Summary: Cartilaginous constructs have been grown in v i f m with use of isolated cells, biodegradable polymer scaffolds, and bioreactors. In the present work, the relationships between the composition and mechanical properties of engineered cartilage constructs were studied by culturing bovine calf articular chondrocytes on fibrous polyglycolic acid scaffolds ( 5 mm in diameter, 2-mm thick; and 97% porous) in three different environments: static flasks, mixed flasks. and rotating vessels. After 6 weeks of cultivation, the composition, morphology, and mechanical function of the constructs in radially confined static and dynamic compression all depended on the conditions of in vitro cultivation. Static culture yielded small and fragile constructs, while turbulent flow in mixed flasks yielded constructs with fibrous outer capsules: both environments resulted in constructs with poor mechanical properties. The constructs that were cultured freely suspended in a dynamic laminar flow field in rotating vessels were the largest, contained continuous cartilage-like extracellular matrices with the highest fractions of glycosarninoglycan and collagen, and had the best mechanical properties. The equilibrium modulus, hydraulic permeability, dynamic stiffness, and streaming potential correlated with the wet-weight fractions of glycosaminoglycan, collagen, and water. These findings suggest that the hydrodynamic conditions in tissue-culture bioreactors can modulate the composition. morphology, mechanical properties, and electromechanical function of engineered cartilage.Articular cartilage derives its form and mechanical function from its matrix. which consists of tissue fluid and a framework of structural macromolecules (collagens, proteoglycans, and noncollagenous proteins and glycoproteins). During the development, maintenance, and remodeling of cartilage, chondrocytes synthesize appropriate types and amounts of macromolecules and assemble thcm into a highly organized matrix (4). Adult articular cartilage has a limited capacity to repair damage resulting from injury or disease, and there have been many different approaches to restore tissue composition, structure, and function. including the development of engineered cartilage for potential implantation (5.27). Cartilaginous constructs have been grown in vitro with use of isolated chondrocytes. biodegradable polymer scaffolds, and bioreactors and implanted in vivo to form subcutaneous cartilage or to promote joint repair (16). Fibrous polyglycolic acid scaffolds permitted chondrocytes to Received April 3,1998; accepted September 30. lYY8. Address correspondence and reprint requests to G. VunjakNovakovic at Massachusetts Institute of Technology E25-342. 45 Carleton Street, Carnbridgc, MA 02139, U.S.A. E-mail: gordana@ rnit.edu maintain their differentiated phenotype and provided a three-dimensional framework for tissue regeneration (14), and bioreactors provided control over the conditions of cell seeding and tissuc cultivation and affectcd construct structures and compositio...

Cartilage was obtained from eight matched knee (tibiofemoral and femoropatellar) and ankle (talocrural) joints of five different donors (both left and right from donors 14, 22, and 38 years of age, and left only from donors 31 and 45 years of age) within 24 hours of death. All cartilage was graded as normal by the macroscopic visual Collins' scale and the histological Mankin scale. Cylindrical disks of cartilage were harvested from 10 sites within the tibiofemoral and femoropatellar joint surfaces and four sites within the talocrural joint, and uniaxial confined compression measurements were performed to quantify a spectrum of physical properties including the equilibrium modulus, hydraulic permeability, dynamic stiffness, streaming potential, electrokinetic coupling coefficient, and electrical conductivity. Matched specimens from the same 14 sites were used for complementary measurements of biochemical composition and molecular interaction, including water content, hypotonic swelling behavior, and sulfated glycosaminoglycan and collagen contents. In comparison of the top 1-mm slices of talar cartilage with the top 1-mm of tibiofemoral cartilage, the talar cartilage appeared denser with a higher sulfated glycosaminoglycan content, lower water content, higher equilibrium modulus and dynamic stiffness, and lower hydraulic permeability. The equilibrium modulus increased with increasing sulfated glycosaminoglycans per wet weight and decreased with increasing water content for all joint surfaces. Multiple linear regression showed that greater than 80% of the variation in the equilibrium modulus could be accounted for by variations in the biochemical parameters (water content, sulfated glycosaminoglycans/wet weight, and hydroxyproline content/wet weight) for each joint surface. Nonhomogeneous depth-dependent changes in the physical properties and biochemical composition of full-thickness distal femoral cartilage were consistent with previous reports. Since the compressive deformation of cartilage during cyclic loading is confined to the more superficial regions, the differences in properties of the upper regions of the talar compared with tibiofemoral or femoropatellar cartilage may be important in the etiology of osteoarthritis.

The structure and function of cartilaginous constructs, engineered in vitro using bovine articular chondrocytes, biodegradable scaffolds and bioreactors, can be modulated by the conditions and duration of tissue cultivation. We hypothesized that the integrative properties of engineered cartilage depend on developmental stage of the construct and the extracellular matrix content of adjacent cartilage, and that some aspects of integration can be studied under controlled in vitro conditions. Disc-shaped constructs (cultured for 5 i 1 days or 5 * 1 weeks) or explants (untreated or trypsin treated cartilage) were sutured into ring-shaped explants (untreated or trypsin treated cartilage) to form composites that were cultured for an additional 1-8 weeks in bioreactors and evaluated biochemically, histologically and mechanically (compressive stiffness of the central disk, adhesive strength of the integration interface). Immature constructs had poorer mechanical properties but integrated better than either more mature constructs or cartilage explants. Integration of immature constructs involved cell proliferation and the progressive formation of cartilaginous tissue, in contrast to the integration of more mature constructs or native cartilage which involved only the secretion of extracellular matrix components. Integration patterns correlated with the adhesive strength of the disc-ring interface, which was markedly higher for immature constructs than for either more mature constructs or cartilage explants. Trypsin treatment of the adjacent cartilage further enhanced the integration of immature constructs.

No abstract