/−) mice are fertile, their litter size is comparable with that of wild-type (WT) mice, and the growth rate of these mice is indistinguishable from that of WT animals. Moreover, the skeletal structure of young OPN−/− mice appears to be radiographically normal (14, 17–19).

With regard to the role of OPN in cartilage, previous studies have demonstrated that OPN is expressed in the calcified zone of hypertrophic cell cartilage, in which type X collagen is also expressed (20, 21). More importantly, it was shown that the expression of OPN proteins and genes correlated well with the severity of OA, such as in the disintegration of the cartilaginous matrix (22, 23). However, the functional roles of OPN in cartilage degradation have not been elucidated.

In the present study, we performed a functional analysis of OPN in cartilage degradation. Specifically, we focused on the functional role of OPN in development of aging-associated and instability-induced OA. Our study demonstrated that OPN was involved in the development of both aging-associated and instability-induced OA. OPN deficiency led to the induction of matrix metalloproteinase 13 (MMP-13), which degrades a major component of the cartilage matrix protein type II collagen. We also found that the release of proteoglycan from cartilage tissue was augmented in the absence of OPN. Both the loss of proteoglycan and the induction of a collagen-degrading enzyme, MMP-13, facilitated the development of OA. Therefore, results from the present study indicate that OPN acts as a key molecule in the regulation of OA development in vivo.

MATERIALS AND METHODS

Animals.

OPN−/− mice were generated as described previously (17). Six-week-old male OPN−/− mice were backcrossed 9 generations to C57BL/6 mice, and age-matched male C57BL/6 WT mice were used as a control. The animal experiments were carried out in the Institute of Animal Experimentation, under the rules and regulations of the Animal Care and Use Committee of Hokkaido University Graduate School of Medicine. In this study, only male OPN−/− and WT mice were used.

Instability-induced and aging-associated OA models.

An instability-induced OA model was created in 8-week-old WT or OPN−/− mice, as illustrated in Figures 1A and B. Animals were anesthetized with isoflurane. The right knee joint was destabilized by transection of the medial collateral ligament followed by removal of the cranial half of the medial meniscus (5). The left knee joint was sham-operated, in that the joint was prepared using the same approach as that for the right knee joint but without any ligament transection and menisectomy. Mice were kept for an additional 8 weeks after surgery.

Figure 1.

Histopathologic features of the knee joints of wild-type (WT) and osteopontin-deficient (OPN−/−) mice in the instability-induced osteoarthritis (OA) model. A and B, Illustration of the surgical method used to produce instability-induced OA. A normal knee joint (A) and an operated knee joint (B) are shown. The operated knee joint was destabilized by transection of the medial collateral ligament (MCL) and removal of the cranial half of the medial meniscus (MM). The operated joints of WT mice (C, E, G, I, and K) and OPN−/− mice (D, F, H, and J) were harvested and evaluated by hematoxylin and eosin staining (C and D), Safranin O–fast green staining (E and F), immunohistochemistry for type II collagen (G and H), immunohistochemistry for type X collagen (I and J), and immunohistochemistry for OPN (K). Insets show higher-magnification views of the boxed areas (original magnification × 400). Arrowheads indicate the level of the tidemark. Arrows indicate hypertrophic cells staining positive for type X collagen in the calcified zone of cartilage. Blue, red, and black bars in the margins indicate the uncalcified zone, calcified zone, and subchondral bone, respectively. Fe = femur; Ti = tibia; Pa = patella. Bar = 100 μm.

For the aging-associated model of OA, WT and OPN−/− mice were kept up to the age of 15 months and were followed up for spontaneous development of OA. (24).

Histopathologic evaluation of OA.

The hind knee joints were dissected and fixed in 10% buffered formalin, and then decalcified in 10% EDTA, pH 7.5, for 1 week. After the tissue was dehydrated and embedded in paraffin, serial sagittal sections (5 μm thick) were cut through the knee joints. Sections were stained with Safranin O–fast green to assess the presence of proteoglycan and to measure the cartilage layer thickness in OA joints. The severity of cartilage erosion was evaluated according to a semiquantitative scoring system, as described previously (25), in which 0 = no damage, 1 = roughened articular surface and small fibrillations, 2 = fibrillations down to the layer immediately below the superficial layer and some loss of surface lamina, 3 = loss of surface lamina and fibrillations extending down to the calcified cartilage, 4 = major fibrillations and cartilage erosion down to the subchondral bone, 5 = major fibrillations and erosion of up to 80% of the cartilage, and 6 = >80% loss of cartilage.

