Advertisement

Inflammatory type 2 conventional dendritic cells contribute to murine and human cholangitis

Open AccessPublished:July 04, 2022DOI:https://doi.org/10.1016/j.jhep.2022.06.025

      Highlights

      • Cholangitis in patients with PSC and mouse models is associated with portal expansion of cDC2s.
      • Cholangitis in mice was aggravated following depletion of dendritic cells sparing cDC2s.
      • Single-cell analyses of hepatic dendritic cells revealed an inflammatory phenotype of accumulated cDC2s.
      • This phenotype corresponded to the previously described inflammatory cDC2B subtype.
      • Expanded hepatic cDC2s in cholangitis promote and maintain Th17 responses of infiltrating T cells.

      Background & Aims

      Primary sclerosing cholangitis (PSC) is a progressive cholangiopathy characterised by fibrotic stricturing and inflammation of bile ducts, which seems to be driven by a maladaptive immune response to bile duct injury. The histological finding of dendritic cell expansion in portal fields of patients with PSC prompted us to investigate the role of dendritic cells in orchestrating the immune response to bile duct injury.

      Methods

      Dendritic cell numbers and subtypes were determined in different mouse models of cholangitis by flow cytometry based on lineage-imprinted markers. Findings were confirmed by immunofluorescence microscopy of murine livers, and liver samples from patients with PSC were compared to control samples from bariatric surgery patients. Using genetic tools, selected dendritic cell subsets were depleted in murine cholangitis. The dendritic cell response to bile duct injury was determined by single-cell transcriptomics.

      Results

      Cholangitis mouse models were characterised by selective intrahepatic expansion of type 2 conventional dendritic cells, whereas plasmacytoid and type 1 conventional dendritic cells were not expanded. Expansion of type 2 conventional dendritic cells in human PSC lesions was confirmed by histology. Depletion studies revealed a proinflammatory role of type 2 conventional dendritic cells. Single-cell transcriptomics confirmed inflammatory maturation of the intrahepatic type 2 conventional dendritic cells and identified dendritic cell-derived inflammatory mediators.

      Conclusions

      Cholangitis is characterised by intrahepatic expansion and inflammatory maturation of type 2 conventional dendritic cells in response to biliary injury. Therefore, type 2 conventional dendritic cells and their inflammatory mediators might be potential therapeutic targets for the treatment of PSC.

      Lay summary

      Primary sclerosing cholangitis (PSC) is an inflammatory liver disease of the bile ducts for which there is no effective treatment. Herein, we show that the inflammatory immune response to bile duct injury is organised by a specific subtype of immune cell called conventional type 2 dendritic cells. Our findings suggest that this cell subtype and the inflammatory molecules it produces are potential therapeutic targets for PSC.

