Beneficial reward-to-risk action of glucosamine during pathogenesis of osteoarthritis

Human chondrocytes derived from normal human articular cartilage were purchased from Cell Applications (San Diego, CA). Cells were cultured in chondrocyte growth medium (Cell Applications). Cells were maintained in a humidified incubator at 37 °C with 5 % CO₂. To check the differential ability of chondrocytes, pellet cultures were applied and stained with Alcian blue and the RNA levels of type II collagen and aggrecan were analyzed before the experiments. For this study, cells were treated with 10, 30, 50 and 100 mM d-(+)-glucosamine hydrochloride for 2 h (short time exposure) or 24 h (long time exposure) unless exposure time is indicated, respectively (Sigma-Aldrich, St. Louis, MO, dissolved in growth media).

For differentiation, alginate beads with encapsulated human chondrocytes (Cell applications, San Diego, CA, USA) were grown in a chondrocyte differentiation medium (Cell applications, San Diego, CA, USA) for 2 weeks. Briefly, human chondrocyte were suspended at a density of 4 × 10⁶ per milliliters in a 1.2 % solution of sterile alginate in 0.15 M NaCl. The cell suspension was slowly expressed through a 22-gauge needle and dropped into a 102-mM CaCl₂ solution. The beads with approximately 40,000 cells/bead (diameter 3 mm) were allowed to polymerize for 10 min and washed twice with 0.15 M NaCl, followed by two washes in DMEM/F12. The beads were then transferred to medium and cultured at 37 °C in a humidified atmosphere of 5 % CO₂. Alcian blue staining was used to detect chondrocyte nodule formation after 2 weeks of alginate beads culture. Alginate beads were rinsed with PBS and fixed in 4 % paraformaldehyde in PBS for 20 min. Alginate Beads were washed with PBS three times and stained in 1 % Alcian blue in 0.1 N HCl for 8 h at room temperature. Alginate Beads were de-stained in 0.1 N HCl two times and stored in water for image capture.

Total RNA was isolated using RNAiso Plus (TaKaRa, Japan) according to the manufacturer’s protocol. Aliquots of total RNA (1 μg) from each sample were reverse-transcribed into cDNA according to the instructions of the PrimeScript 1st strand cDNA Synthesis Kit (TaKaRa, Japan). Quantitative real-time PCR (qRT-PCR) was performed using the StepOnePlus Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). PCR were prepared and heated to 95 °C for 2 min followed by 40 cycles of denaturation at 95 °C for 15 s, annealing at specific Tm for 15 s, and extension at 72 °C for 20 s. Quantification of the PCR signals was achieved by comparing the cycle threshold value (C _t) of the gene of interest with the C _t value of the reference gene GAPDH. The following oligonucleotides were used as primers: MMP13: 5′-TTGCAGAGCGCTACCTGAGATCAT-3′, antisense, 5′-TTTGCCAGTCACCT CTAAGCCGAA-3′, sense; ADAMTS4: 5′-CGCTTTGCTTCACTGAGTAGAT-3′, antisense, 5′-CTGTTAGCAGGTAGCGCTTTAG-3′, sense; PMP70: 5′- CCAGTTGGGTCATATCCT TGAA-3′, antisense, 5′- CTTGCCATCGCC ATTCTTTG-3′, sense; LC3A: 5′- ACAGCATGG TGAGTGTGTC-3′, antisense, 5′- GGGAGGCGTAGACCATATAGA-3′, sense; LC3B: 5′- GCCTTCTTCCTGTTGGTGAA-3′, antisense, 5′- TGGGAGGCATAGACCATGTA-3′, sense; CASP1: 5′-GGCAGGCCTGGA TGATGA-3′, antisense, 5′-ATACCAAGAACT GCCCAAGTTTG-3′, sense, CASP3: 5′-GCGCCCTGGCAGCAT-3′, antisense, 5′-GCCTACAGCCCATTTCTCCAT-3′ sense, CASP9: 5′-AACAGCATTAGCGACCCTAAGC-3′ antisense, 5′-AGCAGTGGGCTCAC TCTGAAG-3′, sense, FAS: 5′-CCAGCATGG TTGTTGAGCAA-3′ antisense, 5′-ACCCGCTCAGTACGGAGTTG-3′, sense, COMP: 5′-TCACAAGCATCTCCCACAAA-3′ antisense, 5′-GACAGTGATGGCGATGGTATAG-3′ sense, ACAN: 5′-TCGAGGGTGT AGCGTGTAGAGA-3′ antisense, 5′-TCGAGGACA GCGAGGCC-3′ sense, GAPDH: 5′-GATCATCAGCAATGCCTCCT-3′, antisense, 5′-TGTGGTCATG AGTCCT TCCA-3′, sense.

