Suramin Inhibits Osteoarthritic Cartilage Degradation by Increasing Extracellular Levels of Chondroprotective Tissue Inhibitor of Metalloproteinases 3

Abstract

Osteoarthritis is a common degenerative joint disease for which no disease-modifying drugs are currently available. Attempts to treat the disease with small molecule inhibitors of the metalloproteinases that degrade the cartilage matrix have been hampered by a lack of specificity. We aimed to inhibit cartilage degradation by augmenting levels of the endogenous metalloproteinase inhibitor, tissue inhibitor of metalloproteinases (TIMP)-3, through blocking its interaction with the endocytic scavenger receptor, low-density lipoprotein receptor-related protein 1 (LRP1). We discovered that suramin (C₅₁H₄₀N₆O₂₃S₆) bound to TIMP-3 with a K_D value of 1.9 ± 0.2 nM and inhibited its endocytosis via LRP1, thus increasing extracellular levels of TIMP-3 and inhibiting cartilage degradation by the TIMP-3 target enzyme, adamalysin-like metalloproteinase with thrombospondin motifs 5. NF279 (8,8'-[carbonylbis(imino-4,1-phenylenecarbonylimino-4,1-phenylenecarbonylimino)]bis-1,3,5-naphthalenetrisulfonic acid hexasodium salt), a structural analog of suramin, has an increased affinity for TIMP-3 and increased ability to inhibit TIMP-3 endocytosis and protect cartilage. Suramin is thus a promising scaffold for the development of novel therapeutics to increase TIMP-3 levels and inhibit cartilage degradation in osteoarthritis.

Fig. 1.

Suramin binds to TIMP-3 and inhibits its cellular endocytosis by LRP1. (A) Glycosaminglycan-binding 96-well plates were coated with suramin (10 µg/ml in PBS, filled circles) or PBS (open circles) and blocked in 0.2% gelatin in PBS. Wells were then incubated with TIMP-3 (0.4–50 nM) and binding was detected using an M2 anti-FLAG antibody (mean ± S.D., n = 3). (B) Medium-binding 96-well ELISA plates were coated with LRP1 (5 nM) and blocked with 10% BSA in TNC buffer. Wells were then incubated with TIMP-3 (0.4–50 nM), either alone (open circles) or preincubated with suramin (filled circles, 200 µg/ml, 1 hour, 37°C) and binding was detected using an M2 anti-FLAG antibody (mean ± S.D., n = 3). (C) HTB94 cells were incubated with recombinant TIMP-3 (1 nM) or TIMP-3 preincubated with suramin (200 µg/ml, 1 hour, 37°C) for 0–8 hours and TIMP-3 remaining in the medium was analyzed by immunoblotting and densitometry (mean ± S.D., n = 4). TIMP-3 (open circles) was taken up from the medium with a half-life of 4.0 ± 1.3 hours, whereas TIMP-3 preincubated with suramin (filled circles) was minimally endocytosed. (D) HTB94 chondrosarcoma cells were incubated with suramin (50–200 µg/ml) in serum-free DMEM for 30 hours. Conditioned media were concentrated by TCA precipitation and TIMP-3 levels were analyzed by immunoblotting and densitometry. Values are expressed relative to the amount of TIMP-3 in the medium of untreated cells, defined as 1 (mean ± S.D., n = 5; ***P ≤ 0.001 by one-way ANOVA with Bonferroni’s correction). (E) HTB94 chondrosarcoma cells were treated with suramin (0–250 µg/ml, 18 hours) and expression of TIMP-3 mRNA was analyzed by quantitative PCR relative to RPLP0. TIMP-3 expression in the absence of suramin was defined as 1 (mean ± S.D., n = 3, P > 0.05 by one-way ANOVA with Bonferroni’s correction). (F) HTB94 cells were treated with suramin (0–250 µg/ml) or sodium nitroprusside (10 mM) for 72 hours and cell viability was assessed using MTS (mean ± S.D., n = 3; ***P ≤ 0.001 by one-way ANOVA with Bonferroni’s correction). (G) Primary chondrocytes were isolated from human OA or porcine cartilage and incubated with suramin (0–250 µg/ml) in serum-free DMEM for 48 hours. Conditioned media were concentrated by TCA precipitation and TIMP-3 levels were analyzed by immunoblotting. (H) Human OA chondrocytes were treated with suramin (0–250 µg/ml, 48 hours) and expression of TIMP-3 mRNA was analyzed by quantitative PCR relative to RPLP0, with expression in the absence of suramin defined as 1 (n = 5 donors, mean ± S.D., P > 0.05 by one-way ANOVA with Bonferroni’s correction). (I) Human OA chondrocytes were treated with suramin (0–250 µg/ml) and/or retinoic acid (1 μM) for 48 hours and cell viability assessed using MTS (mean ± S.D., n = 3 donors, P > 0.05 by two-way ANOVA with Bonferroni’s correction). MTS, 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium; PCR, polymerase chain reaction; RA, retinoic acid; SNP, sodium nitroprusside; TCA, trichloroacetic acid.

