Rapid changes in chromatin structure during dedifferentiation of primary hepatocytes in vitro

Abstract

Primary hepatocytes are widely used in the pharmaceutical industry to screen drug candidates for hepatotoxicity, but isolated hepatocytes quickly dedifferentiate and lose their mature metabolic function in culture. Attempts have been made to better recapitulate the in vivo liver environment in culture, but the full spectrum of signals required to maintain hepatocyte function in vitro remains elusive. Here we studied the dedifferentiation process in detail through RNA-sequencing of hepatocytes cultured over eight days. We identified three distinct phases of dedifferentiation. An early phase, where mature hepatocyte genes are rapidly downregulated in a matter of hours. A middle phase, where fetal genes are activated, leading to hepatocytes with a fetal phenotype. A late phase, where initially rare contaminating non-parenchymal cells over-grow the culture as the hepatocytes gradually die. Using genetically tagged hepatocytes, we demonstrate that the cells reactivating fetal marker alpha-fetoprotein arise from cells previously expressing the mature hepatocyte marker albumin, and not from albumin negative precursor cells, proving that hepatocytes undergo true dedifferentiation. To better understand the signaling events that result in the rapid down-regulation of mature hepatocyte genes, we examined changes in chromatin accessibility of hepatocytes during the first 24h of culture using ATAC-seq. We find that drastic and rapid changes in chromatin accessibility occurs immediately upon start of culture. Using binding motif analysis of the areas of open chromatin sharing similar temporal profiles, we identify several candidate transcription factors potentially involved in the dedifferentiation of primary hepatocytes in culture.

The native liver as inspiration to create superior in vitro hepatic models

“…99,100 The 3D-scaffold model consists of hepatocytes supported by a scaffold material processed into a 3D shape. This 3D-shape is created by a controlled emulsification, lyophilization, solvent casting, salt leaching or gas foaming Primary human hepatocytes • Gold standard 13 • Dedifferentiation 34 36 and 37 • Representing population 33 • Availability 33 • Invasive • Time-consuming isolation 35 • Limited proliferation 33 HepG2…”

Dynamic regulatory module networks for inference of cell type-specific transcriptional networks

“…Dedifferentiation could be due to the changes in the regulatory program over time, however, little is known about the transcriptional and epigenetic changes during this process. To measure transcriptional dynamics during dedifferentiation, Seirup et al generated an RNA-seq and ATAC-seq timecourse dataset of 16 time points from 0 hours to 36 hours [26]. We applied DRMN with k = 5 modules to this dataset and partitioned genes into five levels of expression at all time points with 1 representing the lowest level of expression and 5 the highest level of expression (Figure 6A).…”

“…The dedifferentiation time course consisted of samples were extracted from adult mouse liver and gene expression and chromatin accessibility were assayed by RNA-seq and ATAC-seq, respectively, at 0 hrs, 0.5 hrs, 1 hrs, 2 hrs, 4 hrs, 6 hrs, 8 hrs, 10 hrs, 12 hrs, 14 hrs, 16 hrs, 18 hrs, 20 hrs, 22 hrs, 24 hours, and 36 hrs (16 time points in total, [26]). All sequencing reads were aligned to mouse mm10 reference genome using Bowtie2 [65], and gene expression was quantified using RSEM [66].…”