The repair of focal cartilage defects using human embryonic stem cells

Abstract

Purpose: Damage to articular cartilage occurs frequently as a result of joint trauma and disease. Cell based treatments to repair defects have been developed using autologous chondrocytes and bone marrow stem cells and have shown some success. We have investigated the potential of human embryonic stem cells (hESC) as a source of chondrocytes, as they have capacity for unlimited self-renewal and could provide a ready supply of donor cells. In initial work we developed a 14 day culture protocol using serum free, chemically defined medium and generated chondro-progenitors from hESC, which were up to 97% SOX9 positive, expressing COL2A1 and ACAN genes. This system is chemically defined and scalable and with potential to provide cells for clinical grade use. In this study we developed the protocol further and tested hESC derived chondro-progenitors in vivo in focal defects in immunocompromised nude rats. Methods: Human embryonic stem cells (hESC) expanded on feeder culture were transferred to feeder free/ serum free culture on fibronectin coated plates. After two passages a 14 day chondrogenesis protocol was initiated with a sequential series of growth factors, which drive the hESCs through mesendoderm/mesoderm to chondroprogenitors. These cells were characterised by qRT-PCR using a range of chondrocyte specific marker genes, negative controls and by immunofluorescence for SOX9, a chondrocyte transcription factor. Some chondro-progenitors were also derived from hESC transduced with GFP using lentiviral vectors. To test their capacity for cartilage formation, chondro-progenitors in fibrin gel (3X106 cells/ml) were implanted into osteochondral defects (2mm diameter, 2mm deep, 14 defects in 8 animals) in the patella groove of nude rats. Joint tissue was removed, decalcified, fixed and sectioned for histological and immunochemical analysis after 4 weeks or 12 weeks. Results: The hESC derived chondro-progenitors at the end of the protocol formed cell clusters and showed chondrocyte properties, including high expression of SOX9. They strongly immunostained for SOX9 protein and for collagen II and for aggrecan and expressed collagen II and XI genes, but not collagen I and negligible collagen X, which is a marker for hypertrophic chondrocytes. The expression of matrilin 3 increased and was more than 50 time higher than matrilin 1 at the end of the protocol, again suggesting an articular rather than an epiphyseal phenotype. Also, we found the expression of core band factor beta (CBF-beta) was elevated and ZNF145, ZNF219, p300 and SirT1 were also increased. Defects in nude rat joints were seeded with chondro-progenitors (approx 2 x105cells per defect) in fibrin gel and in a contralateral control, fibrin gel alone. Joint tissue was isolated at 4 weeks and 12 weeks and in these preliminary experiments histology showed evidence of repair cartilage filling in the defect areas in the cell seeded joints (cartilage in 2 from 3 animals at 4 weeks and in 2 from 4 animals at 12 weeks). When GFP cells were implanted they were detected by fluorescence within areas of neo-cartilage formation and immunohistology using anti human vimentin antibody confirmed human cells in the repair tissue, which stained with safranin O for proteoglycan and immunostained for collagen II. Only fibrous tissue was found in the defect areas of joints implanted with fibrin gel only. Conclusions: Human embryonic stem cells in feeder free, serum free, chemically defined medium were differentiated into chondrogenic cells, which when implanted in focal defects in nude rats participated in the formation of cartilage repair tissue assessed up to 12 weeks. The study demonstrates that human embryonic stem cells can be efficiently differentiated to produce chondro-progenitors with a protocol that is suitable for future clinical applications.