Transcriptomic Profile of the Mouse Postnatal Liver Development by Single-Nucleus RNA Sequencing

The liver plays a vital role in maintaining the physiological homeostasis of mammals and is responsible for many biological processes, including detoxification, bile acid synthesis, glycolysis, and lipid metabolism (Ben-Moshe and Itzkovitz 2019). The liver consists of repeating anatomical units termed liver lobules, including parenchymal and non-parenchymal cells. Hepatocyte, the liver parenchymal cells, account for 60% of hepatic cells composition and 80% of liver mass (Godoy et al., 2013). While the liver non-parenchymal cells (NPCs), including bile duct cell (cholangiocyte), liver endothelial cell (LEC), hepatic stellate cells (HSC), Kupffer cell, and other immune cell populations, account for the remaining 20% of liver mass (MacParland et al., 2018; Aizarani et al., 2019). When the cell function is compromised or its composition becomes abnormal, it can cause many diseases, such as fatty liver disease, cirrhosis, and hepatocellular carcinoma (Saviano et al., 2020). During embryonic development, a few cells from the endoderm begin to specialize into hepatoblasts around embryonic day (E) 8.5 - E9.0 in mouse embryos (Tremblay and Zaret 2005). By E10.5, hepatoblast, LEC, and HSC start to organize into primitive sinusoidal capillaries (Gordillo et al., 2015). Finally, hepatoblast begin to differentiate into hepatocyte and cholangiocyte approximately at E13.5 (Roskams and Desmet 2008; Yang et al., 2017). In those developmental processes, several transcription factors and signaling pathways, including Hhex, hepatocyte nuclear factor 4α (Hnf4α), fibroblast growth factors (FGFs), bone morphogenetic proteins (BMPs), Wnt/β-catenin, and Hippo pathway play essential roles (Bort et al., 2006; McLin et al., 2007; Negishi et al., 2010; Alder et al., 2014; Wang et al., 2015; Ober and Lemaigre 2018). Until E16.5, the liver is the main hematopoietic organ in the body, then gradually becomes a metabolic organ (Zaret 2002). After birth, both the nutrient metabolism and the immune system in the liver will undergo significant transformations in response to drastic changes, such as the primary source of energy switching from glucose in the cord blood to lipids in breast milk (Ehara et al., 2015). As a result, metabolic pathways, including fatty acid β-oxidation, gluconeogenesis, and de novo lipogenesis, are upregulated in the neonatal liver (Perichon and Bourre 1995; Sekine et al., 2007). After a short period of adaptation, hepatocytes begin to proliferate and differentiate rapidly around postnatal day 3 (P3), as a result of increased β-catenin signaling (Apte et al., 2007). By P7, the liver lobule structure has been formed (Ober and Lemaigre 2018), becomes more intact, and performs several functions, including xenobiotic metabolism, steroid metabolism, and bile acids biosynthetic (Li et al., 2009; Cui et al., 2012). In recent years, single-cell transcriptome sequencing has been developing rapidly and has been applied in many research fields, such as cell atlas construction, as well as studying embryo development and disease pathogenesis (Granja et al., 2019; Tang et al., 2019; Cao et al., 2020; Trevino et al., 2021). In terms of liver development, single-cell transcriptome sequencing has been used to elucidate the origin of hepatoblasts, the differentiation trajectory of hepatoblasts and NPCs, and the interaction among different cell types (Popescu et al., 2019; Lotto et al., 2020; Mu et al., 2020; Wang et al., 2020). However, few single-cell transcriptomic studies on liver development in mice after birth. A robust and comprehensive bulk RNA sequencing study was performed to characterize the liver development spanning E12.5 to postnatal week 8, which found that many important liver metabolic functions are acquired after birth (Gong et al., 2020). However, bulk RNA sequencing cannot distinguish the gene expression of different cell types and does not provide information about cell-cell interaction and microenvironment composition, thus calling for a more systematic single-cell transcriptome study of postnatal liver development. Moreover, single-nucleus RNA sequencing (snRNA-seq) has an obvious advantage over single-cell RNA sequencing (scRNA-seq) in detecting multiple cell types without any bias (Zeng et al., 2016; Lake et al., 2017; Bakken et al., 2018; Ding et al., 2020). This is crucial for liver tissue because hepatocytes are prone to cell death during liver single-cell isolation, and obtaining HSCs and cholangiocytes remains quite challenging. In addition, it could lead to skewed cell types ratio in scRNA-seq datasets (Donne et al., 2020; Brazovskaja et al., 2021; Guilliams et al., 2022). For this study, we performed snRNA-seq to profile 82,967 nuclei from four key time points of postnatal murine liver development (P0, P3, P7, P14). We identified 28 clusters of hepatic cell types and analyzed the dynamic changes in cells composition and functions and the hepatocyte differentiation trajectories during this process. Interestingly, we found two HSC subtypes specifically expressing some markers of LECs or Kupffer cells. In addition, the ligand-receptor interaction and transcription factor regulative activity analysis significantly increased the reliability of the new cell types.