Materials and Methods

Sample population—Articular cartilage from grossly normal joints and nasal septum cartilage were aseptically harvested from 24 equine cadavers within 12 hours after the horses were euthanized. The horses comprised 6 foals (< 12 months old), 6 yearlings (12 to 24 months old), and 12 adults (2 to 5 years old). Horses did not have evidence of sepsis or musculoskeletal disease.

Cell culture—Articular cartilage obtained from 11 of the equine cadavers was minced with a scalpel and digested overnight in basal digestion media (70mM NaCl, 30mM KCl, 3mM K₂HPO₄, 1mM CaCl₂, 10mM NaHCO₃, 60mM sorbitol, 5 mg of dextrose/mL, and 1 mg of albumin/mL in 25mM HEPES^a [pH, 7.2] supplemented with 30μM tosyl-lysyl-chloromethane^a) and 0.5% collagenase-D.^b After digestion, the isolated chondrocytes were rinsed with PBS solution and counted. Viability was assessed by use of trypan blue staining, and an experiment was continued only when the proportion of viable cells was > 95%. Isolated chondrocytes (10 × 10⁶) were distributed in 30-mm culture dishes treated with poly-2-hydroxyethyl methacrylate^b to prevent cell adhesion and maintain cartilage phenotype.³¹ The suspension cultures were maintained at 37°C with 5% carbon dioxide at saturated humidity in high-glucose Dulbecco modified Eagle media, 10% fetal calf serum, gentamicin (50 μg/mL), amphotericin B (0.25 μg/mL), and ascorbic acid (50 μg/mL). After incubation for 24 hours, the cells were washed twice with serum-free media, which was followed by incubation in serum-free media containing the various treatments. The cells were treated with human recombinant IL-1β^c (5 ng/mL) alone, dexamethasone (10⁻⁶ or 10⁻⁷M) or triamcinolone (10⁻⁶ or 10⁻⁷M) alone, and IL-1β and dexamethasone (10⁻⁶ or 10⁻⁷M) or triamcinolone (10⁻⁶ or 10⁻⁷M). After incubation with the treatment for 24 hours, cells and spent media were harvested.

Explants of nasal septum cartilage—Explants of nasal septum cartilage were obtained from 18 of the equine cadavers. After the skull of each equine cadaver was sectioned longitudinally, the septum was removed and the mucosal surfaces were washed extensively with 2% chlorhexidine solution and rinsed with sterile PBS solution. The mucous membrane was removed, and a 6-mm biopsy punch was used to harvest cartilage disks. The disks were washed 3 times in PBS solution that contained antimicrobials. Viability was assessed via ethidium homodimer–calcein staining,^d and > 90% viable cells was required for further processing of the remaining disks.

The disks were maintained under the same culture conditions as those described for the aforementioned articular cell culture. After 48 hours in culture, the disks were washed twice with serum-free media; they then were incubated in fresh serum-free media containing the various treatments, similar to procedures used for the articular cell cultures. After incubation with the respective treatments for 72 hours, the disks and spent media were harvested. Each disk was blotted dry by use of a standard technique (3 s/side on a fresh paper towel), weighed, and minced. The explants were extracted by incubation for 3 to 5 days in a solution consisting of 1.2 mL of 4M guanidinium HCl and 0.05M sodium acetate (pH, 6.8) that contained protease inhibitors (0.01M EDTA and 0.1M 6-amino hexanoic acid), 0.005M benzamidine HCl, and 0.01M n-ethylmaleimide. Duplicates of explants from 8 equine cadavers were stored in liquid nitrogen at −80°C for RT PCR analysis.

Articular cartilage explants—Articular cartilage explants were obtained from grossly normal stifle joints of 7 equine cadavers. The explants were obtained aseptically with a 6-mm biopsy punch, after which the cartilage was separated from the subchondral bone. Processing, viability assessment, culture conditions, treatments, harvesting, and extraction of GAG were similar to those for nasal septum cartilage explants. Duplicates were obtained and stored in liquid nitrogen at −80°C for RT PCR analysis.

