To identify eRNAs potentially regulated by FXR, we first obtained a list of putative enhancers in mouse liver based on ENCODE H3K27ac ChIP-seq data (ENCFF001KMI, see Materials and methods). Then, we performed ribo-depleted total RNA-seq in the livers from mice treated with a specific FXR agonist, GW4064, to identify transcripts produced from these enhancer regions (see Supplementary file 1a). To avoid confounding issues of RNA signals from genes, we specifically focused on intergenic enhancer regions (±3 kb from H3K27ac peak center) that harbor discernible RNA-seq signals (RPKM >1).

This genomic analysis resulted in identification of 190 high-confidence eRNAs in mouse liver. Among these, 14 eRNAs were upregulated and 5 were downregulated by GW4064 treatment (FDR <0.05, log₂FC >1, Figure 1A) (see Supplementary file 1b). FXR-regulated eRNAs were produced adjacent to many genes with important roles in liver metabolism and disease, e.g., Hes1 (Figure 1—figure supplement 1A and B). One of the most robustly induced eRNAs was Fincor, an unannotated novel transcript located on chromosome 19 (Figure 1A, B, and C). Reverse transcription qPCR (RT-qPCR) confirmed that treatment with GW4064 substantially induced expression of Fincor, more than 10-fold in mouse liver, which is similar to induction of Nr0b2, a well-known FXR target gene (Figure 1D; Claudel et al., 2005; Evans and Mangelsdorf, 2014; Goodwin et al., 2000). Induction of Fincor by GW4064 was transient, peaked within 1 hr and then declined gradually, a pattern similar to that of Nr0b2 (Figure 1—figure supplement 1C). Expression of genes adjacent to Fincor, including Gcnt1, Rfk, Pcsk5, and Prune2, did not change after acute 1 hr FXR activation as shown by RNA-seq (Figure 1C), and confirmed for Gcnt1 by time course qPCR (Figure 1—figure supplement 1C).

We also found that the short-time GW4064 treatment resulted in 590 upregulated genes and 500 downregulated genes (FDR <0.05) (Figure 1—figure supplement 2A). Gene ontology (GO) enrichment analysis of these differentially expressed genes revealed their roles in the regulation of triglyceride, fatty acid, and cholesterol metabolism (Figure 1—figure supplement 2B), which is consistent with known roles of FXR in these physiological processes (Claudel et al., 2005).

We sought to examine: (1) if these eRNAs were directly activated by FXR, and (2) if the activation takes place transcriptionally using Fxr liver-specific knockout (Fxr-LKO) mice that were treated with GW4064 to determine if FXR was required for induction of eRNAs by GW4064. Fxr-LKO mice were generated from Fxr floxed (Fxr-Flox) mice (Figure 2—figure supplement 1A) and transcription of eRNAs was detected by GRO-seq, a widely used method to detect nascent RNA transcription, including eRNAs (Lam et al., 2014; Li et al., 2016; Sartorelli and Lauberth, 2020). GRO-seq showed that FXR-induced eRNAs were activated transcriptionally by GW4064 in Fxr-flox mice, but such induction was abolished in Fxr-LKO mice (Figure 2A). In particular, Fincor is robustly induced in the GRO-seq analysis and its induction is dependent on hepatic FXR (Figure 2B). RT-qPCR confirmed the FXR-dependent expression of Fincor, similar to that of Nr0b2 (Figure 2—figure supplement 1B).

We next examined if FXR binds to the enhancers that produce the identified eRNAs by analyzing published mouse liver ChIP-seq data for FXR (see Data availability) (Lee et al., 2012; Thomas et al., 2010). FXR binding was strongly enriched at the enhancers associated with FXR-induced eRNAs as were the enhancer marks H3K27ac and H3K4me1 (Figure 2—figure supplement 1C). The binding of FXR and the presence of histone marks at the Fincor enhancer region determined by ChIP-seq as compared with nascent transcripts detected by GRO-seq is shown in Figure 2B. These analyses support the conclusion that activation of FXR transcriptionally induces this series of eRNAs via chromatin binding at these enhancers, including Fincor.

