Materials and Methods

The DN powder was prepared by combining NZGLM (Perna canaliculus), abalone (Haliotis sp), SKC (Galorhinus galeus), BO, and honey and flavoring (< 5%) according to a proprietary formulation.^b Biota orientalis was stored at 4°C and warmed to room temperature (approx 21°C) before preparation of the DN powder and BO_sim.

Simulated digestion—The DN_sim, BO_sim, and INDO_sim (indomethacin used as a reference anti-inflammatory drug) were prepared as previously described.⁴ Briefly, DN (0.85 g), BO (2.5 mL; equivalent to 0.85 g of BO), and indomethacin (0.71 g) were agitated individually for 2 hours at 37°C in 35 mL of simulated gastric fluid (37mM NaCl, 0.03 N HCl, and pepsin [3.2 mg/mL]). Acidity was neutralized with NaOH (1.5 mL; concentration of NaOH was 2.2 N), and the suspension was added to 36.15 mL of simulated intestinal fluid (30mM K₂HPO₄, 160mM NaH₂PO₄, and pancreatin [20 mg/mL]; pH adjusted to 7.4) and agitated at 37°C for a further 2 hours. The final suspension was centrifuged at 3,000 × g for 25 minutes at 4°C to remove particulates. The supernatant was filtered (0.22 μm) then fractioned by use of a size-exclusion (50-kd) ultrafiltration centrifuge unit.^c A BL_sim (no product included) was prepared simultaneously by the same method. This BL_sim was the vehicle in which the DN, BO, and indomethacin were suspended and was used to control for effects of the simulated digestion protocol, which are independent of the effects of DN, BO, or indomethacin.

Preparation of explants—Thoracic limbs of market-weight pigs were obtained from a local abattoir that had Canadian Food Inspection Agency authorization to process pigs. Full-thickness cartilage explants were aseptically prepared with a sterile 4-mm disposable biopsy tool from grossly normal cartilage in the weight-bearing region of intercarpal joints, as previously described.⁴ Beginning at culture time 0 hours, the cartilage was cultured in Dulbecco modified Eagle medium with amino acids, sodium selenite, manganese sulfate, NaHCO₃, and ascorbic acid in 24-well tissue culture plates (Appendix). The total culture period was 120 hours.

Cartilage explants were 4 mm in diameter and approximately 0.5 mm in thickness; mean ± SEM weight of the explants was 17.9 ± 0.5 mg. Explants were stabilized in this unconditioned medium for the initial 24 hours, after which time 10 μL of BL_sim (n = 12 pigs), DN_sim (0.06 or 0.18 mg/mL; 12 pigs each), BO_sim (0.06 or 0.18 mg/mL; 9 pigs each), or INDO_sim (0.02 mg/mL; 12 pigs) was added to fresh medium (990 μL) every 24 hours for the duration of the experiment. Explants from 9 pigs were used in all treatment groups; explants from 3 additional pigs were used in each of the BL_sim, DN_sim, and INDO_sim experiments. Recombinant human IL-1β^d (0 [ie, no stimulation] or 10 ng/mL) was added to the conditioned media for the final 48-hour culture period (culture time, 96 to 120 hours).

Every 24 hours beginning immediately after explantation, the entire volume of conditioned medium (ie, 1,000 μL) from each experimental treatment was collected into sterile microcentrifuge tubes containing 10 μL of indomethacin (10 μg of indomethacin/mL of dimethyl sulfoxide, which was added to inhibit postcollection formation of PGE₂) and replaced with 990 μL of fresh, unconditioned medium with 10 μL of DN, BO, or indomethacin. Media collected from explants after 24 and 48 hours of culture were discarded; samples collected at the 72-hour culture time point were considered as prestimulation baseline samples (ie, samples in which the accumulations of PGE₂, GAG, or NO over the 24-hour period [culture time, 48 to 72 hours] immediately preceding stimulation with IL-1 [0 or 10 ng/mL] were assessed). These prestimulation baseline concentrations reflected the effect of conditioning (BL_sim, INDO_sim, DN_sim, or BO_sim) for 48 hours prior to stimulation with IL-1 (0 or 10 ng/mL). Collected samples were immediately frozen at −80°C until analyzed. Following commencement of stimulation with IL-1 (0 or 10 ng/mL), samples of media were collected at 24 hours (culture time, 96 hours) and 48 hours (culture time, 120 hours). At the end of the experiment, explant tissue was collected into sterile microcentrifuge tubes containing sterile PBS solution and stained for determination of cell viability.

