Activation of FXR by GW4064 induces FincoR, a novel eRNA, in mouse liver.

(A) C57BL/6 male mice were fasted overnight and i.p. injected with vehicle or GW4064 (30 mg/kg) for 1 h, and livers were collected (n = 4 mice). RNAs from two mice (per group) were pooled for RNA-seq. Volcano plot showing significantly induced eRNAs (FincoR highlighed) by GW4064 treatment. The x axis denotes log2 fold change (GW4064/Veh) of eRNAs and the y axis denotes -log10 FDR of eRNAs. (B) The bar plot showing the RPMs of up-regulated eRNAs by GW4064 treatment. (C) FincoR induction by GW4064 treatment was shown in the genome browser track. (D) GW4064 induction of FincoR and Shp was measured in the liver by qPCR (n = 4 mice). Data are presented as mean ± SEM. Statistical significance was determined by the two-way ANOVA Sidak’s multiple comparisons test with *p < 0.05 and ***p < 0.001. Veh, vehicle; GW, GW4064; pos, positive strand; neg, negative strand; Refseq, reference sequence.

Ligand-activated FXR directly activates transcription of eRNAs including FincoR in the liver.

(A) Experimental outline: FXR-Flox and LKO male mice were fasted overnight and treated with vehicle or GW4064, livers were collected 1 h later (n = 2 each) and nuclei was isolated for Gro-seq. Box plot shows the up-regulated eRNA GRO-Seq RPKMs in different conditions. (B) Gro-seq profile of FincoR expression and genome browser tracks of H3K4me1, H3K27ac, FXR and RXRα binding peaks at FincoR locus were shown. (C) ChIP assays were performed in liver samples from Fig. 1A to detect FXR, RXRα, and BRD4 occupancy to the FXR binding peak region close to the transcription start site of FincoR. (D) HepG2 cells were transfected with luciferase reporter expressing wild type FXRE or mutant FXRE (see materials and methods) for 24 hrs before further subject to GW4064 for 6 hrs. The relative luciferase activity was shown.

FincoR is a liver-specific nucleus-enriched eRNA.

(A) C57BL/6 male mice were fasted overnight and i.p. injected with vehicle or GW4064 (30 mg/kg) for 1 h. Tissues were collected and FincoR mRNA level was measured (n = 2 mice). (B) C57BL/6 male mice were fasted overnight and i.p. injected with vehicle or GW4064 (30 mg/kg) for 1 h. One small piece of liver was snap-frozen for later RNA isolation and the remaining part was used for immediate primary hepatocyte isolation. Then, RNAs were extracted from liver or primary hepatocytes and FincoR expression was measured. (C) Agarose gel electrophoresis of PCR products generated in 5′ (left) and 3′ (right) RACE of FincoR in liver samples. Primer locations were show. (D) Schematic diagrams of full-length FincoR. (E) PhyloCSF analysis of the coding potential of FincoR. (F) In vitro translation of FincoR using the Promega Transcend Non-Radioactive Translation Detection Systems. Luciferase is used as controls for coding RNA. (G) qPCR analysis of FincoR, 36b4 in Poly(A)+ and Poly(A)-RNA fractions from GW4064 treated mouse liver. (H) FincoR identified in the subcellular fraction using cellular fractionation assays. The primary hepatocytes were isolated from GW4064 or DMSO treated mouse liver and the cytoplasm and nucleus fractions of these hepatocytes were separated and both fractions were subjected to RNA extraction and qPCR. RACE, rapid amplification of cDNA ends.

FincoR is induced by the hammerhead class of non-steroidal FXR agonists, including GW4064 and tropifexor.

(A) The chemical structures of the FXR agonists including the hammerhead class of non-synthetic FXR agonists, non-hammerhead-type non-synthetic agonists, semi-synthetic BA and natural BA. (B) C57BL/6 mice were treated with GW4064 (30 mg/kg) or Cilofexor (30 mg/kg) or Tropifexor (0.5 mg/kg) or Fexaramine (100 mg/kg) or OCA (20 mg/kg) for 1 h after over-night fasting. The liver RNAs were extracted and FincoR expression was measured (n = 3∼4 mice). (C) C57BL/6 mice were treated with OCA (20 mg/kg) or Fexaramine (100 mg/kg) for 4 hrs after over-night fasting. The liver RNAs were extracted and FincoR expression was measured (n = 3 mice). (D) C57BL/6 mice were daily treated with OCA (20 mg/kg) for 7 days after over-night fasting. The liver RNAs were extracted and FincoR expression was measured (n = 5 mice). Shp gene was used as a positive control. (E) C57BL/6 mice were fed with 0.5% cholic acid (CA) diet for 6 hrs after over-night fasting. The liver RNAs were extracted and FincoR expression was measured (n = 3∼4 mice). (B-E) Data are presented as mean ± SEM. Statistical significance was determined by the Student’s t test with *p < 0.05, **p < 0.01 and ***p < 0.001.

Generation of CRISPR/Cas9-mediated FincoR liver-specific knockdown mice.

(A) Male Cas9 mice were infected with adenovirus expressing sgRNA for FincoR or control for 1 week. Then the liver and serum were collected from these mice after 4∼5 hrs of fasting. (B) The expression of FincoR in the liver was measured by qPCR (n = 4 mice). (C) Hepatic triglyceride, cholesterol, bile acid, glycogen and serum NEFA were measured (n = 5∼7 mice) (D) Male Cas9 mice were infected with adenovirus expressing sgRNA for FincoR or control for 1 week. Then these mice were fasted overnight and treated with GW4064 for 3 hrs before tissue collection. RNA-seq profiles of expression of hepatic FincoR and the adjacent genes were shown (n = 2 mice). (E) Genome-wide changes in mRNA expression shown in a volcano plot. The numbers refer to the number of genes up-or down-regulated by 2-fold or more with a p-value < 0.01. (F) Gene ontology analysis of biological pathways using DAVID Tools.

