Secreted 14-3-3ε: discovery by proteomics of a novel biomarker and/or therapeutical target in osteoarthritis

Abstract

s / Osteoarthritis and Cartilage 21 (2013) S63–S312 S225 Figure 2. Assessment of OA severity. (A) Representative joint histology of the medial tibiofemoral contact region in left (DMM surgery) hind limbs (F 1⁄4 femur, M1⁄4 meniscus, T 1⁄4 tibia). Severe cartilage cleft and loss were found in the joints of u6 and SFA mice. Scale bar is 500 mm (B) Modified Mankin scores representative of OA severity. Data was analyzed by two-factor (diet and surgery) repeated measures ANOVA. Both factors and their interaction were significant (p < 0.01). Tukey's HSD post-hoc test on diet showed that the joints from Control and u3 diet had significantly lower score compared to the joints from u6 and SFA diet (p < 0.05). All operated limbs had significantly higher score than their corresponding control non-operated (right) joints (p < 0.01) and the operated joints of u6 and SFA mice also had worse OA as compared to the operated joints of Control and u3 mice (* p < 0.05). n 11 mice per diet group. Data shown as mean standard error. Figure 3. Assessment of joint synovitis. (A) Representative joint histology of the medial femoral condyle of left (DMM surgery) hind limbs (F 1⁄4 femur, M1⁄4 meniscus, T 1⁄4 tibia, S 1⁄4 synovium). Thickened synovium from u6 and SFA diet with a high number of infiltrated cells was observed (indicated by black arrows). Scale bar is 500 mm. (B) Total Joint synovitis scores are representative of the degree of synovial inflammation. Data was analyzed by two-factor (diet and surgery) repeated measures ANOVA. Both factors but not their interaction were significant (p < 0.01). Tukey's HSD post-hoc test on diet showed that the joints from u3 diet had significantly lower synovial inflammation compared to the joints from SFA and u6 diet (* p < 0.05). n 11 mice per diet group. Data shown as mean standard error. Conclusions: Our results indicate that dietary fatty acids differentially influence the risk for development of OA following joint injury, with u6 and SFA diets significantly increasing OA severity following DMM. Despite the fact that mice fed pro-inflammatory high-fat diets showed worse OA, they had similar locomotor activity levels as the control mice, indicating that changes in voluntary exercise did not significantly contribute to OA in obese mice. The significant association between the different diets and OA severity further suggests a role of fatty acids and their lipid mediators in OA development and OA-related joint inflammation. 425 SECRETED 14-3-3e: DISCOVERY BY PROTEOMICS OF A NOVEL BIOMARKER AND/OR THERAPEUTICAL TARGET IN OSTEOARTHRITIS S. Priam, C. Bougault, X. Houard, M. Gosset, F. Berenbaum, C. Jacques. UR 4 Univ. Pierre et Marie Curie, Paris, France Purpose: Several experiments suggest that subchondral bone remodeling triggered by overload could initiate and/or contribute to cartilage loss in osteoarthritis (OA) through a bone/cartilage interplay. In order to find novel biomarkers and/or therapeutical targets, we used a secretomic approach in a novel and unique bone/cartilage communication model. Methods: Murine experiments: Thanks to a three dimensional (3D) culture model, murine osteoblasts were submitted to compression in Abstracts / Osteoarthritis and Cartilage 21 (2013) S63–S312 S226s / Osteoarthritis and Cartilage 21 (2013) S63–S312 S226 Biopress Flexercell plates (1.7 MPa, 1Htz during 24h). Conditioned media from compressed (CM) or uncompressed (UCM) osteoblasts were used to stimulate mouse articular chondrocytes. Then, soluble mediators released by compressed osteoblasts were identified by iTRAQ , a differential secretomic analysis approach. Chondrocyte expression of matrix metalloproteinase 3 (MMP-3) and MMP-13, tissue inhibitors of metalloproteinases (TIMPs), and cartilage extracellular matrix components type II collagen and aggrecan were assessed by RT-PCR, western blot analysis and ELISA. Immunodepletion and blocking antibody experiments were realized using mouse compressed conditioned media. Human experiments: OA tissue was obtained from patients undergoing total knee replacement. Subchondral bone and cartilage from tibial plateaus and femoral condyles were isolated, cut into small pieces, washed and incubated in culture medium. Conditioned media were then collected, centrifuged and used to stimulate human articular chondrocytes. Results: Media from compressed osteoblast (CM) strongly induced MMP-3 and -13 chondrocyte mRNA expression. Consistently, CM osteoblasts-derived media also significantly stimulated the releases of MMP-3 and -13 by chondrocytes. CM osteoblasts-derived media