Sample analysis—Frozen samples of media were thawed at room temperature and analyzed for PGE₂, GAG, and NO concentrations; these concentrations were designated as mPGE₂, mGAG, and mNO concentrations. Each sample underwent a single freeze-thaw cycle.

Assessment of mPGE₂ concentration—The mPGE₂ concentrations in 50-μL volumes of each sample were determined by use of a PGE₂-specific ELISA kit^e; assays were run on coated 96-well microtiter plates according to the manufacturer's protocol. Forty samples in duplicate were run on each plate. Plates were read by use of a microplate reader^f with absorbance set at 450 nm. A best-fit, third-order, polynomial standard curve was developed for each plate (R² ≥ 0.99), and those equations were used to calculate mPGE₂ concentrations for samples from each plate.

Assessment of mGAG concentration—The mGAG concentration in samples was determined by use of a spectrophotometric assay.⁴ For each sample, 75-μL volumes of media samples were added to 96-well plates at 50% dilution and serially diluted 1:2 to a final dilution of 1:64. Guanidine hydrochloride (275 mg/mL) was added to each well followed immediately by addition of 150 μL of dimethylmethylene blue reagent. Plates were incubated in the dark for 10 minutes, and absorbance at 530 nm was measured. Sample absorbance was compared with that of a bovine chondroitin sulfate standard^g (range, 2.12 to 11.43 μg/mL). Each 96-well plate contained a standard curve in duplicate and 6 samples in duplicate. A best-fit, linear standard curve was developed for each plate (R² ≥ 0.99), and these equations were used to calculate mGAG concentrations for samples on each plate. Each sample concentration was compared with the standard curve developed from the plate on which it was analyzed.

Assessment of mNO concentration—Nitrite, a stable oxidation product of NO, was analyzed by use of the Griess reaction.⁴ Volumes (75 μL each) of undiluted samples were added to 96-well plates. Sulfanilamide (0.01 g/mL) and N-(1)-napthylethylene diamine hydrochloride (1 mg/mL) dissolved in phosphoric acid (0.085 g/L) were added to all wells, and absorbance at 530 nm was measured within 5 minutes. Sample absorbance was compared with that of a sodium nitrite standard (range, 0.06 to 7.75 μg/mL). Each 96-well plate contained a standard curve in duplicate and 39 samples in duplicate. A best-fit, linear standard curve was developed for each plate (R² ≥ 0.99), and these equations were used to calculate nitrite concentrations for samples from each plate. Each sample concentration was compared with the standard curve developed from the plate on which it was analyzed.

Cell viability—Stock viability stain was prepared by combining C-AM^h (a stain for cells with biologically active intracellular esterases) and EthD-1ⁱ (a stain for nucleic acids in dead or damaged cells) in sterile PBS solution to concentrations of 4 and 8μM, respectively.⁴ Each explant was washed 3 times in 500 μL of sterile PBS solution, placed aseptically into a sterile 96-well microtiter plate, and incubated in 200 μL of stock viability stain in the dark at room temperature for 40 minutes. Care was taken to place an explant in the center of the well prior to detection of fluorescence. The microplate reader was set to scan each well, beginning at the bottom, with 10 horizontal steps at each of 3 vertical displacements set 0.1 mm apart. This was done to generate measurements of fluorescence at 30 points at 3 depths across the explant. The C-AM fluorescence of live cells and EthD-1 fluorescence of dead or damaged cells in explants were measured with excitation and emission filters of 485 and 530 nm and 530 and 685 nm, respectively. Mean C-AM fluorescence and EthD-1 fluorescence were determined for each individual explant, and a ratio of C-AM fluorescence to EthD-1 fluorescence was determined as a measure of relative viability. These ratios comprised the individual datum points for each explant. Mean ratios were determined for explants from each pig, and these data were statistically analyzed.

Data analysis—Data were reported as mean ± SEM. Means of replicates of each treatment and animal (mPGE₂, mGAG, and mNO concentrations) after the final 48 hours of culture were analyzed by use of a 2-way ANOVA to compare each treatment over time. Viability data were reported as a unitless fluorescence ratio of C-AM (live cells) to EthD-1 (dead cells). Viability data were analyzed by use of a 1-way ANOVA to compare treatments in stimulated and unstimulated explants. When a significant F ratio was obtained, the Holm-Sidak post hoc test was used to identify significantly (P < 0.05) different mean values.

