(T1230-05-28) Characterization of an Inhouse Developed Human Hepatocyte Co-culture Model for Evaluating Induction of Cytochrome P₄₅₀ CYP2C Enzyme Family

Purpose: Regulatory agencies recommend assessment of CYP450 induction potential for new chemical entities (ICH M12 Guidance). While traditional in vitro 2D monoculture models of primary human hepatocytes are currently considered the gold standard and perform appropriately for assessing induction of the 3 major CYP450 enzymes (CYP3A4, CYP2B6 CYP1A2), the limitations of this model, including the narrow dynamic response when assessing induction of CYP2C isoforms are well documented. Hepatocyte monoculture systems rapidly dedifferentiate and lose their phenotype in culture within 24-72 hours. Additionally, the extracellular matrix in monocultures significantly differs from the hepatic microenvironment, affecting gene expression patterns and responses to xenobiotics treatment. To mitigate these limitations, co-culture or triculture models, which incorporate non-parenchymal cells alongside hepatocytes, are a promising alternative that may better recapitulate the hepatic environment and exhibit a more relevant physiological response to xenobiotic treatment (Kukla et al., 2025, Ramsden et al., 2025). In this study, our general objective was to characterize an “in-house developed” long-term human hepatocyte co-culture model for induction studies, with a specific goal to investigate whether the dynamic range response for CYP2C induction can be improved in this model relative to the traditional 2D hepatocyte monoculture.

Methods: Human hepatocytes co-culture was prepared by co-plating human hepatocytes and 3T3-J2 murine embryonic fibroblast. Briefly murine fibroblasts were sub-cultured in a T-75 tissue culture flask and later co-plated with plateable human hepatocytes at a density of 30,000 fibroblasts/well and 30,000 human hepatocytes/well in a 96-well plate. Five days post seeding, hepatocytes co-culture was treated with rifampin, a known CYP2C inducer for 2, 4 and 7 days, after which mRNA samples were extracted and analyzed for CYP2C expression.

Results: Using the hepatocyte co-culture model and mRNA expression as endpoint, rifampin exhibited concentration dependent induction of CYP2C8 and CYP2C19, with estimated induction kinetic parameters (EC₅₀ and Emax) across the three treatment durations ranging from 0.1-0.4 µM and 5.9-6.9 fold for CYP2C8, and 0.3-0.7 µM and 14.6-17.4 fold for CYP2C19, in human hepatocytes lot D. CYP2C9 showed a narrow dynamic response range across treatment concentrations and duration (2, 4 and 7 days), and the induction kinetic parameters could not be estimated. Within each inducible CYP2C isoform, the 4 days treatment duration show slightly higher response than the 2- and 7-days treatment. The E_max values following 2, 4 and 7 days of rifampin treatment were 5.9-, 7.5- and 6.9-fold for CYP2C8, 2.1-, 1.8- and 1.6-fold for CYP2C9 and 14.2-, 17.4- and 16-fold for CYP2C19, respectively. After 2 days of rifampin treatment in the corresponding 2D hepatocyte monoculture, the maximal induction response was 2.3-, 1.2- and 6.7-fold for CYP2C8, CYP2C9 and CYP2C19, respectively.

Conclusion: The in-house developed hepatocyte co-culture model shows improved CYP2C mRNA dynamic response range when compared to the corresponding 2D monoculture, suggesting a potential usefulness of the in vitro murine fibroblast/human hepatocyte co-culture model for evaluating induction mediated DDI of the CYP2C enzyme family. The maximal induction response to rifampin treatment that was observed in co-culture was more than 2-fold higher when compared to the corresponding 2 days monoculture treatment (Fig 1). This co-culture model is useful for preclinical assessment of CYP2C induction and could enable informed decision making regarding the conduct of clinical DDI studies, in line with regulatory guidances. With this promising mRNA endpoint data, additional investigations are ongoing, including testing different hepatocyte lots to assess variability, investigating enzyme activity as induction endpoints, and assessing the response of non-P450 drug metabolizing enzymes and transporters in the co-culture model.

References: 1. ICH M12 Drug Interaction Studies – Guidance for Industry, 2024.
2. Kukla DA et al. 2025. Clearance Prediction with three novel plated human hepatocyte models compared to conventional suspension assays: assessment with 50 compounds and multiple donors. Drug Metab Dispos 53: 100032.
3. Ramsden D et al. 2025. Quantitative clinical risk assessment of CYP2C, UDP-glucuronosyltransferase, P-glycoprotein induction, and complex drug-drug interactions using TruVivo human hepatocytes triculture platforms. Drug Metab Dispos 3: 100052.

Acknowledgements: All authors are employees of AbbVie and may own AbbVie stocks. The design, study conduct, and financial support for this research were provided by AbbVie. AbbVie personnel participated in the acquisition and interpretation of data, review, and approval of this publication. This work was fully sponsored by AbbVie Inc. We thank our colleagues D. Stresser and D. Kukla for providing useful tips regarding setting up of the co-culture protocol.

Fig 1. Response of CYP2C8, CYP2C9 and CYP2C19 in human hepatocytes co-culture vs 2D monoculture following treatment with 10 uM rifampin for 48 hours