Bone resorption and remodeling in murine collagenase-induced osteoarthritis after administration of glucosamine

Abstract

Introduction: Glucosamine is an amino-monosaccharide and precursor of glycosaminoglycans, major components of joint cartilage. Glucosamine has been clinically introduced for the treatment of osteoarthritis but the data about its protective role in disease are insufficient. The goal of this study was to investigate the effect of long term administration of glucosamine on bone resorption and remodeling.

Methods: The effect of glucosamine on bone resorption and remodeling was studied in a model of collagenase-induced osteoarthritis (CIOA). The levels of macrophage-inflammatory protein (MIP)-1α, protein regulated upon activation, normal T-cell expressed, and secreted (RANTES), soluble receptor activator of nuclear factor kappa-B ligand (RANKL), tumor necrosis factor (TNF)-α, and interleukin (IL)-6, 4 and 10 in synovial fluid were measured by enzyme-linked immunosorbent assay (ELISA). Cell populations in synovial extracts and the expression of RANKL, of receptors for TNF-α (TNF-αR) and interferon γ (IFN-γR) on clusters of differentiation (CD) three positive T cells were analyzed by flow cytometry. Transforming growth factor (TGF)-β3, bone morphogenetic protein (BMP)-2, phosphorylated protein mothers against decapentaplegic homolog 2 (pSMAD-2), RANKL and Dickkopf-1 protein (DKK-1) positive staining in CIOA joints were determined by immunohistochemistry.

Results: The administration of glucosamine hydrochloride in CIOA mice inhibited loss of glycosaminoglycans (GAGs) and proteoglycans (PGs) in cartilage, bone erosion and osteophyte formation. It decreased the levels of soluble RANKL and IL-6 and induced IL-10 increase in the CIOA joint fluids. Glucosamine limited the number of CD11b positive Ly6G neutrophils and RANKL positive CD3 T cells in the joint extracts. It suppressed bone resorption via down-regulation of RANKL expression and affected bone remodeling in CIOA by decreasing BMP-2, TGF-β3 and pSMAD-2 expression and up-regulating DKK-1 joint levels.

Conclusions: Our data suggest that glucosamine hydrochloride inhibits bone resorption through down-regulation of RANKL expression in the joints, via reduction of the number of RANKL positive CD3 T cells and the level of sRANKL in the joints extracts. These effects of glucosamine appear to be critical for the progression of CIOA and result in limited bone remodeling of the joints.

Figure 1

Effect of glucosamine on the development of CIOA. CIOA was induced by injection of 1 U/mouse collagenase at Day 0 and Day 2. After seven days, CIOA mice were orally treated with PBS (CIOA; n = 15), with glucosamine hydrochloride (20 mg/kg/daily; CIOA + Glu1; n = 15) or glucosamine sulfate (20 mg/kg/daily; GS, n = 10) for 20 days, or with glucosamine hydrochloride (20 mg/kg/daily) started with the second collagenase injection and lasted 20 days (20 mg/kg/daily; CIOA + Glu2; n = 10). A control group of non-arthritic mice were fed with PBS (healthy; n = 9). (a) Representative joint sections stained with H&E showed fissures and fractures in cartilage and bone matrix of CIOA mice at Day 30, attenuated by Glu1 and Glu2, but not by GS; magnification × 40. Staining with Toluidine blue (b) and Safranin O (c) demonstrated GAGs and PG loss in CIOA that was less pronounced in Glu1-treated CIOA mice; magnification × 100. Glu1 and Glu2 significantly reduced the osteophyte areas in CIOA mice, while GS was not effective (d). Data are expressed as the mean ± SD from the evaluation of five joint sections/group; Student's t-test; **P <0.01 vs CIOA group. (e) Glu1 and Glu2 significantly decreased the total histological score in CIOA mice and nonsignificantly by GS. Total histological score was calculated using the semi-quantitative grading and staging system. Data are expressed as the mean ± SD from the evaluation of 10 joints/group; Student's t-test; ***P <0.001 vs CIOA group.