Results

For purposes of the study, samples collected at the 72-hour culture time point were considered as prestimulation baseline samples. Treatment of explants with IL-1 was commenced, and samples of media were collected at 24 hours (culture time, 96 hours) and 48 hours (culture time, 120 hours) for assessment; these time points were referred to as 24 and 48 hours after stimulation, regardless of whether samples were treated with IL-1 at a concentration of 10 ng/mL or 0 ng/mL (ie, no stimulation).

mPGE₂ concentration—Compared with findings in unstimulated control samples, stimulation with IL-1 (10 ng/mL) significantly increased mPGE₂ concentration in explants treated with the BL_sim after 24 hours (419.4 ± 68.1 pg/mL vs 844.2 ± 102.9 pg/mL;) and after 48 hours (363.0 ± 43.5 pg/mL vs 600.4 ± 36.8 pg/mL; Figure 1). There was no effect of BO_sim on mPGE₂ concentration in unstimulated explants. In IL-1–stimulated explants, BO_sim (0.06 and 0.18 mg/mL) reduced mPGE₂ concentration at 24 hours (398.8 ± 207.8 pg/mL and 427.5 ± 91.5 pg/mL, respectively), compared with stimulated control explants.

Mean ± SEM PGE₂ concentration in samples of tissue culture medium (mPGE₂) from cartilage explants conditioned with BL_sim (simulated digestion fluid only), BO_sim (from oil derived from BO seeds) or a DN_sim (0.06 [A and B] or 0.18 [C and D] mg/mL), or INDO_sim (0.02 mg/mL). Explants were stimulated with IL-1 (10 ng/mL [A and C]) or unstimulated (B and D). In panels A and C, data from unstimulated BL_sim-treated explants are shown for comparison. Explants were conditioned for 48 hours with BL_sim, INDO_sim,DN_sim,or BO_sim and then treated with IL-1 (0 [ie, no stimulation] or 10 ng/mL) for the final 48-hour period of culture; the data were obtained during the final 48 hours. ^a–dFor a given time point, different superscripted letters denote significant (P < 0.05) differences in values among treatments. *Within a given treatment, value is significantly (P < 0.05) different from the 0-hour value.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Mean ± SEM PGE₂ concentration in samples of tissue culture medium (mPGE₂) from cartilage explants conditioned with BL_sim (simulated digestion fluid only), BO_sim (from oil derived from BO seeds) or a DN_sim (0.06 [A and B] or 0.18 [C and D] mg/mL), or INDO_sim (0.02 mg/mL). Explants were stimulated with IL-1 (10 ng/mL [A and C]) or unstimulated (B and D). In panels A and C, data from unstimulated BL_sim-treated explants are shown for comparison. Explants were conditioned for 48 hours with BL_sim, INDO_sim,DN_sim,or BO_sim and then treated with IL-1 (0 [ie, no stimulation] or 10 ng/mL) for the final 48-hour period of culture; the data were obtained during the final 48 hours. ^a–dFor a given time point, different superscripted letters denote significant (P < 0.05) differences in values among treatments. *Within a given treatment, value is significantly (P < 0.05) different from the 0-hour value.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

Mean ± SEM PGE₂ concentration in samples of tissue culture medium (mPGE₂) from cartilage explants conditioned with BL_sim (simulated digestion fluid only), BO_sim (from oil derived from BO seeds) or a DN_sim (0.06 [A and B] or 0.18 [C and D] mg/mL), or INDO_sim (0.02 mg/mL). Explants were stimulated with IL-1 (10 ng/mL [A and C]) or unstimulated (B and D). In panels A and C, data from unstimulated BL_sim-treated explants are shown for comparison. Explants were conditioned for 48 hours with BL_sim, INDO_sim,DN_sim,or BO_sim and then treated with IL-1 (0 [ie, no stimulation] or 10 ng/mL) for the final 48-hour period of culture; the data were obtained during the final 48 hours. ^a–dFor a given time point, different superscripted letters denote significant (P < 0.05) differences in values among treatments. *Within a given treatment, value is significantly (P < 0.05) different from the 0-hour value.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

