Impact of expansion and redifferentiation conditions on chondrogenic capacity of cultured chondrocytes.

Cartilage regeneration based on isolated and culture-expanded chondrocytes is studied in a variety of in vitro models, but with varying morphological quality of tissue synthesized. The goal of the present study was to investigate the extent of the influence of expansion and redifferentiation conditions on final tissue morphology by comparing 2 expansion and redifferentiation methods. Chondrocytes from 9 human donors were expanded in medium without growth factor supplementation (basic expansion condition [BEC]) or in medium with basic fibroblast growth factor (bFGF) supplementation (growth factor supplemented expansion condition [GFSEC]). After expansion, cells were either redifferentiated in pellet culture or seeded on collagen type II-coated filters. Post-expansion mRNA levels of collagen type I and II and Sox-5, -6, and 9, measured by semiquantitative real-time polymerase chain reaction (PCR), suggested that expansion in GFSEC results in increased dedifferentiation compared to BEC. However, after 28 days of redifferentiation culture, morphology of tissue synthesized by GFSEC-expanded chondrocytes scored significantly higher on the Bern scale compared to BEC (6.4 +/- 0.3 points vs. 4.5 +/- 0.3 points in pellet culture and 6.0 +/- 0.4 points vs. 4.5 +/- 0.3 points on collagen-coated filters; p < 0.05). Expansion in GFSEC compared to BEC increased proteoglycan (PG) synthesis rate at day 9 (4.0-fold in pellet culture and 1.9-fold on collagen-coated filters; p < 0.01), PG release (6.7-fold in pellet culture and 3.2-fold on collagen-coated filters; p < 0.001), and final PG content at day 28 (1.6-fold in pellet culture and 1.5-fold on collagen-coated filters; p < 0.05). Redifferentiation on collagen-coated filters compared to pellet culture increased PG synthesis rate at day 9 (5.2-fold in BEC-expanded chondrocytes and 2.6-fold in GFSEC-expanded chondrocytes; p < 0.01), PG release (4.2-fold in BEC-expanded chondrocytes and 3.1-fold in GFSECexpanded chondrocytes; p < 0.01), and final PG content (1.3-fold in BEC-expanded chondrocytes and 1.9- fold in GFSEC-expanded chondrocytes; p < 0.01). Moreover, as visualized via electron microscopy, chondrocytes and organization of extracellular matrix cultured on filters was more similar to those found for hyaline cartilage. In conclusion, chondrocyte expansion in GFSEC and redifferentiation on collagen-coated filters resulted in most optimal chondrogenesis.

Altered in vitro chondrogenic properties of chondrocytes harvested from unaffected cartilage in osteoarthritic joints.

OBJECTIVE: In vitro models of chondrogenesis often depart from chondrocytes harvested from less-affected areas of osteoarthritic joints. However, there are indications that these chondrocytes are phenotypically different from chondrocytes from healthy joints and thus might differ in their capacity to generate hyaline cartilage. The goal of this study was to compare the chondrogenic capacity of chondrocytes from healthy and OA joints. DESIGN: Chondrocytes isolated from nine healthy and nine OA knee joints were expanded in monolayer for two passages. Chondrocytes from passages 1 and 2 were analyzed for expression of (de)differentiation and hypertrophy markers and were seeded at passage 2 on collagen-coated filters for redifferentiation culture to study cartilage matrix formation. RESULTS: The collagen II/I mRNA ratio, reflecting differentiation, decreased from passage 1 to 2 in both chondrocytes from OA joints and chondrocytes from healthy joints (P<0.05), without a significant difference between the two donor types. At passage 1, levels of the cartilage transcription factors Sox-5, Sox-6 and Sox-9 appeared to be higher in chondrocytes from OA joints (n.s.), but this was not seen at passage 2. However, a clear difference was observed in collagen type X expression, which was high in chondrocytes from OA joints at both passages, while undetectable in chondrocytes from healthy joints (P<0.01). Tissue generated by chondrocytes from healthy joints redifferentiated for 28 days, showed a significantly better morphology, as assessed by histological scoring (P<0.01) and higher proteoglycan content (P<0.05), compared to chondrocytes from OA joints. Matrix turnover parameters, i.e., proteoglycan synthesis and degradation rate, were not significantly affected by donor tissue origin. CONCLUSIONS: These results suggest that clear differences between chondrocytes from healthy and OA joints exist and that these are not completely abolished during the process of de- and redifferentiation. Therefore, in vitro cartilage regeneration models, which use chondrocytes from OA joints, should be interpreted with care.