Quantitation of GAG—The GAG content of media and cartilage extracts was measured by use of a dimethylmethylene blue assay. Chondroitin sulfate-C from shark cartilage was used as a standard.³²

Northern blot analysis—Cells were solubilized in monophasic acid phenol.^e Total RNA was extracted, quantitated in a UV spectrophotometer, and separated in 1% denaturing agarose gels (10 μg/lane). The gels were capillary blotted to charged nylon membranes^f and UV cross-linked. Random primed [³²P] deoxycytidine triphosphate–labeled probes were prepared by use of cDNA for MMP-3, MMP-13, ADAMTS4, ADAMTS5, and the housekeeping gene EF1. A commercial buffer^g was used for prehybridization and hybridization at 65°C for 30 minutes and 12 hours, respectively; membranes were then washed twice for 20 minutes at 55°C with 0.5X and 0.25X sodium citrate solution with 0.1% SDS. Membranes were exposed in phosphor-storage cassettes,^h and images were quantified by use of imaging software.ⁱ

Real-time RT PCR assay—Frozen cartilage explants were pulverized with a freezer mill^j and solubilized in monophasic acid phenol,^e and then RNA was extracted. Real-time RT PCR assay^k was performed. The PCR products were detected with a probe format^l on the basis of fluorescence resonance energy transfer by use of 2 sequence-specific probes labeled with different dyes. Primers and probes (Appendix) were designed by use of software^m on the basis of equine ADAMTS4 and ADAMTS5 previously cloned by one of the authors (DWR). Primers and probes for the housekeeping gene equine EF1a were designed from a sequence obtained from the National Center for Biotechnology Information database (accession No. AY237113). Data were quantified by use of a calibrator-normalized, relative quantification methodⁿ that resulted in a normalized ratio for expression of the target gene to the housekeeping gene for each sample.

Western blot analysis—The western blot analysis was conducted as described elsewhere.³³ Briefly, spent media were stored at −20°C after addition of the protease inhibitor phenylmethylsulphonyl fluoride (100mM). Media from the cartilage explants were lyophilized, dissolved in 500 μL of 100mM Tris-acetate (pH, 6.5), and deglycosylated with chondroitinase ABC^b (0.01 U/10 μg of GAG), keratanase^o (0.01 U/10 μg of GAG), and keratanase II^o (0.0001 U/10 μg of GAG) for 2 hours at 37°C. Samples were dialyzed exhaustively against water by use of 12,000 to 14,000 molecular weight dialysis tubing and lyophilized. After reconstitution in loading buffer with 10% β-mercaptoethanol, samples were heated at 100°C for 5 minutes, and gels were loaded on an equal GAG basis (4 μg of GAG/lane; determined by use of dimethylmethylene blue assays). Samples were separated in a 4% to 12% gradient polyacrylamide gel (assayed at 70 mA) and electroblotted overnight at 21 V to polyvinylidene difluoride membranes.^p Membranes were washed in TBSTW and blocked by incubation in 0.1% TBSTW with 5% nonfat milk for 1 hour at 21°C. After 3 washes in TBSTW, the membranes were incubated with primary antibody (mouse monoclonal antibody BC-3^q at a dilution of 1:100) for 2 hours at 21°C. The BC-3 antibody recognizes the new N-terminal sequence ARGSVIL, which is formed after cleavage of aggrecan within the IGD (Glu³⁷³-Ala³⁷⁴ site) by aggrecanase.³¹ After 3 washes with TBSTW, secondary antibody (horseradish peroxidase–conjugated goat anti-mouse IgG^r at a dilution of 1:10,000) was added and membranes were incubated for 1 hour at 21°C. Development conjugate dilution buffer was added, and the chemiluminescent procedure as described for a commercially available system^s was used. Chemifluorescent detection was performed thereafter with fluorescent scanning on an imager.^h

ELISA for measurement of aggrecanase activity—Spent media were stored at −80°C after addition of the protease inhibitor leupeptin hemisulfate (10mM). A sensitive aggrecanase activity ELISA^t was used in accordance with the manufacturer's instructions. Duplicate samples (5 μL) of spent media were incubated with a human recombinant aggrecan substrate for 15 minutes at 37°C in a 96-well plate containing a monoclonal antibody that recognized the new N-terminal sequence ARGSVIL; proteolytic reactions were stopped by the addition of 150 μL of EDTA dilution buffer, and samples then were incubated for 90 minutes at 25°C. Wells were emptied, washed 3 times with wash buffer, and incubated with a secondary antibody (antibody peroxidase conjugate) for an additional 90 minutes at 25°C. Wells were emptied and washed 5 times; 100 mL of detection solution^u was added, and wells were incubated in the dark at 21°C for 30 minutes. Peroxidase reactions were stopped by the addition of 100 μL of sulfuric acid. Absorbance measurements were determined in a microplate reader at a wavelength of 450 nm, and the amount of N-terminus peptide was calculated from a standard curve³⁴ and compared with the amount for the control treatment.

