Hepatic progenitor cell-originated ductular reaction facilitates liver fibrosis through activation of hedgehog signaling

Background: It is poorly understood what cellular types participate in ductular reaction (DR) and whether DR facilitates recovery from injury or accelerates hepatic fibrosis. The aim of this study is to gain insights into the role of hepatic progenitor cell (HPC)-originated DR during fibrotic progression. Methods: DR in liver specimens of PBC, chronic HBV infection (CHB) or NAFLD, and four rodent fibrotic models by different pathogenic processes was evaluated. Gli1 expression was inhibited in rodent models or cell culture and organoid models by AAV-shGli1 or treating with GANT61. Results: Severity of liver fibrosis was positively correlated with DR extent in patients with PBC, CHB or NAFLD. HPCs were activated, expanded, differentiated into reactive cholangiocytes and constituted “HPC-originated DR”, accompanying with exacerbated fibrosis in rodent models of HPC activation & proliferation (CCl4/2-AAF-treated), Μdr2-/- spontaneous PSC, BDL-cholestatic fibrosis or WD-fed/CCl4-treated NASH-fibrosis. Gli1 expression was significantly increased in enriched pathways in vivo and in vitro. Enhanced Gli1 expression was identified in KRT19+-reactive cholangiocytes. Suppressing Gli1 expression by administration of AAV-shGli1 or GANT61 ameliorated HPC-originated DR and fibrotic extent. KRT19 expression was reduced after GANT61 treatment in sodium butyrate-stimulated WB-F344 cells or organoids or in cells transduced with Gli1 knockdown lentiviral vectors. In contrast, KRT19 expression was elevated after transducing Gli1 overexpression lentiviral vectors in these cells. Conclusions: During various modes of chronic injury, Gli1 acted as an important mediator of HPC activation, expansion, differentiation into reactive cholangiocytes that formed DR, and subsequently provoked hepatic fibrogenesis.

from paracancerous liver tissues based on SR staining.Immunofluorescent costaining of KRT19 and Gli1 was performed with paraffin sections, and immunofluorescent co-staining of Epcam and KRT7, as well as KRT19 and α-SMA, was performed in cryosections.
All experimental animals were housed or bred, and maintained under specific pathogen-free (SPF) conditions in the Animal Experiment Center of Shanghai University of Traditional Chinese Medicine (Shanghai, China), with constant temperature and humidity and a 12-hour light/dark cycle.Animals were fed standard chow and had access to water ad libitum.
Rats were gavaged with 2-AAF and simultaneously injected subcutaneously with 30% CCl 4 in olive oil solution in order to promote massive proliferation of HPCs during continuous development of hepatic fibrogenesis.Another group of rats received an equal amount of subcutaneous olive oil injection and equal amount of normal saline by gavage as control (Oil group).Rats were euthanized under sodium pentobarbital i.p. injection at the end of 9 th week, and liver tissues were collected for analysis (Fig.

3A).
Mdr2 -/-spontaneous PSC mice: Due to a deficiency of canalicular phospholipid translocase in Mdr2 -/-mice, the lack of phospholipids in the bile results in spontaneous liver injury, and consequently progresses to primary sclerosing cholangitis [3].Mdr2 -/- mice were randomly divided into the control adeno-associated virus vector (AAV-NC) group and AAV-shGli1 group, or the solvent-treated group and the GANT61-treated group (GANT61, 50 mg/kg of body weight; gavage, once every other day) at 8 weeks of age.Mice were all euthanized at the end of the 11 th week, and liver tissues were collected for analysis.

BDL-cholestatic fibrosis in rats:
SD rats were randomly divided into the shamoperated group (Sham group) and the BDL group.All rats were anesthetized via inhalation of 5% isoflurane, and anesthesia was maintained with 1.5% isoflurane.
Abdominal fur was shaved, and the shaved abdominal skin was sterilized in rats.The abdomen was opened with a midline laparotomy, and the common bile duct, left hepatic duct, and right hepatic duct were separated.The abdomen was then closed for the Sham group.For the BDL group, the left and the right hepatic ducts were ligated.
Additionally, the upper end of the common bile duct near the confluence of hepatic ducts was ligated once and the lower end near the duodenum was ligated, and then the abdomen was closed.One week post BDL establishment, rats were randomly divided into the GANT61-treated group (GANT61, 25 mg/kg of body weight; gavage, once every other day) and the model group.The rats were euthanized at the end of the fourth week, and liver tissues were collected for analysis.WD-fed/CCl 4 -treated NASH-fibrosis model: C57BL/J mice were randomly divided into a normal control group and a WD-fed/CCl 4 -treated group.WD-fed/CCl 4 -treated mice were given high-fat, high-fructose, and high-cholesterol diet as previously described [4], which contained 21.1% fat, 41% sucrose, and 1.25% cholesterol (Teklad diets, TD, 120528), in addition to a high sugar solution containing 23.1 g/L d-fructose (Sigma-Aldrich, F0127, Shanghai, China) and 18.9 g/L d-glucose (Sigma-Aldrich, G8270).Diet administration and subcutaneous injections of 10% CCl 4 in olive oil solution (0.02 mL/10 g of body weight; once per week) started simultaneously.Mice were euthanized at the end of 16 th week, and liver tissues were collected for analysis.