Conditioning of explants with INDO_sim significantly inhibited PGE₂ release at all time points in both stimulated and unstimulated explants. Treatment with the DN_sim (0.06 and 0.18 mg/mL) significantly inhibited IL-1–induced PGE₂ release at all time points; mPGE₂ concentration was also significantly lower in explants conditioned with DN_sim (0.06 and 0.18 mg/mL) in unstimulated explants. In IL-1–treated explants at baseline and at 24 and 48 hours after stimulation, inhibition of PGE₂ release by DN_sim was comparable with that associated with INDO_sim. The mPGE₂ concentration in unstimulated explants conditioned with 0.06 mg of DN_sim/mL was not different from the concentration in explants conditioned with INDO_sim at the 24- and 48-hour time points; the mPGE₂ concentration in unstimulated explants conditioned with 0.18 mg of DN_sim/mL was not different from the concentration in explants conditioned with INDO_sim at all time points.

mGAG concentration—Prestimulation baseline mGAG concentrations (at 72 hours of culture) in unstimulated and stimulated control samples (ie, those conditioned with BL_sim) were not different (126.3 ± 9.3 μg/mL vs 142.1 ± 15.8 μg/mL). Compared with findings in unstimulated explants treated with BL_sim, stimulation with IL-1 (10 ng/mL) significantly increased mGAG concentration at 24 hours (75.5 ± 6.5 μg/mL vs 169.9 ± 18.6 μg/mL) and 48 hours (55.65 ± 2.9 μg/mL vs 120.1 ± 9.8 μg/mL; Figure 2). Conditioning of IL-1–stimulated explants with BO_sim and DN_sim (0.06 mg/mL) resulted in a significant increase in mGAG concentration, compared with the respective prestimulation baseline values (an increase that was not seen in stimulated control explants); however, mGAG concentration in these 2 treatment groups was not different from that in the stimulated control explants at any time point. There was no effect of BO_sim, INDO_sim, or treatment with 0.18 mg of DN_sim/mL on mGAG concentration from IL-1–stimulated explants.

Mean ± SEM GAG concentration in samples of tissue culture medium (mGAG) from cartilage explants conditioned with BL_sim, BO_sim or a DN_sim (0.06 [A and B] or 0.18 [C and D] mg/mL), or INDO_sim (0.02 mg/mL). Explants were stimulated with IL-1 (10 ng/mL [A and C]) or unstimulated (B and D). In panels A and C, data from unstimulated BL_sim-treated explants are shown for comparison. Explants were conditioned for 48 hours with BL_sim, INDO_sim, DN_sim, or BO_sim and then treated with IL-1 (0 [ie, no stimulation] or 10 ng/mL) for the final 48-hour period of culture; the data were obtained during the final 48 hours. See Figure 1 for key.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Mean ± SEM GAG concentration in samples of tissue culture medium (mGAG) from cartilage explants conditioned with BL_sim, BO_sim or a DN_sim (0.06 [A and B] or 0.18 [C and D] mg/mL), or INDO_sim (0.02 mg/mL). Explants were stimulated with IL-1 (10 ng/mL [A and C]) or unstimulated (B and D). In panels A and C, data from unstimulated BL_sim-treated explants are shown for comparison. Explants were conditioned for 48 hours with BL_sim, INDO_sim, DN_sim, or BO_sim and then treated with IL-1 (0 [ie, no stimulation] or 10 ng/mL) for the final 48-hour period of culture; the data were obtained during the final 48 hours. See Figure 1 for key.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

Mean ± SEM GAG concentration in samples of tissue culture medium (mGAG) from cartilage explants conditioned with BL_sim, BO_sim or a DN_sim (0.06 [A and B] or 0.18 [C and D] mg/mL), or INDO_sim (0.02 mg/mL). Explants were stimulated with IL-1 (10 ng/mL [A and C]) or unstimulated (B and D). In panels A and C, data from unstimulated BL_sim-treated explants are shown for comparison. Explants were conditioned for 48 hours with BL_sim, INDO_sim, DN_sim, or BO_sim and then treated with IL-1 (0 [ie, no stimulation] or 10 ng/mL) for the final 48-hour period of culture; the data were obtained during the final 48 hours. See Figure 1 for key.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

