Purpose: We identified Prohibitin (PHB1) in the transcriptional repressor complex turning-off PITX1 gene expression in OA cartilage. PITX1 gene encodes a transcription factor, which plays an essential role during development for the formation of joints and skeletal elements of the lower limbs, as well as to maintain bone and cartilage homeostasis. PITX1 expression was completely lost in OA patients. In the present study, we investigate the role of SUMOylation in the nuclear accumulation and trapping of PHB1 in primary OA. Methods: Cartilage specimens were obtained from tibial plateaus and femoral condyle tissues from 27 OA patients (10 male, 17 female; mean age 65 ± 20 years) undergoing total knee joint replacement. For comparison, we included normal cartilage samples of 4 trauma patients without osteoarthritis and who reported no family history of OA (2 male, 2 female; mean age 44 ± 28 years). All the samples were collected with the written consent of the patients. The Institutional Research Ethics Boards of Sainte-Justine University Hospital and Maisonneuve-Rosemont Hospital, Montreal, Canada, approved the study protocol. To study the intracellular localization of PHB1, PML, SUMO1, SUMO2/3 and UBC9 in OA patients, we carried out immunofluorescence (IF) staining of human articular chondrocytes in OA patients and control subjects with corresponding antibodies. To investigate whether the nuclear trapping of PHB1 is triggered by its SUMOylation, we first carried-out classical in vitro SUMOylation assays. In order to identify the source of increased SUMOylation in primary knee joint OA, using different complementary approaches, we investigated the contribution of UBC9, which is the unique E2 ligase involved in the SUMOylation pathway. Results: We detected an overall increase in small ubiquitin-related modifier (SUMO) conjugation activity in knee/hip OA cartilage. Co-localization experiments with antibodies against promyelocytic leukemia (PML) proteins and either SUMO1 or SUMO2/3, confirmed that SUMO proteins were primarily localized in the PML nuclear bodies (NBs) of the nuclei of OA chondrocytes, when compared to control chondrocytes. Interestingly, the size of PML NBs is increased in OA chondrocytes and could be explained by a recent report showing a significant decrease in the expression of SENP6 in OA hip cartilage relative to non-OA hip control cartilage. Indeed, SENP6 depletion drastically increases the size of PML NBs and preferentially results in the formation of SUMO2-3 conjugates with PML. We did not find any evidence that PHB1 is directly SUMOylated. However, by generating several PHB1 constructs, we could conclusively demonstrate that PHB1 can interact with SUMO1-SUMOylated nuclear partners through a SUMO-binding motif (SBM). Interestingly, Ubc9 transgenic mice in which Ubc9 is strongly expressed in all tissues under the chicken β-actin promoter, showed an OA-like phenotype at their knee joints when compared to age- and gender matched normal mice. Conclusions: Our current evidence suggests that mitochondrial depletion of PHB1 through its nuclear sequestration could provide a first mechanism explaining concomitantly the mitochondrial dysfunction associated with OA and the repression of PITX1 induced by nuclear PHB1. Although PHB1 nuclear localization has previously been reported in many cell types, this study is the first to propose a molecular mechanism involving SUMOylation in PHB1 nuclear accumulation and its contribution to OA pathogenesis. We presented evidence that PHB1 nuclear accumulation in OA chondrocytes requires its interaction with SUMO1-modified proteins, since the removal of its SBM abrogated this event. Understanding the molecular mechanisms by which elevation of SUMOylation is triggered and then utilizes nuclear PHB1 in governing chondrocyte hypertrophy, is likely to result in the development of predictive diagnostic tools and therapeutic strategies to overcome OA onset and its progression.