Optimizing nutrient channel spacing and revisiting TGF-beta in large engineered cartilage constructs

Cartilage tissue engineering is a promising approach to treat osteoarthritis. However, current techniques produce tissues too small for clinical relevance. Increasingly close-packed channels have helped overcome nutrient transport limitations in centimeter-sized chondrocyte-agarose constructs, yet optimal channel spacings to recapitulate native cartilage compositional and mechanical properties in constructs this large have not been identified. Transient active TGF-β treatment consistently reproduces native compressive Young׳s modulus (EY) and glycosaminoglycan (GAG) content in constructs, but standard dosages of 10 ng/mL exacerbate matrix heterogeneity. To ultimately produce articular layer-sized constructs, we must first optimize channel spacing and investigate the role of TGF-β in the utility of channels. We cultured ∅10 mm constructs with 0, 12, 19, or 27 nutrient channels (∅1 mm) for 6-8 weeks with 0, 1, or 10 ng/mL TGF-β; subsequently we analyzed them mechanically, biochemically, and histologically. Constructs with 12 or 19 channels grew the most favorably, reaching EY=344±113 kPa and GAG and collagen contents of 10.8±1.2% and 2.2±0.2% of construct wet weight, respectively. Constructs with 27 channels had significantly less deposited GAG than other groups. Channeled constructs given 1 or 10 ng/mL TGF-β developed similar properties. Without TGF-β, constructs with 0 or 12 channels exhibited properties that were indistinguishable, and lower than TGF-β-supplemented constructs. Taken together, these results emphasize that nutrient channels are effective only in the presence of TGF-β, and indicate that spacings equivalent to 12 channels in ∅10 mm constructs can be employed in articular-layer-sized constructs with reduced dosages of TGF-β.