Endogenous and xenobiotic metabolic stability of primary human hepatocytes in long-term 3D spheroid cultures revealed by a combination of targeted and untargeted metabolomics

Abstract

Adverse reactions or lack of response to medications are important concerns for drug development programs. However, faithful predictions of drug metabolism and toxicity are difficult because animal models show only limited translatability to humans. Furthermore, current in vitro systems, such as hepatic cell lines or primary human hepatocyte (PHH) 2-dimensional (2D) monolayer cultures, can be used only for acute toxicity tests because of their immature phenotypes and inherent instability. Therefore, the migration to novel phenotypically stable models is of prime importance for the pharmaceutical industry. Novel 3-dimensional (3D) culture systems have been shown to accurately mimic in vivo hepatic phenotypes on transcriptomic and proteomic level, but information about their metabolic stability is lacking. Using a combination of targeted and untargeted high-resolution mass spectrometry, we found that PHHs in 3D spheroid cultures remained metabolically stable for multiple weeks, whereas metabolic patterns of PHHs from the same donors cultured as conventional 2D monolayers rapidly deteriorated. Furthermore, pharmacokinetic differences between donors were maintained in 3D spheroid cultures, enabling studies of interindividual variability in drug metabolism and toxicity. We conclude that the 3D spheroid system is metabolically stable and constitutes a suitable model for in vitro studies of long-term drug metabolism and pharmacokinetics.-Vorrink, S. U., Ullah, S., Schmid, S., Nandania, J., Velagapudi, V., Beck, O., Ingelman-Sundberg, M., Lauschke, V. M. Endogenous and xenobiotic metabolic stability of primary human hepatocytes in long-term 3D spheroid cultures revealed by a combination of targeted and untargeted metabolomics.

Keywords: 3D cell culture; cytochrome P450 enzymes; drug metabolism; hepatic metabolism; mass spectrometry.

Figure 1.

Hepatic expression signatures are preserved for multiple weeks in 3D PHH spheroid culture. A) Bright-field images depicting the temporal development of morphologic PHH phenotypes in 2D monolayer and 3D spheroid culture. Expression of phase I (CYP1A2, CYP2C8, CYP2C9, CYP2D6, CYP3A4; B) and phase II drug metabolizing enzymes (UGT1A1, UGT2B15, and GSTP1; C) drug and bile transporters (ABCC2, ABCC3, SLCO1B1 and ABCB11; D), xenobiotic sensors (CAR, PXR; E), and hepatic markers (ALB, HNF4A; E) remain close to physiologic levels in 3D PHH culture, whereas expression was mostly lost in 2D culture of cells from the same donors (n = 3). The data are presented on semilog plots showing expression fold changes (FC) compared to freshly isolated cells. Dashed lines: the evolution of fold changes between 2- and 3D culture over time. Error bars = sem.

Figure 2.

3D spheroid culture significantly improves the functional activity of major human CYP enzymes. A) Chromatograms of primary metabolites (acetaminophen, hydroxymidazolam, desethylamodiaquine, dextrorphan, and hydroxytolbutamide) of 5 CYP probe substrates in a calibrator (8 ng/ml) and in an incubated sample. B–F) Column plots showing the levels of the metabolic activities of CYP1A2 (B), CYP3A4 (C), CYP2C8 (D), CYP2D6 (E), and CYP2C9 (F) from 3 donors cultured in 2D monolayer and 3D spheroid culture. Dashed line: metabolite levels compared to freshly isolated cells (FICs). Error bars = sd. G) Line plot of fold changes between 2- and 3D cultures of the same donors (n = 3) demonstrate that metabolic activities are significantly elevated in 3D PHH spheroids.

Figure 3.

Interindividual differences in metabolic patterns are reflected in spheroid culture. A) Scheme visualizing different metabolic fates of dextromethorphan. K_m values were obtained from another publication (70). B) Metabolic profiles of dextromethorphan metabolism in PHH from 3 different donors in freshly isolated cells (FICs) and spheroids after 3 wk in culture. CYP2D6 and CYP3A4 genotypes were determined with a CYP+ panel. CYP2D6 and CYP3A4 were phenotyped directly after isolation by rate of formation of dextrorphan and 6β-hydroxytestosterone, respectively. Activity is presented in picomoles of produced metabolite per minute per million cells as provided by supplier. Donors were classified into extensive (EM) and poor (PM) metabolizers on the basis of phenotypic data.

Figure 4.

Intra- and extracellular metabolomes of 3D PHH spheroids remain stable over multiple weeks. A, B) Scatterplots of log intracellular (A) and extracellular (B) metabolite abundances at d 21 in 3D culture and in freshly isolated cells. For each metabolite, the average abundance of n = 6 biologic replicates is plotted. Red dots: probe substrate metabolites, unambiguously identified with internal standards. Dashed line: bisectrix corresponding to perfect correlation. The Pearson correlation coefficients indicate that metabolic profiles were stable over the course of 3 wk in culture. C) Venn diagram depicting the overlap between intracellular and extracellular compounds.

Figure 5.

Concentrations of important endogenous metabolites remain stable in 3D PHH spheroid cultures. A, B) Log scatterplots of 56 endogenous hepatic metabolites in 3D spheroid (A) and 2D (B) culture. Concentrations in freshly isolated cells (FICs; x axis) are correlated with levels after 4 h, 24 h, 7 d, 14 d, and 21 d, and the Pearson correlation coefficient on log-transformed data is indicated. C) Heat map showing the temporal evolutions of metabolite levels of 56 basic physiologic compounds in 2D monolayer and 3D spheroid culture. Color coding depicts fold changes compared to freshly isolated cells. Metabolites are sorted in descending order of absolute concentration.