To determine whether the OA changes in mice were associated with loss of chondrocytes, the cartilage cellularity was quantified by counting the number of chondrocytes in a microscopic field. In tibial or femoral cartilage from the mice, 4 images were obtained by microscopy under 400× magnification (26). The total number of chondrocytes was evaluated in each image. In addition, the proteoglycan content of the articular cartilage was graded on Safranin O–stained sections using categories of Mankin scores of Safranin O staining (range 0–4), in which 0 = normal, 1 = slight reduction in staining, 2 = moderate reduction in staining, 3 = severe reduction in staining, and 4 = no dye noted (27, 28).

To evaluate the thickness of the cartilage, sections from each stratum of each knee joint were photographed, and the digital images were analyzed using Adobe Photoshop 7.0 (Adobe Systems, San Jose, CA). The cartilage thickness was evaluated using Adobe Photoshop selection tools, which allowed us to select an area that spanned 450 μm across the surface of the cartilage and extended down to the surface of the subchondral bone. The number of pixels within the area was converted to μm2, and this converted value was divided by 450 μm to estimate the mean cartilage thickness within the selected area (29, 30).

Assessment of subchondral bone density in the knee joint.

The bone density of the subchondral bone at the tibial condyle was analyzed on soft x-ray radiographs. Subchondral bone was defined as a sclerotic area from the bone surface above the growth plate, as identified in the range of interest (ROI) on soft x-ray radiographs (see Figure 5F). Lateral images of the knee joints were obtained with an aluminum step phantom under consistent conditions (30 kV, 4 mA, for 70 seconds) (Tanaka SRO-M50D; MFG, Tokyo, Japan). The bone density of the ROI in the tibial condyle was defined as the value equivalent to the aluminum width (expressed in mmAL Eq.), which was calculated with the use of NIH Image software from the approximated curve of the pixel density of the aluminum step phantom (31).

Figure 5.

Comparison of OA severity between WT and OPN−/− mice in the aging-associated OA model, as assessed by A, scores of cartilage erosion, B, number of chondrocytes, C, scores of proteoglycan loss by Safranin O staining, and D, cartilage layer thickness in the tibia. E and F, The knee joints were also assessed by soft x-ray radiography in both WT and OPN−/− mice (E), and subchondral bone density was defined as a sclerotic area from the bone surface above the growth plate of the tibia on the soft x-ray radiograph, indicated as the range of interest (ROI) (F). The column on the left of F shows soft x-ray of the aluminum step phantom, with density ranges from 0 to 2.0 mmAL. The density in the ROI is based on the value equivalent to that of the aluminum phantom (mmAL Eq.). G, The subchondral bone density was compared between WT and OPN−/− mice. Bars show the mean and SD of 10 mice per group. ∗ = P < 0.05; ∗∗ = P < 0.01. NS = not significant (see Figure 1 for other definitions).

Immunohistochemistry for the detection of OPN protein and type II and type X collagens in articular cartilage.

Sections of the knee joints were stained with anti-mouse OPN rabbit IgG (O-17; IBL, Gunma, Japan), followed by a Dako EnVision+ system (Dako, Carpinteria, CA) as specified by the manufacturers. Type II and type X collagens were identified using polyclonal rabbit antibodies against rat type II and type X collagens (LSL, Tokyo, Japan).

Proteoglycan release from cultured cartilage explants.

To determine the role of OPN in cartilage degradation under in vitro conditions, total proteoglycan release from the articular cartilage was quantified. Femoral head articular cartilage was harvested from 4-week-old WT and OPN−/− mice, and the articular cartilage was separated from the underlying subchondral bone. Cartilage samples were cultured as explants for 48 hours at 37°C in a humidified atmosphere of 5% CO2 and 95% air in Dulbecco's modified Eagle's medium (DMEM) containing 1% antibiotic/antimycotic solution (Sigma, St. Louis, MO), 2 mM glutamine, 10 mM HEPES, 50 μg/ml ascorbate, and 10% fetal bovine serum. For the induction of cartilage degradation, the explants were then washed 3 times and cultured for an additional 72 hours in serum-free DMEM. Conditioned medium was collected at the end of the culture period (32, 33).