BODIPY 493/503 or BODIPY 665/676 (Molecular Probes, Carlsbad, CA, USA), was diluted in PBS at a concentration of 1 mg/mL. Following fixation with 4 % paraformaldehyde (PFA) for 10 min and staining with 4,6-diamidino-2-phenylindole (DAPI) for identification of nuclei, the samples were washed 3 times in PBS for 10 min and stained with BODIPY. The samples were mounted in VECTASHIELD (Vector Laboratories), covered with glass cover slips, and digital images were obtained with an Olympus Fluoview FV100 confocal laser scanning microscope under epifluorescent optics.

Human chondrocytes were transfected with the mRFP-EGFP-SKL plasmid. Two days after the first transfection, the cells were cultured for another 24 h in the presence of lysosomal inhibitors, 120 μM leupeptin (Sigma-Aldrich), and 2 μM E-64 (Enzo Life Sciences). After incubation, the cells were washed in PBS and fixed with 4 % PFA in PBS for 15 min at RT.

Human chondrocytes were transfected with LC3-GFP plasmid. Transfection efficiency of the plasmid was assessed visually (>60 % and identical for all wells). After 2 days, the cells were starved for 3 h in the presence of 20 μM chloroquine (InvivoGen) and permeabilized with 0.025 % digitonin in PBS for 5 min to wash out the cytosolic LC3 protein and enrich for the vesicle-associated LC3. Subsequently, the cells were washed in PBS and fixed with 4 % PFA in PBS for 15 min at RT.

Total lipids were extracted using a chloroform/methanol (2:1 v/v) mixture. The extracted lipids were separated on a Sep-Pak Silica Cartilage column (Waters, Milford, MA, USA) and transmethylated with 0.5 M CH3ONa in methanol by heating in a sealed tube at 70 °C for 1 h under nitrogen. The fatty acid methyl esters were extracted with hexane. Subsequent gas chromatography–mass spectrometry (GC–mass) analysis was performed according to the standard protocol. Briefly, the sample was injected into a Finnigan MAT-8430 mass spectrometer connected to an HP-5890 gas chromatography (Finnigan MAT, Bremen, Germany) equipped with a DB-5 capillary column.

Apoptosis and autophagy profiling was conducted using the RT² Profiler PCR Array Kit (QIAGEN, GmbH, Hilden, Germany), which included specific primers for apoptosis (PAHS-012Z) or autophagy (PAHS-084Z). Each profiling array was performed according to the manufacturer’s instructions. cDNA was synthesized with the RT² First Strand Kit (Qiagen, #330401) and qRT-PCR was performed using the RT² qPCR SYBR Green Fluor qPCR Master Mix (Qiagen, #330512) according to the manufacturer’s instructions. Data were analyzed and gene expression data visualized using GenEx (Weihenstephan, Germany).