      Graphical abstract

      Keywords

      Introduction

      Primary sclerosing cholangitis (PSC) is a progressive cholestatic liver disease that is characterised by inflammation and fibrotic stricturing of intra- and extrahepatic bile ducts.
      • Dyson J.K.
      • Beuers U.
      • Jones D.E.J.
      • Lohse A.W.
      • Hudson M.
      Primary sclerosing cholangitis.
      There is no effective treatment that can arrest PSC progression; liver transplantation is the only known cure for advanced PSC. Hence, patients with PSC are at a high risk of developing cirrhosis and cholangiocarcinoma. Therefore, it is of great importance to better understand the elusive aetiology and pathogenesis of PSC.
      It is believed that cholangitis is initiated by stress or damage of cholangiocytes caused by heterogeneous insults, including toxicity or infection.
      • Lazaridis K.N.
      • LaRusso N.F.
      Primary sclerosing cholangitis.
      Such bile duct injury then induces a reactive phenotype of cholangiocytes and the recruitment of various immune cell types, promoting acute peribiliary inflammation.
      • Lazaridis K.N.
      • LaRusso N.F.
      Primary sclerosing cholangitis.
      ,
      • Azad A.I.
      • Krishnan A.
      • Troop L.
      • Li Y.
      • Katsumi T.
      • Pavelko K.
      • et al.
      Targeted apoptosis of ductular reactive cells reduces hepatic fibrosis in a mouse model of cholestasis.
      Both innate and adaptive immune cells seem to contribute to cholangitis and drive PSC pathogenesis. Indeed, myeloid-derived innate immune cell populations are present in PSC lesions and have been demonstrated to contribute to PSC pathogenesis.
      • Chen Y.Y.
      • Arndtz K.
      • Webb G.
      • Corrigan M.
      • Akiror S.
      • Liaskou E.
      • et al.
      Intrahepatic macrophage populations in the pathophysiology of primary sclerosing cholangitis.
      • Guicciardi M.E.
      • Trussoni C.E.
      • Krishnan A.
      • Bronk S.F.
      • Lorenzo Pisarello M.J.
      • O'Hara S.P.
      • et al.
      Macrophages contribute to the pathogenesis of sclerosing cholangitis in mice.
      • Reuveni D.
      • Gore Y.
      • Leung P.S.C.
      • Lichter Y.
      • Moshkovits I.
      • Kaminitz A.
      • et al.
      The critical role of chemokine (C-C Motif) receptor 2-positive monocytes in autoimmune cholangitis.
      Moreover, PSC pathogenesis has been linked with IL-17-secreting T helper (Th)17 cells that are also present in the hepatic PSC lesions.
      • Katt J.
      • Schwinge D.
      • Schoknecht T.
      • Quaas A.
      • Sobottka I.
      • Burandt E.
      • et al.
      Increased T helper type 17 response to pathogen stimulation in patients with primary sclerosing cholangitis.
      • Oo Y.H.
      • Banz V.
      • Kavanagh D.
      • Liaskou E.
      • Withers D.R.
      • Humphreys E.
      • et al.
      CXCR3-dependent recruitment and CCR6-mediated positioning of Th-17 cells in the inflamed liver.
      • Nakamoto N.
      • Sasaki N.
      • Aoki R.
      • Miyamoto K.
      • Suda W.
      • Teratani T.
      • et al.
      Gut pathobionts underlie intestinal barrier dysfunction and liver T helper 17 cell immune response in primary sclerosing cholangitis.
      Intriguingly, cholangitis seems to be perpetuated by mutual cross-communication between biliary epithelial cells, myeloid-derived cells and Th17 cells.
      • Jeffery H.C.
      • Hunter S.
      • Humphreys E.H.
      • Bhogal R.
      • Wawman R.E.
      • Birtwistle J.
      • et al.
      Bidirectional cross-talk between biliary epithelium and Th17 cells promotes local Th17 expansion and bile duct proliferation in biliary liver diseases.
      ,
      • Kunzmann L.K.
      • Schoknecht T.
      • Poch T.
      • Henze L.
      • Stein S.
      • Kriz M.
      • et al.
      Monocytes as potential mediators of pathogen-induced Th17 differentiation in patients with primary sclerosing cholangitis (PSC).
      Note, however, that the relevant myeloid-derived cell populations are difficult to define, because these cells display a high degree of plasticity, with mostly ambiguous markers, and diverse functions that can either promote inflammation or regeneration.
      • Weston C.J.
      • Zimmermann H.W.
      • Adams D.H.
      The role of myeloid-derived cells in the progression of liver disease.
      • Bassler K.
      • Schulte-Schrepping J.
      • Warnat-Herresthal S.
      • Aschenbrenner A.C.
      • Schultze J.L.
      The myeloid cell compartment-cell by cell.
      • Guilliams M.
      • Ginhoux F.
      • Jakubzick C.
      • Naik S.H.
      • Onai N.
      • Schraml B.U.
      • et al.
      Dendritic cells, monocytes and macrophages: a unified nomenclature based on ontogeny.
      Thus far, macrophages and monocyte-derived cells have attracted the most attention, whereas the role of dendritic cells (DCs) in PSC is only poorly defined.
      • Weston C.J.
      • Zimmermann H.W.
      • Adams D.H.
      The role of myeloid-derived cells in the progression of liver disease.
      However, we reasoned that DCs might be important contributors to PSC pathogenesis, because they are sentinel cells that reside within the portal field, which can readily recruit other inflammatory cells upon activation, and act as potent antigen-presenting cells that can effectively activate the adaptive immune system.
      • Durai V.
      • Murphy K.M.
      Functions of murine dendritic cells.
      Moreover, a recent publication has described type 1 conventional DCs (cDC1s) as drivers of liver pathology in non-alcoholic steatohepatitis,
      • Deczkowska A.
      • David E.
      • Ramadori P.
      • Pfister D.
      • Safran M.
      • Li B.
      • et al.
      XCR1+ type 1 conventional dendritic cells drive liver pathology in non-alcoholic steatohepatitis.
      suggesting that specific DC subsets might also contribute to PSC and other liver pathologies.
      Based on ontogeny, 3 major DC types can be distinguished: the conventional DCs type 1 and type 2 (cDC1s, cDC2s), and plasmacytoid DCs (pDCs).
      • Guilliams M.
      • Ginhoux F.
      • Jakubzick C.
      • Naik S.H.
      • Onai N.
      • Schraml B.U.
      • et al.
      Dendritic cells, monocytes and macrophages: a unified nomenclature based on ontogeny.
      At baseline, DCs promote immune tolerance; however, they become inflammatory upon sensing tissue damage or infection. Functionally, cDC1s are specialised in antigen cross-presentation and activation of CD8 T cells, as well as Th1 cells; whereas cDC2s mainly present exogenous antigens and activate CD4 T-cell responses, in particular Th2 and Th17 responses.
      • Durai V.
      • Murphy K.M.
      Functions of murine dendritic cells.
      The pDCs are major producers of type I interferons.
      • Reizis B.
      Plasmacytoid dendritic cells: development, regulation, and function.
      In inflammatory settings, monocytes can also develop into monocyte-derived DCs (moDCs) that resemble cDC2s, but can be distinguished based on their CD64 expression.
      • Langlet C.
      • Tamoutounour S.
      • Henri S.
      • Luche H.
      • Ardouin L.
      • Grégoire C.
      • et al.
      CD64 expression distinguishes monocyte-derived and conventional dendritic cells and reveals their distinct role during intramuscular immunization.
      However, this concept has recently been questioned by demonstrating that bona fide cDC2s can also acquire CD64 expression,
      • Bosteels C.
      • Neyt K.
      • Vanheerswynghels M.
      • van Helden M.J.
      • Sichien D.
      • Debeuf N.
      • et al.
      Inflammatory Type 2 cDCs acquire features of cDC1s and macrophages to orchestrate immunity to respiratory virus infection.
      raising the possibility that inflammatory myeloid cells that have previously been thought to be moDCs might actually be inflammatory cDC2s.
      • Bosteels C.
      • Neyt K.
      • Vanheerswynghels M.
      • van Helden M.J.
      • Sichien D.
      • Debeuf N.
      • et al.
      Inflammatory Type 2 cDCs acquire features of cDC1s and macrophages to orchestrate immunity to respiratory virus infection.
      Moreover, the transcriptional regulation of cDC2 heterogeneity has recently been characterised, identifying a proinflammatory Tbet- cDC2 subtype that can readily express monocyte markers.
      • Brown C.C.
      • Gudjonson H.
      • Pritykin Y.
      • Deep D.
      • Lavallée V.P.
      • Mendoza A.
      • et al.
      Transcriptional basis of mouse and human dendritic cell heterogeneity.
      Herein, we explore the role of DCs in the pathogenesis of cholangitis and particularly in its initiation, using different mouse models. Mice that are fed with 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC) develop acute cholangiopathy that is characterised by cholangitis, periductal fibrosis and ductular reaction.
      • Fickert P.
      • Stöger U.
      • Fuchsbichler A.
      • Moustafa T.
      • Marschall H.U.
      • Weiglein A.H.
      • et al.
      A new xenobiotic-induced mouse model of sclerosing cholangitis and biliary fibrosis.
      Due to its timed inducibility, this model is particularly well suited to study early cholangiopathogenesis. Mice lacking the Mdr2 gene (also known as Abcb4) also develop chronic cholangiopathy with similarity to PSC.
      • Fickert P.
      • Fuchsbichler A.
      • Wagner M.
      • Zollner G.
      • Kaser A.
      • Tilg H.
      • et al.
      Regurgitation of bile acids from leaky bile ducts causes sclerosing cholangitis in Mdr2 (Abcb4) knockout mice.
      This model is highly relevant for its spontaneous, chronic bile toxicity-related pathogenesis. We characterised the DCs in these 2 models, applying a flow cytometry strategy that is based on a set of lineage-imprinted markers that enable the identification of the major DC subtypes.
      • Guilliams M.
      • Dutertre C.A.
      • Scott C.L.
      • McGovern N.
      • Sichien D.
      • Chakarov S.
      • et al.
      Unsupervised high-dimensional analysis aligns dendritic cells across tissues and species.