We validated FXR binding at the enhancer region that produces Fincor using mouse liver ChIP (Figure 2C). We also examined binding of a well-known DNA binding partner of FXR, retinoid X receptor alpha (RXRα/NR2B1) (Evans and Mangelsdorf, 2014; Zheng et al., 2018), and bromodomain-containing protein 4 (BRD4), an acetylated histone reader protein that often binds at active enhancers (Chen et al., 2016; Li et al., 2016; Rahnamoun et al., 2018; Sartorelli and Lauberth, 2020) and a transcriptional coactivator of FXR (Jung et al., 2020). GW4064 treatment resulted in substantial increases in recruitment of both FXR and RXRα to the enhancer region close to the transcription start site of Fincor (arrow shown below in Figure 2B), whereas binding was not detected at a control region (Figure 2C). We also found that BRD4 occupancy was increased at this enhancer region after GW4064 treatment (Figure 2C). These results indicate that GW4064 activation of FXR leads to increased occupancy of the FXR/RXRα heterodimer and BRD4 to the enhancer region to upregulate Fincor eRNA in the liver.

We identified an inverted repeat1 (IR1) motif that is known to bind FXR (Calkin and Tontonoz, 2012; Lee et al., 2006) within the major FXR binding peak near the start site of Fincor (arrow shown below in Figure 2B), which we also refer to as FXRE (Figure 2B and C). We examined the functionality of this IR1 motif for mediating transcriptional activation by GW4064 using reporter assays (Figure 2D). We cloned the region containing the IR1 motif into the pGL4.23 luciferase reporter and generated a mutated IR1 motif construct as a comparison (Figure 2D). After transfection into human hepatic HepG2 cells, GW4064 treatment significantly elevated the luciferase activity of the reporter with the wild-type IR1 motif, but not with the mutated IR1 motif (Figure 2D). Together, these results suggest that GW4064-activated FXR directly upregulates Fincor expression.

Because enhancers and eRNAs generally act in a tissue-specific manner (Li et al., 2016; Sartorelli and Lauberth, 2020), we examined the tissue-specific expression of Fincor in mice. Strikingly, Fincor is highly expressed in the liver and it is expressed at extremely low levels in most other tissues, except for a detectable, but still fairly low, level in the lung (Figure 3A). GW4064 treatment resulted in induction of Fincor specifically in the liver (Figure 3A). The level of Fincor detected in primary mouse hepatocytes (PMHs) isolated from GW4064-treated mouse liver was similar to that in the liver tissue from the same mouse, suggesting that the majority of Fincor is present in hepatocytes (Figure 3B).

By using the 5’ and 3’ rapid amplification of cDNA ends (RACE), we identified one transcript of Fincor that is approximately 3.7 kb in length (Figure 3C). However, based on RNA-seq, the length of Fincor is over 10 kb (Figure 1C), suggesting there are likely additional multiple RNA isoforms that we were not able to identify by RACE. We next analyzed the coding potential of Fincor utilizing a comparative genomic program, PhyloCSF. While adjacent genes Gcnt1 and Prune2 (Figure 3D and E) were correctly predicted to encode proteins, Fincor did not contain a potential protein-coding open reading frame (Figure 3E). Consistent with this bioinformatic prediction, a vector expressing Fincor failed to produce any proteins in an in vitro transcription/translation assay (Figure 3F), confirming that the Fincor transcript is an ncRNA. Fincor transcripts were enriched by binding to oligo-dT beads, suggesting that Fincor is 3’ polyadenylated (Figure 3G). Further, Fincor was detected in the nuclear compartment and GW4064 treatment increased the nuclear abundance of Fincor (Figure 3H), consistent with its potential transcriptional regulatory function. Together, these results from molecular biochemical characterization studies reveal that Fincor is a liver-enriched nuclear polyadenylated eRNA.

