Matrilin-3 induction of IL-1 receptor antagonist is required for up-regulating collagen II and aggrecan and down-regulating ADAMTS-5 gene expression

Abstract

Introduction: Deletion or mutation of the gene encoding the cartilage extracellular matrix (ECM) protein matrilin-3 (MATN3) results in the early onset of osteoarthritis (OA), suggesting chondroprotective properties of MATN3. To understand the mechanisms underlying these properties, we determined the effects of MATN3 protein on the expression of several key anabolic and catabolic genes involved in chondrocyte homeostasis, and the dependence of such regulation on the anti-inflammatory cytokine: IL-1 receptor antagonist (IL-1Ra).

Methods: The effects of recombinant human (rh) MATN3 protein were examined in C28/I2 immortalized human chondrocytes, primary human chondrocytes (PHCs), and primary mouse chondrocytes (PMCs). Messenger RNA levels of IL-1Ra, COL2A1, ACAN, MMP-13, and ADAMTS-4 and -5 were determined using real-time RT-PCR. Knocking down IL-1Ra was achieved by siRNA gene silencing. IL-1Ra protein levels were quantified by ELISA and the Bio-Plex Suspension Array System. COL2A1 protein level was quantified using Western blot analysis. Statistic analysis was done using the two-tailed t-test or one-way ANOVA.

Results: rhMATN3 protein induced gene expression of IL-1Ra in C28/I2 cells, PHCs, and PMCs in a dose- and time-dependent manner. Treatment of C28/I2 cells and PHCs with MATN3 protein stimulated gene expression of COL2A1 and ACAN. Conversely, mRNA levels of COL2A1 and ACAN were decreased in MATN3 KO mice. MATN3 protein treatment inhibited IL-1β-induced MMP-13, ADAMTS-4 and ADAMTS-5 in C28/I2 cells and PHCs. Knocking down IL-1Ra abolished the MATN3-mediated stimulation of COL2A1 and ACAN and inhibition of ADAMTS-5, but had no effect on MATN3 inhibition of MMP-13 mRNA.

Conclusion: Our findings point to a novel regulatory role of MATN3 in cartilage homeostasis due to its capacity to induce IL-1Ra, to upregulate gene expression of the major cartilage matrix components, and to downregulate the expression of OA-associated matrix-degrading proteinases in chondrocytes. The chondroprotective properties of endogenous MATN3 depend partly on its induction of IL-1Ra. Our findings raise a possibility to use rhMATN3 protein for anti-inflammatory and chondroprotective therapy.

Figure 1

Matrilin-3 MATN3) enhances IL-1Ra expression by chondrocytes. Recombinant human (rh) MATN3 protein enhances the gene expression of IL-1Ra in a dose-dependent manner at 12 hours post cell culture treatment as seen here in C28/I2 cells (A). Recombinant MATN3 also enhances IL-1Ra expression in primary mouse chondrocytes after 24 hours of treatment (B). MATN3-induced IL-1Ra gene expression is evident in primary human chondrocytes (PHCs) at 24 hours post treatment in both the presence and absence of IL-1β (C). PHCs treated with both MATN3 and IL-1β for 24 hours exhibit significantly higher concentrations of IL-1Ra protein in their cell media, relative to the media of cells treated with IL-1β alone (D). In these experiments, rh MATN3 protein is used at one of two concentrations: lower dose (+) of 100 ng/ml, or higher dose (++) of 200 ng/ml. Rh IL-1β protein treatment is always 5.0 ng/ml. *Significant differences (P ≤ 0.05) from the untreated control group; ^#significant differences (P ≤ 0.05) from the IL-1β only treated group. Data are representative of three individual experiments.

Figure 2

Matrilin-3 (MATN3) maintains the expression of key cartilage extracellular matrix (ECM) genes. Recombinant human (rh) MATN3 treatment maintains type II collagen (COL2A1) gene expression in the presence and absence of IL-1β in C28/I2 cells (A) and primary human chondrocytes (PHCs) (B) at 8 hours. MATN3 also inhibits the IL-1β-induced downregulation of aggrecan (ACAN) gene expression in PHCs at 36 hours (C). Whole limb mRNA analysis of newborn MATN3 knockout (KO) mice of the C57BL/6J background reveal lower basal gene expression of Col21a and Acan compared to wild-type mice of the same genetic background (D). For all cell culture experiments, rh MATN3 protein is used at 200 ng/ml. Rh IL-1β protein treatment is always 5.0 ng/ml. *Significant differences (P ≤ 0.05) from the untreated control group; ^#significant differences (P ≤ 0.05) from the IL-1β only treated group. For cell culture experiments, data are representative of three individual experiments (n = 5 for mouse whole limb mRNA studies).