Conditioning of unstimulated explants with INDO_sim resulted in significantly higher mGAG concentration, compared with findings in unstimulated control explants, at 24 hours (175.7 ± 15.5 μg/mL) and 48 hours (133.1 ± 18.1 μg/mL) after commencement of the stimulation period (Figure 2). There was no effect of BO_sim or DN_sim (0.06 and 0.18 mg/mL) on mGAG concentration in unstimulated explants.

mNO concentration—Compared with findings in unstimulated control explants, mNO concentration was significantly increased in control explants after stimulation with IL-1 (10 ng/mL) for 24 hours (0.28 ± 0.04 μg/mL vs 1.26 ± 0.13 μg/mL) and for 48 hours (0.19 ± 0.04 μg/mL vs 1.04 ± 0.08 μg/mL; Figure 3).

Mean ± SEM NO concentration in samples of tissue culture medium (mNO) from cartilage explants conditioned with BL_sim, BO_sim oraDN_sim (0.06 [A and B] or 0.18 [C and D] mg/mL), or INDO_sim (0.02 mg/mL). Explants were stimulated with IL-1 (10 ng/mL [A and C]) or unstimulated (B and D). In panels A and C, data from unstimulated BL_sim-treated explants are shown for comparison. Explants were conditioned for 48 hours with BL_sim, INDO_sim,DN_sim,orBO_sim and then treated with IL-1 (0 [ie, no stimulation] or 10 ng/mL) for the final 48-hour period of culture; the data were obtained during the final 48 hours. See Figure 1 for key.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Mean ± SEM NO concentration in samples of tissue culture medium (mNO) from cartilage explants conditioned with BL_sim, BO_sim oraDN_sim (0.06 [A and B] or 0.18 [C and D] mg/mL), or INDO_sim (0.02 mg/mL). Explants were stimulated with IL-1 (10 ng/mL [A and C]) or unstimulated (B and D). In panels A and C, data from unstimulated BL_sim-treated explants are shown for comparison. Explants were conditioned for 48 hours with BL_sim, INDO_sim,DN_sim,orBO_sim and then treated with IL-1 (0 [ie, no stimulation] or 10 ng/mL) for the final 48-hour period of culture; the data were obtained during the final 48 hours. See Figure 1 for key.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

Mean ± SEM NO concentration in samples of tissue culture medium (mNO) from cartilage explants conditioned with BL_sim, BO_sim oraDN_sim (0.06 [A and B] or 0.18 [C and D] mg/mL), or INDO_sim (0.02 mg/mL). Explants were stimulated with IL-1 (10 ng/mL [A and C]) or unstimulated (B and D). In panels A and C, data from unstimulated BL_sim-treated explants are shown for comparison. Explants were conditioned for 48 hours with BL_sim, INDO_sim,DN_sim,orBO_sim and then treated with IL-1 (0 [ie, no stimulation] or 10 ng/mL) for the final 48-hour period of culture; the data were obtained during the final 48 hours. See Figure 1 for key.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

Conditioning of explants with INDO_sim resulted in significant increases in mNO concentration at baseline and 24 hours after commencement of the stimulation period, compared with findings in control explants, in both stimulated and unstimulated conditions. Compared with stimulated control explants, DN_sim significantly reduced mNO concentrations in IL-1–stimulated explants at 24 hours (0.91 ± 0.11 μg/mL and 0.92 ± 0.10 μg/mL for 0.06 and 0.18 mg of DN_sim/mL, respectively) and 48 hours (0.61 ± 0.10 μg/mL and 0.58 ± 0.09 μg/mL for 0.06 and 0.18 mg of DN_sim/mL, respectively). Conditioning of unstimulated explants with DN_sim significantly increased mNO concentrations at the 24-hour stimulation time point (0.66 ± 0.18 μg/mL and 0.67 ± 0.14 μg/mL for 0.06 and 0.18 mg of DN_sim/mL, respectively), compared with findings in unstimulated controls. Conditioning with BO_sim (0.06 or 0.18 mg/mL) had no effect on NO production by IL-1–stimulated or unstimulated explants.

Cell viability—There was no significant effect of IL-1 stimulation (10 ng/mL) on cell viability in the presence or absence of INDO_sim (Figure 4). At a concentration of 0.18 mg/mL, DN_sim and BO_sim each resulted in a significant increase in cell viability in IL-1–stimulated explants, compared with their effects in stimulated control explants. At a concentration of 0.06 mg/mL, there was no effect of DN_sim or BO_sim on cell viability.