also affected cartilagematrix proteins expressions by downregulating type II collagen mRNA. Effects of CM osteoblasts-derived media on cytosolic type II collagen protein amounts were confirmed by western blot. So, the ability of soluble mediators released by loaded osteoblasts to induce a chondrocyte pro-catabolic phenotype was demonstrated. In order to identify osteoblast-derived soluble mediators responsible for this chondrocyte phenotype, osteoblast-derived conditioned media were analyzed by iTRAQ . This sophisticated proteomic technique allowed identification of 105 proteins secreted by osteoblasts among which only 10% were upregulated in response to compression. Among them, secreted 14-3-3e (s14-3-3 e) dose-dependently induced the release of pro-catabolic factors (MMP-3, MMP-13) by mouse chondrocytes. s14-3-3e dramatically mimicked the effects of CM osteoblastsderived media. MMP-3 and MMP-13 chondrocyte expressions were strongly inhibited by a s14-3-3 blocking antibody or by immunodepleting s14-3-3 when added into the CM osteoblasts-derived media. Furthermore, s14-3-3ewas strongly released by human OA subchondral bone and paralleled with the increase of MMP-3 expression in these human samples. MMP-3 expression was dose dependently stimulated by recombinant s14-3-3e in human OA chondrocytes. Conclusions: We have identified s14-3-3e as a novel soluble mediator, critical in the communication between subchondral bone and cartilage, that opens new perspectives in the fields of biomarkers and therapeutical targets in osteoarthritis. 426 WNT5A/CAMKII PATHWAY IS ACTIVATED IN OSTEOARTHRITIS AND PROMOTES LOSS OF CHONDROCYTE PHENOTYPE G. Nalesso y, B.L. Thomas y, S.E. Eldridge y, K. Wagner y, J. Sherwood z, J. Bertrand z, T. Pap z, C. Pitzalis y, F. Dell'Accio y. yWilliam Harvey Res. Inst., London, United Kingdom; z Inst. fur Experimentelle Muskuloskelettale Medizin, Muenster, Germany Purpose of the study: Disruption of Wnt signalling results in osteoarthritis both in humans and experimental models. However, the role of Wnt pathways in the pathogenesis of this disease has not been fully characterized yet. We have recently demonstrated that Wnt-dependent activation of the Calcium Calmodulin Kinase II (CaMKII) is detrimental for the maintenance of articular chondrocyte homeostasis in vitro (Nalesso et al., JCB 2011). The aim of this study was then to determine the role of the Wnt/CaMKII pathway in human articular cartilage and in the development of osteoarthritis. Methods: Cartilage explants were removed from the knees of patients undergone joint replacement for osteoarthritis. Osteoarthritis was induced in 8 weeks old 129/Sv mice by destabilization of the medial meniscus (DMM) and the mice were sacrificed 8 weeks post-surgery. The expression level of the different CaMKII isoforms in articular cartilage was evaluated by PCR and immunohistochemistry. CaMKII phosphorylation was assessed by immunofluorescence. Articular chondrocytes were isolated by enzymatic digestion from the metatarsal joint of a bovine hoof obtained from a local abattoir. The chondrocytes were stimulated with recombinant Wnt5a in presence/absence of KN93, a CaMKII inhibitor. Calciummobilization was detected by cellular accumulation of FURA-2 dye and phosphorylation of CaMKII by Western blotting analysis. Gene expression analysis of chondrocyte markers including Col2A1, Aggrecan and Sox9, was performed by quantitative real time PCR. Results: Only CaMKIIg and -d were expressed in human and murine articular cartilage. Phospho-CaMKII (pCaMKII), the activated form of CaMKII, was expressed only in the cartilage of osteoarthritic patients and was totally absent in normal cartilage. pCaMKII was also significantly upregulated in mice upon DMM in comparison with the contralateral sham-operated joint. The expression of Wnt5a, known to modulate the CaMKII pathway in different biological systems, was upregulated in OA samples. Wnt5a stimulation promoted calcium mobilization and CaMKII phosphorylation in articular chondrocytes. Wnt5a induced loss of chondrocyte phenotype which was rescued by CaMKII inhibition. Conclusions: These results showed that activation of the Wnt5a/ CaMKII pathway correlates with the development of osteoarthritis both in human and in animal models. CaMKII blockade rescued the loss of chondrocyte phenotype induced by Wnt5a in articular chondrocytes, therefore suggesting that the inhibition of this signalling cascade might represent a new therapeutic strategy useful to tackle the development/ progression of osteoarthritis. 427 ANNEXIN A6: A NOVEL THERAPEUTIC TARGET FOR THE TREATMENT OF OSTEOARTHRITIS? T. Kirsch, T. Minashima, K. Campbell, Y. Zhang. NYU Sch. of Med.,