Search

A hierarchial co-culture, in which rat hepatocytes and non-parenchymal liver cells (NPLCs) were separated by a collagen layer and which was designed to mimic the in vivo microenvironment, was carried out with the aim of developing a module for bio-artificial liver support. Compared with a monolayer co-culture and hepatocytes cultured alone in a monolayer, higher urea synthesis activity was maintained for 6 d in the hierarchical co-culture. When a rat hepatoma cell line H4-II-E-C3, which retains the induction of tyrosine aminotransferase (TAT), was co-cultured in a monolayer with NPLCs, dose-dependent stimulation of TAT induction was observed. In a hierarchical co-culture, NPLCs further stimulated TAT induction in H4-II-E-C3 cells. Since peritoneal macrophages could stimulate TAT induction in hepatocytes in both monolayer and hierarchical co-cultures, bone marrow cells, which can proliferate and differentiate into macrophages in vitro, were investigated as a possible substitute for NPLCs. Bone marrow cells isolated from rat femurs were cultivated in the presence of IL-3 and macrophage colony-stimulating factor (M-CSF), and co-cultured with hepatocytes. Urea synthesis and TAT induction of hepatocytes were stimulated in the co-culture. The co-culture of bone marrow and H4-II-E-C3 cells, both of which have proliferation ability in vitro, was also shown to be effective in stimulating liver functions. The hierarchical configuration, in which two cell types can communicate with the soluble factor(s) through a collagen layer, was found to be more effective than a monolayer in long-term co-culture.

Free amino groups of chitosan, a substance which has previously been shown to be a good scaffold for hepatocyte attachment, were covalently modified with fructose. The modification significantly increased the number of cells that could be attached on the surface of chitosan gel. Rat hepatocytes cultivated on fructose-chitosan behaved similarly to those on unmodified chitosan, i.e., they retained the spherical shape they have in vivo, and released much less lactate dehydrogenase than cells attached on a collagen-coated surface. The modification with fructose did not alter the important characteristics of chitosan for hepatocyte culture: liver-specific functions such as urea synthesis and drug metabolism were stably maintained for 5 d in the hepatocytes cultured on fructose-chitosan. In sharp contrast, hepatocytes attached on a collagen-coated surface underwent a severe morphological change, from spherical to flat, and lost almost all their lidocaine-removal activity within 5d. A very thin fructose-chitosan layer was also applied onto the collagen-coated surfaces of polystyrene plates and a dextran microcarrier by crosslinking free amino groups in the chitosan and collagen with glutaraldehyde to fix the thin layer. Hepatocytes on the fructose-chitosan-coated surface retained their spherical shape, masking the cell-flattening effect of the collagen layer. Perfusion culture was then carried out using a hollow-fiber cartridge in which hepatocytes attached on fructose-chitosan-coated microcarriers were suspended in the extracapillary space: the liver-specific functions were stably maintained during 4d of the culture. A fructose-chitosan-coated surface thus appears to be a very promising scaffold for hepatocyte attachment which can be used in cellular biological studies of liver functions, especially in relation to cytochrome P450, as well as in bioartificial liver support systems.

The effectiveness of chitosan, a biocompatible polymer derived by the deacetylation of chitin, as a scaffold of hepatocyte attachment, was examined. Since chitosan gel was too fragile to use for cell culture, its free amino groups were crosslinked by glutaraldehyde to increase its strength. Rat hepatocytes seeded onto glutaraldehyde-crosslinked chitosan (GA-chitosan) gel could stably attach to the surface, retaining its spherical form, the same as in vivo, and then release a very small amount of lactate dehydrogenase during the 5 d culture period. By contrast, hepatocytes on a collagen-coated surface spread flat, and they released much more lactate dehydrogenase than those on the GA-chitosan gel. Hepatocytes on GA-chitosan also retained higher urea synthesis activity, a liver-specific function, than those on the collagen-coated surface. These results indicate that chitosan is a promising biopolymer as a scaffold of hepatocyte attachment, which can be applied to an effective bioartificial liver support system.

Multicellular spheroids of hepatocytes are known to maintain liver functions for a long period. Rat hepatocytes were isolated to form spheroids by rotation culture and immobilized within calcium alginate. Immobilized spheroids had a much higher extent of tyrosine aminotransferase induction, which is one of the liver-specific differentiated functions, than immobilized non-aggregated cells, while the spheroids secreted significantly less prothrombin than non-aggregated cells. Co-culture of hepatocytes and non-parenchymal liver cells in a monolayer enhanced tyrosine aminotransferase induction and suppressed prothrombin secretion, while conditioned medium prepared from non-parenchymal cells greatly stimulated tyrosine aminotransferase induction and suppressed the prothrombin secretion and DNA synthesis in monolayer-cultured hepatocytes. Prothrombin secretion in hepatocytes was subjected to cell-density-dependent regulation. In a similar manner to other growth-related functions, prothrombin secretion was stimulated at low cell density. It has been reported that thrombin activates the zymogen of hepatocyte growth factor activator [Shimomura, T., Kondo, J., Ochiai, M., Naka, D., Miyazawa, K., Morimoto, Y. & Kitamura, N. (1993) J. Biol. Chem. 268, 22,927-22,932]. Therefore, prothrombin secretion could be one of the growth-related functions and involved in wound healing and liver regeneration.