In diet-induced NASH mice, tropifexor-mediated beneficial effects on reducing hepatic steatosis are largely independent on FincoR.

(A-E) Male Cas9 mice were fed with NASH diet for 12 weeks. Then these mice were randomly assigned to 3 groups and infected with adenovirus expressing sgRNA for FincoR or control. Three days later, the mice were treated with Tropifexor (0.3 mg/kg) for 12 days. On the tissue collection day, the mice were given the last shot of Tropifexor or vehicle and fasted for 4 hrs before tissues were collected (n = 6 ∼ 7 mice). (A) Experimental scheme. (B) Hepatic FincoR expression was measured (n = 6∼7 mice). (C) Oil Red O staining of liver sections. Scale bar (50 μm). (D) Hepatic TG, hepatic cholesterol, gallbladder BA and hepatic BA levels were measured (n = 6 ∼ 7 mice). (E) mRNA levels in the liver of the indicated genes involved in bile acid regulation and lipid regulation (n = 6 ∼ 7 mice). (B, D and E) Data are presented as mean ± SEM. Statistical significance was determined by the one-way ANOVA (Sidak’s multiple comparisons test) with *p < 0.05, **p < 0.01 and ***p < 0.001. Ad, adenovirus; H & E, hematoxylin and eosin; TG, triglyceride; Veh, vehicle; ns, not significant.

In diet-induced NASH mice, tropifexor-mediated beneficial effects on reducing liver fibrosis and inflammation are diminished by FincoR downregulation.

H & E, F4/80, Sirius Red and TUNEL staining of liver sections from Figure 6. Scale bar (50 μm). Image analyses were done using Image J and the area of collagen staining, TUNEL and F4/80 levels were quantified (n = 5/group). (B) Serum ALT level was measured (n = 5/group). (C) IL-1β and CCL2 levels in the liver tissues were determined by ELISA (n = 5/group). (D) mRNA levels in the liver of the indicated genes involved in inflammation, fibrosis and cell death (n = 5/group). (E) Model: FincoR is a liver-specific lncRNA that is induced specifically by hammerhead-type FXR agonists (top). In diet-induced NASH mice, FincoR is required for tropifexor-mediated beneficial effects on reducing hepatic inflammation, fibrosis, and cell death with the mechanisms to be explored further (bottom). (A-D) Data are presented as mean ± SEM. Statistical significance was determined by the one-way ANOVA (Sidak’s multiple comparisons test) with *p < 0.05, **p < 0.01 and ***p < 0.001.

(A, B) Examples of FXR-regulated eRNAs produced near the genes Hes1 (A) and Slc35g1 (B) were shown. (C) Time course expression of FincoR. C57BL/6 male mice were fasted overnight and i.p. injected with vehicle or GW4064 (30 mg/kg) for 1, 3, 6,12 h. Livers were collected at indicated time points (n = 3 mice) and FincoR, Shp and Gcnt1 were measured. Data are presented as mean ± SEM. Statistical significance was determined by the two-way ANOVA Sidak’s multiple comparisons test with *p < 0.05 and ***p < 0.001.

(A) Volcano plot showing the differential expressed genes (DEGs) by GW4064 treatment (DESeq2 FDR<0.05). The numbers refer to the number of genes up-or down-regulated. (B) Bar plot showing the enriched Gene ontology in terms of biological processes for up-regulated genes.

(A) FXR protein levels in the livers isolated from FXR-Flox and FXR-LKO mice were shown. (B) Validation of hepatic FXR-dependent induction of FincoR by qPCR (n = 3 mice). Shp gene was used as a control. Data are presented as mean ± SEM. Statistical significance was determined by the two-way ANOVA Tukey’s multiple comparisons test with *p < 0.05, **p < 0.01 and ***p < 0.001. (C) Metagene plots showing H3K27ac, H3K4me1, FXR, and RXRα ChIP-Seq profiles centered on up-regulated eRNAs.

(A) The scheme for the FincoR loss of function experiments. (B) The genomic DNAs from liver, spleen, intestine, brain, heart, muscle, kidney, lung and adipose tissue were isolated and PCR was performed using the primers (Supplementary Table S5E) to verify tissue-specific knock-out. (C-G) RNA-seq profiles of expression of hepatic Prune2 (C), PPP1r3g (D), Igfbp2 (E), Eda2r (F) and Fndc1 (G) were shown (n = 2 mice).

The effects of FincoR downregulation on NASH pathologies.

Male Cas9 mice were fed with NASH diet for 12 weeks. Then these mice were randomly assigned to 2 groups and infected with adenovirus expressing sgRNA for FincoR or control. The tissues were collected 2 weeks later. Liver histology analysis was performed. Scale bar (50 μm).

Hepatic genome browser tracks of FXR, RXRα, LXR, PPARα, and HNF4α binding peaks at FincoR locus.

PPARα occupnacy in the FincoR enhancer region.

C57BL/6 male mice were fasted overnight with or without refed for 3 hrs and then sacrificed. ChIP assays were performed in liver samples to detect PPARα occupancy to the FXR binding peak region close to the transcription start site of FincoR.

Sequencing data generated in this study

Public ChIP-Seq dataset

Primer sequences used in this study