Statistical analysis—Data were analyzed by use of a mixed-model ANOVA.^v Cartilage type (articular or nasal), gene (MMP-3, MMP-13, ADAMTS4, ADAMTS5, or EF1), and treatment (IL-1β, IL-1β with 10⁻⁶ or 10⁻⁷M dexamethasone, IL-1β with 10⁻⁶ or 10⁻⁷M triamcinolone, 10⁻⁶ or 10⁻⁷M dexamethasone, or 10⁻⁶ or 10⁻⁷M triamcinolone) were the class variables. Dependent variables were micrograms of GAG in media per milligrams of ww cartilage; micrograms of GAG per milligrams of ww cartilage; ARGSVIL (ie, aggrecanasegenerated epitope) peptide per milligrams of ww cartilage (compared with results for the control treatment); and expression ratios of ADAMTS4 to EF1, ADAMTS5 to EF1, MMP-3 to EF1, and MMP-13 to EF1 (compared with results for the control treatment). A plot of the residual versus predicted values was used to test for equality of variance and normal distribution of the data. When there was an increase in variance with higher values, data were normalized via logarithmic transformation (natural logarithm [base e]) and reanalyzed. Back transformation of data was performed, and values were reported as mean ± SD. When the association between outcome and treatment was significant, individual treatments were compared by use of the F-protected least significance difference method. Significance was defined as values of P < 0.05.

Results

Cartilage specimens—Cartilage (nasal, articular, or both) was obtained from 24 equine cadavers. Mean ± SD weight of the nasal explants was 0.071 ± 0.022 g of ww cartilage, and mean weight of the articular explants was 0.068 ± 0.017 g of ww cartilage.

Quantitative analysis of GAG content in spent media and cartilage—The GAG content of spent media, nasal cartilage explants, and articular cartilage explants was measured in samples obtained from 24 (19 nasal and 8 articular), 18, and 6 equine cadavers, respectively (Figures 1 and 2). Treatment with IL-1β significantly increased the GAG content in the spent media for both cartilage types, compared with the GAG content for the control treatment; there was a concurrent decrease of GAG content in the cartilage explants. Addition of glucocorticoids to IL-1β–treated cartilage did not significantly change GAG content in media or cartilage. Glucocorticoids alone significantly increased GAG release in spent media from both cartilage types, but the release was still significantly (P < 0.001) less than the GAG release after IL-1β treatment. Glucocorticoids alone did not significantly alter the GAG content in cartilage. Articular and nasal cartilage responded similarly to treatment, but the loss of GAG content in articular cartilage was not as pronounced, probably because of uneven guanidinium HCl extraction, and no significant difference was detected. There was no significant association between type or dose of glucocorticoid with GAG content in spent media or cartilage.

View Full Size

Mean ± SD GAG content in spent media for articular (light gray bars) and nasal (dark gray bars) cartilage explants treated with IL-1β alone, IL-1β and glucocorticoids, or glucocorticoids alone for 72 hours. Data reported represent results for 27 experiments performed on samples obtained from 24 equine cadavers. Nasal cartilage explants had a more pronounced response to treatment with IL-1β, compared with the response for articular cartilage; therefore, results were analyzed for each cartilage type separately. Treatment with IL-1β resulted in a significant (P < 0.001) increase in GAG content in the media, compared with that for the control treatment, which was not significantly altered by addition of glucocorticoids. Treatment with glucocorticoids alone resulted in an increase in GAG content in the media, which was significantly (P < 0.001) more than the value for the control treatment but significantly (P < 0.001) less than the increase after treatment with IL-1β. Both nasal and articular cartilage explants responded similarly to treatment. ^a–cWithin nasal cartilage explants, values with different letters differed significantly (P < 0.001). ^A–CWithin articular cartilage explants, values with different letters differed significantly (P < 0.001). Dex = Dexamethasone. Tri = Triamcinolone.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.176

