Xeno-sensing activity of the aryl hydrocarbon receptor in human pluripotent stem cell-derived hepatocyte-like cells

Abstract

Although hepatocyte-like cells derived from human pluripotent stem cells (hPSC-HLCs) are considered a promising model for predicting hepatotoxicity, their application has been restricted because of the low activity of drug metabolizing enzymes (DMEs). Here we found that the low expression of xenobiotic receptors (constitutive androstane receptor, CAR; and pregnane X receptor, PXR) contributes to the low activity of DMEs in hPSC-HLCs. Most CAR- and PXR-regulated DMEs and transporters were transcriptionally down-regulated in hPSC-HLC. Transcriptional expression of CAR and PXR was highly repressed in hPSC-HLCs, whereas mRNA levels of aryl hydrocarbon receptor (AHR) were comparable to those of adult liver. Furthermore, ligand-induced transcriptional activation was observed only at AHR in hPSC-HLCs. Bisulfite sequencing analysis demonstrated that promoter hypermethylation of CAR and PXR was associated with diminished transcriptional activity in hPSC-HLCs. Treatment with AHR-selective ligands increased the transcription of AHR-dependent target genes by direct AHR-DNA binding at the xenobiotic response element. In addition, an antagonist of AHR significantly inhibited AHR-dependent target gene expression. Thus, AHR may function intrinsically as a xenosensor as well as a ligand-dependent transcription factor in hPSC-HLCs. Our results indicate that hPSC-HLCs can be used to screen toxic substances related to AHR signaling and to identify potential AHR-targeted therapeutics.

Figure 1. Characterization of hPSC-HLCs.

All subsequent experiments were performed on both hESC-HLCs and hiPSC-HLCs at day 18, respectively. (a) Relative expression level of hepatic marker genes, ALB, AAT, HNF4A, and AFP in differentiated cells at day 0 and day 18 and PHHs were measured by qRT-PCR. Results represent mean ± SD (n = 3). ^*p < 0.05, significant values in comparison with PHH; ^§§§ p < 0.001, significant values in comparison with hESC-HLCs. (b, c) Immunofluorescence for ALB, AAT, HNF4A, and AFP in hESC-HLCs (b) and hiPSC-HLCs (c). Nuclei were counterstained with DAPI. Scale bar = 100 μm. (d, e) Flow cytometry analysis of ALB and AAT in hESC-HLCs (d) and hiPSC-HLCs (e). Blue line, isotype control; red line, target antibody. (f, g) Periodic acid-Schiff staining indicated hESC-HLCs (f) and hiPSC-HLCs (g) exhibiting cytoplasmic glycogen storage. Nuclei (light blue) were counterstained with hematoxylin. (h, i) Acetylated-low-density lipoprotein (Ac-LDL)-positive cells were detected in hESC-HLCs (h) and hiPSC-HLCs (i). (j) Albumin secretion by hPSC-HLCs was measured in the conditioned media by ELISA assay. Values represent means ± S.D (n = 3).

Figure 2. Quantitative gene expression analysis of genes regulated by CAR, PXR, and AHR.

Quantitative comparison of CAR (a), PXR (b), and AHR (c) target genes in fetal liver, hESC-HLCs, and hiPSC-HLCs normalized to human adult liver (dotted line at value of 10⁰). The microarray data for adult and fetal liver are averaged across two independent sources, and for hESC-HLCs and hiPSC-HLCs are averaged across two different passages of origins, respectively. Gene expression levels in hPSC-HLCs ranged from 0.001- to 100-fold (a, c) and from 0.001- to 10-fold (b) as relative to adult liver levels. Ph-I, phase-I enzymes; Ph-II, phase-II enzymes; DT, drug transporters.

Figure 3. Changes in CAR, PXR, and AHR mRNA levels during hepatic differentiation of hESCs and hiPSCs.

Expression levels of CAR (a), PXR (b), and AHR (c) mRNA in differentiated hESCs and hiPSCs at day 0, 5, 10, and 18 and in human adult and fetal liver were measured by qRT-PCR. Results represent mean ± SD (n = 3). ^**p < 0.01, ^***p < 0.001, significant values in comparison with adult liver. AL, adult liver; FL, fetal liver.

Figure 4. DNA methylation and CAR, PXR, and AHR mRNA expression in hPSC-HLCs and PHHs.

DNA methylation status in the regulatory regions of CAR (a), PXR (b), and AHR (c) in hESC- and hiPSC-HLCs, and PHHs were analyzed by bisulfite sequencing. Each diagram represents the investigated locations containing CpG islands in promoter and gene body region. Each row represents the methylation status of each CpG in a series of 9 – 10 bacterial clones. The methylated and unmethylated CpG dinucleotides are represented as filled and open circles, respectively. (d) Expression levels of CAR, PXR and AHR mRNA in hESC- and hiPSC-HLCs and PHHs were measured by qRT-PCR. Results are the mean ± S.D (n = 3). ^***p < 0.001, significant values in comparison with PHHs. PHHs, primary human hepatocytes.

Figure 5. Ligand-induced transcriptional activation of CAR, PXR, and AHR in hPSC-HLCs.

Inductions of CAR-, PXR-, and AHR-target genes in hESC-HLCs (a) and hiPSC-HLCs (b) treated with receptor-specific ligands for 6 h and 24 h were determined by qRT-PCR. CITCO (100 nM), SR12813 (200 nM), and ITE (500 nM) were used as agonists of CAR, PXR, and AHR, respectively. Results represent mean ± SD (n = 3). ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, significant values in comparison with 0.1% DMSO control (Con).

Figure 6. Ligand-dependent recruitment AHR to CYP1A1 and CYP1B1 promoter regions in hPSC-HLCs.

hESC- and hiPSC-HLCs were treated with 1 μM BaP, 10 nM TCDD, 1 μM 3-MC, or DMSO (0.1%) for indicated amounts of time. ChIP assay were performed with primer pairs specific to the XRE-containing promoter regions of CAP1A1 (a), CYP1A2 (b), and CYP1B1 (c) in hESC-HLCs (left panels) and hiPSC-HLCs (right panels) using real-time PCR. Primer pairs that amplify XRE-containing promoter regions as indicated in each diagram (upper panels). Numbers indicate nucleotide positions in relation to the transcription start site ( + 1, an arrow). Results are mean ± SD (n = 3) and are presented relative to isotype control (IgG = 1). ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, significant values in comparison with DMSO.

Figure 7. Ligand-dependent CYP1A1 and CYP1B1 induction in hPSC-HLCs.

Cells were treated with 1 μM BaP, 10 nM TCDD, 1 μM 3-MC, or DMSO (0.1%) for indicated amounts of time. Ligand-induced CYP1A1 (a, c) and CYP1B1 (b, d) expression in hESC-HLCs (a, b) and hiPSC-HLCs (c, d) were determined by qRT-PCR analysis. Results are presented as the mean ± SD (n = 3). ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, significant values in comparison with DMSO.

Figure 8. TMF antagonized AHR-mediated target gene expression.

PHHs and hESC- and hiPSC-HLCs were treated with vehicle (DMSO), 1 μM BaP, 10 nM TCDD, and 1 μM 3-MC with or without 10 μM TMF for 6 h. Expression levels of AHR-responsive CYP1A1 and CYP1B1 in PHHs (a), hESC-HLCs (b), and hiPSC-HLCs (c) were determined by qRT-PCR analysis. Results are presented as the mean ± S.D (n = 3). *p < 0.05, ^**p < 0.01, ^***p < 0.001, significant values in comparison with cells treated with indicated compounds without TMF.