To quantify the proteoglycan release from cartilage explants in areas of cartilage degradation, the proteoglycan content in the conditioned medium was determined as the amount of sGAG measured on a dimethylmethylene blue assay (34), using chondroitin sulfate C from shark cartilage (Sigma) as a standard. To investigate the effect of MMP-13 on the release of proteoglycan from the cultured cartilage, we used 10 nM or 100 nM of a selective MMP-13 inhibitor (Calbiochem, Darmstadt, Germany) and 10 μM of a broad-spectrum MMP inhibitor (GM6001; Calbiochem). These inhibitors were added to serum-free DMEM after preculturing the cartilage tissue for 2 days; DMSO alone was added as a negative control.

Complementary DNA (cDNA) array.

We collected femoral head articular cartilage that had been cultured in DMEM (as described above) or that had been left uncultured (as a control). Total RNA was extracted from the samples using TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. After synthesis of the cDNA and biotin-dUTP labeling, a GEArray Q-Series Mouse Extracellular Matrix and Adhesion Molecules gene array (SuperArray Bioscience, Frederick, MD) was used to analyze the gene expression profile of MMPs, according to the manufacturer's protocol. The results were analyzed using a LAS-1000plus (Fuji Photo Film, Tokyo, Japan). The results were normalized to the values for the GAPDH housekeeping gene.

Real-time quantitative reverse transcription–polymerase chain reaction (RT-PCR) analysis.

To monitor the expression of MMP-13 messenger RNA (mRNA), cultured and uncultured cartilage from WT and OPN−/− mice was prepared. Total RNA was extracted from the articular cartilage, and first-strand cDNA was generated with a First-Strand cDNA synthesis kit (Amersham Biosciences, Uppsala, Sweden). Real-time quantitative RT-PCR was performed on a LightCycler-FastStart DNA Master SYBR Green 1 System (Roche Diagnostics, Basel, Switzerland). The specific primers used in the RT-PCR were as follows: for MMP-13, sense 5′-CATCCATCCCGTGACCTTAT-3′ and antisense 5′-GCATGACTCTCACAATGCGA-3′; for GAPDH, sense 5′-ACCACAGTCCATGCCATCAC-3′ and antisense 5′-TCCACCACCCTGTTGCTGTA-3′. MMP-13 gene expression levels were normalized to those of GAPDH in each cartilage sample.

Statistical analysis.

Results are presented as the mean ± SD. Significant differences between groups were determined using Student's unpaired t-test. One-way analysis of variance was conducted to examine the differences in mean values between multiple groups. Post hoc tests were used to identify potentially significant intragroup differences. The significance level was set at P values less than or equal to 0.05.

RESULTS

Histopathologic features of the sham-operated joints.

No gross histologic changes were seen in the knee articular cartilage at 8 weeks after sham operation in WT mice, nor were there significant structural alterations or loss of proteoglycans detected in the sham-operated WT joints (Figures 2A and C). Similarly, no significant OA changes were detected at 8 weeks after sham operation in OPN−/− mice (Figures 2B and D). In the sham-operated joints of WT and OPN−/− mice, the surface cartilage consisted of uncalcified and calcified zones underneath existing subchondral bone (indicated by the blue, red, and black bars, respectively, in Figure 2). The integrity of the tidemark (indicated by the arrowheads in Figures 2A–D) was well preserved, and thus there was clear separation of the uncalcified zones from the calcified zones in both WT and OPN−/− mice.

Figure 2.

Histopathologic features of the sham-operated knee joints of WT and OPN−/− mice. The sham-operated joints of WT mice (A, C, E, G, and I) and OPN−/− mice (B, D, F, and H) were harvested and evaluated by hematoxylin and eosin staining (A and B), Safranin O–fast green staining (C and D), immunohistochemistry for type II collagen (E and F), immunohistochemistry for type X collagen (G and H), and immunohistochemistry for OPN (I). Insets show higher-magnification views of the boxed areas (original magnification × 400). Arrowheads indicate the tidemark. Arrows indicate hypertrophic cells staining positive for type X collagen in the calcified zone of cartilage. Blue, red, and black bars in the margins indicate the uncalcified zone, calcified zone, and subchondral bone, respectively. Bar = 100 μm. See Figure 1 for definitions.