      Materials and methods

      Human samples

      Liver biopsies from 10 patients with PSC, 5 with primary biliary cholangitis (PBC) and 5 with autoimmune hepatitis (AIH) were used for histochemistry; as controls, liver specimens from adenoma resection margins of 3 patients and liver specimens from 2 patients undergoing bariatric surgery were used. For immunofluorescence, explanted livers from 7 patients with PSC and end-stage disease and 5 control samples of patients undergoing bariatric surgery were used. Clinical data are shown in Table S1. The use of human liver tissue samples was approved by the responsible ethics committee (Ethikkomission der Landesärztekammer Hamburg; PV3912) and patients gave informed consent.

      Animals

      Inbred C57Bl/6J, B6.Cg-Abcb4tm1Bor/J (Mdr2-/-),
      • Fickert P.
      • Stöger U.
      • Fuchsbichler A.
      • Moustafa T.
      • Marschall H.U.
      • Weiglein A.H.
      • et al.
      A new xenobiotic-induced mouse model of sclerosing cholangitis and biliary fibrosis.
      B6.129S(C)-Batf3tm1Kmm/J (Batf3-/-),
      • Hildner K.
      • Edelson B.T.
      • Purtha W.E.
      • Diamond M.
      • Matsushita H.
      • Kohyama M.
      • et al.
      Batf3 deficiency reveals a critical role for CD8alpha+ dendritic cells in cytotoxic T cell immunity.
      and Cd11c.DOG mice
      • Hochweller K.
      • Striegler J.
      • Hämmerling G.J.
      • Garbi N.
      A novel CD11c.DTR transgenic mouse for depletion of dendritic cells reveals their requirement for homeostatic proliferation of natural killer cells.
      were bred and maintained in the animal care facility of the University Medical Centre Hamburg-Eppendorf under specific pathogen-free conditions. Cholangitis was induced in C57Bl/6J, Batf3-/-, or Cd11c.DOG mice by feeding with a diet containing 0.1% DDC (Sigma-Aldrich) for 36 hours. Seven- to sixteen-week-old mice of both sexes were used in different experiments, but each experiment was performed with 1 sex only and age- and sex-matched controls. Experiments were reproduced at least twice. Alternatively, spontaneous cholangitis was assessed in female Mdr2-/- mice at an age of 5-6 weeks. The effect of cDC1 deficiency was analysed in Batf3-/- mice in comparison to Batf3-proficient littermates. Depletion of DCs in CD11c.DOG mice carrying the transgenic diphtheria toxin receptor (DTR) (+/T) was induced by daily intraperitoneal injection of 24 μg/kg DT, starting 1 day before DDC feeding. Non-specific effects of DT application were controlled for by using non-transgenic littermates (+/+). In all animal experiments, clinical and behavioural humane endpoints were applied to reduce pain and distress. All animal experiments were carried out in accordance with the principles of the Basel Declaration, the European Directive 2010/63/EU and FELASA recommendations, and had been approved by the animal experimentation review board of the State of Hamburg.

      Isolation of hepatic non-parenchymal cells

      Livers were perfused via the portal vein with collagenase D in Gey’s balanced salt solution and further digested for 25 min at 37 °C with collagenase (Nordmark) and DNAse I (Roche), followed by filtration through a 100 μm mesh to obtain a single-cell suspension. After removal of hepatocytes by repeated centrifugation at 40 g, non-parenchymal cells (NPCs) were separated with an 29% Optiprep-gradient (Stemcell Technologies).

      Flow cytometry

      DC subtypes were identified as described
      • Guilliams M.
      • Dutertre C.A.
      • Scott C.L.
      • McGovern N.
      • Sichien D.
      • Chakarov S.
      • et al.
      Unsupervised high-dimensional analysis aligns dendritic cells across tissues and species.
      with some minor modifications. Briefly, immunofluorescence staining of cells was performed with antibodies listed in the CTAT table. Dead cells were stained with Pacific Orange-NHS (ThermoFisher Scientific). For intranuclear staining of transcription factors, the Foxp3 Transcription Factor Staining Kit (eBioscience) was used according to the manufacturer’s protocol. Flow cytometry data was analysed with FACS Diva Software (BD Biosciences).

      Magnetic-activated cell sorting

      Hepatic NPCs were isolated as described above and samples were then cleared with Debris Removal Solution according to the manufacturer's protocol (Miltenyi), followed by blocking with anti-CD16/32 antibody. For negative selection of DCs, NPCs were incubated with lineage-specific antibodies (F4/80-PE; CD3-PE; CD19 PE; NK1.1-PE) and Anti-PE-Beads (Miltenyi) to facilitate the retention of these cell lineages on LD columns according to the manufacturer’s protocol. The cells in the flow-through were incubated with Anti-CD11c-Beads (Miltenyi), and positive selection of CD11c+ cells with MS columns was performed according to the manufacturer’s protocol (Miltenyi).

      T-cell stimulation

      Unprimed CD4+ T cells (5 × 105 per well) from spleen were cultured with 2 × 104 liver DCs from C57BL/6J mice fed either with standard diet or with DDC diet, and stimulated under non-polarising conditions with soluble anti-CD3 (145-2C11; 2 μg/ml; BD Biosciences) in 200 μl of IMDM, supplemented with 5% FCS and 1% P/S. After 2 days, 10 ng/ml of IL-2 was added, and Th17 differentiation was assessed on day 5 by flow cytometry, following stimulation with 50 ng/ml of PMA and 5 μg of ionomycin and 5 μg/ml of Golgi Stop for 4 hours at 37 °C.

      CITE-Seq

      For cellular indexing of transcriptomes and epitopes by sequencing (CITE-Seq), NPCs were isolated from 4 DDC-fed and 7 control mice as described above and pooled, followed by negative selection with magnetic-activated cell sorting (Miltenyi), using lineage-specific antibodies (F4/80-PE; CD3-PE; CD19 PE; Ly6G.PE) as described above. Thereafter, TotalSeq antibodies (BioLegend) to CD11b, CD11c, CD103, CD19, CD64, B220, CD172a, XCR1, F4/80, I-A/I-E (MHC-II), NK1.1, and Siglec H, together with fluorochrome-conjugated antibodies to CD45, CD11c, I-A/I-E, F4/80, CD3, CD19, and Ly6G were added. DCs were then sorted into RPMI + 25% FCS, using a FACS Aria III sorter (BD Biosciences). Thirty thousand purified DCs were then further processed with the Chromium Single Cell 3ʹ Reagent Kit v3 from 10xGenomics according to the manufacturer. Quantification and quality control was done with the Qubit dsDNA HS Assay Kit and Agilent High Sensitivity DNA Assay according to the manufacturers’ protocols. The antibody-derived tags (ADT) library preparation was done according to the BioLegend protocol. After dilution of the ADT libraries, all libraries had a concentration of 3-4 nM and the corresponding sample and ADT libraries were pooled in a 4:1 ratio. The libraries were sequenced on a NovaSeq 6000 system; 600 M reads/sample on an S4 flow cell (Illumina, USA).

