key: cord-0013824-a1arjs6g
authors: Wang, Qiming; Sun, Da; Liang, Zhen; Wang, Junyi; Zhong, Xinxing; Lyu, Yulin; Cao, Junning; Lin, Zhongqing; Du, Yuanyuan; Miao, Zhenchuan; Lu, Shichun; Li, Cheng; Xu, Jun; Shi, Yan; Deng, Hongkui
title: Generation of human hepatocytes from extended pluripotent stem cells
date: 2020-03-09
journal: Cell Res
DOI: 10.1038/s41422-020-0293-x
sha: e304e1b6f3430f37e030ca394fca1f7e03e7392f
doc_id: 13824
cord_uid: a1arjs6g

nan

(c) Dynamic analysis of ALB secretion in EPS-Heps since day 0 post maturation in hepatic maturation medium. n = 3. (d) Bile acids secretion in EPS cells, EPS-Heps and PHHs. n = 3. Data are presented as the mean ± SEM. For all measurements, 'n' represents the number of biological replicates.

Primary human hepatocytes were isolated as previously described. Briefly, human liver tissue was first perfused with PBE buffer (9 g/L NaCl, 0.42 g/L KCl, 2.1 g/L NaHCO3, 0.9 g/L glucose, 4.78 g/L HEPES and 0.37 g/L EDTA in sterilized water) for 0.5 to 2 hours, and the liver tissue was further perfused with PBCD buffer (9 g/L NaCl, 0.42 g/L KCl, 2.1 g/L NaHCO3, 0.9 g/L glucose, 4.78 g/L HEPES, 0.25 g/L collagenase and 0.25 g/L dispase in sterilized water) until the liver tissue was incompact. The loosened liver tissue was separated with tweezers, and the human hepatocytes were collected for further experiments.

Human fetal liver tissue was obtained from aborted tissue with informed patient consent.

The fetal liver tissue was shredded and incubated in 1 mg/ml collagenase IV for 20 minutes at 37°C. The tissue fragments and cells were plated on Matrigel-coated dishes in hepatic progenitor expansion medium, and the medium was changed every day. Hepatic progenitor expansion medium included 50% DMEM/F12 (Thermo Fisher Scientific), 50% William's E Medium (Thermo Fisher Scientific) supplemented with 1% penicillin-streptomycin (P/S, Thermo Fisher Scientific), 2% B27 (without VA, Thermo Fisher Scientific), 5 mM nicotinamide (Sigma-Aldrich), 200 μM 2-phospho-L-ascorbic acid (pVc, Sigma-Aldrich), 3 μM CHIR99021(Tocris), 5 μM SB431542 (Selleck), 0.5 μM sphingosine-1-phosphate (S1P, Sigma-Aldrich), 5 μM lysophosphatidic acid (LPA, Aladdin), 50 ng/ml EGF (Peprotech) and 20 μM forskolin (Tocris). After the hepatic progenitors migrated out from the fetal liver tissue and the cells were confluent, the cells and tissue fragments were digested with Accutase (Millipore).

The tissue fragments were discarded with a sifter. The hepatic progenitors were collected for further experiments. The EPS cells used in this study were EPS1 cells and EPS2 cells, as previously described 1 .

EPS cells were cultured on mitomycin C (Sigma-Aldrich) inactivated mouse embryonic fibroblast (MEF) feeder cells (3×10 4 cells per cm 2 ) with EPS culture medium (48.25% DMEM/F12 (Thermo Fisher Scientific), 48.25% Neurobasal medium (Thermo Fisher Scientific), 0.5% N2 supplement (Thermo Fisher Scientific), 1% B27 supplement (Thermo Fisher Scientific), 1% GlutaMAX, 1% MEM NEAA, 1% P/S, 10 ng/ml recombinant human LIF (PeproTech), 1 μM CHIR99021, 2 μM (S)-(+)-dimethindene maleate (Tocris) and 2 μM minocycline hydrochloride (Santa Cruz Biotechnology). The EPS cells were passaged using 0.05% trypsin-EDTA (Thermo Fisher Scientific) at a proportion from 1:3 to 1:10.

To induce EPS cells into a primed liked state in stage 1, EPS cells were digested into single cells with 0.05% Trypsin-EDTA and seeded at 5×10 4 cells/cm 2 in TeSR™2 medium for 48 hours on Matrigel-coated plates. To generate definitive endoderm in stage 2, differentiated cells were treated with MCDB medium (Thermo Fisher Scientific) containing 1% B27 (without vitamin A), 100 ng/ml Activin A (Stemimmune LLC), 0.25 mM pVc, 25 ng/ml Wnt3a (R&D), and 0.05 µM PI103 (Selleck) for 1 day and were then treated with MCDB medium containing 100 ng/ml Activin A and 0.25 mM pVc for 3 days. Next, to generate posterior foregut in stage 3, the differentiated cells were cultured in modified MCDB medium containing 1% B27 (without vitamin A), 50 ng/ml KGF (Stemimmune LLC), 0.25 mM pVc and 10 μM SB431542 (Selleck) for 3 days. Next, in stage 4, to generate hepatic progenitors from the anterior foregut cells, the differentiated cells were first cultured in DMEM medium (Thermo Fisher Scientific) containing 1% B27 (without vitamin A), 20 ng/ml KGF, 10 ng/ml bFGF (Peprotech), 20 ng/ml BMP2 (Stemimmune LLC) and 50 ng/ml BMP4 (Stemimmune LLC) for 4 days. Next, the cells were replated at a ratio of 1:3 and cultured in modified William's E Medium (Beijing Vitalstar Biotechnology) for 8 days. After that, the cells were further cultured in hepatic expansion medium (HEM) containing 49% William's E Medium, 49% DMEM/F12, 2% B27, 0.25 mM pVc, 5 μM SB431542, 3 μM CHIR99021, 0.5 μM S1P, 5 μM LPA, 40 ng/ml EGF and 10 mM Nicotinamide for at least 10 days. In this stage, the cells were passaged when they were confluent. In stage 5, hepatocyte maturation medium (HMM), which contained William's E Medium, 2% B27, 1% GlutaMAX, 10 μM SB431542 and 50 μM forskolin, was used to generate mature hepatocytes from hepatic progenitors. These EPS-Heps could be matured and maintained in HMM. All media were changed every day. Accutase was used for passaging in all stages. All the molecular and functional test of EPS-Heps were preformed 4 to 5 weeks post maturation.