On immunohistochemical analysis, type II collagen, a cartilage-specific collagen, was strongly stained in both the uncalcified and calcified zones of the articular cartilage, both above and below the tidemark, but was not evident in the subchondral bone in the sham-operated joints of WT and OPN−/− mice (Figures 2E and F). Type X collagen, a specific marker of hypertrophic chondrocytes (indicated by the arrows in Figures 2G and H), was faintly stained in the calcified cartilage, below the tidemark, in the sham-operated articular cartilage of WT and OPN−/− mice. It should be noted that there was no significant difference in the staining of type II collagen and type X collagen between WT and OPN−/− mice. OPN was detected in the calcified zones but was absent in the uncalcified zones of surface cartilage after sham operation (Figure 2I).

Appearance of severe OA changes in OPN−/− mice in the instability-induced OA model.

We analyzed the development of instability-induced OA changes in WT and OPN−/− mice. In the joints of WT mice at 8 weeks after transection/menisectomy, OA changes developed as expected, which included a loss of the uncalcified zone (indicated by the blue bar in Figure 1C), a reduced number of chondrocytes, and alteration of the tidemark integrity (indicated by the arrowheads in Figure 1C) (see Figure 2A for comparison with sham-operated joints). In the absence of OPN, however, more severe OA changes were detected. In the operated joints of OPN−/− mice, the uncalcified zone was completely lost and the number of chondrocytes was significantly reduced (Figure 1D). Of note, cartilage erosion extended into the calcified zone, and thus the tidemark was completely lost (Figure 1D).

We quantified the following parameters of OA severity in the instability-induced OA model: cartilage erosion, chondrocyte number, and cartilage thickness (inversely related to a loss of the uncalcified zone), and comparisons between groups were made, as shown in Figure 3. The extent of cartilage erosion at 8 weeks after transection/meniscectomy was significantly higher in OPN−/− mice than in WT mice (mean ± SD erosion score 3.3 ± 0.5 in OPN−/− mice versus 2.3 ± 0.5 in WT mice; P < 0.01 [n = 6 per group]) (Figure 3A).

Figure 3.

OA severity in WT and OPN−/− mice in the instability-induced OA model compared with the sham-operated (Sham-Op) joints of these mice, as assessed by A, scores of cartilage erosion, B, number of chondrocytes, C, scores of proteoglycan loss by Safranin O staining, and D and E, cartilage layer thickness in the tibia. In D, the selected area for determination of cartilage layer thickness is marked by a thin black line. The same mice as used in the histopathologic experiments (shown in Figures 1 and 2) were analyzed. Bars show the mean and SD of 6 mice per group. ∗ = P < 0.05; ∗∗ = P < 0.01. NS = not significant; SCB = subchondral bone; M = meniscus (see Figure 1 for other definitions).

Cellularity of the cartilage was quantified to determine whether the OA changes in OPN−/− mice were associated with a loss of chondrocytes. The number of chondrocytes per defined cartilage area in the tibia was significantly lower in OPN−/− mice than in WT mice (mean ± SD 26.7 ± 4.1 in OPN−/− mice versus 41.3 ± 9.4 in WT mice; P < 0.05 [n = 6 per group]) (Figure 3B).

Finally, the thickness of the cartilage layer in the tibia was significantly reduced in OPN−/− mice as compared with that in WT mice (38.6 ± 10.3 μm in OPN−/− mice versus 57.0 ± 7.2 μm in WT mice; P < 0.05 [n = 6 per group]) (Figure 3E).

Another OA parameter, loss of proteoglycans as judged by a reduction in the area of positive staining by Safranin O, was also quantified. Although proteoglycan loss was evident in WT mice (Figure 1E) (see Figure 2C for comparison with sham-operated joints) at 8 weeks after transection/menisectomy, the same degree of proteoglycan loss was detected in OPN−/− mice (Figure 1F). Based on the results of Safranin O staining, there was no statistically significant difference in proteoglycan loss between the 2 groups (Figure 3C).