      Single-cell data analysis

      The raw sequencing data were aligned to the mouse reference genome (mm10, Ensemble 84) using cellranger (version 3.1.0, 10X Genomic, USA) in Feature Barcode mode. All further analysis steps were performed in R (version 3.6.2; The R Foundation for Statistical Computing, Vienna, Austria) with the Seurat package (version 3.1).
      • Stuart T.
      • Butler A.
      • Hoffman P.
      • Hafemeister C.
      • Papalexi E.
      • Mauck 3rd, W.M.
      • et al.
      Comprehensive integration of single-cell data.
      First, we filtered out low quality cells by keeping cells with mitochondrial gene expression percentage <5%, number of expressed genes between 750 and 3,000 and a minimum number of 1,000 unique molecular identifiers. Overall, 3,615 standard diet cells and 3,529 DDC-fed cells remained. To normalise the RNA expression for downstream analysis, we used Seurat’s SCTransform function and included regression variables for cell cycling classes (vars.to.regress = c(“S.Score”, “G2M.Score”)), based on the cell cycling genes provided in Seurat’s cc.genes list after mapping them to unique homologous mouse genes. For visualisation of RNA expression, we used Seurat’s LogNormalize method and ADT expression was normalised with the centred log ratio transformation.
      To combine the cells of both samples we used Seurat’s anchor-based integration method. After data integration we used the first 20 principal components for the uniform manifold approximation and projection (UMAP) embedding and in the graph-based clustering with a resolution of 0.05. Cluster markers were calculated with the FindAllMarkers function using the logistic regression method with the sample names as latent variables on the normalised RNA expression. To find differentially expressed genes within each cluster between both samples we used the Wilcoxon rank sum test. In both cases we used a false discovery rate value <0.1 as a cut-off for significance. To create the ADT gating-based clustering we exported the normalised ADT expression to Flow Cytometry Standard (fcs) files with the flowCore package,
      • Hahne F.
      • LeMeur N.
      • Brinkman R.R.
      • Ellis B.
      • Haaland P.
      • Sarkar D.
      • et al.
      flowCore: a Bioconductor package for high throughput flow cytometry.
      performed classical marker gating in FlowJo and imported the resulting labels into the Seurat object. Differential abundance of cell types between feeding conditions was assessed with a Chi-squared test.
      To perform the reference mapping onto the Brown et al. data we downloaded the mouse spleen cluster annotation and raw count tables from GEO accessions GSE137710, GSM4085510 and GSM4085511. After performing the same pre-processing steps as previously described, we used this dataset as the reference data and our pDCs, cDC1s and cDC2s as the query data set when following Seurat’s mapping and annotating query datasets vignette (https://satijalab.org/seurat/articles/integration_mapping.html).
      Overrepresentation analysis of differentially expressed genes was performed with WebGestalt using the protein-coding genome as background and a false discovery rate <0.1 as a significance cut-off.
      • Liao Y.
      • Wang J.
      • Jaehnig E.J.
      • Shi Z.
      • Zhang B.
      WebGestalt 2019: gene set analysis toolkit with revamped UIs and APIs.

      Real-time quantitative PCR

      Total RNA was extracted from liver tissue or magnetic-activated cell sorting-isolated liver DCs using the NucleoSpin RNA Kit (Macherey-Nagel) according to the manufacturer’s protocol. In case of less than 5 × 105 cells, the RNeasy Plus Micro Kit (Qiagen) was used according to the manufacturer’s protocol. cDNA was reverse-transcribed from total RNA (High Capacity cDNA Reverse Transcription Kit; ThermoFisher Scientific). The expression of Ccl2 (Mm00441242_m1), Cxcl2 (Mm00436450_m1), Il6 (Mm00446190_m1), Il1b (Mm00434228_m1), Tnf (Mm00443258_m1) and S100a4 (Mm00803372_g1) was measured using the Taqman Universal PCR master mix and Taqman gene expression kits (ThermoFisher Scientific). Target gene expression was normalised to Hprt (Mm03024075_m1).

      Histology and serum liver enzymes

      H&E (Roth) staining were performed on formalin-fixed liver sections. In addition, cryo-fixated murine liver sections were stained with antibodies to CD11c (BioLegend) and MHC-II (ThermoFisher Scientific); nuclei were stained with Hoechst 33342. Mouse serum liver enzymes (alanine aminotransferase [ALT] and aspartate aminotransferase) were measured at the Institute of Experimental Immunology and Hepatology, University Medical Centre Hamburg-Eppendorf, using a COBAS Mira System (Roche Diagnostic). To stain cDC2s in human liver samples, sections were stained with antibodies to HLA-DR and CD1c (BioLegend).

      Statistics

      Differences between 2 experimental groups were assessed for statistical significance by the Mann-Whitney test. Differences between more than 2 groups were assessed using the Kruskal-Wallis test and Dunn’s post hoc test (p <0.05 was considered as significant).