Total RNA was isolated with RNeasy Mini Kit (QIAGEN) following the manufacturer's instructions. RNA (500 ng) was reverse-transcribed to cDNA with TransScript First-Strand cDNA Synthesis SuperMix (TransGen Biotech). KAPA SYBR ® FAST Universal qPCR Mix (KAPA Biosystems) was used for RT-qPCR analysis, which was performed on a BIO-RAD CFX384 TM Real-Time System. All relative expression levels were normalized to the housekeeping gene RRN18S. The RT-qPCR primer sequences are provided in the following 

For immunofluorescence staining, cells were fixed with 4% paraformaldehyde (DingGuo)

for 15 minutes and blocked with PBST (0.25% Triton X-100 and 5% normal donkey serum in PBS). Then, the cells were incubated with primary antibodies at 4°C overnight followed by the appropriate secondary antibodies for 1 hour at room temperature. DAPI (Roche) was used to indicate nuclei. The information of antibodies are listed in the following table. 

For flow cytometry analysis, cells were released into single-cell suspensions with Accutase and fixed with Fixation/Permeabilization solution (BD) for 20 min at 4°C. Then, the cells were incubated with primary antibodies diluted in 1X BD Perm/Wash buffer at 4°C for 2 hours followed by the appropriate secondary antibodies for 1 hour at 4°C. Finally, the cells were resuspended in BD Perm/Wash buffer and analyzed on a CytoFLEX (Beckman Coulter) flow cytometry system. The data were analyzed with CytExpert software. The antibodies used in flow cytometry analysis were the same as those used for immunofluorescence staining.

We assessed the quality of the raw RNA sequencing data with FastQC software. The raw fastq files were trimmed with Trimmomatic software using the following parameters:

ILLUMINACLIP:/path/to/adapters/TruSeq3-PE-2.fa:2:30:7:1:true LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 HEADCROP:10 MINLEN:36. The trimmed clean data were then aligned to the human reference genome hg19 with STAR software with the default parameters.

Next, the gene count matrix of all samples was generated with the featureCounts function of the R package Rsubread. Finally, normalization and variance-stabilizing transformation were performed on the gene count matrix with the R package DESeq2. The differentially expressed genes were also determined with DESeq2.

To compare our transcriptome data with the annotated epiblast single-cell RNA sequencing data (GSE109555), we reanalyzed the public data with the R package Seurat and simulated the bulk RNA sequencing data of epiblast cells by averaging the gene expression levels of all epiblast cells in the same day.

We performed hierarchical clustering with the R function hclust. The distance between two samples was defined as one minus the Pearson correlation between the z-scores of the gene vectors. The sample distance definition above was also used to draw heatmaps with the R package pheatmap.

We performed CellNet analysis on our data and some public RNA sequencing data (GSE103078 and GSE98710). The RNA sequencing data were analyzed with the R package CellNet. We followed the official CellNet pipeline to analyze all the sequencing data and visualize the results.

Tet-uPA/Rag2 -/-/γc -/-(URG) mice on a BALB/c background were purchased from Beijing Vitalstar Biotechnology. For transplantation, cells were dissociated to single cells with Accutase and suspended in HCM TM medium (Lonza) at a final concentration of 10 7 cells/ml. URG mouse was injected with 200 μl suspension into the spleen. Eight weeks post injection, mouse was sacrificed for immunofluorescence staining. The liver of the mouse was fixed with 4% paraformaldehyde and dehydrated with a 30% sucrose solution. Then, the liver tissue was embedded in OCT compound (Sakura) and frozen in liquid nitrogen. Cryosections were generated using a cryostat (Leica) for immunofluorescence staining. The repopulation rate of human ALB positive cells was evaluated with Vectra Polris (PerkinElmer) using 6 random cryosections.

The experiment on mouse model was approved by the Institutional Animal Care and Use

Committee of Peking University, and were performed according to NIH guidelines.

EPS-Heps, HepG2 cells and freshly isolated primary human hepatocytes were dissociated and suspended to measure their CYP3A4 and CYP1A2 activities. One 500 μL reaction contained 2.5 × 10 5 cells and 200 μM testosterone or 200 μM phenacetin as substrate of CYP3A4 or CYP1A2, respectively. Reaction without cells as well as reaction without testosterone or phenacetin were also performed to exclude reaction background. After incubation for 15 minutes at 37 °C, reactions were stopped by 1.5 mL methanol containing

isotope-labeled reference metabolite, 6β-hydroxytestosterone-[D7] for CYP3A4 or Acetomidophenol-[13C2, 15N] for CYP1A2, for further mass spectrometry (ultra-performance liquid chromatography-tandem mass spectrometry, UPLC/MS/MS) analysis. The drugmetabolic product of CYP3A4 and CYP1A2 reaction for UPLC/MS/MS analysis were 6β-Hydroxytestosterone and Acetaminophen