Maintenance of cytochrome p-450 in cultured adult human hepatocytes☆

The first-order degradation rate constant (k_deg) of cytochrome P450 (CYP) enzymes is a known source of uncertainty in the prediction of time-dependent drug–drug interactions (DDIs) in physiologically-based pharmacokinetic (PBPK) modelling. This study aimed to measure CYP k_deg using siRNA to suppress CYP expression in primary human hepatocytes followed by incubation over a time-course and tracking of protein expression and activity to observe degradation. The magnitude of gene knockdown was determined by qPCR and activity was measured by probe substrate metabolite formation and CYP2B6-Glo™ assay. Protein disappearance was determined by Western blotting. During a time-course of 96 and 60 h of incubation, over 60% and 76% mRNA knockdown was observed for CYP3A4 and CYP2B6, respectively. The k_deg of CYP3A4 and CYP2B6 protein was 0.0138 h⁻¹ (±0.0023) and 0.0375 h⁻¹ (±0.025), respectively. The k_deg derived from probe substrate metabolism activity was 0.0171 h⁻¹ (±0.0025) for CYP3A4 and 0.0258 h⁻¹ (±0.0093) for CYP2B6. The CYP3A4 k_deg values derived from protein disappearance and metabolic activity were in relatively good agreement with each other and similar to published values. This novel approach can now be used for other less well-characterised CYPs.

The ability to predict circulating human metabolites of a candidate drug before first-in-man studies are carried out would provide a clear advantage in drug development. A recent report demonstrated that while in vitro studies using human liver preparations reliably predict primary human metabolites in plasma, the predictability of secondary metabolites, formed by multiple reactions, was low, with total success rates of < 65%. Here, we assess the use of chimeric mice with humanized liver as an animal model for the prediction of human metabolism in vivo. Metabolism studies with debrisoquine and (S)-warfarin demonstrated significantly higher concentrations of their primary human abundant metabolites in serum or plasma in chimeric mice than in control mice. Humanized chimeric mice were also capable of producing human-specific metabolites of several in-house compounds which were generated through more than one metabolism reaction. This model is closer to in vivo human physiology and therefore appears to have an advantage over in vitro systems in predicting complex metabolites in human plasma. However, prediction of human metabolites failed for other compounds which were highly metabolized in mice. Although requiring careful consideration of compound suitability, this model represents a potential tool for predicting human metabolites in combination with conventional in vitro systems.

The aim of this study is to maintain and prolong the viability and metabolic functions of hepatocytes isolated from hepatitis C virus (HCV)-infected human livers. Collagen-coated surface and 10% fetal bovine serum (FBS) supplemented culture media were studied. Metabolic functions and attachment ability of human hepatocytes in these conditions were compared with normalized conditions (Petri dish, medium without FBS). Our results showed that collagen-coated dishes with 10% FBS-supplemented culture medium increased urea production and albumin synthesis. Urea concentration remained stable (30 pg/cell/h) during the first 3-days when human hepatocytes were cultured in collagen-coated dishes. These hepatocytes showed higher urea production when compared with hepatocytes cultured in normal Petri dishes producing 36.4 and 31.7 pg/cell/h concentrations of urea, respectively. Albumin secretion rate was maximal on day 3 after plating (0.86 pg/cell/h). When using culture medium supplemented with 10% FBS, the metabolic functions of hepatocytes were higher compared to when using normal culture medium. This study demonstrates that human hepatocytes can be isolated from HCV-infected livers via surgical biopsies with high cell yield rate, and their metabolic functions remain stable when cultured in collagen-coated dishes with FBS-rich medium for several days. Base on these results, HCV-infected human hepatocytes can be used in the study of clinical diseases and in the performance of physiological and pharmacological liver experiments in vitro.