The staining pattern and intensity of staining for type II and type X collagens in the cartilage were indistinguishable between the 2 groups (Figures 1G–J). As compared with the cartilage of WT mice, a more severe reduction in the area that stained positive for type II collagen was observed in OPN−/− mice (Figure 1H), which could be attributable to the severe loss of the uncalcified zone (Figure 1D). The intensity of staining for type X collagen in the operated joints of WT and OPN−/− mice (Figures 1I and J) was much stronger than that in the sham-operated joints of each group (Figures 2G and H). Type X collagen was strongly stained in and around chondrocytes in the calcified zone in the operated joints of both WT and OPN−/− mice (Figures 1I and J). Moreover, expression of OPN was detected along the tidemark of OA joints (Figure 1K), which was significantly stronger than that in sham-operated joints (Figure 2I).

Appearance of OA changes in OPN−/− mice in the aging-associated spontaneous OA model.

In mice at 2 months of age, there were no significant OA changes in the knee articular cartilage of either WT mice or OPN−/− mice (n = 8 per group) (results not shown). In 15-month-old mice, slight, but significant, OA changes were detected in the knee joints of WT mice (Figure 4A) as compared with that in 4-month-old WT mice (Figure 2A). There was mild, superficial cartilage erosion in 15-month-old WT mice (Figure 4A). However, this was indistinguishable from that in 15-month-old OPN−/− mice (Figure 4B).

Figure 4.

Histopathologic features of the knee joints of WT and OPN−/− mice in the aging-associated OA model. Joints were harvested from WT mice (A, C, E, and G) and OPN−/− mice (B, D, F, and H) at 15 months of age and evaluated by hematoxylin and eosin staining (A and B), Safranin O–fast green staining (C and D), immunohistochemistry for type II collagen (E and F), and immunohistochemistry for type X collagen (G and H). Insets show higher-magnification views of the boxed areas (original magnification × 400). Arrowheads indicate the tidemark. Blue, red, and black bars in the margins indicate the uncalcified zone, calcified zone, and subchondral bone, respectively. Bar = 100 μm. See Figure 1 for definitions.

There was no statistically significant difference in the scores of cartilage erosion and femoral cartilage thickness (Figures 5A and D, respectively) between OPN−/− and WT mice at age 15 months. The tidemark integrity was well preserved in these aged WT and OPN−/− mice (Figures 4A and B). We observed a stronger reduction in the chondrocyte number in OPN−/− mice as compared with that in WT mice, and this difference was statistically significant (mean ± SD 71.9 ± 14.3 in OPN−/− mice versus 90.2 ± 13.0 in WT mice; P < 0.01 [n = 10 per group]) (Figure 5B). The reduction in chondrocyte number in OPN−/− mice was much milder in the aging-associated spontaneous OA model (Figure 5B) than in the instability-induced OA model (Figure 3B). The proteoglycan content in the femur, as evaluated by Safranin O staining, was significantly reduced in 15-month-old OPN−/− mice as compared with that in age-matched WT mice (Figures 4C and D) (mean ± SD score 3.2 ± 0.6 in OPN−/− mice versus 1.9 ± 0.9 in WT mice; P < 0.01 [n = 10 per group] [Figure 5C]).

Similar to the observations in the instability-induced OA model (Figures 1G and H), type II collagen stained strongly in all zones of the articular cartilage, both above and below the tidemark, in the aging-associated spontaneous OA model, and there was no apparent difference between OPN−/− and WT mice (Figures 4E and F). It should be noted that, unlike that in the instability-induced OA model (Figures 1H and I), type X collagen stained faintly in the calcified cartilage, below the tidemark, in the articular cartilage of both OPN−/− and WT mice in the aging-associated OA model (Figures 4G and H). In addition, the subchondral bone density of aged OPN−/− mice was significantly higher than that of aged WT mice (mean ± SD 1.10 ± 0.12 mmAL Eq. in OPN−/− mice versus 0.96 ± 0.08 mmAL in Eq. WT mice; P < 0.05 [n = 6 per group]) (Figures 5F and G). This suggests that subchondral bone sclerosis is more severe in OPN−/− mice. Osteophyte formation was not visible in the knee joints of either OPN−/− mice or WT mice (Figure 5E). These results indicate that the progression of spontaneous OA lesions is accelerated by OPN deficiency.