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Mean ± SD GAG content in articular and nasal cartilage explants in response to treatment with IL-1β alone, IL-1β with glucocorticoids, and glucocorticoids alone for 72 hours. Data represent results of 18 nasal and 6 articular cartilage experiments performed on samples obtained from 24 equine cadavers. Nasal and articular cartilage explants were analyzed separately. Treatment with IL-1β resulted in a decrease in GAG content in nasal cartilage, compared with that for the control treatment, which was not significantly altered by addition of glucocorticoids. The GAG content in cartilage after treatment with glucocorticoids alone was not significantly altered, compared with GAG content for the control treatment. Articular and nasal cartilage had a similar response to treatment, but the response to treatment with IL-1β was significantly (P < 0.001) different only for nasal cartilage. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.176

• Download Figure
• Download figure as PowerPoint slide

Northern blot analysis—A total of 47 northern blots were performed with RNA obtained from 11 equine cadavers (Figure 3). There was a significant (all values, P < 0.01) association between treatment and expression of MMP-3, MMP-13, ADAMTS4, and ADAMTS5. Treatment with IL-1β for 24 hours resulted in significant upregulation of MMP-3, MMP-13, and ADAMTS4 mRNA (Figure 4). Treatment with IL-1β did not increase ADAMTS5 mRNA. Addition of glucocorticoids to IL-1β–treated chondrocytes decreased the expression of MMP-3 and MMP-13 mRNA, independent of type and concentration of glucocorticoid. Expression of ADAMTS4 was low in normal cartilage, which made it difficult to detect differences for the northern blots. Quantification and statistical analysis performed on control- and IL-1β–treated chondrocytes revealed that ADAMTS4 expression was significantly (P = 0.04) increased with IL-1β. Addition of glucocorticoids to IL-1β–treated chondrocytes resulted in a significant upregulation of ADAMTS5 mRNA, compared with upregulation for IL-1β treatment alone. Treatment with glucocorticoids alone resulted in a significant increase in ADAMTS5 mRNA and downregulation of MMP-3 and MMP-13 mRNA. We did not detect a significant difference between type or dose of glucocorticoid and gene expression.

View Full Size

Gene expression of MMP-13, MMP-3, ADAMTS4, and ADAMTS5 detected by use of northern blot analysis of chondrocytes in cell cultures treated with IL-1β alone, IL-1β with glucocorticoids, or glucocorticoids alone for 24 hours. Northern blot analysis of MMP-3 was performed in 15 experiments (n = 8 equine cadavers), MMP-13 analysis was performed in 10 experiments (8), ADAMTS4 analysis was performed in 9 experiments (7), and ADAMTS5 analysis was performed in 15 experiments (9). Retinoic acid was used as a positive control sample, but data were not included in the final analysis. Expression of MMP-13, MMP-3, and ADAMTS4 was increased significantly (P < 0.044) after treatment with IL-1β, whereas ADAMTS5 expression was unchanged. Addition of glucocorticoids significantly (P < 0.003) downregulated MMP-13 and MMP-3 expression, whereas ADAMTS5 expression was significantly (P = 0.004) upregulated. Treatment with glucocorticoids alone significantly (P < 0.001) decreased expression of MMP-13 and MMP-3 and caused significant (P < 0.02) upregulation of ADAMTS5 expression. Baseline gene expression of ADAMTS4 was low, which made it more difficult to interpret the northern blots. kb = Kilobase.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.176

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Mean ± SD expression of MMP-3 (15 experiments [n = 8 equine cadavers]; A), MMP-13 (10 experiments [8]; B), and ADAMTS5 (15 experiments [9]; C) determined by use of northern blot analysis performed on chondrocytes in cell culture after treatment with IL-1β alone, IL-1β with glucocorticoids, or glucocorticoids alone for 24 hours. Values reported represent the fold increase, compared with expression for the control treatment. Treatment with IL-1β resulted in a significant (P < 0.044) upregulation of MMP-3 (A) and MMP-13 (B) expression, compared with results for the control treatment, whereas ADAMTS5 expression was unchanged (C). Addition of glucocorticoids significantly (P < 0.003) decreased the expression of MMP-3 (A) and MMP-13 (B) and resulted in significant (P = 0.004) upregulation of ADAMTS5 expression (C), compared with results for IL-1β alone. Treatment with glucocorticoids alone resulted in significant (P < 0.001) downregulation of MMP-3 (A) and MMP13 (B) expression, whereas ADAMTS5 expression was significantly (P < 0.02) upregulated (C). There was no significant association between gene expression and type or dose of glucocorticoid. ^a–eWithin a gene, values with different superscript letters differ significantly (P < 0.02). See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.176

