BECs accumulate lipids.

(A) Schematic depicting fatty acid (FA) treatment of BEC-organoids in vitro. (B) Representative brightfield and immunofluorescence (IF) images of lipids (BODIPY) in control (BSA) and FA-treated organoids. Close-up IF images were digitally zoomed in four times. n=3. (C) Cell-titer Glo and Caspase 3/7 activity measurement for viability and apoptosis detection relative to panel A. n=3. (D) Representative Nucgreen Dead 488 staining as composite images from brightfield and fluorescent microscopy. n=3. (E-F) Quantification of Scd1 (E) and Hmgcs2, Pdk4, and Aldh1a1 (F) mRNA in control (BSA) and FA-treated organoids. n=3. (G) Schematic depicting CD and HFD feeding in vivo. (H) Representative images for co-staining of BODIPY and PANCK, relative to panel G. Close-up IF images were digitally zoomed in four times. n=5.

Data are shown as mean ± SD. Absence of stars or ns, not significant (p > 0.05); *p < 0.05; ****p < 0.0001; one-way ANOVA with Dunnett’s test (C), and Fisher’s LSD test (E, F) was used. PV, portal vein. Arrowheads mark bile ducts. Scale bars, 200 μm (B-brightfield), 100 μm (B-IF, D) and 20 μm (H-I).

The following figure supplements are available for figure 1:

Figure supplement 1. Further characterization of lipid accumulation in BECs.

HFD feeding induces EPCAM+ BEC proliferation.

(A) Scheme depicting the isolation of EPCAM+ BECs from CD- and HFD-fed mice by FACS. (B) Gene set enrichment analysis (GSEA) of Gene Ontology (GO) terms. Top 15 upregulated biological processes (BP), ordered by normalized enrichment score (NES). q-value: false discovery rate adjusted p-values. NES: normalized enrichment score. (C-F) Representative co-staining images (C, E) and quantification (D, F) of BECs stained for OPN and Ki67 (C-D), and OPN and EdU (E-F) in livers of CD/HFD-fed mice. n=10 for Ki67 and n=5 for EdU. (G) Schematic depicting BEC-organoid formation in vitro from CD/HFD-fed mouse livers. (H) Images of organoid colonies formed 6 days after seeding, and quantification of organoids per well. n=5.

Violin graphs depict the distribution of data points i.e the width of the shaded area represents the proportion of data located there. Other data are shown as mean ± SD. **p < 0.01; ****p < 0.0001; unpaired, two-tailed Student’s t-test was used. PV, portal vein. Arrowheads mark bile ducts. Scale bars, 20 μm (C, E), 200 μm (H).

The following figure supplements are available for figure 2:

Figure supplement 1. RNA-seq analysis of EPCAM+ BECs upon HFD.

E2Fs are enriched in DDC and HFD datasets and mediate DR initiation in vivo.

(A) Over-representation analysis results. Top 13 enriched biological processes (BP) upon HFD (own data) and DDC (GSE125688) treatment. q-value: false discovery rate adjusted p-values, counts: number of found genes within a given gene set. (B-C) Enriched transcription factors (TFs) of upregulated genes identified by over-representation analysis in HFD (own data) and DDC (GSE125688) treatment (B), and during the process of organoid formation from single BECs (Organoids vs T0) (GSE123133) (C). Asterisk (*) marks TFs of the “TF_ZHAO” gene set. (D) Schematic depicting in vivo E2F1 analysis. (E-F) Representative images of PANCK/Ki67 co-staining in livers of E2f1+/+ and E2f1-/- mice fed with CD or HFD (E) and quantification of proliferative BECs in the indicated mice (F). For CD, n=5 for E2f1+/+ and E2f1-/-. For HFD, n=7 for E2f1+/+, and n=8 for E2f1-/-.

Violin graphs depict the distribution of data points i.e the width of the shaded area represents the proportion of data located there. ns, not significant; **p < 0.01; two-way ANOVA with Tukey’s test was used. PV, portal vein. Arrowheads mark bile ducts. Scale bars, 20 μm (E).

The following figure supplements are available for figure 3:

Figure supplement 1. Extended analysis of BECs upon HFD, DDC, and during BEC-organoid formation.

E2Fs promote glycolysis in BEC-organoids.

(A-B) Seahorse Substrate Oxidation Assay using UK5099, a mitochondrial pyruvate carrier inhibitor (A), and assessment of the glucose dependency (B) in CD-derived BEC organoids. n=7 for control (Ctrl), n=8 for UK5099. (C) Scheme depicting the treatment of CD/HFD-derived BEC-organoids with FA mix. (D-E) Proton efflux rate (PER) (D), and basal and compensatory glycolysis (E) measured using Seahorse XF Glycolytic Rate Assay. Relative to panel C. n=13. (F) Scheme depicting the treatment of CD-derived BEC-organoids with E2F inhibitor, HLM006474. (G) RT-qPCR of selected cell cycle and glycolytic genes, relative to panel H. n=4. (H-I) PER during the Seahorse XF Glycolytic Rate Assay (H), and basal and compensatory glycolysis (I), relative to panel F. n=10 for control (Ctrl), n=11 for HLM006474.