Fig. 2.

Suramin does not impair inhibitory activity of TIMP-3. (A) HTB94 cells were incubated with suramin (0–250 µg/ml, 18 hours) and expression of TIMP-1 mRNA was analyzed by quantitative PCR relative to RPLP0. TIMP-1 expression in the absence of suramin was defined as 1 (mean ± S.D., n = 3, P > 0.05 by one-way ANOVA with Bonferroni’s correction). (B) HTB94 cells were incubated with suramin (0–250 µg/ml, 18 hours) and expression of TIMP-2 mRNA was analyzed by quantitative PCR relative to RPLP0. TIMP-2 expression in the absence of suramin was defined as 1 (mean ± S.D., n = 3, P > 0.05 by one-way ANOVA with Bonferroni’s correction). (C) HTB94 cells were incubated with suramin (0–250 µg/ml, 18 hours) and expression of LRP1 mRNA was analyzed by quantitative PCR relative to RPLP0. LRP1 expression in the absence of suramin was defined as 1 (mean ± S.D., n = 3; ***P ≤ 0.001 by one-way ANOVA with Bonferroni’s correction). (D) HTB94 cells were treated with suramin (0–250 µg/ml, 30 hours), and conditioned media were concentrated by TCA precipitation and analyzed by immunoblotting using an 8G1 anti-LRP1 antibody. (E) ADAMTS-5 (0.5 nM) was incubated (1 hour, 37°C) with TIMP-3 (0.5–5 nM) and combinations of suramin (0.05 µg/ml) or PPS (0.05 µg/ml). Residual activity against a fluorescent peptide substrate was determined, and K_i(app) values (expressed in nM) were calculated from equilibrium rates of substrate hydrolysis using the tight binding equation (mean ± S.D., n = 4–5; *P ≤ 0.05 by one-way ANOVA with Bonferroni’s correction). (F) TIMP-3 (0.3–50 nM) was incubated (1 hour, 37°C) with MMP-1 (0.5 nM) or MMP-3 (1 nM) in the presence or absence of suramin (0.05 µg/ml). Residual activity against a fluorescent peptide substrate was determined, and K_i(app) values (expressed in nM) were calculated from equilibrium rates of substrate hydrolysis using the tight binding equation (mean ± S.D., n = 3; P > 0.05 by Student’s t test). ns, not significant.

Fig. 3.