• Download Figure
• Download figure as PowerPoint slide

Real-time RT PCR assay—Real-time RT PCR assay of ADAMTS4 (Figure 5) and ADAMTS5 (Figure 6) was performed on 8 nasal and 7 articular cartilage explants; all explants had a similar response to treatment. Expression of ADAMTS4 was low, whereas ADAMTS5 was constitutively expressed. Treatment with IL-1β resulted in significant (P < 0.001) upregulation of ADAMTS4 and ADAMTS5 mRNA. The increase of ADAMTS4 mRNA was significantly (P = 0.008) higher than the increase of ADAMTS5 mRNA. Addition of glucocorticoids to IL1β–treated explants resulted in significant (P = 0.002) downregulation of ADAMTS4 mRNA, but gene expression remained significantly (P < 0.001) higher than values for the control treatment. In contrast to ADAMTS4 expression, ADAMTS5 expression was not diminished by the addition of glucocorticoids to IL-1β–treated cartilage. Treatment with glucocorticoids alone did not significantly alter gene expression of ADAMTS4, whereas ADAMTS5 mRNA was significantly increased. Although ADAMTS5 mRNA concentrations were increased by glucocorticoids alone, the values remained significantly lower than those for cartilage treated with IL-1β. There was no significant association between expression of ADAMTS4 or ADAMTS5 and type or dose of glucocorticoid.

View Full Size

Mean ± SD expression of ADAMTS4, determined by use of a real-time RT PCR assay, in articular and nasal cartilage explants following treatment with IL-1β alone, IL-1β with glucocorticoids, or glucocorticoids alone for 72 hours. Data represent results of 7 articular cartilage experiments (n = 7 equine cadavers) and 8 nasal cartilage experiments (8). Treatment with IL-1β resulted in a significant (P < 0.001) increase in expression of ADAMTS4, compared with the response for the control treatment. Addition of glucocorticoids resulted in a significant (P < 0.001) downregulation, but expression was still significantly (P < 0.001) higher than values for the control treatment. Treatment with glucocorticoids alone did not significantly alter ADAMTS4 gene expression. †Nasal and articular cartilage had a similar response to treatment; therefore, a single letter represents significant differences for both cartilage types for a specific treatment. ^a–cValues with different superscript letters differ significantly (P < 0.001). See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.176

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Mean ± SD expression of ADAMTS5, determined by use of a real-time RT PCR assay, in articular and nasal cartilage explants following treatment with IL-1β alone, IL-1β with glucocorticoids, or glucocorticoids alone for 72 hours. Data represent results of 7 articular cartilage experiments (n = 7 equine cadavers) and 8 nasal cartilage experiments (8). Treatment with IL-1β significantly (P < 0.001) increased expression of ADAMTS5, compared with values for the control treatment. Addition of glucocorticoids did not significantly alter the expression of ADAMTS5, but treatment with glucocorticoids alone significantly (P < 0.001) increased the expression of ADAMTS5, compared with values for the control treatment. Expression of ADAMTS5 after treatment with glucocorticoids alone was still significantly (P < 0.001) less than the expression after IL-1β treatment. See Figures 1 and 5 for remainder of key.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.176

• Download Figure
• Download figure as PowerPoint slide

Western blot analysis—Western blot analysis with monoclonal antibody BC-3 was performed on spent media of nasal and articular cartilage (Figure 7). Gels were loaded on an equal GAG basis to identify changes in cleavage patterns of the fragments released. The band at 62 to 70 kDa is C-terminally truncated, ADAMTS-cleaved aggrecan released in both control and treated cultures. Treatment with IL-1β increased the fragments with larger molecular weight (less C-terminally truncated and therefore more important in cartilage degradation). Combined with an increase in GAG release in the media caused by IL-1β, these results were indicative of an increase in aggrecanase-mediated cleavage within the IGD. Addition of glucocorticoids in IL-1β–treated cartilage caused no change (articular) or a small increase (nasal) in density of bands, which indicated no change (articular) or an increase (nasal) in aggrecanase activity. However, the high density of bands evident after IL-1β treatment may have prevented detection of a further increase in articular cartilage. With glucocorticoids alone, the density of mainly large molecular size bands increased in both cartilage types, which indicated an increase in aggrecanase activity, compared with results for the control treatment.