      Results

      To explore the possible role of DCs in PSC, we stained liver biopsies from 10 patients with PSC, 5 with PBC and 5 with AIH, as well as sections from 6 control livers, for CD11c and determined the number of CD11c+ cells per portal field. We found that DCs were significantly expanded in the portal fields of PSC livers, compared to control tissue; representative sections and the mean numbers per portal field are shown in Fig. 1A. By tendency, PBC and AIH specimens also had elevated DC numbers, compared to controls, but the elevation was not significant, due to high variability. Next, using immunofluorescence for CD11c and MHC-II, we assessed DC expansion in liver cryosections from different mouse cholangitis models (Fig. 1B). Whereas CD11c+/MHC-II+ DCs were sparse in healthy control livers, DCs were clearly expanded in mice that were fed for 36 hours with DDC (DDC cholangitis), as well as in 5-week-old Mdr2-/- mice (Mdr2-/- cholangitis). Thus, DC expansion was a histological feature shared between human PSC and mouse cholangitis models.
      Figure thumbnail gr1
      Fig. 1Increase of portal hepatic dendritic cells in patients with PSC and murine cholangitis.
      (A) Biopsies of 10 PSC livers, 5 PBC livers, 5 AIH livers and 6 control livers from bariatric surgery patients (control) were stained for CD11c; representative sections are shown. The numbers of CD11c+ dendritic cells were counted in 2 to 10 portal fields of each individual, depending on the size of the liver specimens. The graph shows the mean number of dendritic cells per portal field in each individual, represented as a dot, indicating increased dendritic cell numbers in PSC. (B) Mouse livers of untreated C57BL/6J mice (control), or C57BL/6J mice with cholangitis induced by feeding with 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC cholangitis), or C57BL/6J mice with cholangitis due to Mdr2 gene deficiency (Mdr2-/- cholangitis), stained for CD11c (green) and MHC-II (red), indicating increased dendritic cell numbers in murine cholangitis. ∗∗∗p <0.001 (Kruskal-Wallis and Dunn’s post hoc test). AIH, autoimmune hepatitis; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; PBC, primary biliary cholangitis; PSC, primary sclerosing.
      To study this DC expansion in the different PSC mouse models, we characterised the major DC populations in the liver by flow cytometry with a method that has been developed based on lineage-specific markers;
      • Guilliams M.
      • Dutertre C.A.
      • Scott C.L.
      • McGovern N.
      • Sichien D.
      • Chakarov S.
      • et al.
      Unsupervised high-dimensional analysis aligns dendritic cells across tissues and species.
      the gating strategy is shown in Fig. S1. Briefly, after exclusion of dead cells and doublets, we identified DCs as CD45+ CD11c+ MHC-II+ and lineage- cells, the excluded lineages being F4/80+ macrophages, CD3+ T cells, CD19+ B cells, and Nk1.1+ natural killer cells. These liver DCs were further subdivided into CD317+ B220+ SiglecH+ pDCs, XCR1+ cDC1s, some of which also express CD103, and into IRF8- CD172a+ cDC2s, some of which co-express CD11b or CD64.
      This gating strategy was first applied to the DDC model, as early as 36 hours after replacement of the standard diet with DDC chow. At this early time-point, peribiliary inflammation was evident on histology (Fig. S2A), and serum ALT levels were already significantly elevated (Fig. S2B). As shown in Fig. 2A, flow cytometry confirmed a significant increase in liver DCs in DDC-fed mice, both by proportion and absolute numbers. The increase of DCs related entirely to the significant expansion of cDC2s, both CD11b+ and CD64+, whereas the proportion of intrahepatic pDCs and cDC1s was reduced (Fig. 2A). To confirm our findings in another cholangitis model, we analysed 5-6-week-old Mdr2-/- mice, which showed histological evidence of cholangitis (Fig. S3A) and serological ALT elevation (Fig. S3B). Flow cytometry confirmed a significant expansion of hepatic DCs, which was again caused by significant proportional and absolute expansion of both CD11b+ and CD64+ cDC2s; in contrast, the proportion of hepatic pDCs and cDC1s was again reduced (Fig. 2B). To further confirm that cDC2s were selectively expanded in the livers in both cholangitis models, we determined the percentage of cDC2s among CD45+ cells and found that cDC2s increased from below 2% in control mice to about 4% both in DDC-fed mice and Mdr2-/- mice (Fig. S4).
      Figure thumbnail gr2
      Fig. 2Hepatic expansion of cDC2s in cholangitis.
      (A) and (B) Using the gating strategy of Fig. S1, hepatic CD45+ CD11c+ Lineage- MHC-II+ dendritic cells (black headings) and their subtypes CD317+ B220+ SiglecH+ pDCs (green headings), XCR1+ cDC1s, some of which express CD103 (turquoise headings), and IRF8- CD172a+ cDC2s, some of which express either CD11b or CD64 (red headings) were characterised by flow cytometry. (A) Shown is the comparison of liver dendritic cells of mice fed with standard chow (w/o DDC) or mice fed for 36 h with 3,5-diethoxycarbonyl-1,4-dihydrocollidine (36 h DDC). (B) Shown is the comparison of liver dendritic cells of WT or Mdr2-/- mice. (C) Human liver samples were stained for the human cDC2 marker CD1c and MHC-II, and the number of cDC2s per portal field was counted in liver sections from patients with PSC or control livers sections from patients undergoing bariatric surgery (Control). Also shown are representative stainings of a section taken from a bariatric surgery patient with evident steatohepatitis (Control) and of a section from a patient with PSC. (D) Liver DCs isolated from DDC-fed or standard diet-fed mice were used to provide co-stimulation to unprimed CD4 T cells stimulated in vitro with soluble anti-CD3 antibody, and subsequently, the frequency of induced Th17 cells was compared. (E) The hepatic frequency of Th17 cells following DDC diet was determined ex vivo without further re-stimulation ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001 (Mann-Whitney test for 2 group comparisons; Kruskal-Wallis and Dunn’s post hoc test for 3 group comparisons; n = 5–10). cDC, conventional dendritic cell; DC, dendritic cell; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; pDC, plasmacytoid dendritic cell; PSC, primary sclerosing cholangitis; WT, wild-type.
      To learn whether cDC2 expansion is also a feature of human PSC, we stained human liver sections derived from explanted PSC livers or control livers of patients undergoing bariatric surgery, either with evident hepatic steatosis or with healthy liver appearance, for the human cDC2 marker CD1c together with HLA-DR. As shown in Fig. 2C, PSC livers had significantly more cDC2s than control livers; representative staining of cDC2s is also depicted in Fig. 2C. Taken together, hepatic expansion of cDC2s is a feature of human PSC and murine cholangitis. As cDC2s are particularly effective in activating Th17 responses,
      • Durai V.
      • Murphy K.M.
      Functions of murine dendritic cells.
      which have been linked to PSC pathogenesis,
      • Katt J.
      • Schwinge D.
      • Schoknecht T.
      • Quaas A.
      • Sobottka I.
      • Burandt E.
      • et al.
      Increased T helper type 17 response to pathogen stimulation in patients with primary sclerosing cholangitis.
      • Oo Y.H.
      • Banz V.
      • Kavanagh D.
      • Liaskou E.
      • Withers D.R.
      • Humphreys E.
      • et al.
      CXCR3-dependent recruitment and CCR6-mediated positioning of Th-17 cells in the inflamed liver.
      • Nakamoto N.
      • Sasaki N.
      • Aoki R.
      • Miyamoto K.
      • Suda W.
      • Teratani T.
      • et al.
      Gut pathobionts underlie intestinal barrier dysfunction and liver T helper 17 cell immune response in primary sclerosing cholangitis.