Figure 3

Matrilin-3 MATN3) inhibits the expression of several osteoarthritis (OA)-associated proteases. Recombinant human (rh) MATN3 protein inhibits the gene expression of metalloproteinase (MMP)-13 in a dose-dependent manner in the presence and absence of IL-1β in C28/I2 cells (A). MATN3 inhibits IL-1β-induced MMP-13 mRNA levels (B) and protein levels in primary human chondrocyte (PHC)-conditioned media (C). MATN3 inhibits the mRNA expression of OA-associated proteases ADAMTS-4 and ADAMTS-5 in C28/I2 cells (D) and PHCs (E). MMP-13 mRNA and protein analysis were conducted after MATN3 treatment for 24 hours. In these experiments, rh MATN3 protein is used at one of two concentrations: lower dose (+) of 100 ng/ml, or higher dose (++) of 200 ng/ml. Rh IL-1β protein treatment is always 5.0 ng/ml. *P ≤ 0.05, **P ≤ 0.01, for statistically significant differences from the untreated control group; ^#P ≤ 0.05, ^##P ≤ 0.01 for statistically significant differences from the IL-1β only treated group. Data are representative of three individual experiments.

Figure 4

Small interfering RNA significantly suppresses mRNA and protein levels of IL-1Ra. Knocking down all isoforms of IL-1Ra via IL1RN siRNA treatment reduces the IL-1β induction of this gene as observed at two time points (24, 36 hours) post cell culture treatment (A, B). This also results in significant reduction of soluble IL-1Ra protein levels in primary human chondrocyte (PHC) cell supernatants after 24 hours (C). PHCs treated with a nonspecific scrambled siRNA construct is used as the control group. For these experiments, recombinant human (rh) matrilin-3 (MATN3) protein is used at 200 ng/ml. Rh IL-1β protein treatment is always 5.0 ng/ml. *P ≤ 0.05 for statistically significant differences between groups. Data are representative of three individual experiments.

Figure 5

Matrilin-3 MATN3) regulation of type II collagen (COL2A1) and aggrecan (ACAN) gene expression, as well as COL2A1 protein expression requires IL-1Ra. Knocking down IL-1Ra abolishes the ability of recombinant human (rh) MATN3 protein to maintain COL2A1 gene expression in primary human chondrocytes (PHCs) that are challenged with IL-1β (A). A significant reduction in the basal gene expression of COL2A1 is also observed in IL-1Ra knock-down PHCs. PHCs cultured for 48 hours in media supplemented with rh MATN3 protein (200 ng/ml) exhibit increased COL2A1 protein levels; however, upon suppressing IL-1Ra via siRNA-based gene knock-down, MATN3 enhancement of collagen is diminished according to western blot analysis (B). Relative quantification of band intensity is the accumulated result of three individual experiments (n = 3). Knocking down of IL-1Ra also abolishes MATN3 stimulation of ACAN gene expression (C). The concentration of rh MATN3 protein used was 200 ng/ml and of IL-1β was 5.0 ng/ml for all experiments. (A, C) Gene expression analysis was conducted 36 hours post exposure to cell culture treatment conditions. ^#P ≤ 0.05 for statistically significant differences from the untreated group; *P ≤ 0.05 for statistically significant differences between groups. (B) *P ≤ 0.05 for statistically significant differences from the untreated control group. Individual experiments were done in biological triplicate per patient sample. Data are representative of five individual experiments.

Figure 6

Matrilin-3 MATN3) inhibition of ADAMTS-5 gene expression require IL-1Ra. Knocking down IL-1Ra in primary human chondrocytes (PHCs) only partially affects MATN3-induced inhibition of ADAMTS-5 (A) and does not seem to significantly affect MMP-13 gene expression (B). The concentration of recombinant human MATN3 protein used was 200 ng/ml and of IL-1β was 5.0 ng/ml. Gene expression analysis was conducted 36 hours post exposure to cell culture treatment conditions. ^#P ≤ 0.05 for statistically significant differences from the untreated group; *P ≤ 0.05 for statistically significant differences between groups. Individual experiments were done in biological triplicate per patient sample. Data are representative of five individual experiments.