Elodie Faure , Julien Wegrzyn 2, Ilaria Bernabei 1, Nicolas Bertheaume 3, Tristan Pascart 3, Thomas Hugle 1, Nathalie Busso 1, Sonia Nasi 1
1 Service of Rheumatology, 2 Service of Orthopedics, Department of Musculoskeletal Medicine, Centre Hospitalier Universitaire Vaudois and University of Lausanne, Switzerland; 3 Mablab, UFRS 3, University of Lille, Lille, France
Background: Pathologic calcification (PC) of cartilage is a hallmark of osteoarthritis (OA). Calcification is a complex biological process initiated by chondrocytes, the only cells present in cartilage. Calcification is represented by deposition of calcium-containing crystals encompassing BCP (basic calcium phosphate) and CPPD (calcium pyrophosphate dihydrate) crystals. Objectives: To set-up a new ex vivo human model of OA to evaluate the effects of ongoing cartilage calcification.
Methods: Explants from femoral and tibial cartilage obtained from 11 OA patients at day of knee replacement surgery were cultured with medium alone (DMEM high glucose, control, NT) or medium supplemented with ascorbic acid and β-glycerophosphate (calcifying medium, CM). CT-scan analysis of the explants were performed at day of explant isolation (day 0) and after 21 days in culture (T21). Histological analysis of the explants was performed at day 21 using Safranin-O staining. In explant supernatants IL-6, MMP-3, -13 were measured by ELISA and glycosaminoglycans (GAG) was analyzed by dimethyl-methylene blue staining. Primary human OA chondrocytes from 3 independent OA patients were cultured in control medium or calcifying medium. Crystals were identified by RAMAN spectrometry.
Results: We found small calcifications at the onset of the experiment (day 0) which in the CM group increased significantly at day 21 (CT scan Bone Volume day 0=1, day 21=1.43, N=11 patients) while there was no increase in the NT group (CT scan Bone Volume day 0=1, day 21=1.09, N=11). Crystals were mainly localized in cartilage superficial layer, as evidenced by CT-scan and Alizarin-red staining of histological sections of explants at day 21. We reproduced the pro-calcifying effect of CM in isolated OA chondrocytes and assessed by Raman spectrometry that hydroxyapatite crystals were formed under this condition. Similar percent of apoptotic chondrocytes (NT = 72±13% versus CM = 77±9%) was found in explants cultured in NT or CM, ruling out apoptosis as a calcification trigger in this model. We next examined IL-6 secretion by NT and CM-treated cartilage explants. At both 10 and 21 days, IL-6 was significantly increased in CM (day 10: 6±5pg/mg tissue in NT group versus 11±8pg/mg tissue in CM group; day 21: 18±14pg/mg tissue in NT versus 24±16pg/mg tissue in CM), further stressing the importance of this cytokine in calcification. Similarly, using primary OA chondrocytes, we also found that CM increased IL-6 secretion. We next asked if CM could impact cartilage structure. Histological examination of cartilage explant sections stained by Safranin-O revealed loss of proteoglycans (loss of Safranin-O staining) in CM treated tissues. In agreement with this latter result GAG content was increased in supernatants of CM-treated tissues. We hypothesized that these effects could be accounted for by increased metalloprotease production such as MMP-3-, and -13. Indeed, we found increased mean of both MMP-3 and -13 in CM-treated explants although these increases did not reach significancy. Using primary OA chondrocytes, we also found that CM increased significantly MMP-13 secretion.
Conclusion: We described here a new model to study de novo cartilage calcification. We showed that these new calcifications increased IL-6 secretion and had deleterious effects on cartilage proteoglycans content. This new model will allow identification of new drugs that could prevent or reduce cartilage calcification in OA.