CYP3A is responsible for approximately 50% of the therapeutic drug-metabolizing activity in the liver. The present study was undertaken to establish the CYP3A4 inducible model for analysis of human drug metabolism using a bioartificial liver composed of the functional hepatocellular carcinoma cell (HCC) line FLC-5. A radial-flow bioreactor (RFB), which is a carrier-filled type bioreactor, was used for 3-dimensional perfusion culture of FLC-5 cells. The CYP3A4 messenger RNA (mRNA) expression level 48 hours after rifampicin treatment in the RBF was approximately 100 times higher than that in a monolayer culture. Western blot analysis also demonstrated an increase in expression of the CYP3A protein. When testosterone, a substrate for CYP3A4, was added to the rifampicin-treated cell culture, 6β-hydroxy testosterone as a metabolite was formed. Electrophoretic mobility shift assay (EMSA) with a CYP3A4 ER6 probe demonstrated that relatively high molecular weight complex containing pregnane X receptor (PXR)/retinoid X receptorα(RXRα), compared with that in the monolayer culture, is possibly generated in the RFB culture of FLC-5 treated with rifampicin. Similarly, the assay with a probe of HNF-4α-binding motif indicated the formation of a large protein complex in the RFB culture. Because it is known that PXR transactivates CYP3A4 gene via its response element and expression of PXR is regulated by HNF-4α, the large complexes binding to response elements of PXR or HNF-4α in the RFB culture may contribute to up-regulation of CYP3A4 mRNA. In conclusion, the bioartificial liver composed of human functional HCC cell line was useful in studying drug interactions during induction of human CYP3A4. (HEPATOLOGY 2003;37:665-673.)

Toxicity of bgugaine, a pyrrolidine alkaloid extracted from the tubers of Arisarum vulgare, was studied in three different liver cell culture models: (1) the rat hepatocyte primary culture; (2) a liver epithelial cell line; and (3) the human hepatoblastoma cell line HepG₂. Cytotoxicity was evaluated by LDH release, MTT reduction and MDA production. DNA fragmentation was analysed by flow cytometry or DNA gel-electrophoresis. In hepatocyte and epithelial cell cultures, drug toxicity appeared at 30 μM and was evaluated by an increase in LDH release, a decrease in MTT reduction and a higher level of MDA production. Bgugaine concentrations lower than 30 μM did not induce changes in these parameters. In HepG₂ cells, bgugaine treatment also induced LDH release at concentrations of 40 and 50 μM. DNA fragmentation, analysed in the HepG₂ cell line by flow cytometry, was observed in cultures exposed to 50 μM bgugaine. However, using DNA gel-electrophoresis, we demonstrated that lower bgugaine concentrations (10, 20 and 30 μM) also induced DNA damage. Our results show that: (1) bgugaine induces an important hepatotoxicity; (2) bgugaine toxicity is not mediated by a metabolic derivative; and (3) bgugaine induces a significant DNA damage. Therefore, our data suggest that the alkaloid bgugaine contained in Arisarum vulgarae may be involved in the toxicologic symptoms observed after consumption of this plant tubers by humans and animals.

Background/Aims: Liver cirrhosis and carcinogenesis are accompanied by an alteration in extracellular matrix material. Histological studies reveal upregulation of the intermediate filaments cytokeratins 8 and 18 and de novo synthesis of vimentin, and cytokeratin 7 or 19 in hepatocytes. The aim of this study was to investigate how these two processes are linked.

Methods: Human hepatocytes were seeded: (i) on the matrix components collagen I, IV, laminin, or fibronectin; (ii) on stoichiometrically different complete matrices, derived from human placenta (matrix I) or the Englebreth-Holm-Swarm tumor (matrix II), and (iii) inside a three-dimensional collagen I sandwich. Filament expression and assembly were measured by cytofluor analysis or confocal laserscan microscopy.

Results: The matrix components or complete matrices triggered enhancement of cytokeratins 8 and 18 and de novo synthesis of cytokeratins 7, 19 and vimentin in a characteristic way. Confocal images demonstrated a dense and uniform network of cytokeratin 18 in freshly isolated cells, which was “replaced” by a few, thick protein bundles within 20 days. Interestingly, newly synthesized cytokeratin 19 structurally resembled the cytokeratin 19 organization in biliary epithelial cells. Marked cytokeratin alterations could be partially prevented when hepatocytes were grown in a three-dimensional collagen sandwich.

Conclusions: Pathological alterations to the chemical composition, molecular structure, or spatial arrangement of the liver matrix lead to specific changes in the intermediate filament pattern in human hepatocytes. We assume that degradation of the matrix results in pathological alterations to the hepatocyte-receptor matrix-ligand ratio, followed by a switch from physiological to pathological cell-activation.