Figure 2

Effect of glucosamine on the synovial level of soluble RANKL, IL-6 and IL-10. Joint extracts were collected from individual mice treated with PBS (healthy, n = 9), non-arthritic mice treated with glucosamine hydrochloride for 20 days (20 mg/kg/daily; healthy + Glu, n = 10), CIOA mice (n = 15) and CIOA mice treated with glucosamine hydrochloride for 20 days, starting from Day 7 (20 mg/kg/daily; CIOA + Glu1; n = 15). The levels of particular cytokine (MIP-1α, RANTES, soluble RANKL, TNF-α, IL-6, IL-4 and IL-10) on Day 14 (insert) and on Day 30 of CIOA were determined by ELISA and expressed as picograms/ml supernatant. Data represent the mean ± SD of two independent experiments involving from 9 to 15 animals/group; *P < 0.05; ***P < 0.001 CIOA vs CIOA+Glu1; ^###P < 0.001 vs healthy. Student's t-test.

Figure 3

Phenotype of cells in the synovial extract at Day 30 of CIOA. Joint extracts were collected from mice treated with PBS (healthy, n = 5), nonarthritic mice treated with glucosamine hydrochloride for 20 days (20 mg/kg/daily; healthy + Glu, n = 5), CIOA mice (n = 5) and CIOA mice treated with of glucosamine hydrochloride for 20 days, starting from Day 7 (20 mg/kg/daily; CIOA + Glu1; n = 5) or with GS (20 mg/kg/daily; CIOA + GS, n = 5). Samples were obtained separately from each mouse. Individual data were presented in dot-plot graphs. The median value is shown in the graphs with line. *P < 0.05; **P < 0.01; ***P < 0.001 vs healthy; ^###P < 0.001 vs CIOA group; ⁺⁺P < 0.01; ⁺⁺⁺P < 0.001 CIOA + Glu1 group vs GS. Student's t-test.

Figure 4

Glucosamine up-regulated TNF-αR1 and down-regulated IFN-γR1 expression in CIOA on synovial CD3 T cells. Synovial cells were obtained at Day 30 from joint extracts of healthy (n = 5) mice, CIOA mice treated with PBS (CIOA; n = 5) and CIOA mice treated with glucosamine (CIOA + Glu1, n = 5). Synovial cells were stained with PE-labeled antibody against CD3 and with biotinylated antibodies against TNF-αR1 and IFN-γR1 followed by avidin-FITC staining and were subjected to flow cytometry. Data represent the mean of fluorescence expression ± SD from two independent experiments involving five mice/group; Student's t-test; ***P < 0.001.

Figure 5

Effect of glucosamine on the percentage of RANKL positive synovial and peripheral CD3 T cells. Synovial cells and PBMCs were obtained at Day 30 from healthy (n = 5) mice, CIOA mice treated with PBS (CIOA; n = 5) and CIOA mice treated with glucosamine hydrochloride (CIOA + Glu1, n = 5). The cells were stained with PE-labeled antibody against CD3 and with biotinylated antibody against RANKL followed by avidin-FITC staining and were subjected to flow cytometry (a) Representative data showing the inhibitory effect of glucosamine on the number of RANKL positive synovial CD3 T cells, (b) Glucosamine suppressed the RANKL expression of synovial CD3 T cells. Data represent the mean of positive cells ± SD from two independent experiments involving five mice/group; Student's t-test; ***P < 0.001, (c) Representative data showing increased percentage of RANKL positive peripheral CD3 T cells in glucosamine-treated CIOA mice, (d) Glucosamine increased the RANKL expression of peripheral CD3 T cells. Data represent the mean of positive cells ± SD from two independent experiments involving five mice/group; Student's t-test; **P < 0.01, ***P < 0.001.

Figure 6

Glucosamine altered the expression of bone resorption and bone remodeling markers in the CIOA joints. Joints of healthy mice, CIOA mice treated with PBS (CIOA) and CIOA mice treated with glucosamine hydrochloride (CIOA + Glu1) were fixed, decalcified, dehydrated and embedded in paraffin. The sections (n = 10/mice) were stained with antibodies against RANKL, BMP-2, TGF-β3, pSMAD-2 and DKK-1. The specific staining was detected using a peroxidase-DAB staining kit. The positive staining for RANKL in the joints (a), BMP-2 in osteophyte areas (b) and TGF-β3 (c), pSMAD-2 (d) and DKK-1 (e) in cartilage of CIOA mice untreated or treated with glucosamine was indicated with arrows; magnification × 100.