(J-K) PER during the Seahorse XF Glycolytic Rate Assay (J), and basal and compensatory glycolysis (K), relative to panel C and treatment with HLM006474. n=12 for control (Ctrl), n=11 for HLM006474.

Data are shown as mean ± SD (SEM for A, D, H, J). Absence of stars or ns, not significant (p > 0.05); *p < 0.05; **p < 0.01; ***p <0.001; ****p < 0.0001; unpaired, two-tailed Student’s t-test (G), two-way ANOVA with Sidak’s test (B, E, I, K) were used.

The following figure supplements are available for figure 4:

Figure supplement 1. E2F activation correlates with increased glycolysis.

Further characterization of lipid accumulation in BECs.

(A) Scheme depicting FA mix treatment for already grown BEC-organoids for 4 days. (B) Representative brightfield and immunofluorescence (IF) images of lipids (BODIPY) and ACTIN co-staining, relative to panel A. n=4. (C) Measurement of triglyceride concentration upon FA mix treatment relative to panel A, normalized by protein amount. n=4. (D) Cell-titer Glo measurement for viability detection relative to panel A. n=4. (E) Representative images of CD/HFD-fed mouse livers and hematoxylin and eosin (H&E) staining and the final body-weight measurement at the end of the experiment. n=10. (F) Representative Oil Red O staining, relative to panel E. n=8. (G- H) Representative cleaved caspase 3 stainings, relative to panel E (G) and quantification of apoptotic cells in bile ducts (H). n=3.

Data are shown as mean ± SD. Absence of stars or ns, not significant (p > 0.05); *p < 0.05; ***p < 0.001; one-way ANOVA with Dunnet’s test (C, D) or unpaired, two-tailed Student’s t-test (E) were used. PV, portal vein. Arrowheads mark bile ducts. Scale bars, 100 μm (B-E), 50 μm (G-IF), and 20 μm (F, G-brightfield).

RNA-seq analysis of EPCAM+ BECs upon HFD.

(A) FACS gating strategy for isolation of CD45−/CD11b−/CD31−/EpCAM+ BECs: individual cells were sequentially gated based on cell size (FSC versus SSC) and singlets. BECs were then selected based on EPCAM positivity after excluding leukocytes (CD45+), myeloid cells (CD11b+), and endothelial cells (CD31+), yielding a population of single CD45-/CD11b-/ CD31-/EPCAM+ BECs.

(B) Principal component analysis (PCA) of mRNAs measured in mice fed CD or HFD by RNA- seq. n= 5 for CD, n=7 for HFD. (C) Volcano plot of HFD vs CD differential analysis. Top 20 differentially expressed genes were labeled. Blue dots represent downregulated genes (log2(FC) < −1 & adj. p-value < 0.05). Red dots represent upregulated genes (log2(FC) > 1 & adj. p-value < 0.05). Grey dots represent genes not changing significantly. (D) Box plot representing the differential gene expression of Ncam1. n= 5 for CD, n=7 for HFD. The Y-axis depicts log2(cpm+1) values. (E) Gene set enrichment analysis (GSEA) of KEGG terms. Top 15 enriched pathways (sorted by q-value). q-value: false discovery rate adjusted p-values. NES: normalized enrichment score.

Data are shown as mean ± SD. Unpaired two-tailed Student’s t-test was used.

Extended analysis of BECs upon HFD, DDC, and during BEC-organoid formation.

(A) Over-representation analysis results. Top 15 enriched biological processes (BP) upon HFD (own data) and DDC (GSE125688) treatment in BECs. q-value: false discovery rate adjusted p-values, counts: number of found genes within a given gene set. (B) Heatmap of E2F1-4 target genes from TF gene sets. Genes with absolute fold change lower than 0.5 were not shown. (C) GO over-representation analysis of upregulated BP during the process of organoid formation from single BECs (Organoids vs T0) (GSE123133).

E2F activation correlates with increased glycolysis.

(A) Over-representation analysis results. Top 22 enriched KEGG and EHMN pathways during the process of organoid formation from single BECs (Organoids vs T0) (GSE123133). q-value: false discovery rate adjusted p-values, counts: number of found genes within a given gene set. (B-C) Seahorse Substrate Oxidation Assay using BPTES, n=8 (B), or Etomoxir, n=7 for control (Ctrl) and n=8 for inhibitor conditions (C) in CD-derived BEC-organoids. (D-E) Scheme depicting Seahorse Glycolytic Rate Assay (D), and Mito Stress Test (E). (F-G) Oxygen consumption rate (OCR) (F), and basal and maximal respiration (G) measured using Seahorse XF Mito Stress Test. Relative to panel C. n=14.

Data are shown as mean ± SD (SEM for F). absence of stars or ns, non-significant (p-value > 0.05);

****p < 0.0001; two-way ANOVA with Sidak’s test (B, C, G) was used.

Primers used for qPCR analysis.