To determine whether induction of the FXR-induced Fincor is ligand-specific, we examined the effects of several hammerhead-type synthetic FXR agonists, including GW4064, cilofexor, and tropifexor; a semi-synthetic agonist, OCA; and a non-hammerhead-type gut-specific agonist, fexaramine (Figure 4A; Downes et al., 2003; Fang et al., 2015). Remarkably, treatment with each of the hammerhead-type agonists for 1 hr resulted in a robust induction of Fincor (Figure 4B), whereas Fincor levels were unchanged after treatment with OCA or fexaramine for 1 hr (Figure 4B). Treatment with OCA for 4 hr or even 1 week treatment with OCA failed to induce hepatic Fincor in mice, while expression of Nr0b2 was significantly induced (Figure 4C and D). Acute feeding with a diet supplemented with 0.5% cholic acid (CA), a primary BA, for 6 hr also failed to induce Fincor expression (Figure 4E). Collectively, these results demonstrate that Fincor is induced specifically by hammerhead-type FXR agonists.

To explore the functional role of hepatic Fincor, we utilized the CRISPR/Cas9 technique to generate Fincor liver-specific knockdown (Fincor-LKD) mice (Figure 5—figure supplement 1A and B). Adenoviral-mediated sgRNA expression in Cas9 mice resulted in downregulation of Fincor specifically in the liver by about 60%, but not in other tissues (Figure 5B). As FXR is a key regulator of BA, cholesterol, lipid, and glucose metabolism and Fincor is specifically regulated by FXR, Fincor may have a role in the metabolic process in physiology and disease. We examined liver triglyceride, cholesterol, BA, glycogen, and serum non-esterified fatty acids (NEFA) but Fincor downregulation did not result in any significant changes under physiological conditions (Figure 5C).

To explore the molecular signatures and pathways affected by Fincor, we examined global gene expression by RNA-seq analysis in mouse liver after Fincor knockdown (Figure 5D). While Fincor was markedly downregulated, the neighboring genes, such as Gcnt1, Rfk, Pcsk5, and Prune2, were largely unchanged (Figure 5D; Figure 5—figure supplement 1C). The RNA-seq analysis revealed 18 upregulated genes and 53 downregulated genes in Fincor-LKD liver (Supplementary file 2; Figure 5E). GO analysis of those downregulated genes indicated that these genes were enriched in pathways involved in fatty acid oxidation, organelle organization, and metabolic process (Figure 5F). Among the downregulated genes, Ppp1r3g, which has a role in controlling glycogen synthesis, and Igfbp2, which functions in insulin resistance, were markedly reduced (Supplementary file 2; Figure 5—figure supplement 1D and E). Among the upregulated genes, expression was substantially increased for Eda2r, which is a member of the tumor necrosis factor receptor superfamily and involved in inflammation, the immune response, and development (Figure 5—figure supplement 1F). Fndc1, which is involved in fibronectin matrix remodeling, was also suppressed by Fincor (and thus upregulated upon its knockdown, Figure 5—figure supplement 1G). These studies suggest that Fincor has a role in modulating metabolic homeostasis by regulating genes involved in metabolism and inflammation.

Tropifexor, also known as LJN452, is a highly potent hammerhead-type FXR agonist that is currently under clinical trials for NASH and PBC patients (Kremoser, 2021; Sanyal et al., 2023; Tully et al., 2017). Because Fincor is induced specifically by the hammerhead class of FXR agonist (Figure 4) and has a potential role in the regulation of metabolism and inflammation (Figure 5), we hypothesized that Fincor may play a role in tropifexor-mediated beneficial effects on reducing NASH pathologies in mice.