Ratio of C-AM to EthD-1 fluorescence (mean ± SEM values) in cartilage explants conditioned for 96 hours with BL_sim (simulated digestion fluid only), BO_sim or a DN_sim (0.06 or 0.18 mg/mL), or INDO_sim (0.02 mg/mL). Explants were conditioned for 48 hours with BL_sim, INDO_sim, DN_sim, or BO_sim and then treated with IL-1 (10 [A] or 0 [no stimulation; B] ng/mL) for the final 48-hour period of culture. Explants were collected at the end of the experiment and immediately stained with C-AM and EthD-1 for determination of cell viability. *Value is significantly (P < 0.05) different from the control BL_sim value.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Ratio of C-AM to EthD-1 fluorescence (mean ± SEM values) in cartilage explants conditioned for 96 hours with BL_sim (simulated digestion fluid only), BO_sim or a DN_sim (0.06 or 0.18 mg/mL), or INDO_sim (0.02 mg/mL). Explants were conditioned for 48 hours with BL_sim, INDO_sim, DN_sim, or BO_sim and then treated with IL-1 (10 [A] or 0 [no stimulation; B] ng/mL) for the final 48-hour period of culture. Explants were collected at the end of the experiment and immediately stained with C-AM and EthD-1 for determination of cell viability. *Value is significantly (P < 0.05) different from the control BL_sim value.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

Ratio of C-AM to EthD-1 fluorescence (mean ± SEM values) in cartilage explants conditioned for 96 hours with BL_sim (simulated digestion fluid only), BO_sim or a DN_sim (0.06 or 0.18 mg/mL), or INDO_sim (0.02 mg/mL). Explants were conditioned for 48 hours with BL_sim, INDO_sim, DN_sim, or BO_sim and then treated with IL-1 (10 [A] or 0 [no stimulation; B] ng/mL) for the final 48-hour period of culture. Explants were collected at the end of the experiment and immediately stained with C-AM and EthD-1 for determination of cell viability. *Value is significantly (P < 0.05) different from the control BL_sim value.

Citation: American Journal of Veterinary Research 69, 12; 10.2460/ajvr.69.12.1560

• Download Figure
• Download figure as PowerPoint slide

Discussion

Results of the present study supported the hypotheses that BO_sim and DN_sim modulate response of articular cartilage explants to exogenous IL-1 without adversely affecting viability or metabolism of chondrocytes. The use of indomethacin as a reference anti-inflammatory drug provided an allopathic benchmark against which the effects of these 2 nutraceuticals could be compared. As expected, INDO_sim caused pronounced inhibition of PGE₂ release in both stimulated and unstimulated explants. However, the apparent adverse effects of INDO_sim on cartilage metabolism, as evidenced by INDO_sim-related increases in mGAG and mNO concentrations in our study and by INDO_sim-related inhibition of proteoglycan synthesis in other studies,^10,11 suggest that it is an unsuitable treatment for articular inflammation.

Simulated digests of NZGLM, SKC, and abalone (the 3 constituents of the DN other than BO) are known to have anti-inflammatory or cartilage-sparing effects in IL-1–stimulated cartilage explants.⁴ The data obtained in the present study indicated that the oil from the seed of BO also partially inhibits IL-1–dependent PGE₂ production by cartilage explants. There was considerable variability in mean PGE₂ concentration from explants conditioned with BO_sim, which resulted in a conservative estimation of the PGE₂-inhibiting potential of BO_sim. This may result, in part, from the difficulty in emulsifying BO in the simulated gastric and intestinal fluids. Pancreatic enzymes (including α-amylase, trypsin, lipases, ribonucleases, and proteases) were added to the simulated intestinal fluid, and although emulsification was improved, miscibility was still incomplete. In the experience of one of the authors (WP), the inclusion of bile acids and Tween 20 in the simulated digest vehicle stimulate PGE₂ production in cartilage explants; thus, bile acids and Tween 20 are not appropriate emulsification agents.^j Investigation of other possible emulsifiers or processes to improve miscibility of the BO seed oil in this type of experiment should be undertaken; this may include sonicating the oil with RBC membranes with or without plasma proteins prior to simulated digestion, as is used to improve emulsification of silicone oil.¹²