      we explored whether cDC2 expansion was of functional relevance for the Th17 response in cholangitis. To that end, we used liver DCs either from DDC-fed or from control mice to stimulate unprimed CD4 T cells in vitro; T-cell activation was induced with soluble anti-CD3 antibody, so that co-stimulation depended on the DCs. As shown in Fig. 2D, priming on liver DCs from DDC-fed mice induced a significantly higher frequency of IL-17-producing CD4 T cells than priming on liver DCs derived from mice on standard diet, indicating that cDC2-enriched liver DCs induced increased Th17 differentiation. To confirm this notion, we next directly analysed ex vivo the hepatic frequency of IL-17-producing CD4 T cells in response to DDC feeding; note that IL17 production was determined without re-stimulation, thus reflecting IL-17 production as it occurred in vivo. Following a change to DDC feeding, the frequency of liver Th17 cells gradually increased significantly (Fig. 2E), suggesting that the expanded cDC2 population in the liver of DDC-fed mice served to maintain a hepatic Th17 response.
      The question then was whether the observed cDC2 expansion or the proportionate cDC1 contraction contributed to the progression of cholangitis. Note that a protective role of cDC1s against progression of steatosis towards steatohepatitis has previously been suggested, based on a more aggressive disease course in mice deficient in cDC1s.
      • Heier E.C.
      • Meier A.
      • Julich-Haertel H.
      • Djudjaj S.
      • Rau M.
      • Tschernig T.
      • et al.
      Murine CD103+ dendritic cells protect against steatosis progression towards steatohepatitis.
      Therefore, we first explored whether the observed hepatic contraction of the cDC1 population might contribute to cholangitis. To that end, we applied the DDC model to Batf3-/- mice, which are deficient in cDC1s.
      • Hildner K.
      • Edelson B.T.
      • Purtha W.E.
      • Diamond M.
      • Matsushita H.
      • Kohyama M.
      • et al.
      Batf3 deficiency reveals a critical role for CD8alpha+ dendritic cells in cytotoxic T cell immunity.
      We found that both Batf3-/- mice and their Batf3-proficient littermates exhibited a similar degree of cholangitis upon histology (Fig. 3A), and a similar elevation of serum ALT (Fig. 3B). Except the expected absence of cDC1s in the livers of Batf3-/- mice, there was no difference in the overall numbers of liver DCs, pDCs or cDC2s (Fig. 3C). Moreover, we did not find any significant difference in the expression of selected inflammatory mediators, such as the chemokines CCL2 and CXCL2, or the cytokines IL-1b, TNF or IL6 (Fig. 3D). Thus, the complete absence of cDC1s did not seem to have an aggravating effect on DDC-induced cholangitis, indicating that cDC1s are probably not protective in this condition. Hence, we focused on the functional role of the cDC2 expansion observed in murine cholangitis and human PSC.
      Figure thumbnail gr3
      Fig. 3Absence of cDC1s does not influence DDC-induced cholangitis.
      Batf3-deficient mice lacking cDC1s (Batf3-/-) and their Batf3-proficient littermates featuring cDC1s (Batf3+/-) were compared for cholangitis development 36 hours after DDC feeding. (A) Representative histological presentation of Batf3+/- and Batf3-/- mice, showing a similar degree of tissue damage. (B) Similar elevation of serum ALT levels in response to DDC in Batf3+/- and Batf3-/- mice. (C) Similar numbers of dendritic cells in Batf3+/- and Batf3-/- mice, except for the expected lack of cDC1s in Batf3-/- mice. (D) Similar levels of selected inflammatory mediators in Batf3+/- and Batf3-/- livers. ∗∗∗∗p <0.0001 (Mann-Whitney test for 2 group comparisons; n = 5–11). ALT, alanine aminotransferase; cDC, conventional dendritic cell; DC, dendritic cell; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; pDC, plasmacytoid dendritic cell.
      For want of a reliable method to selectively deplete cDC2s, we resorted to a DTR-transgenic mouse model, which was originally developed to deplete all DCs, i.e. the CD11c.DOG mouse.
      • Hochweller K.
      • Striegler J.
      • Hämmerling G.J.
      • Garbi N.
      A novel CD11c.DTR transgenic mouse for depletion of dendritic cells reveals their requirement for homeostatic proliferation of natural killer cells.
      Heterozygous CD11c.DOG+/T mice, which express the human DTR under control of the CD11c-promoter, or their CD11c.DOG+/+ littermates, which do not feature the DTR, were treated with a total of 3 DT injections and fed with DDC for 36 h. As shown in Fig. 4A, there was no apparent histological difference between DTR-expressing mice or their non-expressing littermates at 36 h after replacement of the standard diet with DDC chow. However, we found significantly elevated serum ALT in the CD11c.DOG+/T mice (Fig. 4B). The overall frequency of hepatic DCs was significantly reduced, but not as strongly as expected (Fig. 4C). Whereas pDCs and cDC1s were effectively depleted, cDC2s surprisingly remained largely unaffected by DT-induced depletion. Indeed, cDC2s in CD11c.DOG+/T mice were not ablated and accounted for the large majority of the remaining DCs in CD11c.DOG+/+ littermates (Fig. 4C). Thus, the hepatic cDC2 population either seemed to be resistant to DT-induced depletion or rapidly replenished by moDCs. Note that more than 80% of the hepatic DCs expressed the monocyte marker CD64 (Fig. 4C), indeed suggesting that these DT-resistant cells might be moDCs. However, most of these cells co-expressed the DC marker CD26 (Fig. 4D) and not the additional monocyte marker CD88 (Fig. 4E), indicating that the DT-resistant cDC2s were rather bona fide DCs, and not derived from monocytes. Moreover, we observed a downregulation of CD11c molecules on non-depleted cDC2s in CD11c.DOG mice (Fig. S5), offering a possible explanation for the apparent DT-resistance of the cDC2s, as the DTR in these mice is also under CD11c promoter control. Either way, the CD11c.DOG mouse model unexpectedly allowed us to specifically explore the role of cDC2s in murine cholangitis. Strikingly, expansion of these cDC2s was associated with significant differences in the hepatic expression of selected inflammatory mediators, such as the chemokines CCL2 and CXCL2, or the cytokines IL-1b, TNF or IL6 (Fig. 4F). However, it should also be noted that DDC-fed CD11c.DOG mice had increased hepatic numbers of monocytes, macrophages and neutrophils following DC depletion (Fig. S5). Nonetheless, these findings altogether indicated that cDC2s, unlike cDC1s, promoted DDC-induced cholangitis.
      Figure thumbnail gr4
      Fig. 4DC depletion sparing cDC2s aggravates DDC-induced cholangitis.
      CD11c.DOG mice conditionally expressing the human diphtheria receptor in DCs (CD11c.DOG+/T) and their littermates lacking expression of the diphtheria receptor (CD11c.DOG+/+) were treated with diphtheria toxin and compared for cholangitis development 36 hours after DDC feeding. (A) Representative histological presentation of CD11c.DOG+/+ and CD11c.DOG+/T mice, showing similar degree of tissue damage. (B) Increased elevation of serum ALT levels in response to DDC in CD11c.DOG+/T mice. (C) Effective diphtheria toxin-mediated depletion of pDCs and cDC1s, but unexpected increase of cDC2s in CD11c.DOG+/T mice. (D) The majority of the cDC2s escaping depletion express the DC marker CD26; (E) whereas only a small fraction of these cells express the monocyte marker CD88. (F) Increased levels of selected inflammatory mediators in CD11c.DOG+/T livers. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001 (Mann-Whitney test for 2 group comparisons; n = 5–20). ALT, alanine aminotransferase; cDC, conventional dendritic cell; DC, dendritic cell; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; pDC, plasmacytoid dendritic cell.