We utilized a mouse model that had been fed the amylin liver NASH-promoting (AMLN) diet (Hernandez et al., 2019; Sun et al., 2022; Zhao et al., 2018), and examined the potential impact of liver-specific downregulation of Fincor on tropifexor’s effects on NASH pathology. Cas9 mice fed the AMLN diet for 12 weeks were injected via tail veins with adenovirus expressing control sgRNA or Fincor sgRNA, respectively, and then treated daily with tropifexor (0.3 mg/kg) for 12 days (Figure 6A). In these mice, as a technical validation of RNA induction and knockdown, Fincor levels were significantly increased by FXR agonist tropifexor, and the increase was blocked by adenovirus expressing sgRNA for Fincor (Figure 6B).

We then examined the effect of tropifexor treatment and Fincor downregulation on hepatic steatosis in these mice. Tropifexor treatment markedly reduced neutral lipids determined by Oil Red O staining of liver sections (Figure 6C) and liver TG levels (Figure 6D), and these beneficial effects on reducing fatty liver were not altered by Fincor downregulation. Also, Fincor downregulation had little effect on liver cholesterol and gallbladder BA levels, although gallbladder BA levels were reduced by tropifexor (Figure 6D).

Consistent with the phenotypes, hepatic expression of key genes involved in BA synthesis was dramatically reduced by tropifexor treatment (i.e. Cyp7a1 and Cyp8b1), which is consistent with decreased gallbladder BA levels mediated by this agonist (Figure 6E). Similarly, tropifexor also lowered lipid synthesis genes (Srebp1c, Lpin1, Scd1), consistent with decreased liver TG levels. However, downregulation of Fincor did not result in changes in mRNA levels of these genes (Figure 6E). These results indicate that tropifexor-mediated beneficial effects on reducing hepatic steatosis are independent of Fincor.

Tropifexor ameliorated fibrotic NASH pathologies in preclinical studies (Hernandez et al., 2019; Tully et al., 2017) and has recently concluded phase 2 clinical trials for NASH patients (Sanyal et al., 2023). We, therefore, further examined the effects of Fincor downregulation on altering other NASH pathologies, including hepatocellular apoptosis, liver fibrosis, and inflammation.

In the same AMLN diet-fed mice as described above (Figure 6), analyses of liver sections revealed that tropifexor treatment reduced hepatocyte swelling/ballooning (hematoxylin and eosin [H&E] staining), decreased numbers of apoptotic cells (TUNEL staining), alleviated fibrosis (Sirius Red staining), and lowered infiltration of macrophages (F4/80 staining) (Figure 7A). Remarkably, these tropifexor-mediated beneficial effects on NASH pathologies were all markedly diminished by Fincor downregulation (Figure 7A). This was consistently found in various liver lobes and two representative pictures from two different lobes are shown (Figure 7A). In control experiments, Fincor downregulation in vehicle-treated mice did not result in marked changes in NASH pathologies (Figure 7—figure supplement 1). These results demonstrate that Fincor is required for tropifexor-mediated beneficial effects on reducing NASH pathologies, specifically reducing hepatic inflammation, fibrosis, and hepatocyte apoptosis. Consistent with these results, serum ALT and AST levels, indicators of liver damage, were significantly elevated after Fincor downregulation (Figure 7B). Protein levels of key inflammatory markers, IL1β and CCL2, in liver extracts were also elevated after Fincor downregulation (Figure 7C).

Consistent with the phenotypes from histological analyses, expression of hepatic genes involved in the above pathological process was altered by Fincor knockdown. For example, tropifexor treatment reduced mRNA levels of several genes that promote fibrosis (Col1a1, Col1a2, Acta2) and hepatic inflammation (Eda2r, Ifng, Ccl3), whereas these reductions were largely reversed by Fincor downregulation (Figure 7D). We also detected increased expression of inflammatory genes (Ccl2, Ccr2, Lcn2) and an extracellular matrix remodeling gene (Fndc1) in liver with Fincor knockdown (Figure 7D). Tropifexor treatment suppressed hepatic apoptosis by reducing pro-apoptotic genes (Ctsb, Ctss) and upregulating anti-apoptotic genes such as Bcl2 (Warren et al., 2019). Importantly, these effects were significantly reversed by Fincor downregulation (Figure 7D).