The heterogeneous composition of the DN (ie, NZGLM, SKC, abalone,⁴ and BO) resulted in significantly improved NO- and PGE₂-inhibiting properties, compared with the effects of the individual ingredients alone at the same dose. This synergistic effect may result from each ingredient influencing divergent PGE₂- or NO-synthetic pathways. In the DN, the primary putative bioactive constituents from NZGLM, SKC, and abalone are hexosamine-GAGs, omega-3 fatty acids, and proteins.^4,13,14 Several hypotheses have been proposed for the mechanisms by which exogenous hexosamine and GAG salts exert anti-inflammatory effects, including specific inhibition of COX-2 via interference with post translational glycosylation,¹⁵ provision of substrate for assembly of new proteoglycan through the hexosamine pathway,¹⁶ and interference with IL-1 signal-transduction–associated cascades.¹⁷ Specific COX-2 inhibition combined with omega-3 fatty acids results in a synergistic reduction of lipopolysaccharide-dependent tumor necrosis factor-α production in murine macrophages,¹⁸ a mechanism that may have contributed to the synergistic reduction in mNO and mPGE₂ concentrations detected in the present study.¹⁹ Indeed, if such a synergistic decrease in tumor necrosis factor-α production occurred through provision of the DN, this may have contributed to the increase in cell viability in DN_sim- and BO_sim-conditioned explants.²⁰

If the effects of BO_sim and DN_sim on IL-1–induced PGE₂ release were mediated through a mechanism similar to that of indomethacin, we would predict that BO_sim and DN_sim, like INDO_sim, would not protect cartilage against IL-1–induced NO release and would augment GAG release in cartilage explants. And indeed, in unstimulated explants, conditioning with both doses of DN_sim resulted in increases in mNO concentration. However, in explants conditioned with DN_sim, IL-1–induced NO release was significantly reduced. This apparent dichotomy may result from action of the DN via an IL-1–independent pathway; however, it may also be explained on the basis of the associative regulation of iNOS and COX, the enzymes that catalyze formation of PGE₂ and NO, respectively. Cellular expressions of iNOS and COX-2 are differentially regulated through activation of at least 2 mitogen-activated protein kinases.²¹ Net expressions of iNOS and COX-2 are at least partially dependent on the relative amounts of pericellular NO and PGE₂.²² Thus, products that increase pericellular NO concentration can effectively down-regulate expression of COX-2 and vice versa.^22,23 This may provide some explanation as to why the DN_sim had a significant inhibitory effect on NO release in IL-1–stimulated explants but caused an increase in mNO concentration in unstimulated explants and also supports a mechanism of COX-2 inhibition by the DN. Further research should investigate cell signaling pathways, particularly p42/44 and p38,²¹ that may be involved in the effect of the DN on production of NO and PGE₂ protein.

Because at least 2 possible cell signaling pathways implicated in IL-1 stimulation of COX-2 expression are immediately associated with expression of catabolic matrix-degrading enzymes,^24,25 and because compounds capable of inhibiting COX-2 are associated with decreased expression of matrix mettalloproteinases,^25,26 we predicted that the DN would reduce mGAG concentration (an effect similar to that associated with NZGLM or abalone⁴), thereby providing evidence for a cartilage-sparing effect. However, results of the present study did not indicate any significant effect of BO_sim or DN_sim on mGAG concentration. Because NZGLM and abalone both inhibit mGAG production in IL-1–stimulated explants,⁴ further research endeavors should investigate the net flux of cartilage proteoglycan in the presence and absence of DN_sim to determine whether the apparent protective effect of NZGLM or abalone is lost when applied in an NZGLM-SKC-abalone-BO preparation or whether the DN is able to stimulate proteoglycan synthesis in cartilage explants.