View Full Size

Aggrecanase-generated IGD metabolites detected by use of western blot analysis in nasal (A) and articular (B) cartilage explants after treatment with IL-1β alone, IL-1β with glucocorticoids, or glucocorticoids alone for 72 hours. Western blot analysis was performed in 7 nasal cartilage experiments (n = 7 equine cadavers) and 2 articular cartilage experiments (2). Retinoic acid was used as a positive control sample, but this treatment was not performed in all experiments; therefore, data were not included in the final analysis. Gels were loaded on an equal GAG basis to identify a change in the cleavage pattern of the fragments released; results were interpreted in combination with GAG data. The pronounced band at 62 to 70 kDa is C-terminally truncated, ADAMTS-cleaved aggrecan released in both control and treated cultures. A heterogeneous population of aggrecan exists in cartilage because of C-terminal truncation, which after cleavage of aggrecan, results in catabolites of different molecular sizes (45 to 250 kDa). Treatment with IL-1β increased the fragments with larger molecular weight (less C-terminally truncated and therefore more important in cartilage degradation). In combination with an increase in GAG release in the media caused by IL-1β, these results were indicative of an increase in aggrecanase-mediated cleavage within the IGD to cause the GAG release. Addition of glucocorticoids revealed that BC-3 catabolites were mildly increased (A) or not diminished (B). Glucocorticoids alone resulted in a distinct increase of BC-3–positive fragments, which indicated an increase in aggrecanase activity. Articular and nasal cartilage had a similar response to treatment. Values reported represent the number of kilodaltons. MWM = Molecular weight marker. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.176

• Download Figure
• Download figure as PowerPoint slide

ELISA—For further quantification of total production of aggrecanase-generated neoepitope unrelated to size, an ELISA was performed on 4 nasal and 5 articular cartilages. The amount of ARGSVIL-detected peptide was the combined product of residual aggrecanase activity in spent media and aggrecan neoepitopes generated by treatments, which represented overall aggrecanase activity. Both cartilage types responded similarly to treatments. Treatment with IL-1β significantly (P < 0.001) increased aggrecanase activity in both cartilage types, compared with results for the control treatment (Figure 8). Addition of triamcinolone to IL-1β–treated cartilage significantly (P = 0.03) increased aggrecanase activity, compared with results for IL-1β alone. Addition of dexamethasone to IL-1β–treated cartilage resulted in a mild (but not significant) increase in aggrecanase activity. Glucocorticoids alone did not result in a significant change in aggrecanase activity, nor was there any significant effect of type of cartilage, type of glucocorticoid, or dose.

View Full Size

Mean ± SD aggrecanase-generated ARGSVIL peptide production detected with an ELISA in spent media of articular and nasal cartilage explants after treatment with IL-1β alone, IL-1β with glucocorticoids, or glucocorticoids alone for 72 hours. Data represent results of 5 articular cartilage experiments (n = 5 equine cadavers) and 4 nasal cartilage experiments (4). Values reported represent the fold increase, compared with production for the control treatment. Response to treatment did not differ significantly between nasal and articular cartilage; therefore, results were analyzed for both cartilage types combined. Treatment with IL-1β significantly (P < 0.001) increased aggrecanase-generated peptide, compared with values for the control treatment. Treatment with IL-1β and glucocorticoids caused an increase in aggrecanase activity, compared with results for IL-1β treatment alone, but this was a significant (P < 0.03) difference for only triamcinolone. Treatment with glucocorticoids alone caused a mild increase in generated ARGSVIL peptide, but this was not significantly different from the value for the control treatment. ^a–dValues with different superscript letters differ significantly (P < 0.03) between treatments for both cartilage types combined. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 71, 2; 10.2460/ajvr.71.2.176

• Download Figure
• Download figure as PowerPoint slide

Discussion

Treatment with IL-1β resulted in an increase in GAG release in the media with a concurrent increase in aggrecanase activity determined by use of results of ELISA and western blot analysis, which is consistent with results in other reports.^3,4,20 Even though aggrecanase activity has been reported to be a major contributor to GAG degradation,^3,4,6,7 the exact contribution of MMPs and aggrecanases to GAG degradation is unknown. In the study reported here, transcription of both MMPs and aggrecanases was significantly upregulated by IL-1β. However, IL-1β treatment resulted in a greater increase in ADAMTS4 expression (mean, 27-fold increase) than in ADAMTS5 expression (mean, 13-fold increase), which indicated differences in transcriptional regulation. Results of studies^{20–25,35,36} in which investigators have reported the influence of IL-1 on gene expression of aggrecanases are inconsistent. Our results suggested that both ADAMTS4 and ADAMTS5 contribute to GAG degradation in IL-1β–treated cartilage. It has been proposed^27–29 that ADAMTS5 is the major aggrecanase in osteoarthritis in vivo, but other studies have revealed that there can be aggrecanolysis in ADAMTS5-knockout mice²⁵ and that ADAMTS4 contributes to aggrecan degradation.^30–37 These results indicate that the exact role of each aggrecanase in cartilage degradation remains to be determined.

