Suitability and Limitations of Primary Cells as Physiologic Models

Growing numbers of hits from primary high-throughput discovery programs generate the need for early, fast and robust lead candidate identification. In order to avoid creating a bottleneck at this stage, decisions on compound selection/rejection must be made before animal testing, with the required throughput and minimal substance use. For the accurate prediction of compound specificity and related toxicity, these assays must rely on physiological models that clearly reflect the situation in the target tissue. Miniaturized culture systems ofprimary cells or tissue samples - preferably of human origin - should best suit this purpose. Indeed, the human primary hepatocyte is one of the most extensively used in-vitro models in toxicology because of its central role in drug metabolism and detoxification and as a sensitive detector for generalized cytotoxicities relevant to other organs [1]. Since human hepatocytes are obtained from donor livers not used in transplantation, the availability of fresh cells is unpredictable and limited ( Frozen hepatocytes have become increasingly available as a more reliable source, and have also improved in quality due to sophisticated cryopreservation protocols [2]. However, they still suffer from highly variable attachment to the culture dish and changes in gene activity affecting apoptosis, differentiation, and metabolic enzymes [3]. The expression pattern is not just different between the liver and the cultured hepatocyte; rather, it changes progressively, with the time between isolation and analysis further complicating validation [4]. Many of these changes result from interactions with immune cells in response to injury during the isolation procedure itself. Liver slices which a priori would be expected to reflect in-vivo gene expression better, as they maintain much of the tissue architecture, strongly induce proinflammatory genes and do not reflect compound-induced changes better than primary hepatocytes [5]. Assay variation is also affected by gender and age differences between donors, as well as polymorphisms determining the activity of metabolic enzymes such as cytochrome P450s. Up to 25 variants have been found for a single enzyme (CYP2D6) [6]. These challenges might in time direct the focus to rodent hepatocytes, where a homogenous supply can be assured.

However, even when the changes in gene expression are taken into account, human primary cells often reproduce the patient's response better than rodent cells or rodent in-vivo models. For instance, compounds that act as agonists for PPARalpha (peroxisome proliferator- activated receptor) which regulate the expression of lipid and xenobiotic metabolism genes induce hepatic and pancreatic tumors in rats and mice. However, tumorigencity is absent in man and monkey because of low receptor density and reduced response of regulated genes (for a review, see [7]).

0 0

Post a comment