Cell culture and SB-induced WB-F344 cell differentiation
Hepatic progenitor cell line WB-F344 was purchased from Xiangf Bio (Shanghai, China).Cells were grown in DMEM high glucose medium (Thermo Fisher, 11965092) containing 10% fetal bovine serum (FBS) (Thermo Fisher, 10100154) and cultured at 37°C with 5% CO 2 concentration and 95% relative humidity in an incubator.Once cells were in 80-90% confluence, they were passaged, cryopreserved, or inoculated to multiple-well plates for experiments.

Co-culture system of WB-F344 cells and HSCs
Transwell (0.4 μm, Corning, USA) was used for co-culture of WB-F344 cells with LX2 cells, a human HSC line.WB-F344 cells were cultured in the lower chamber of the transwell, and SB was used to induce WB-F344 cell dedifferentiation.After two days of treatment with SB, LX-2 cells, which was pre-cultured in upper chamber of transwell for 24 h, ,were co-cultured in upper chamber with WB-F344 cells in lower chamber for another two days in transwells and were collected, separately.All cell experiments were repeated three times using independent cell cultures.

Edu staining
The Cell-LightTM Edu Apollo®488 In Vitro Imaging Kit (Ribobio, Guangzhou, China) was used to assess cell proliferation.Briefly, cells were subjected to Edu labeling for 2 hours following the manufacturer's protocol.Visualization of Edu incorporation was performed using the ArrayScan VTI HCS Reader and quantitation was carried out using Cell Health Profiling BioApplication Software.

Hydroxyproline (Hyp) content measurement
Liver tissues (100 mg) were measured for Hyp content based on the improved method by Jamall et al [5].In brief, liver tissue homogenates were hydrolyzed in hydrolyzing solution at 110°C for 18 hours.After the homogenate was filtered through filter paper, chloramine-T was added to the hydrolysate, and mixed.Then dimethylaminobenzaldehyde at 410 mM (Sinopharm, Shanghai, China) was added and the solution was mixed, followed by incubation at 60°C for 30 min.The OD value at 560 nm was recorded after the samples cooled to room temperature.As an indirect measurement of tissue collagen content, hepatic Hyp content was presented as wet weight (mg/g).The Hyp standard was purchased from Sigma-Aldrich.

Histopathology and immunohistochemical analysis
Liver sections were embedded in paraffin, and were cut at a thickness of about 4 μm.
Sections were stained with H&E (lot.20161225, NJBI, Nanjing, China) and SR to evaluate liver injury and fibrosis status, respectively.As described previously, immunohistochemical staining was performed using paraffin-embedded liver sections or 8-μm thick OCT-embedded frozen liver sections.Briefly, sections were deparaffinized and washed, and the antigens were retrieved, and endogenous peroxidase activity was blocked.Sections were then blocked with 10% goat serum and incubated with primary antibodies overnight at 4°C.After washing, sections were incubated with the corresponding secondary antibodies conjugated with HRP (GTVision III Immunohistochemical Detection Kit, HRP/DAB, anti-mouse/rabbit IgG, two-step, GK5005/5007, Gene Tech, Shanghai, China).The slides were then washed, colored with DAB, counterstained with hematoxylin, washed, dehydrated and sealed.
Stained liver tissue sections were scanned by a Leica SCN 400 slide scanner (Leica Microsystems Ltd., Mannheim, Germany).The antibodies used for this study are listed in Table S1.
Cells were observed and photographed by a DP80 fluorescence inverted microscope (Olympus, Beijing, China) or laser scanning confocal microscope (Olympus, Beijing, China).The antibodies used for this study are listed in Table S1.

Western blot and densitometric analysis
Liver tissue (20-50 μg) was used for Western blot analysis, homogenized with RIPA buffer (Beyotime Biotechnology, P10013B, Shanghai, China) supplemented with protease and phosphatase inhibitor cocktail (Beyotime Biotechnology, P1045).For cellular experiments, 6-well plates were used for culture, and RIPA buffer was added to lyse cells.An Odyssey 2.1 software of Odyssey infrared scanner (LI-COR Biosciences, Lincoln, USA) was employed for scanning, and Image-J software was used for densitometric analysis.All antibodies used for this study are listed in Table S1.

