Prediction of PK parameters of human using human chimeric mice

The prediction of drug metabolism and phramacokinetics (PK) in human is important during pre-clinical stage in pharmaceuticals. Several prediction methods have been developed using cryopreserved human liver microsomes or hepatocytes. However, accurate human clearances prediction is difficult since a lot of new drug candidates with diversities of chemical structures have increased to be metabolized by non-CYP enzymes. In the present study, we used chimeric mice with humanized liver (PXB mice, PhoenixBio, Co., Ltd, Hirohisma, Japan), whose livers are replaced with approximately 80% of human hepatocytes. The expression levels and metabolic activities of CYP and non-CYP in liver of PXB mice are similar to the patterns of human. In this study, we attempted to verify the predictability of PK parameters in PXB mice using several model compounds which have been known to be mainly metabolized by CYP and/or non-CYP in liver. (1) In vitro metabolic study; Fresh hepatocytes were isolated from PXB mice using in situ collagenase perfusion method and purified by removal of mice hepatocytes. Suspended hepatocytes were incubated for 2hrs after treatment of each drug. In vitro intrinsic clearances were calculated from time course of disappearance of parent compounds after incubation in human chimeric hepatocytes. (2) In vivo PK study; PXB mice substituted with human hepatocytes were administrated bolus intravenously (i.v.) of drugs at 0.3-5mg/kg body weight. (3) Human PK data for verification; total clearance (CLt) and plasma unbound fraction (fu) in actual human are applied from literature values. In vivo intrinsic clearance (CLint,in vivo) were calculated from the CLt, fu, and average hepatic blood flow (Qh) using the analysis based on well-stirred model. Directly comparison between in vitro intrinsic clearance (CLint,in vitro) of PXB mice hepatocytes and CLint,in vivo calculated by well-stirred model in observed in human, which showed the moderate correlation. On other hand, we compared CLt between human and PXB mice, which was showed the good correlation. Additionally, there was also good correlation of half life (t1/2) after i.v. administration between human and PXB mice. In this study, the predictability of hepatocytes from PXB mice was similar to that of reported using cryopreserved human hepatocytes. However, abundant fresh hepatocytes are available again by transplantation of donor hepatocytes. Consequently, this in vitro approach may be utility for screening the metabolic stability of chemical library. Furthermore, in vivo experiments should be also useful for optimization of new chemical entities with good bioavailability and stable plasma level in human.