Augmented proteoglycan release from cultured cartilage explants derived from OPN−/− mice.

To understand the mechanism of the decreased proteoglycan levels in the OA joints of OPN−/− mice, we performed in vitro experiments. Cartilage explants were maintained in culture, and the release of proteoglycan into the medium was determined using a dimethylmethylene blue assay. Under in vitro conditions, a significant increase in total proteoglycan release was observed in OPN−/− cartilage compared with cartilage from WT mice (mean ± SD 3.9 ± 1.2 μg/ml sGAG/mg cartilage in OPN−/− mice versus 2.6 ± 1.1 μg/ml sGAG/mg cartilage in WT mice; P < 0.05 [n = 8 per group]) (Figure 6A). Histologic examination of the explants also demonstrated that Safranin O staining for proteoglycans was reduced in OPN−/− cartilage (Figure 6B).

Figure 6.

Proteoglycan release and matrix metalloproteinase 13 (MMP-13) expression in the joints of WT and OPN−/− mice. A, Total proteoglycan release was assessed in cultured articular cartilage obtained from OPN−/− and WT mice after 72 hours. Bars show the mean and SD level of sulfated glycosaminoglycan (sGAG) in the cartilage of 8 mice per group. B, Joint sections of cultured cartilage from OPN−/− and WT mice were stained with Safranin O–fast green; representative histologic findings are shown. Bar = 100 μm. C, Quantitative real-time reverse transcription–polymerase chain reaction was used to analyze MMP-13 mRNA expression on days 0 and 5 of articular cartilage degradation in OPN−/− and WT mice. D, Effects of MMP inhibitors (broad-spectrum GM6001 and 2 concentrations of an MMP-selective inhibitor) on the release of proteoglycan from cultured cartilage explants were assessed. Bars show the mean and SD level of sGAG in the cartilage of 3 mice per group. ∗ = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001. NS = not significant; Cont = uncultured control (see Figure 1 for other definitions).

Augmented expression of MMP-13 mRNA in articular cartilage of OPN−/− mice during in vitro culture.

To investigate the mechanisms underlying the accelerated development of OA changes in OPN−/− mice, we carried out cDNA SuperArray and real-time RT-PCR analyses. Among 18 distinct MMP genes examined in this study (MMPs 1a, 2, 3, 7–17, 19, 20, 23, and 24), the mRNA expression levels for the MMP-3 and MMP-13 genes were found to be elevated during in vitro culture. In OPN−/− mice, a significant up-regulation in the MMP-13 gene in cartilage was detected, as compared with that in WT mice.

Real-time RT-PCR analysis confirmed that cartilage MMP-13 mRNA expression was significantly increased in OPN−/− mice as compared with WT mice on day 5 of in vitro culture (mean ± SD 3.0 ± 0.8 in OPN−/− mice versus 1.1 ± 0.7 in WT mice; P < 0.05 [n = 3 per group]) (Figure 6C). There was no apparent difference in MMP-13 mRNA expression between WT mice and OPN−/− mice on day 0, indicating that the absence or presence of OPN does not simply determine the level of MMP-13 expression.

To determine whether MMP-13 expression is related to proteoglycan release from cartilage, cartilage explants were cultured in the presence of a selective MMP-13 inhibitor, and proteoglycan release into the medium was determined. The MMP-13 inhibitor significantly inhibited the release of proteoglycan at concentrations of both 10 nM and 100 nM (mean ± SD 2.57 ± 0.13 μg/ml sGAG/mg in control cartilage versus 2.04 ± 0.08 μg/ml sGAG/mg in cartilage treated with 10 nM of the selective inhibitor [P < 0.01] and 1.34 ± 0.02 in cartilage treated with 100 nM of the selective inhibitor [P < 0.001] [n = 3 per group]) (shown in Figure 6D as the amount of sGAG with and without inhibitors). A broad-spectrum MMP inhibitor, GM6001, further inhibited the release of proteoglycans when applied at a concentration of 10 μM, indicating the possible involvement of other MMPs in proteoglycan release from cartilage (mean ± SD 2.57 ± 0.13 μg/ml sGAG/mg in control cartilage versus 1.33 ± 0.02 μg/ml sGAG/mg in cartilage treated with GM6001; P < 0.001 [n = 3 per group]) (shown in Figure 6D as the amount of sGAG).