In the present study, cell viability subsequent to IL-1 challenge was significantly increased by BO_sim and DN_sim. These increases were not a result of reduced apoptosis, as might have been predicted from the NO inhibitory properties of the DN, because the increases in viability resulted from an increase in C-AM staining (ie, staining of live cells) but not a decrease in EthD-1 staining (ie, staining of dead cells); cell viability in stimulated BO_sim-treated and DN_sim-treated explants was greater than the corresponding findings in unstimulated control explants. The primary mechanism for chondrocyte growth in cartilage has been attributed to the binding of PGE₂ to 1 of its 4 cell-surface receptors (ie, EP1, EP2, EP3, or EP4), and specifically to EP1^27,28 or EP2.²⁹ The DN_sim, perhaps through provision of BO_sim (which had almost equal effect on cell viability), may produce an agonist effect on the prostanoid receptor involved in chondrocyte proliferation, resulting in an increase in live-cell staining. This would also provide a basis for the concurrent inhibition of PGE₂ release by DN_sim. Substances that act as agonists for PGE₂ receptors produce an effect similar to that associated with increasing the amount of PGE₂ in the pericellular environment. Prostaglandin E₂ production from IL-1–stimulated cells is inhibited by increasing extracellular concentrations of PGE₂ through a negative feedback mechanism.³⁰ Further research is warranted to investigate preferential inhibition of the 4 prostanoid receptors in the presence of DN_sim to determine whether this mechanism is associated with the bioactivity of DN_sim in vitro.

Although the data obtained from experiments involving simulated digests of nutraceuticals and cartilage explants stimulated with IL-1 as a model of articular inflammation may more accurately represent the in vivo situation than data obtained from experiments involving traditional cartilage explants alone, the former should still be viewed with some caution from a clinical perspective until in vivo studies to assess safety and efficacy of the test materials are carried out in a target species.

In addition to investigations of other emulsifiers as alternatives to the simulated digestion process, there are several factors that could be examined to further increase the in vivo applicability of data derived from experiments involving the method used in the present study. For example, the bioavailability of the test material of interest should be investigated. In the procedures used in the present study, the digestible portion of the DN product was assumed to be 100% in a treated animal. Pharmacokinetic information on the candidate bioactive constituents within the test material would be useful in assessing the bioavailability of the product. Species differences in absorption and metabolism of the test material should be examined, as these differences are currently assumed to be of negligible importance within the context of the current model. An interesting possibility for further research would be to extract gastrointestinal juices from the stomach and the proximal portion of the small intestine from other animals, including horses and dogs, for direct digestion of nutraceuticals ex vivo to assess differences in the bioactivity between those simulated digests and the digests used in the present study. Furthermore, the digested and absorbed nutraceutical product is assumed to be evenly distributed throughout the total body water compartment of an animal and not preferentially sequestered into any particular tissue or cell type. Substantial biochemical characterization of the ingredients of a test material, followed by pharmacokinetic studies in the target species, would be required to address this factor.

Another factor not considered in the procedures used in our study is biotransformation of test products in the liver. In the study technique, the active constituents of nutraceuticals are assumed to be substantively unaltered by transformation in the liver. The introduction of an S9 fraction of liver microsomes or intact liver microsomes into the preparation of simulated digests may provide additional information on biotransformation products of a test material.

The experimental protocol used in the present study assumed that there are no interactions between other structures and tissues of the joint or cells within synovial fluid that could impact the bioavailability of the nutraceuticals in different components of a joint. Development of cartilage coculture methods, which may include synovial, subchondral bone, or ligamentous explants with or without neutrophils, would provide interesting new information on this aspect of nutraceutical bioactivity.

Another factor of interest is the effect of dynamic compression on cartilage matrix. In the experimental method used in the present study, it was assumed that dynamic compression does not affect the movement of nutraceutical constituents into the cartilage matrix. Further research may integrate the simulated digestion-ultrafiltration protocol with a compression model of cartilage damage,³¹ which may alter the pharmaco-dynamics of the nutraceutical constituents. Additionally, there may be differences in explant cellularity. The viability assay developed for the present study cannot account for these small differences in explant cellularity resulting from random sampling across the surface of the intercarpal joint, from differences in explant thickness, or from orientation of the explant within the microtiter plate. Thus, viability data may be vulnerable to random sampling bias or to orientation of the explant within the well. The impact of both of these factors could be reduced by increasing the number of individual measurements across a single explant to capture both the articulating surface of the explant and the cartilage zone that interfaces with calcified cartilage or by incorporating a larger number of animals into a study of this nature.

A final limitation of the present study is that it was assumed that there are no constituents in the 50-kd fraction of a digest would undergo extensive physiologic regulation in vivo (eg, blood glucose), resulting in rapid removal of those constituents from the extra-cellular fluids. The impact of this on bioactivity of a nutraceutical product will require in vivo studies to elucidate.