Collectively, these results demonstrate that in diet-induced NASH mice, pharmacological activation of FXR by tropifexor reduced fibrosis, apoptosis, and inflammation, which was dependent, at least in part, on the induction of Fincor.

To determine whether expression of Fincor is altered in chronic liver disease, we utilized mouse models of NAFLD/NASH and cholestatic liver injury. Hepatic Fincor levels were significantly increased in mice fed with a high-fat diet (HFD) for 12 weeks (Figure 8A) and in mice fed an HFD with high fructose in drinking water for 12 weeks (Figure 8B). Elevated hepatic Fincor levels were also observed in mice treated with α-naphthylisothiocyanate (ANIT), a chemical inducer of liver cholestasis (Jung et al., 2020; Kim et al., 2016; Figure 8C), and in mice with bile duct ligation (BDL), a surgical method to induce cholestatic liver injury (Figure 8D).

The sequence of Fincor is moderately conserved between mice and humans as displayed in the UCSC genome browser (Figure 8E). Annotation in the NCBI genome data viewer of the human sequence region with similarity to mouse Fincor revealed an functionally uncharacterized human lncRNA, XR_007061585.1, in this region (Figure 8F). However, whether this sequence (or an as yet to be identified lncRNA) is functional or not has not been determined. To explore the potential changes or role of lncRNA XR_007061585.1 in human liver pathological conditions, we measured hepatic levels of the transcripts in PBC and NAFLD patients. Compared to normal individuals, hepatic lncRNA XR_007061585.1 levels were elevated in patients with PBC or NAFLD, but not in severe NASH-fibrosis patients (Figure 8G and H). These results demonstrate that hepatic levels of a potential human analog of Fincor are elevated in NAFLD and PBC patients, as Fincor is in mouse models of chronic liver disease with hepatic inflammation and liver injury. However, whether human lncRNA XR_007061585.1 is analogous to mouse Fincor in terms of functions and mechanisms, and whether elevated XR_007061585.1 levels may have a role in the disease progression or may be an adaptive response to liver injury remain to be determined.

All animal studies were performed according to procedures approved by the Institutional Animal Care & Use Committee at the University of Illinois at Urbana-Champaign (protocol # 17009) and were in accordance with National Institutes of Health guidelines. Mice were maintained in 12/12 hr light/dark cycles and fed standard rodent chow. Fxr-LKO mice were generated by breeding Fxr-Flox mice with Albumin-Cre mice (The Jackson Lab). Fincor-LKD mice were generated as previously reported (Zhao et al., 2018). Briefly, Cas9 transgenic mice (JAX #024858) were injected via the tail vein with adenoviruses (approximately 5×10⁸ PFU) expressing two sgRNAs targeting Fincor (sgRNA1: GGGTTAAGAGCTGTAGGCTG and sgRNA2: ACTTCTATGTCCAACAACCG). The sequences of sgRNAs were designed using a CRISPR design tool (http://crispr.mit.edu/).

Mice were given a single dose of vehicle or 30 mg/kg GW4064 (in corn oil, Tocris Bioscience, #2473) after overnight fasting. Mice were treated with 0.5 mg/kg tropifexor (in corn oil, MedChem Express, HY-107418), 30 mg/kg cilofexor (in corn oil, MedChemExpress, HY-109083), 100 mg/kg fexaramine (in 0.5% methylcellulose, MedChem Express, HY-10912), and 20 mg/kg OCA (in 0.5% methylcellulose, MedChem Express, HY-12222) as indicated. C57BL6 mice were fed with a chow diet containing 0.5% CA for 6 hr (Jung et al., 2020).