Cells were lysed in RIPA buffer (50 mM Tris–HCl, pH 7.4, containing 150 mM NaCl, 1 % Nonidet P-40, 0.1 % SDS, 0.1 % deoxycholic acid, 10 mM NaF, 10 mM Na₄P₂O₇, 0.4 mM Na₃VO₄, and protease inhibitors) for 30 min on ice. The total protein content of the cells was determined by BCA Protein Acid Reagent (bicinchoninic acid) (Pierce Biotechnology Inc., Rockford, MN, USA). Proteins (30 μg) were separated by 10 % polyacrylamide gel electrophoresis containing 0.1 % SDS and transferred to nitrocellulose membranes (GE Healthcare Life Sciences, Uppsala, Sweden). The membranes were incubated for 1 h at room temperature in a blocking buffer (20 mM Tris–HCl, 137 mM NaCl, pH 8.0, containing 0.1 % Tween and 3 % skimmed dry milk), and probed with antibodies against Beclin, LC-3B and GAPDH (Cell Signaling, Beverly, MA, USA). The blots were developed with an HRP-conjugated secondary antibody and reacted proteins were visualized using an electrochemiluminescence (ECL) system (GE Healthcare Life Sciences).

Human articular chondrocytes were cultured in growth media and their chondrogenic capacity analyzed. Pellet culture of human chondrocytes showed an increase in Alcian blue staining and the transcription of a well-known cartilage matrix, cartilage oligomeric matrix protein (COMP) was increased as culture progressed. As previously mentioned [17‐20], there is a significant controversial opinion on glucosamine effect, particularly with high dose of glucosamine. To verify the role of high-dose glucosamine in OA pathogenesis, chondrocytes were treated with increasing doses of glucosamine (10–100 mM). The number of chondrocytes was presenting severely degenerative morphology under glucosamine treatment and this degeneration was worsening in a dose-dependent manner (Fig. 1a). Consistent with morphological changes, the RNA levels of MMP-13 and ADAMTS4 were significantly increased in 50 mM-treated chondrocytes (Fig. 1b). To investigate the involvement of chondrocyte apoptosis in high-dose treatment of glucosamine, cells were stained with annexin-V/PI and analyzed by flow cytometry. After treatment with glucosamine for 24 h, the population of cells indicated a shift from viable cells to early and late stage apoptosis in a dose-dependent manner (Fig. 1c). The average percentage of late apoptotic cell populations increased to 15.25 and 19.54 %, respectively, for cells treated with 50 or 100 mM glucosamine suggesting that high dose of glucosamine may trigger chondrocyte apoptosis. Since mitochondrial dysfunction is closely related to cell death, we further assessed alterations in mitochondria potential (Fig. 1d). The average percentage of depolarized dead cells increased to 34.5 and 38.9 %, respectively, for cells treated with 50 or 100 mM glucosamine.

Recently, our laboratory suggested an interconnection between mitochondria and peroxisome during OA pathogenesis [20]. To investigate whether high-dose glucosamine induced peroxisomal dysfunction as well as mitochondrial dysfunction, normal chondrocytes were stained with BODIPY (493/503). We found that significantly large amounts of lipid were accumulated in cells treated with 50 and 100 mM glucosamine (Fig. 2a, left panel), indicating a possibility that high-dose glucosamine may induce dysfunction of peroxisome. To evaluate peroxisomal function, we determined the expression levels of PMP70/ABCD3, a70 kDa peroxisomal membrane protein and acyl-CoA oxidase 1 (ACOX1), which catalyzes the first step of peroxisomal fatty acid β-oxidation by qRT-PCR (Fig. 2b). The expression levels of PMP70 and ACOX1 were significantly decreased in cells treated with either 50 or 100 mM glucosamine.

