Background Osteoarthritis (OA) is a multifaceted joint disease defined by cartilage degradation leading to joint destruction. Ageing and mechanical overload are the two main risk factors of OA. Primary cilium, located at the surface of chondrocytes, is an organelle that may act as a mechanosensor. Galectin 3 (gal-3) is a 30-kDa lectin expressed in several cell types, including cartilage and bone cells. Gal-3 has essential role in ciliogenesis and primary cilium function. In kidney epithelial cells, it is associated with the centrosome, notably at the basis of primary cilia, and major defects in ciliogenesis occur in its absence. Moreover, these defects have functional consequences in gal-3 KO mice: impaired cilia beating performance of the tracheal epithelial cells, associated with defects in cilium formation.

Objectives To evaluate the role of gal-3 in chondrocyte primary cilium formation and function and its role in OA

Methods In vivo, mouse OA was induced by medial meniscectomy (MNX) of the right knee of WT and gal-3 null mice (KO). OA lesions were assessed according to OARSI recommendations using conventional histology and safranin-O staining. In vitro, primary chondrocytes were isolated from the knee and hip of WT and gal-3 KO new born mice. Chondrocytes apoptosis was assessed by TUNEL labelling, caspase 3 activity and cytochrome c release. Chondrocyte primary cilium was assessed by confocal microscopy using antibodies against gamma-tubulin and acetylated alpha-tubulin. Primary cilium of knee cartilage chondrocytes was assessed on microtome-cryostat sections (non-operated and MNX of WT and gal-3 KO mice)

Results In vivo, OA induced by MNX was more severe in gal-3 KO mice compared to WT mice. Cartilage lesions (OARSI score 6,5±4,8 vs 2,5±2,3, p=0,05, n=8) and proteoglycan loss were more important in gal-3 KO mice than WT mice. Expression of gal-3 dramatically decreased in non-calcified cartilage of WT mice. In OA cartilage, chondrocyte apoptosis was more frequent in gal-3 KO cartilages compared to WT samples (27.7%±6.2 vs 15.6%±3.3 p<0.02). Similarly, in vitro, primary chondrocyte apoptosis induced by actynomycin D was more important in gal-3 KO chondrocytes than WT chondrocytes (TUNEL labeling chondrocyte 21.8% vs 10.4%; caspase 3 activity/μg of protein 5220±847 vs 3512±608; cytochrome c release 45.5% vs 25.9%, all p<0.05, respectively). In vitro, gal-3 was co-localized with gamma-tubuline at the basal foot of primary cilium of WT chondrocytes. In the absence of gal-3 primary cilium was more frequently abnormal (29.4±8.4% vs 12.3±3.9%). Several types of abnormal primary cilia were observed: stunted or duplicated, S-form axonema. In vivo, in non-operated cartilages, WT chondrocytes had their primary cilium situated in 17.4%, 49.9% and 32.5% at the articular surface side of the cell, subchondral bone side and lateral side, respectively, while gal-3 KO chondrocytes had their primary cilium situated in 31.8%, 37.5% and 30.5%, respectively. After MNX, the proportion of primary cilium situated at the articular surface side increased in both WT and gal-3 KO mice.

Conclusions Gal3 is involved in chondrocyte primary cilia formation and regulates chondrocyte apoptosis through intrinsic pathway. Its role in mechanical sensing is ongoing.

Disclosure of Interest None declared