To induce cholestasis, mice were treated by gavage with 75 mg/kg ANIT for 48 hr as previously reported (Jung et al., 2020; Kim et al., 2016; Kim et al., 2020). Cholestasis was also induced by BDL or sham operation in mice for 24 hr or 72 hr (Li et al., 2018).

To induce dietary obesity, mice were fed an HFD (TD88137; Harlan Teklad) or an HFD with 25% fructose in water (high fat/high fructose) for 12 weeks (Seok et al., 2021).

To investigate the effect of liver-specific downregulation of Fincor on NASH, male Cas9 mice were fed the AMLN diet (Research Diets, D09100310, 40 kcal% fat, 2% cholesterol, 20 kcal% fructose) for 12 weeks and then, were injected with adenovirus expressing control sgRNA or sgRNA targeting Fincor. Administration of tropifexor was started 3 days later and given at 0.3 mg/kg dissolved in corn oil.

For all the mice experiments, the mice were randomly assigned to control group or treatment groups as needed.

C57BL/6 mice were fasted overnight and i.p. injected with vehicle or GW4064 (30 mg/kg) for 1 hr, and livers were collected (n=4 mice for either vehicle or treated group). Total RNA from each liver was extracted by RNeasy kit (QIAGEN), and two randomly selected mice liver RNAs were pooled for RNA-seq (two RNA-seq reactions from four mice livers for either vehicle or treated group). RNA-seq was performed as previously described (Byun et al., 2018; Byun et al., 2020; Seok et al., 2018). Ribosomal RNA was removed with the Ribozero HMR Gold kit (Illumina). The sequencing library was generated by the following methods described below.

Construction of the RNA-seq libraries and sequencing on the Illumina NovaSeq 6000 were performed at the Roy J. Carver Biotechnology Center at the University of Illinois at Urbana-Champaign. After DNase digestion, purified total RNAs were analyzed on a Fragment Analyzer (Agilent) to evaluate RNA integrity. The total RNAs were converted into individually barcoded polyadenylated mRNA-seq libraries with the Kapa HyperPrep mRNA kit (Roche). Libraries were barcoded with Unique Dual Indexes which have been developed to prevent index switching. The adaptor-ligated double-stranded cDNAs were amplified by PCR for eight cycles with the Kapa HiFi polymerase (Roche). The final libraries were quantitated with Qubit (Thermo Fisher) and the average cDNA fragment sizes were determined on a Fragment Analyzer. The libraries were diluted to 10 nM and further quantitated by qPCR on a CFX Connect Real-Time qPCR system (Bio-Rad) for accurate pooling of barcoded libraries and maximization of number of clusters in the flowcell.

The barcoded RNA-seq libraries were loaded on one SP lane on a NovaSeq 6000 for cluster formation and sequencing. The libraries were sequenced from one end of the fragments for a total of 100 bp. The fastq read files were generated and demultiplexed with the bcl2fastq v2.20 Conversion Software (Illumina, San Diego, CA, USA). The quality of the demultiplexed fastq files was evaluated with the FastQC software, which generates reports with the quality scores, base composition, k-mer, GC and N contents, sequence duplication levels, and overrepresented sequences.