Next, we investigates the possibility that this peroxisomal dysfunction by glucosamine could be resulted in lipid accumulation due to lipid metabolism impairment. Lipidomics analysis (Table 1) showed that treatment with 50 mM glucosamine lead to the accumulation of very long chain fatty acids (VLCFA) in normal chondrocytes (Fig. 3a). By adding exogenous lignoceric acid to the cells, we evaluated the effects of VLCFA on lipid metabolism. Cells treated with VLCFA showed a severely degenerative morphology (Fig. 3b, upper panel). Furthermore, cells treated with lignoceric acid demonstrated significantly increased MMP-13 RNA levels (Fig. 3b, lower panel). Moreover, levels of pro-apoptotic genes including FASL, FAS and BAX, as well as the percentage of apoptotic cell populations, were increased with lignoceric acid treatment (Fig. 3c). These data suggest that high-dose glucosamine-induced VLCFA accumulation may be responsible for chondrocyte apoptosis and MMP-13 induction.

To evaluate whether autophagic cell death is involved in glucosamine-induced apoptosis, the expression levels of genes involved in apoptosis (Fig. 4a) and autophagy (Fig. 4b) were analyzed. Both levels of apoptotic genes and autophagic genes assessed here were significantly increased to more than threefold in cells treated with 50 mM glucosamine. RNA levels of caspase-1, -3, -6, -7, and -9 demonstrated a 15.5-, 9.8-, 9.7-, 13.2-, and 10.9-fold increase, respectively (Fig. 4a). RNA levels of ATG3, ATG5, ATG7, ATG12, and ATG16L1 showed a 3.6-, 7.4-, 12.3-, 5.2-, and 12.2-fold increase, respectively (Fig. 4b). Moreover, expression levels of beclin-1 were increased 4.3-fold in cells treated with 50 mM glucosamine. To confirm that glucosamine induced the hyper-activation of autophagic response, normal chondrocytes were treated with glucosamine in the absence or presence of MG132, a cell-permeable proteasome inhibitor. In the presence of MG132, the cells exposed to glucosamine showed slightly decreased levels of autophagy markers becin-1 and LC-3B, suggesting that the treatment with high-dose glucosamine may actually decrease autophagic responses (Fig. 4c). In addition, the pattern of microtubule-associated LC3, as assessed by immunostaining, indicated inactivation of autophagic response under glucosamine treatment (Fig. 4d, upper panel). Next, we used the tandem fluorochrome pexophagy assay with mRFP-EGFP-SKL as a reporter. Cells treated with glucosamine did not display the characteristic “red only” structure as a consequence of EGFP fluorescence quenching by acid pH after the delivery of peroxisome to lysosome, which would indicate the stimulation of autophagic responses (Fig. 4d, lower panel).

As mentioned earlier, the effect of glucosamine in OA is still controversial. Whilst some reports showed that high-dose glucosamine may stop the degradation of cartilage by acting as a chondro-protective agent [21], others have reported a negative effect of high-dose glucosamine during OA pathogenesis [22], suggesting the possibility that the effects of glucosamine may depend on treatment length or duration. Therefore, we analyzed microscopy-based GFP-LC3 puncta formation in cells exposed to 50 mM glucosamine, to address this possibility. At exposure times of 2 h (short time exposure), glucosamine was found to induce GFP-LC3 puncta. However, glucosamine significantly suppressed GFP-LC3 puncta at 24 h exposure (Fig. 5a). Next, we analyzed whether the difference in the effects of glucosamine based on the exposure time (short time exposure vs. long time exposure) is a result of peroxisomal dysfunction. Peroxisomal oxidation was not affected during short time exposure of glucosamine, although during long time exposure of glucosamine, we observed a significant reduction in peroxisomal oxidation (Fig. 5b). Consistent with these findings, the expression levels of acyl-CoA-binding protein (ACBD5), a typical marker for pexophagy, or acyl-CoA oxidase 1 (Acox-1), an initial enzyme in β-oxidation, increased during short time exposure of glucosamine but decreased during long time exposure. A typical cartilage matrix gene, COMP was dramatically decreased during short time exposure (Fig. 5c). Moreover, when normal chondrocyte were exposed with low concentration, we found that 5 mM glucosamine increased transcription of cartilage matrix gene, COMP and suppressed apoptotic genes such as CASP1 and FAS (Fig. 6).