SOX9 is a key factor for the postnatal maturation of the intrahepatic bile duct network.

It is widely recognized that sex-determining region Y-box 9 (SOX9) plays a critical role in the development of intrahepatic bile ducts (IHBDs) during the embryonic stage. IHBDs are composed of two distinct networks: a hierarchical network and a homogeneous network. The hierarchical network consists of large bile ducts that run along the portal veins (PVs) and branch into smaller bile ducts, forming a reticular structure around the PVs. By contrast, the homogeneous network is an early tubular structure composed of bile ductules surrounding the PVs and serves as a precursor to the mature hierarchical network. This study examined the role of SOX9 in the postnatal development of these networks using three-dimensional imaging analysis and a mouse model in which SOX9 deficiency predominates after birth. Our findings indicate that SOX9 is essential for the maturation and structural organization of the IHBD, particularly in terms of facilitating the proper connection between bile ductules and bile canaliculi. Furthermore, transcriptional changes in Sox9 conditional knockout mice activated compensatory pathways involved in bile acid transport and metabolism, while downregulating intercellular adhesion pathways, thereby impairing cholangiocyte adhesion and structural integrity. These results indicate that SOX9 plays a more pivotal role in the development of the bile duct network than previously thought. C57BL/6J mice (Japan SLC, Shizuoka, Japan), Sox9flox/flox mice (strain #013106, The Jackson Laboratory, Bar Harbor, ME) and Albumin (Alb)-Cre mice (strain #003574, The Jackson Laboratory) were used in this study. To delete Sox9 in liver epithelial cells, Sox9flox/flox mice were crossed with Alb-Cre mice to give Alb-Cre+/+;Sox9 flox/flox (Sox9 cKO) mice. For RNA-seq analysis of mouse liver samples (CT n=3, SOX9 cKO n=3), total RNA was prepared using Trizol extraction (Thermo Fisher, Waltham, MA, USA), and the quality of the total RNA was confirmed using a BioAnalyzer 2100 (RNA integrity number > 9). DNA libraries were prepared according to the Illumina TruSeq protocol using a TruSeq Stranded mRNA LT sample prep kit (Illumina, San Diego, CA, USA) and sequenced using an Illumina NextSeq 500 (Illumina) with a NextSeq 500/550 High Output v2 kit (Illumina) to obtain single-end, 75-nt reads. RNA-seq data were aligned to the human genome (UCSC mm10) using STAR ver. 2.6.0a after trimming to remove adapter sequences and removal of low-quality ends using Trim Galore! v0.5.0 (cutadapt v1.16). Gene expression levels are reported as transcripts per million, as determined using RSEM v1.3.1.