To harvest the nuclei from mouse liver cells, the liver was harvested at indicated time and washed with a cold swelling buffer (10 mM Tris-HCl, pH 7.5, 2 mM MgCl₂, 3 mM CaCl₂, 2 U/ml Superase-In). The nuclei were prepared by Dounce homogenization in cold swelling buffer and filtered using a cell strainer (100 μm, BD Biosciences). Nuclei were collected by centrifugation at 400 × g for 10 min, then resuspended in the lysis buffer (swelling buffer with 10% glycerol and 1% IGEPAL) and incubated on ice for 5 min. Nuclei were washed twice with the lysis buffer and resuspended at a concentration of 10⁸ nuclei/ml in the freezing buffer (50 mM Tris-HCl, pH 8.3, 40% glycerol, 5 mM MgCl₂, 0.1 mM EDTA). We then followed our previous method (Li et al., 2013; Oh et al., 2021) to conduct nuclear run-on and GRO-seq library preparation. Briefly, the nuclei in freezing buffer were subjected to the nuclear run-on reaction by mixing with an equal volume of run-on buffer (10 mM Tris-HCl, pH 8.0, 5 mM MgCl₂, 1 mM dithiothreitol, 300 mM KCl, 20 units of Superase-In, 1% sarkosyl, 500 µM ATP, GTP, Br-UTP, and 2 µM CTP) for 5 min at 30°C. The BrU-labeled run-on RNAs were extracted by TRIzol and purified by anti-BrdU agarose beads (Santa Cruz Biotech, sc-32323 AC). The run-on RNAs were then subjected to end repair by T4 PNK, poly-adenylation, and then to cDNA first strand synthesis by a custom primer (oNTI223) that allows circularization of the cDNA. This cDNA then was re-linearized by Ape1 (NEB), size selected by TBE gel electrophoresis, and the products of the desired size were excised (~320–350 bp) for final library prep and sequencing. GRO-seq samples were run on a NextSeq 500 sequencer from Illumina with a single-end 80 nt model.

For histology, tissues were dissected and immediately fixed in 10% formalin overnight and processed for paraffin embedding and H&E staining. Paraffin-embedded liver sections were incubated with F4/80 antibody, and antibody was detected using a peroxidase-based method (Abcam, ab64238). Liver collagen was detected by Sirius Red staining (Abcam, ab246832) and apoptosis was detected by TUNEL staining (Millipore, S7100). For Oil Red O staining, liver tissue was frozen in OCT compound (Sakura Finetek, 4583), sectioned, and stained. Liver sections were imaged with a NanoZoomer Scanner (Hamamatsu) and quantification was done using NIH ImageJ.

Hepatic levels of TG (Sigma, MAK266), cholesterol (Sigma, MAK043), glycogen (Biovision, K646-100), total BA levels (Diazyme, DZ042A), serum NEFA (Sigma, MAK044), serum ALT (Sigma, MAK052), and serum AST (Sigma, MAK055) were determined according to the manufacturer’s instructions. Mouse liver IL-1β (R&D Systems, MLB00C) and CCL2 (R&D Systems, DY479-05) were detected by commercially available ELISA kit.

Liver specimens from normal organ donors, and patients with PBC or NAFLD were obtained from the Liver Tissue Procurement and Distribution System. The samples were unidentifiable, and thus, ethical approval was not required. Hepatic XR_007061585.1 levels were measured by qPCR.

Liver ChIP assay was performed as described previously (Jung et al., 2020; Seok et al., 2018). Briefly, chromatin extracts were prepared from Fxr-Flox and Fxr-LKO mouse livers after treatment of the mice with GW4064 which was followed by preclearing and immunoprecipitation using control IgG or FXR antibody (Novus Biologicals, NBP2-16550; Santa Cruz, sc-25309), RXRα antibody (Proteintech, catalog no. 21218-1-AP), or BRD4 antibody (Bethyl Laboratories, catalog #A301-985A50). Enrichment in chromatin precipitates of gene sequences was measured by qPCR using primers listed (Supplementary file 5).

Primary hepatocytes were isolated from C57BL/6 mice by collagenase (Worthington Biochemical Corporation, LS004188) perfusion and maintained in William’s E Medium with primary hepatocyte maintenance supplements (Gibco #CM4000) in six-well plates as described previously (Jung et al., 2020; Kim et al., 2015). HepG2 cells were purchased from the American Type Culture Collection and grown in Dulbecco’s modified eagle medium containing 10% FBS and 1% penicillin and streptomycin. This cell line was not authenticated after purchase but routinely tested negative for mycoplasma contamination.