DISCUSSION

OPN is classified not only as an ECM component but also as a cytokine and is involved in various physiologic and pathologic conditions (13, 35–38). With regard to the association of OPN with cartilage degradation, it has been shown that the expression of OPN protein and mRNA increases along with the severity of OA (22). In addition, OPN has been known to act as a regulator of pathologic mineralization in articular cartilage (39). However, other functional roles of OPN in OA pathogenesis have not been elucidated.

In this study, we have clarified, in part, the functional role of OPN in the development of both instability-induced and aging-associated spontaneous OA. Lack of OPN resulted in significant progression of cartilage degradation in 2 distinct OA models. In the instability-induced OA model, structural alterations, reflected by the reduction in cartilage thickness, loss of the uncalcified zone, and a reduction in chondrocyte numbers, were significantly more severe in OPN−/− mice than in WT mice. In contrast, in the aging-associated OA model, significant proteoglycan loss in articular cartilage was evident, while surface cartilage structural integrity was well preserved in OPN−/− mice.

We previously demonstrated that OPN specifically binds to heparin sulfate proteoglycan (16), suggesting the possibility that the cartilage matrix fails to retain proteoglycan in the absence of OPN. Taken together, our results support the idea that OPN plays a pivotal role in the progression of both instability-induced and aging-associated spontaneous OA. However, more importantly, the mechanistic role of OPN during the development of OA might be different between instability-induced OA and aging-associated spontaneous OA.

The present study also demonstrated that in OPN−/− mice, MMP-13 gene expression was significantly up-regulated during in vitro cartilage degradation, as compared with that in WT mice. Our results are consistent with the previous finding that MMP-13 production is up-regulated in OA chondrocytes and synovial cells and is thought to play a critical role in cartilage degradation (40–43). We previously demonstrated that an anti-OPN antibody (M5), which interferes with the binding of OPN to integrin receptors, up-regulated MMP-13 expression in cultured tendon fibroblasts of WT mice (14). In addition, we found that MMP-13 is involved in proteoglycan release from cultured cartilage explants (Figures 6C and D). Collectively, these results suggest that OPN inhibits the up-regulation of MMP-13 during the process of cartilage degradation, and that this inhibitory function may, in turn, inhibit OA progression in vivo. Nevertheless, we could not rule out the possibility that those OA changes are indirect consequences of OPN deficiency and are part of the common cartilage degradation mechanisms that are triggered by diverse etiologic mechanisms.

Kamekura et al showed that MMP-13 expression was induced in instability-induced OA cartilage, and that MMP-13 was colocalized with type X collagen in hypertrophic chondrocytes (7). Those authors suggested that up-regulated expression of MMP-13 in abnormal hypertrophic chondrocytes plays a key role in cartilage degradation. In fact, we found that type X collagen, a specific marker of hypertrophic chondrocytes, exhibited strong staining in and around chondrocytes in the surface layer of cartilage of mice with instability-induced OA. However, type X collagen expression was not different between WT and OPN−/− mice in the instability-induced OA model (Figures 1I and J). In addition, type X collagen was only faintly stained in the surface layer of cartilage in the aging-associated spontaneous OA model (Figures 4G and H). Therefore, future studies are needed to elucidate how the expression of OPN, MMP-13, and type X collagen is related to each other.

This study is the first to describe the effects of OPN deficiency on OA pathogenesis using both instability-induced and aging-associated OA models. The results indicate that this protein plays an important role in preventing OA progression. However, it appears that OPN is a critical intrinsic regulator of cartilage degradation via its inhibitory effect on MMP-13 expression on chondrocytes and is a potential molecule for preventing cartilage degradation. Future studies will be needed to develop a novel strategy for the treatment of OA based on the emerging evidence.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Uede had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Yuichiro Matsui, Iwasaki, Denhardt, Rittling, Uede.

Acquisition of data. Yuichiro Matsui, Kon, Takahashi, Morimoto, Yutaka Matsui.

Analysis and interpretation of data. Yuichiro Matsui, Iwasaki, Minami.

Acknowledgements

The authors thank Dr. Yuichiro Abe for his assistance in preparation of the figures.

Ancillary