Glucocorticoids have the potential to alter the rate of proteoglycan synthesis and degradation.^38–47 Content of GAGs increases in synovial fluid of normal joints after administration of glucocorticoids, which indicates an increase in aggrecan turnover,^38,39 but the effects in normal and abnormal joints differ.^41,42 Aggrecan turnover may be influenced by proteoglycan degradation and proteoglycan synthesis. Several studies^40,42–44 have revealed a decrease in proteoglycan synthesis after glucocorticoid treatment, but no change or an increase in synthesis has also been reported.^39,45,46 The effects of glucocorticoids on proteoglycan degradation are also inconsistent.^44,47 Minimal differences between types of glucocorticoids have been reported,⁴⁷ but results are not consistent⁴⁸ and comparative studies are lacking.

Analysis of our results revealed that addition of glucocorticoids to cytokine-stimulated cartilage did not counteract the negative effects of IL-1β on GAG degradation, and treatment with glucocorticoids alone increased GAG release in the media. There was no significant difference on the basis of type of glucocorticoid (dexamethasone vs triamcinolone). Treatments with glucocorticoids resulted in a decrease in MMP expression similar to that described in another report,⁴⁸ whereas aggrecanase activity determined by use of an ELISA and western blot analysis was increased. Degradation of GAG was unchanged (with IL-1β) or increased (without IL-1β), which indicated a change in relative contribution of MMPs and aggrecanases. Although changes at a transcriptional level do not necessarily reflect similar changes at the protein level, the increase in aggrecanase activity combined with a decrease in transcription of ADAMTS4 are suggestive of a major role of ADAMTS5. Treatment with glucocorticoids alone resulted in an increase in aggrecanase activity (determined by use of western blot analysis), whereas ADAMTS4 mRNA was unchanged and ADAMTS5 mRNA was upregulated, which further substantiated a major role of ADAMTS5.

Potential mechanisms of action of glucocorticoids on aggrecanases include a difference in regulation through transcription factor NF-κB. Regulation of transcription of several MMPs is dependent on NF-κB,^49,50 but information about involvement of NF-κB on transcription of ADAMTS4 and ADAMTS5 is limited. A report¹⁹ for human synovial fibroblasts indicates that ADAMTS4 expression is dependent on NF-κB, whereas ADAMTS5 expression is not. Glucocorticoids can inhibit NF-κB activity by direct and indirect mechanisms.^51–54 This could cause inhibition of gene expression of MMPs and ADAMTS4, whereas ADAMTS5 expression is unaffected. In contrast, inhibition of NF-κB in bovine cartilage explants did not alter expression of either ADAMTS4 or ADAMTS5,²⁴ and the authors of that study proposed a potential role of NF-κB in ADAMTS5 activity. It is evident that further investigation of the role of NF-κB is necessary to determine its influence on expression and activity of ADAMTS4 and ADAMTS5.

Another possible mechanism of influence of glucocorticoids on aggrecanase activity is the influence of Mt4-MMP (a glycosylphosphatidylinositol-anchored MMP) and possibly other MMPs on activation of ADAMTS4. Aggrecanases contain a zinc catalytic domain followed by a disintegrin domain, a thrombospondin domain, a cysteine-rich domain, and a spacer domain (in ADAMTS5, a second thrombospondin domain follows the spacer domain). Noncatalytic ancillary domains play a role in regulating enzyme activity, and deletion of the spacer domain results in an increase in activity of ADAMTS4.¹⁵ Increased activity of ADAMTS4 has been reported^17,55 after Mt4-MMP–mediated processing, which results in an ADAMTS4 protein of similar molecular size as that for a C-terminal spacer domain deletion mutant. The spacer domain does play a role in ADAMTS5 activity, but deletion of the cysteine-rich domain causes the main change in activity.¹⁶ In addition, a study²⁶ in transgenic mice revealed that ADAMTS5 activity is not controlled by Mt4-MMP. The effect of glucocorticoids on Mt4-MMP is unknown, but our results and the results from other studies^48,56 indicate that most MMPs are downregulated with glucocorticoids. A possible decrease in Mt4-MMP with glucocorticoids can result in a decrease in ADAMTS4 activity, whereas ADAMTS5 activity is not affected.