      To better understand the functions of the various DC subsets in cholangitis, we next performed a CITE-Seq analysis of purified DCs from DDC-fed mice and from mice fed with standard diet. After qualitative filtering and normalisation, the transcriptional data of 3,529 DCs from DDC-fed mice and 3,615 DCs from control mice were used for graph-based clustering, and UMAP was used for visualisation. We then used the ADT to assign the RNA expression clusters with a colour code representing various DC subsets (Fig. 5A). We could thus distinguish 4 major separate clusters, which represented pDCs (green), cDC1s (turquoise), cDC2s (red) and a minor population of cells, marked in purple, that seemed to represent CD11c- and MHC-II-expressing natural killer cells. As this cluster did not seem to represent DCs, it was henceforth not further analysed. The distinction between the 2 conditions, with or without DDC feeding, is shown in Fig. 5B. There was a high congruence between RNA-based clusters and ADT-based gating, as illustrated by the Venn diagrams in Fig. S6A-C.
      Figure thumbnail gr5
      Fig. 5CITE-Seq analysis of purified DCs confirms DDC-induced cDC2 expansion.
      The single-cell transcriptional data of 3,529 DCs from DDC-fed mice and 3,615 DCs from control mice was used for graph-based clustering, and UMAP was used for visualisation. (A) Antibody-derived tags were used to assign the RNA expression clusters with a colour code representing various DC subsets. (B) All subclusters were found in both conditions, i.e. with and without DDC. (C) A cell density plot revealed an increase in cDC2s after DDC feeding, which was confirmed by quantification of the cell count proportion (D). (E) Correct assignment of RNA expression clusters to the 3 major DC subtypes was confirmed by expression of typical cluster marker genes corresponding to the subsets, and (F) an extended set of cluster marker genes. cDC, conventional dendritic cell; CITE-Seq, cellular indexing of transcriptomes and epitopes by sequencing; DC, dendritic cell; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; pDC, plasmacytoid dendritic cell; UMAP, uniform manifold approximation and projection.
      Of note, our previous observation that pDCs and cDC1s were proportionally contracted and cDC2 expanded in the DC population of DDC-fed mice was confirmed by UMAP density plot (Fig. 5C) and quantification of the cell count proportion (Fig. 5D). The correct assignment of the RNA expression clusters to the 3 major DC subtypes was confirmed by expression of typical cluster marker genes corresponding to the subsets (Fig. 5E) and an extended set of cluster marker genes (Fig. 5F). Note, however, that the cDC2 cluster also contained cells that expressed monocyte marker genes, such as Fcgr1 (CD64) and C5ar1 (CD88) (Fig. 5E). Nonetheless, the transcriptional profiles of these cDC2s were highly similar, irrespective of whether they expressed CD64 or CD88 or not. This transcriptional similarity indicated that this cluster represented bona fide cDC2s rather than a mixed cluster of cDC2s and monocyte-derived cells.
      In accordance with our previous observation that cDC2s played a major role in cholangitis development, cDC2s showed the highest number of differentially expressed genes (Fig. S7A-C), as summarised in the Venn diagram in Fig. 6A. The differentially expressed genes in the 3 main clusters are indicated in the heat maps shown in Fig. S8-10. Using WebGestalt (WEB-based Gene SeT AnaLysis Toolkit), we performed an overrepresentation analysis of the differentially expressed genes. The most prominent pathways altogether were related to regeneration, response to wounding and regulation of inflammation (Fig. 6B).
      Figure thumbnail gr6
      Fig. 6CITE-Seq analysis reveals inflammatory activation of the Tbet- cDC2B subcluster.
      (A) A Venn diagram showing the number of DDC-induced differentially expressed genes in the major DC subtypes. (B) Overrepresentation analysis of the DDC-induced differentially expressed genes indicated the most prominent pathways within the cDC2 cluster to be related with regeneration, wounding response and inflammation. (C) The DC subclusters of a published dataset by Brown et al. were visualised and used as a reference in Seurat’s reference mapping and UMAP projection approach to classify our DCs (D). (E) This analysis revealed that the large majority of the cDC2s in our dataset corresponded to proinflammatory Tbet- cDC2Bs, as defined by Brown et al. (F) Heatmap of differentially expressed genes in response to DDC feeding of the Tbet- cDC2B subset with highlighted genes involved in phagocytosis, antigen presentation, pattern recognition and inflammation. (G) Overrepresentation analysis of the differentially expressed genes in the Tbet- cDC2B subset indicated their activation and inflammatory maturation. (H) Confirmation of DDC-induced expression of selected inflammatory genes in DCs isolated from independent liver samples. ∗p <0.05, ∗∗p <0.01 (Mann-Whitney test for 2 group comparisons; n = 6). cDC, conventional dendritic cell; CITE-Seq, cellular indexing of transcriptomes and epitopes by sequencing; DC, dendritic cell; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; pDC, plasmacytoid dendritic cell; UMAP, uniform manifold approximation and projection.
      To better understand the role of the expanded cDC2s and their possible relation to moDCs, we took advantage of a previously published dataset by Brown et al., describing 2 major cDC2 subsets, Tbet+ anti-inflammatory and regenerative cDC2s, and Tbet- inflammatory cDC2s.
      • Brown C.C.
      • Gudjonson H.
      • Pritykin Y.
      • Deep D.
      • Lavallée V.P.
      • Mendoza A.
      • et al.
      Transcriptional basis of mouse and human dendritic cell heterogeneity.
      Based on the data provided by Brown et al., we visualised their DC subclustering (Fig. 6C) and then used this data as a reference to classify our DCs (Fig. 6D). This analysis revealed that the large majority of the cDC2s in our dataset corresponded to proinflammatory Tbet- cDC2Bs, as defined by Brown et al. (Fig. 6E). This finding added further evidence to the notion that the expanded cDC2s in the livers of DDC-fed mice are bona fide DCs and not monocyte-derived cells. As the dataset from Brown et al. suggested that these Tbet- cDC2Bs are proinflammatory, we wanted to confirm that the hepatic Tbet- cDC2s induced by DDC were likewise proinflammatory. To that end, we separately analysed the differential gene expression of this subpopulation upon 36 h DDC feeding compared to standard diet. As indicated by the heatmap in Fig. 6F, this Tbet- cDC2 subcluster was particularly responsive to DDC-induced biliary damage, resulting in differential expression of many genes involved in phagocytosis, antigen presentation, pattern recognition and inflammation, as indicated. WebGestalt overrepresentation analysis (Fig. 6G) confirmed the activation of these pathways, indicating DC activation and inflammatory maturation. Among the differentially regulated genes in Tbet- cDC2s were several genes that indicated inflammatory activity and maturation (Cxcl2, Aif, Ccl2, Il1b, Spp1, Sat1, S100a4, Lst1, Rgs1, Plin2, Eno1, Pkm, Calm1, Lgals3, Ftl1, Fth1, CD300c, Mapkapk2), inflammasome activation (Cxcl2, Ctsb, Lgals3, Clec4n), pattern recognition and phagocytosis (C1qb, C1qc, Clec4e, Clec4n, Clec4d, Clec4a1, CD14, Fcgr3, CD68, C3, Fcgr1, CD300lb) and increased antigen processing and presentation (Ctsd, Ctsb, Ctss, Ctsz, Litaf, Gabarap, Plin2, Capza2). To confirm these findings of the CITE-Seq experiments in independent liver samples, we isolated DCs from unrelated samples of DDC-fed or control mice, allowing us to test the expression of a few selected inflammatory candidate genes by quantitative RT-PCR. Indeed, we found that the cDC2 cluster marker S100a4 and the inflammatory mediators Cxcl2, Ccl2, Il1b were significantly upregulated after DDC feeding (Fig. 6H), confirming the validity of the RNA expression data generated by CITE-Seq. Taken together, our findings demonstrated that murine cholangitis and human PSC are characterised by intrahepatic expansion and inflammatory maturation of a Tbet- cDC2 subpopulation, which might thus be a potential treatment target in PSC.

