Isolation of hepatocyte-like cells from mouse embryoid body cells.

We previously reported that embryoid body (EB) cells derived from embryonic stem (ES) cells are capable of differentiating into functional hepatocyte-like cells both in vitro and in vivo. Because transplantation of EB-derived cells into the liver via the spleen resulted in a low incidence of teratoma formation, purification of hepatocyte-like cells is required to prevent teratoma formation. The aim of this study was to purify hepatocyte-like cells from cultured EBs. For the isolation of hepatocyte-like cells, EBs cultured for 15 days were treated with trypsin-EDTA. The disaggregated cells were plated on a gelatin-coated dish as a monolayer. These cells were separated by Percoll gradient centrifugation, enriched by magnetic cell sorting, and purified by FACS. The purified hepatocyte-like cells in monolayer cultures were positive for immunostaining for albumin and expressed albumin mRNA, but not Oct3/4 mRNA. Transplantation of the purified hepatocyte-like cells derived from mouse ES cells might be an effective treatment for liver failure.

Teratoma formation and hepatocyte differentiation in mouse liver transplanted with mouse embryonic stem cell-derived embryoid bodies.

We previously reported that mouse embryonic stem (ES) cells are capable of differentiating into hepatocytes in cultured embryoid bodies (EBs) and that hepatocytes generate in the recipient liver injected with cultured day-9 EB cells via spleen without the formation of a teratoma. Because ES cells frequently form teratomas in recipient mice, we investigated incidence of teratoma formation when day-9 EBs derived from ES cells were transplanted directly into the subcapsule of mouse liver. In contrast to injection of day-9 EB cells through the portal vein via the spleen, direct subcapsular injection of cultured day-9 EB cells into liver, and even of cultured day-15 EBs, resulted in an high incidence of teratoma in the liver. In teratomas of livers injected directly with day-15 EBs, hepatocytes were detected singly and in clusters. These results imply that undifferentiated cells capable of developing into teratomas exist in cultured EBs, and even in cultured day-15 EBs containing differentiated hepatocytes.

Differential expression of S-adenosylmethionine synthetase isozymes in different cell types of rat liver.

Mammalian S-adenosylmethionine (AdoMet) synthetase exists as two isozymes, liver-type and nonhepatic-type enzymes, which are the products of two different genes. It is known that the liver-type isozyme is only expressed in adult liver. Whereas, the nonhepatic-type isozyme is widely distributed in various tissues. In addition to the liver-type isozyme, a minor amount of the nonhepatic-type isozyme is also detected in adult liver. To investigate the distribution of these two isozymes in the liver in detail, the localization of these two isozymes was examined in each cell type of liver using a combination of cell fractionation technique and Western blot analysis. In the parenchymal cells, the liver-type isozyme protein was predominantly expressed, and a small amount of the nonhepatic-type isozyme protein was also detected. On the other hand, in the stellate cells the nonhepatic-type isozyme protein was exclusively or only expressed. Interestingly, a large amount of both isozymes were present in endothelial and Kupffer cell fraction. Using both antibodies to anti-rat nonhepatic-type and liver-type isozymes, respectively, immunohistochemical analysis clearly confirmed these results. In addition, in cultured hepatocellular carcinoma cells (FAA-HTC1), the nonhepatic-type isozyme protein only was detected, and the liver-type isozyme protein completely disappeared. This result indicates that the changes in the isozyme expression is regulated within the parenchymal cells. Administration of hepatotoxic drug carbon tetrachloride (CCl4) to rats resulted in about 40% to 50% reduction of enzyme activity in parenchymal cells and stellate cells compared with those of control rats. However, enzyme activity in endothelial and Kupffer cell fraction was not changed.

Regulation of methionine adenosyltransferase activity by the glutathione level in rat liver during ischemia-reperfusion.

Hepatic ischemia was induced by clamping the hepatic artery, portal vein, and bile duct. After 15 min of ischemia, the hepatic glutathione (GSH) content rapidly decreased. On the other hand, after the start of reperfusion, the hepatic GSH levels promptly increased and reached a peak at about 1 h, and thereafter decreased to a minimum level by 2 h. Under such conditions, we examined the changes in the methionine adenosyltransferase (MAT) activity in the liver. Though the time course of MAT activity was somewhat delayed compared with that of the hepatic GSH levels, both patterns were substantially similar during ischemia-reperfusion. In contrast to the changes in the MAT activity during ischemia-reperfusion, the levels of MAT protein were unchanged during these periods. When endogenous antioxidant coenzyme Q(10) (CoQ(10)) was administered to rats prior to ischemia, both the reduction in the MAT activity and hepatic GSH levels induced by ischemia-reperfusion were protected. Our findings suggest that CoQ(10) may posttranslationally regulate the MAT activity via the changes in the GSH level in the liver.

Correlation between the expression of methionine adenosyltransferase and the stages of human colorectal carcinoma.

Methionine adenosyltransferase (MAT) catalyzes the synthesis of S-adenosylmethionine (AdoMet) from ATP and L-methionine. AdoMet is the major methyl donor in most transmethylation reactions in vivo, and it is also the propylamino donor in the biosynthesis of polyamines. In the present study, we assessed MAT activity in human colons with colorectal carcinoma and the values were compared with those of morphologically normal adjacent mucosa. Higher levels of MAT activity were observed in the colorectal carcinoma than in the normal colon. The ratio of MAT activity in tumor tissue versus normal tissue seemed to be correlated well will the stage of the colorectal tumor. Furthermore, immunoblot analysis showed that the high levels of MAT activity observed in colorectal carcinoma were due to the increased amounts of MAT protein. Immunohistochemical analysis revealed that MAT was most abundant in goblet cells, particularly in granules in the supranuclear area of these cells. In the colorectal carcinoma tissues, MAT was strongly stained in the cancerous cells and localized in granules in the supranuclear region. The results of this preliminary study suggest that determination of the relative ratio of MAT activity in both normal and tumor regions in human colorectal carcinoma could be a clinically useful tool for determining the stage of malignancy of colorectal carcinomas.