Suramin inhibits aggrecan degradation in retinoic acid–stimulated porcine and human cartilage explants. (A) Human OA cartilage explants were stimulated with IL-1 (10 ng/ml) in the presence of suramin (0–100 µg/ml) for 48 hours and degraded aggrecan fragments released into the conditioned medium quantified using the DMMB assay (mean ± S.D., n = 3; *P ≤ 0.05; ***P ≤ 0.001 by two-way ANOVA with Bonferroni’s correction). (B) Porcine cartilage explants were stimulated with IL-1 (10 ng/ml) in the presence of suramin (0–250 µg/ml) for 48 hours and degraded aggrecan fragments released into the conditioned medium were quantified using the DMMB assay (mean ± S.D., n = 3; ***P ≤ 0.001 by two-way ANOVA with Bonferroni’s correction). (C) Human OA cartilage explants were stimulated with retinoic acid (1 µM) in the presence of suramin (0–200 µg/ml) for 48 hours and degraded aggrecan fragments released into the conditioned medium were quantified using the DMMB assay (mean ± S.D., n = 3; ***P ≤ 0.001 by two-way ANOVA with Bonferroni’s correction). (D) Porcine cartilage explants were stimulated with retinoic acid (1 µM) in the presence of suramin (0–250 µg/ml) for 48 hours and degraded aggrecan fragments released into the conditioned medium were quantified using the DMMB assay (mean ± S.D., n = 3; ***P ≤ 0.001 by two-way ANOVA with Bonferroni’s correction). (E) Conditioned media from (D) were analyzed using neo-epitope antibodies that recognize the ¹⁷⁷²AGEG or ³⁷⁴ARGSV termini generated by ADAMTS cleavage of aggrecan. (F) Porcine cartilage was incubated with retinoic acid (1 µM, 30 hours) in the presence of either a TIMP-3 antibody (MAB973, 50 µg/ml) or an isotype control (mouse IgG1, 50 µg/ml). Aggrecan degradation was quantified using the DMMB assay (mean ± S.D., n = 4; **P ≤ 0.01; ***P ≤ 0.001 by two-way ANOVA with Bonferroni’s correction). ns, not significant; RA, retinoic acid.

Fig. 4.

Suramin analog NF-279 shows improved ability to block TIMP-3 uptake and to protect cartilage. (A) Porcine cartilage explants were treated with retinoic acid (1 µM) and/or suramin analogs (200 µg/ml) for 48 hours. Cartilage degradation was assessed by quantifying aggrecan fragments (mean ± S.D., n = 3; ***P ≤ 0.001 by two-way ANOVA with Bonferroni’s correction) released into the medium using the DMMB assay. (B) Glycosaminglycan-binding 96-well plates were coated with 10 µg/ml suramin (red circles, EC₅₀ = 1.90 ± 0.21 nM), NF279 (green triangles, EC₅₀ = 0.85 ± 0.09 nM), NF110 (filled boxes, EC₅₀ = 6.88 ± 0.96 nM), NF157 (filled triangles, EC₅₀ = 1.55 ± 0.11 nM), NF449 (filled diamonds, EC₅₀ = 4.66 ± 0.69 nM), NF023 (open circles, EC₅₀ = 26.5 ± 7.5 nM), NF340 (open squares, EC₅₀ > 100 nM), or NF546 (open triangles, EC₅₀ > 100 nM) in PBS and blocked in 0.2% gelatin in PBS. Wells were then incubated with TIMP-3 (0.4–50 nM) and binding was detected using an M2 anti-FLAG antibody (mean ± S.E., n = 3 technical repeats). (C) HTB94 chondrosarcoma cells were cultured in the presence of suramin analogs (200 µg/ml) for 36 hours and TIMP-3 levels in the conditioned medium were evaluated by Western blotting and quantified by densitometry (mean ± S.D., n = 4, suramin defined as 100%; ***P ≤ 0.001 by one-way ANOVA with Bonferroni’s correction). (D) For each analog, TIMP-3 accumulation [from (C)] was plotted against the log₁₀ of analog affinity for TIMP-3 [EC₅₀ from (B)] and Pearson correlation coefficients were calculated using GraphPad Prism. (E) For each analog, TIMP-3 accumulation [from (C)] was plotted against aggrecan release [from (A)] and Pearson correlation coefficients were calculated using GraphPad Prism. (F) Structural formulae of suramin and its analog, NF279. RA, retinoic acid.