A third possibly relevant mechanism involves the interaction between fibronectin and ADAMTS4. Glucocorticoids can alter the expression of fibronectin,^48,57 and binding of fibronectin to the spacer domain of ADAMTS4 results in inhibition of ADAMTS4 activity.¹⁴

Assays for aggrecanase activity used in the study reported here were based on detection of aggrecanase-generated epitopes within the IGD. The major aggrecanase in cleavage within the IGD is ADAMTS5,²⁵ and the overall activity of ADAMTS5 has been reported to be 1,000-fold as high as that for ADAMTS4¹⁶; therefore, results of our assays may be more representative for ADAMTS5 activity than for ADAMTS4 activity. This would illustrate the potential of glucocorticoids to increase ADAMTS5 activity. Results of the BC-3 western blot analysis and ELISA varied, but they were generally consistent. Western blot analysis revealed an increase primarily in the larger fragments after treatments, whereas the ELISA detected the total amount of generated fragments unrelated to size. The ELISA samples were not deglycosylated; therefore, it is possible that larger fragments (chondroitin-sulfate and keratan-sulfate rich) were not captured and that the results were more representative for an increase in smaller ARGSVIL aggrecan fragments. This could explain minor variations between results for the western blot analysis and ELISA, but further investigation and identification of detected fragments are required. Other explanations could include the difference in aggrecan substrate (equine native vs human recombinant), but results of another study⁵⁸ indicate that subtle changes in aggrecan sequences between human and equine aggrecan do not result in loss of immunoreactivity. A third explanation could be differences in aggrecanase activity in situ versus in solution.

Differences were detected between results for northern blot analysis of cell cultures and real-time RT PCR assays of cartilage explants. The reason for these differences is unclear, but they were most likely attributable to a difference in culture and technique. Transcriptional changes in cartilage explants may be more representative of changes in vivo than are changes for isolated cell cultures, but the overall results clearly indicated that both IL-1β and glucocorticoids affected ADAMTS4 and ADAMTS5 differently.

One limitation of this study was the use of aggrecanase assays that only detected aggrecanase activity at the classic site within the IGD. Both aggrecanases have aggrecanolytic activities at sites within the chondroitinsulfate–rich domain of aggrecan,^59,60 and additional assays with antibodies specific for these sites would have provided additional information. A second limitation was that the relative contribution of ADAMTS4 and ADAMTS5 in aggrecanase activity was suggested on the basis of changes at the transcriptional level; additional assays to determine protein concentrations would provide further information. Our attempts to use western blot analyses for the ADAMTS proteins were unsuccessful. A third potential limitation was the variation in age of horses used for the study. Glycosylation of aggrecan plays a role in aggrecanase activity,^15,16,59 and glycosylation changes with age,⁶¹ which indicates that age may be a factor in aggrecanase-mediated cartilage degradation.¹⁰ Furthermore, the dose range of glucocorticoids used was too limited to detect a potential dose-dependent effect, which has been reported for other studies.^40,44,48,56 These doses were selected because they are within the therapeutic range and preliminary experiments revealed effects on MMP and ADAMTS5 transcription.

Matrix metalloproteinase-3, MMP-13, and ADAMTS4 are regulated differently than ADAMTS5. Content of ADAMTS5 mRNA was increased by IL-1β, but significantly less than the increase for ADAMTS4 mRNA, which suggested a role for both ADAMTS4 and ADAMTS5 in degradation of cytokine-stimulated cartilage. Despite downregulation of MMP-3 and MMP-13 mRNA, glucocorticoids did not alter GAG degradation. Aggrecanase activity was increased with glucocorticoids, whereas ADAMTS4 mRNA was downregulated and ADAMTS5 gene expression was maintained or increased, which suggested a major role of ADAMTS5 in aggrecanase activity. The effect of glucocorticoids on aggrecanase activity has major implications in terms of treatment and helps to explain why degradation of matrix takes place despite inhibition of MMPs and ADAMTS4.