IL-37 diminishes proteoglycan loss in human OA cartilage: donor-specific link between IL-37 and MMP-3  E.W. van Geffen, A.P.M. van Caam, W. Schreurs, F.A. van de Loo, P.L.E.M. van Lent, M.I. Koenders, C.S. Thudium, A.C. Bay-Jensen, E.N. Blaney Davidson, P.M. van der Kraan  Osteoarthritis and Cartilage  Volume 27, Issue 1, Pages 148-157 (January 2019) DOI: 10.1016/j.joca.2018.08.016 Copyright © 2018 Osteoarthritis Research Society International Terms and Conditions

Fig. 1 Decrease in sGAG release from human OA cartilage explants by rhIL-37. Quantification of the sGAG release in supernatant of cartilage explants cultures in presence of rhIL-37 (1, 10, 100 ng/ml) by the DMMB-assay. Data are represented as mean ± (S.E.M), n = 10; * = P < 0.05, ** = P < 0.001 as determined via a one-way analysis of variance with Bonferonni multiple comparison test on day 6 time point. Osteoarthritis and Cartilage 2019 27, 148-157DOI: (10.1016/j.joca.2018.08.016) Copyright © 2018 Osteoarthritis Research Society International Terms and Conditions

Fig. 2 No effect of rhIL-37 on 35S-sulphate incorporation and proteoglycan expression in human OA cartilage explants. A) Quantification of the 35S content of cartilage explants as measured by liquid scintillation counting. Explants were incubated for 48 h with 1, 10 or 100 ng/ml rhIL-37 and subsequently labeled for 4 h with 35S-sulphate. Data are represented as individual data points with unique markers for each individual with 95% CI n = 10. B) Relative gene expression of the large proteoglycans aggrecan, versican and perlecan in human OA cartilage explants after incubation for 48 h with rhIL-37 (1, 10 or, 100 ng/ml) as determined by qPCR. Data are represented as individual data points with unique markers for each individual with 95% CI + = mean, n = 10. C) Relative gene expression of the small leucine-rich proteoglycans (SLRPs): decorin, biglycan, fibromodulin and lumican in human OA cartilage explants after incubation for 48 h with rhIL-37 (1, 10 or 100 ng/ml) as determined by qPCR. No changes were observed. Data are represented as individual data points with unique markers for each individual with 95% CI, n = 10. Osteoarthritis and Cartilage 2019 27, 148-157DOI: (10.1016/j.joca.2018.08.016) Copyright © 2018 Osteoarthritis Research Society International Terms and Conditions

Fig. 3 RhIL-37 decreases loss of 35S-sulphate labeled sGAGs and MMP expression in human OA cartilage explants. A) Quantification of the 35S content of cartilage explants pre-labeled for 4 h with 35S-sulphate and subsequently incubated for 48 h with rhIL-37 (1, 10 or 100 ng/ml) as determined by liquid scintillation counting. Data are represented as individual data points with unique markers for each individual with 95% CI, n = 10. B) Relative gene expression of the proteolytic enzymes MMP-3, MMP-13 and ADAMTS-5 in human OA cartilage explants after incubation for 48 h with rhIL-37 (1, 10 or 100 ng/ml) as determined by qPCR. Data are represented as individual data points with unique markers for each individual with 95% CI, n = 10. C and D) Protein expression of MMP-3 and MMP-13 in cartilage explants after 48 h incubation with rhIL-37 (1, 10 or 100 ng/ml) as determined by Western blot. Quantification of the Western Blot was performed by Image J. Data are represented as individual data points with unique markers for each individual with 95% CI, n = 5; * = P < 0.05 as determined via one-way analysis of variance with Bonferonni multiple comparison test. Osteoarthritis and Cartilage 2019 27, 148-157DOI: (10.1016/j.joca.2018.08.016) Copyright © 2018 Osteoarthritis Research Society International Terms and Conditions

Fig. 4 Positive correlation between rhIL-37 and MMP-3-inhibition, but not ADAMTS-5-inhibition, in sGAG release from human OA cartilage explants. A) Quantification of the sGAG release from human OA explants incubated for 6 days with the MMP-3-, MMP-13-, and ADAMTS-5-inhibitor as determined by the DMMB-assay. B) Quantification of the sGAG release from cartilage explants that do respond to both MMP-3-inhibitor and rhIL-37 (1 ng/ml) as determined by the DMMB-assay. C) Quantification of the sGAG release from cartilage explants on day 6 of culture, that do respond to the ADAMTS-5-inhibitor as determined by the DMMB-assay. D) Quantification of the sGAG release from human OA explants incubated for 6 days with rhIL-37 (1 ng/ml) and the MMP-3-inhibitor as determined by the DMMB-assay. Data are represented as individual data points with unique markers for each individual with 95% CI n = 10; *P < 0.05 as determined via one-way analysis of variance with Bonferonni multiple comparison. Osteoarthritis and Cartilage 2019 27, 148-157DOI: (10.1016/j.joca.2018.08.016) Copyright © 2018 Osteoarthritis Research Society International Terms and Conditions

