UDP-glucuronosyltransferase : purification and activities in rat and human hepatocytes
The UDP-glucuronosyltransferases (GT) represent a major family of drug-metabolising enzymes, but little is known about their multiplicity in man. The aims of this project were to purify and characterise a GT isozyme from human liver, and to investigate glucuronidation in rat and human hepatocytes, with the aim of using human hepatocytes in primary culture as an in vitro model for the study of human drug metabolism. Chromatofocusing of human liver microsomes produced separation of GT isozymes, providing evidence for heterogeneity. However, purification in an active form was not achieved, due to lability in the presence of detergent. Rat liver 17β-hydroxysteroid-GT was purified, and antibodies raised against this protein recognised a single protein in human liver microsomes. Rates of glucuronidation of 1-naphthol and phenolphthalein were significantly higher in rat hepatocyte homogenates than in hepatocytes, the magnitude of the difference being greater for 1-naphthol. This was attributed to the presence of excess UDPGA in homogenate assays and the limitation imposed by lipophilicity on substrate uptake into cells by passive diffusion. In contrast, the rate of bilirubin glucuronidation was greater in hepatocytes, possibly as a result of intact carrier-mediated uptake mechanisms, combined with a suitable environment for efficient delivery of bilirubin to the endoplasmic reticulum by membrane-membrane transfer. The same three substrates were glucuronidated at a reduced rate in human compared with rat hepatocytes, apparently due to a lower intracellular UDPGA level and isozyme-specific differences in intrinsic activity and latency. As a result, the rate of glucuronidation of all these substrates was greater in human hepatocyte homogenates than in hepatocytes. Isozyme-specific changes were observed in GT activities in human hepatocytes in primary culture, indicating the need to develop culture systems allowing stable expression of these enzymes before such a model could be used for predictive human drug metabolism.