The enhancer region containing the FXR binding element was amplified by PCR and cloned into the pGL4.23 vector (Promega). HepG2 cells were transiently transfected with FXR (100 ng/well) and RXRα (5 ng/well) in combination with reporters containing the wild-type FXRE or mutated FXRE (200 ng/well) (Supplementary file 5). β-Galactosidase plasmid (200 ng/well) was also transfected as an internal control. Cells were treated with GW4064 or DMSO after transient transfection. Six hours later, the cells were harvested. All reporter assays were repeated at least three times in triplicates.

5’ and 3’ RACE assays were performed using a SMARTer RACE kit (Clontech) according to the manufacturer’s instructions. The resulting PCR products were separated by electrophoresis in agarose gels and cloned into the pRACE vector provided by the kit. The transcription start sites and end sites of Fincor were determined by sequencing. The gene-specific primers used for 5’ and 3’ RACE are listed in Supplementary file 5.

Expression plasmids for luciferase and Fincor were mixed with a Coupled Reticulocyte Lysate System (Promega). After incubating at 30°C for 60 min, translated products were separated on 4–20% gradient SDS polyacrylamide gels and transferred to PVDF membranes. Chemiluminescent detection of in vitro translated protein was performed following the manufacturer’s protocol (Promega).

Total RNA was extracted using the RNeasy Mini Kit (QIAGEN, 74104) and 2 µg of RNA was reverse-transcribed and RNA expression was normalized relative to that of 36B4. The qPCR primers are shown in Supplementary file 5.

Using a Cytoplasmic and Nuclear RNA Purification Kit (Norgen, Thorold, ON, Canada), the cytoplasm and nucleus fractions from primary hepatocytes isolated from livers of mice treated with GW4064 or DMSO were separated, and both fractions were subjected to RNA extraction and qRT-PCR.

Total liver lysates were prepared as described before (Sun et al., 2022). Antibodies for immunoblotting for β-ACTIN (#4970) were purchased from Cell Signaling Technology. Antibodies for immunoblotting for FXR (sc-25309) were purchased from Santa Cruz.

Total RNA was prepared using the RNeasy mini prep kit (QIAGEN). Poly(A)+ and poly(A)- RNA was separated using a Dynabeads mRNA Purification Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. Briefly, total RNA was incubated with the Dynabeads/binding buffer suspension at room temperature for 5 min and the reaction tubes were placed on a magnet until solution was clear. The supernatant containing poly(A)- RNA was saved. The beads with poly(A)+ RNA were washed three times with washing buffer provided by the kit. RNA was extracted from the supernatant and beads respectively using TRIzol. qPCR was performed to analyze levels of Fincor and 36b4 in the poly(A)+ and poly(A)- RNA fractions.

RNA-seq and GRO-seq reads were mapped to the mouse reference genome mm10 with STAR aligner (Dobin et al., 2013). Transcript quantifications were done with the HOMER tool set (Heinz et al., 2010) and enhancer RNAs were identified based on H3K27ac ChIP-seq in mouse liver. Briefly, intergenic H3K27ac ChIP-seq peaks were selected as putative enhancer regions in mouse liver. Then ±3 kb regions around putative enhancers with RNA-seq signal (>1 RPKM [reads per kbp per million]) were considered as putative eRNAs in mouse livers. Overlapped eRNA regions were merged, and redundant ones were removed. In addition, any ±3 kb extended eRNA regions that overlapped with protein-coding genes were further removed to avoid transcriptional readthrough from genes. We used DESeq2 (Love et al., 2014) to identify significantly regulated eRNAs with a cutoff of (FDR <0.05, log₂FC >1). Public ChIP-seq datasets were obtained from ENCODE or GEO (see Data availability).

Statistical analysis was performed using GraphPad Prism 9. Statistical differences were evaluated using the two-tailed unpaired Student’s t-test for comparisons between two groups, or ANOVA and appropriate post hoc analyses for comparisons of more than two groups. Statistical methods and corresponding p-values for data shown in each panel are indicated in figure legends.