Supplementary Fig. 1 Schematic overview of the structure of the proteoglycan aggrecan and the preferable cleavage sites of the MMP and ADAMTS enzymes. Proteoglycans are essential structural components of the cartilage extracellular matrix. The most abundant proteoglycan on weight base in articular cartilage is aggrecan. Aggrecan is bound to a hyaluronan molecule in the extracellular matrix of the cartilage via a link protein. This link protein is bound to the G1 domain of the core protein of the aggrecan molecules. Between the G2 and G3 domain of the core protein, keratin sulphate and chondroitin sulphate groups are present. During OA, proteoglycan degradation takes place. ADAMTS en MMP enzymes cleave the core protein preferably between the G1 and G2 domain. MMP enzymes generate the C-terminal VDIPEN neo-epitope and release the FFGV neo-epitope, whereas ADAMTS enzymes generate the C-terminal NITEGE neo-epitope and release the ARGS neo-epitope1. Osteoarthritis and Cartilage 2019 27, 148-157DOI: (10.1016/j.joca.2018.08.016) Copyright © 2018 Osteoarthritis Research Society International Terms and Conditions

Supplementary Fig. 2 Functionality of the MMP-3- MMP-13- and ADAMTS-5-inhibitor. Quantification of the sGAG release from human OA explants after 2 h pre-incubation with MMP-3-, MMP-13-, and ADAMTS-5- inhibitor followed by a 48 h stimulation with APMA (1 mM) and TNFα (50 μg/ml). APMA induces sGAG release from human OA cartilage explants. This induction was reduced by the MMP-3-inhibitor. The induction in sGAG release by TNFα could be blocked by both the ADAMTS-5 and MMP-13-inhibitor. Results demonstrate the functionality of the used inhibitors. Data are represented as individual data points with unique markers for each individual with 95% CI, n = 5; * = P < 0.05 as determined via one-way analysis of variance with Bonferonni multiple comparison test. Osteoarthritis and Cartilage 2019 27, 148-157DOI: (10.1016/j.joca.2018.08.016) Copyright © 2018 Osteoarthritis Research Society International Terms and Conditions

Supplementary Fig. 3 Histological safranin O/ fast green staining of human OA cartilage explants. To investigate if human OA cartilage explants lose their GAG content in our culture system, human OA cartilage explants were fixed immediately at day 0 or after 6 days of in vitro culture for histological staining with safranin O/ fast green. Safranin O is a basic dye, for staining of proteoglycans. The intensity of the safranin O staining, correlates with the proteoglycan content40,41. Explants at day 0 show an intense red staining of the cartilage. Proteoglycan depletion and cartilage fissures can be observed at the superficial layer of the cartilage. After 6 days of culture, loss of safranin O is present, indicating proteoglycan depletion in the cartilage explants after culturing. Osteoarthritis and Cartilage 2019 27, 148-157DOI: (10.1016/j.joca.2018.08.016) Copyright © 2018 Osteoarthritis Research Society International Terms and Conditions

Supplementary Fig. 4 Neo-epitope measurement of MMP and ADAMTS-generated aggrecan fragments in supernatant of cartilage explants cultures. A reduction in proteoglycan release can occur when the balance between proteoglycan synthesis or degradation is disturbed. To investigate if rhIL-37 inhibits the degradation of sGAGs, aggrecan degradation fragments were measured in the culture media of explants incubated with rhIL-37 (1, 10 or 100 ng/ml) for 6 days by ELISA. Aggrecan neo-epitope expression of the ADAMTS-generated NITEGE- and ARGS neo-epitopes and the MMP-generated FFGV neo-epitope were measured. A trend towards decrease in the ADAMTS-generated ARGs and NITEGE neo-epitope levels was observed after addition of rhIL-37. Furthermore, rhIL-37 induced a trend towards decrease in the MMP-generated FFGV neo-epitope levels in the supernatant. Although, no changes were observed, these results indicate for a reduction in MMP and ADAMTS activity in cartilage explants by rhIL-37. Data are represented as individual data points with unique markers for each individual with 95% CI, n = 10. Osteoarthritis and Cartilage 2019 27, 148-157DOI: (10.1016/j.joca.2018.08.016) Copyright © 2018 Osteoarthritis Research Society International Terms and Conditions

Supplementary Fig. 5 sGAG release measurement of cartilage explants derived from hip (n = 10) or knee (n = 10) joint after incubation with rhIL37, MMP3-inhibitor, MMP13-inhibitor or ADAMTS5-inhibitor.A reduction in sGAG release was observed in both hip and knee cartilage, after incubation of the explants for 6 days with the MMP-3 inhibitor (1 μM) or ADAMTS-5 inhibitor (50 μM). No effect was observed from the MMP-13-inhibitor (1 μM). Between hip and knee cartilage, no prove for differences were found in their response to the different inhibitors used. Data are represented as individual data points with unique markers for each individual with 95% CI, n = 10. Osteoarthritis and Cartilage 2019 27, 148-157DOI: (10.1016/j.joca.2018.08.016) Copyright © 2018 Osteoarthritis Research Society International Terms and Conditions