Utilization of a Human Liver Tissue Chip for Drug-Metabolizing Enzyme Induction Studies of Perpetrator and Victim Drugs

Abstract

Polypharmacy-related drug-drug interactions (DDIs) are a significant and growing healthcare concern. Increasing number of therapeutic drugs on the market underscores the necessity to accurately assess the new drug combinations during pre-clinical evaluation for DDIs. In vitro primary human hepatocyte (PHH)-only models are commonly used only for perpetrator DDI studies due to their rapid loss of metabolic function. But co-culturing non-human cells with human PHHs can stabilize metabolic activity and be utilized for both perpetrator and victim studies, though this raises concerns about human specificity for accurate clinical assessment. In this study, we evaluated Liver Tissue Chip (LTC) with PHH-only liver microphysiological system (MPS) for DDI induction studies. Liver MPS from three individual donors maintained their functional and metabolic activity for up to 4 weeks demonstrating suitability for long-term pharmacokinetics (PK) studies. The responses to rifampicin induction of three PHH donors were assessed using CYP activity and mRNA changes. Additionally, victim PK studies were conducted with midazolam (high clearance) and alprazolam (low clearance) following rifampicin-mediated induction which resulted in a 2-fold and a 2.6-fold increase in midazolam and alprazolam intrinsic clearance values respectively compared to the untreated liver MPS. We also investigated the induction effects of different dosing regimens of the perpetrator drug (rifampicin) on CYP activity levels, showing minimal variation in the intrinsic clearance of the victim drug (midazolam). This study demonstrates the utility of the LTC for in vitro liver-specific DDI induction studies, providing a translational experimental system to predict clinical clearance values of both perpetrator and victim drugs.

Figure 1:

Human liver tissue chip (LTC) system. The LTC system includes (A) LTC with key design features that enables long term hepatic culture and (B) Multiplex controller that controls the medium recirculation of up to 4 connected LTCs at predetermined flow rate.

Figure 2:

Biological characterization of PHHs from three donors (2214423, 1045, Hu8339) maintained in the LTC. A) Optimized experimental workflow for long-term liver MPS maintenance in LTC. B) Brightfield images of PHH (Donor 1045) showing tissue morphology that was maintained for at least 28 days. Liver MPS from all three donors sustained long-term (C) functional activity quantified by measuring albumin and urea production until day 29 and (D) metabolic activity measured by incubating the liver MPS with 5-in-1 probe substrates cocktail for CYP3A4, CYP2C9, CYP2C19, CYP1A2, and CYP2D6 and quantifying associated metabolites until at least day 15. Plotted data represents mean + SD of 3–4 biological replicates.

Figure 3:

Investigating induction effects of rifampicin on LTC across different rifampicin dosing regimens (PHH donor: 1045) (A) Experimental workflow of dosing regimens: daily spiking, daily fresh dose, and single dose. (B) Activity levels of CYP3A4, CYP2C9, CYP2C19, CYP1A2, and CYP2D6 enzymes were quantified using 5-in-1 probe substrate cocktail assay after rifampicin treatment across different dosing regimens. Compared to the vehicle control, the liver MPS showed dose regimen- dependent variation only for CYP2C9 levels and no significant differences among regimens for other measured CYP enzymes (significance level not shown). (C) shows the rifampicin exposure to liver MPS for different dosing regimens represented as AUC of rifampicin concentration-time curve obtained during three days of induction pre-treatment period. Statistical significance is displayed relative to control determined using un-equal variance t-test (* .001 < p < 0.05; ** .0001 < p < .001; *** .00001 < p < .0001; **** p < .00001; no significance (ns) p > 0.05). Plotted data represents mean + SD of 3–4 biological replicates.

Figure 4:

Comparison of rifampicin-mediated induction effects on Liver MPS across different donors (2214423, 1045, Hu8339) (A) Experimental workflow with optimized daily spiking dosing regimen. (B) Enzyme activity levels of CYP3A4, CYP2C9, CYP2C19, CYP1A2, and CYP2D6 quantified using 5-in-1 probe substrate cocktail assay shows rifampicin-mediated induction compared to the vehicle control for all three donors (C) Fold induction changes for mRNA and enzyme activity levels across different donors demonstrate donor variability observed in induced levels of enzyme activity from control. Statistical significance is displayed relative to control determined using an un-equal variance t test (* .001 < p < 0.05; ** .0001 < p < .001; *** .00001 < p < .0001; **** p < .00001; no significance (ns) p > 0.05). Plotted data represents mean + SD of 3–4 biological replicates.

Figure 5:

Time course and enzyme activity levels of liver MPS during long-term DDI study involving rifampicin mediated induction, victim PK study, and de-induction after cessation of rifampicin (PHH donor: 1045). (A) Experimental workflow of DDI victim-perpetrator study with recovery time course. (B) Time course plots of CYP activity levels (CP3A4, CYP2C9, CYP2C19, CYP1A2, and CYP2D6) shows increase in activity during rifampicin treatment (day 3 – day 9) compared to the vehicle control followed by de-induction that is observed by activity levels returning to baseline levels after cessation of rifampicin treatment on day 9. Statistical significance is displayed relative to control determined using un-equal variance t-test (* .001 < p < 0.05; ** .0001 < p < .001; *** .00001 < p < .0001; **** p < .00001; no significance (ns) p > 0.05). Plotted data represents mean + SD of 3–4 biological replicates.

Figure 6:

Drug-Drug Interaction (DDI) observed in LTC during long-term victim-perpetrator study (PHH donor: 1045). (A) Experimental workflow of victim-perpetrator PK study with daily spiking of rifampicin regimen during induction pre-treatment. The PK profile of both (B) high-turnover drug midazolam and (C) low-turnover drug alprazolam shows increased depletion rate in rifampicin induced liver MPS compared to the vehicle control. (D) shows increased production of primary metabolite (1-OH midazolam) of midazolam in induced liver MPS compared to the control. (E) Unbound in vitro intrinsic clearance of victim drugs from induced and control liver MPS. Statistical significance is displayed relative to control determined using un-equal variance t-test (* .001 < p < 0.05; ** .0001 < p < .001; *** .00001 < p < .0001; **** p < .00001; no significance (ns) p > 0.05). Plotted data represents mean + SD of 3–4 biological replicates