Quantitative reverse transcript polymerase chain reaction (qRT-PCR)
Total RNA from frozen liver tissues or cells was isolated by MagExtractor Total RNA Purification Kit (TOYOBO, NPK-201, Shanghai, China).ReverTra Ace qPCR RT Master Mix with gDNA Remover (TOYOBO, FSQ-301) was used for reverse transcription.SYBR Green Realtime PCR Master Mix (TOYOBO, QPK-201) was employed for qRT-PCR.The primer sequences used are listed in Table S2.

RNA sequencing and bioinformatic analysis
Total RNA was extracted and then used to constructed cDNA libraries.The libraries were sequenced on an Illumina Novasep 6000 platform and 150 bp paired-end reads were generated.FPKM of each gene was calculated using Cufflinks, and the read counts of each gene were obtained by HTSeq-count.Differential expression analysis was performed using the DESeq2 R (v3.2.0).p value < 0.05 and fold-change > 2 or fold-change < 0.5 was set as the threshold for significantly differential expression.
Hierarchical cluster analysis of differentially expressed genes (DEGs) was performed to demonstrate expression pattern of genes in different groups and samples.GO enrichment and KEGG pathway enrichment analysis of DEGs were performed respectively using R based on the hypergeometric distribution.The transcriptome sequencing and analysis were conducted by OE Biotech Co., Ltd.(Shanghai, China).

Gene set enrichment analysis (GSEA)
GSEA was implemented on the Java GSEA platform.For each GO enrichment and KEGG pathway, genes involved were defined as a gene set, and a ranked list and a 'gene set' permutation were then generated.NES values were normalized enrichment scores.A p value < 0.05 was set as the threshold for the enriched and trusted gene set.

Figure S9 .
Figure S9.Expression of Gli1 signaling pathway during the differentiation of WB-F344 cells into cholangiocytes.(A) Western blot and immunofluorescence staining of KRT19 (scale bar = 100 µm).(B) Gene expression of Krt19, Dhh, Gli2, Gli3, and Smo was determined by qRT-PCR.(C) The effect of GANT61 on the viability of WB-F344 cells.(D) Immunoblotting for KRT19 and TNF-α.GAPDH was used as loading control.(E) Representative images of WB-F344 cells stained with Edu (scale bar = 100 µm) with or without GANT61 and Edu + cells per HPF.(F) Gene expression of Krt19 and Krt7 of WB-F344 organoids.(G) Gli1 mRNA expression was determined by qRT-PCR.All mRNA values were normalized against Gapdh levels and are shown relative to expression level in the control group.(H) Representative images of WB-F344 cells stained with Gli1 after Gli1 knockdown (scale bar = 33.3µm).*, p < 0.05; **, p < 0.01.Ctrl: the control group; SB: the sodium butyrate-treated group; shNC: the negative lentivirus group; shGli1: the Gli1 knockdown group.

Figure S10 .
Figure S10.Heat map of the significantly enriched items in livers of Μdr2 -/-mice.(A) Heat map of cell differentiation.(B) Heat map of positive regulation of cell migration.(C) Heat map of cAMP signaling pathway.(D) Heat map of positive regulation of cell population differentiation.(E) Heat map of response to wounding.Heat map showed that Gli1 expression increased in all of these significantly enriched items (red arrow).

Figure S11 .
Figure S11.Heat map of the significantly enriched items in WB-F344 cells, treated with or without SB.(A) Heat map of response to wounding.(B) Heat map of positive regulation of cell migration.(C) Heat map of epidermal cell differentiation.(D) Heat map of cAMP signaling pathway.Heat map showed that Gli1 expression increased in all of these significantly enriched items (red arrow).

Figure S12 .
Figure S12.The extended figures of Figure 4E.Confocal analysis of co-staining for KRT19 (red) and Gli1 (green) of paraffin sections in hemangioma/paracancerous tissues of patients with hepatic hemangioma and HCC (scale bar = 100 µm and 25 µm).Nuclei counterstained with DAPI (blue).Negative pictures showed the serial sections with negative staining.

Figure S13 .
Figure S13.Confocal analysis of co-staining for KRT19 (red) and Gli1 (green) of frozen sections of Mdr2 -/-mice after injected with AAV-NC or AAV-shGli1 (scale bar = 100 µm).Higher magnification of the white box area (scale bar = 25 µm) is shown in the right-most column.Nuclei counterstained with DAPI (blue).

Figure S14 .
Figure S14.The effect of GANT61 on DR and the degree of hepatic fibrosis in Mdr2 -/-mice.(A) Confocal analysis of co-staining for KRT19 (red) and Gli1 (green) (scale bar = 100 µm).(B) Confocal analysis of co-staining for Epcam (red) and KRT7