Objective: The purposes of this project were to quantify the cells expressing the mesenchymal stem cell (MSC) marker CD271 in synovial membranes from human osteoarthritic (OA) and healthy joints, and to determine if those CD271 cells were involved in spontaneous human cartilage repair and were beneficial for the repair of human articular cartilage defects.
Methods: The coexpression of CD44/CD271, CD90/CD271, and CD105/CD271 antigens was determined by immunofluorescence in OA and healthy synovial membranes and during spontaneous cartilage repair. Isolated MSCs from the bone marrow of four OA patients (mean age: 64 years) were magnetically separated into MSC CD271+ and MSC CD271- subsets. The separated cell subsets were then implanted into 2 mm focal defects of articular cartilage. These implants were cultured in chondrogenic differentiation medium supplemented with recombinant human transforming growth factor-beta3 for 8 weeks. The repair tissues were analyzed by histochemistry (hematoxylin-eosin and safranin O) and immunohistochemistry for collage types I and II.
Results: Cells expressing the CD271 antigen were diffusely distributed in OA synovial membranes and localized in the subintimal zone in healthy synovial membranes. The number of cells expressing MSC markers was higher in OA synovial membranes than in synovia from healthy joints, corresponding to the highest level of coexpression of CD90/CD271 antigens (9.8% vs. 2.6%). Spontaneous repair tissue showed more cells expressing the CD271 antigen (9.9% ± 4.0%). The highest levels of expression were found to be associated with CD44; 64% of positive CD271 cells coexpressed the CD44 antigen. In both implant cell types, the repair tissue morphology resembled articular cartilage, having an extracellular matrix with a hyaline aspect and numerous lacunae containing cells, and was immunopositive for collagen types I and II. Statistical analyses of the repair tissue demonstrated that the implantation of MSC CD271+ provided such benefits as a greater filling of the chondral defect and better integration between the repair tissue and native cartilage. Safranin O staining of repair tissue was negative in implants of MSC CD271- but more positive in implants with MSC CD271+. The overall histologic score for CD271- implants was 9.5 ± 0.89 and 12.19 ± 1.01 for CD271+ implants.
Conclusions: Synovial membranes from OA patients contain more cells expressing CD271 antigen than those from healthy joints, and spontaneous cartilage repair tissue contains cells positive for CD271 antigen. These data suggest the involvement of CD271 antigen in spontaneous cartilage repair and indicate that the cell subset MSC CD271+ provides higher quality chondral repair than the CD271- subset.