      Discussion

      Biliary inflammation in PSC leads to aberrant regeneration characterised by periductular fibrosis and cholangiocarcinogenesis.
      • Dyson J.K.
      • Beuers U.
      • Jones D.E.J.
      • Lohse A.W.
      • Hudson M.
      Primary sclerosing cholangitis.
      ,
      • Lazaridis K.N.
      • LaRusso N.F.
      Primary sclerosing cholangitis.
      The causes for persistent cholangitis in PSC are not clear, but it seems to be fuelled, at least in part, by cholangiocyte damage and a Th17 immune response to the biliary microbiome.
      • Dyson J.K.
      • Beuers U.
      • Jones D.E.J.
      • Lohse A.W.
      • Hudson M.
      Primary sclerosing cholangitis.
      ,
      • Lazaridis K.N.
      • LaRusso N.F.
      Primary sclerosing cholangitis.
      Herein, we tested the hypothesis that DCs, which are sentinel cells residing in the hepatic portal tracts, contribute to biliary inflammation in PSC, and initiate and organise the immune response to biliary injury. Indeed, we found that DCs are expanded in the portal fields of patients with PSC and of mice in models of acute DDC-induced cholangitis and persistent cholangitis in Mdr2-/--mice (Fig. 1). Flow cytometry revealed that DC expansion in both cholangitis models was entirely owed to a selective intrahepatic expansion of cDC2s, whereas cDC1s and pDCs were not expanded (Fig. 2). Importantly, immunofluorescence staining of human end-stage PSC liver sections confirmed that human PSC was likewise characterised by cDC2 expansion in portal tracts (Fig. 2). Collectively, these findings indicated that intrahepatic cDC2 expansion is a feature of cholangitis in mice and humans. Of note, cDC2 expansion in cholangitis could be linked to induction and maintenance of a Th17 response in cholangitis (Fig. 2).
      Genetic depletion of cDC1s in Batf3-/- mice confirmed that cDC1s, which have been reported to be protective in steatohepatitis,
      • Heier E.C.
      • Meier A.
      • Julich-Haertel H.
      • Djudjaj S.
      • Rau M.
      • Tschernig T.
      • et al.
      Murine CD103+ dendritic cells protect against steatosis progression towards steatohepatitis.
      did not seem to influence cholangitis (Fig. 3). In contrast, DT-induced DC depletion sparing cDC2s in CD11c.DOG mice indicated that cDC2s seemed to promote biliary inflammation (Fig. 4). This resistance of cDC2s to depletion was unexpected, but, as the majority of the remaining cDC2s expressed CD64, a marker expressed by monocyte-derived cells,
      • Langlet C.
      • Tamoutounour S.
      • Henri S.
      • Luche H.
      • Ardouin L.
      • Grégoire C.
      • et al.
      CD64 expression distinguishes monocyte-derived and conventional dendritic cells and reveals their distinct role during intramuscular immunization.
      a possible explanation was that these cells were derived from infiltrating monocytes. However, the majority of the remaining cDC2 population did not express the monocyte marker CD88, but instead the DC marker CD26.
      • Bosteels C.
      • Neyt K.
      • Vanheerswynghels M.
      • van Helden M.J.
      • Sichien D.
      • Debeuf N.
      • et al.
      Inflammatory Type 2 cDCs acquire features of cDC1s and macrophages to orchestrate immunity to respiratory virus infection.
      Thus, it was likely that the intrahepatic cDC2 population rather represented bona fide DCs that had acquired monocyte features and inflammatory activity, recently described as ‘inflammatory cDC2’.
      • Bosteels C.
      • Neyt K.
      • Vanheerswynghels M.
      • van Helden M.J.
      • Sichien D.
      • Debeuf N.
      • et al.
      Inflammatory Type 2 cDCs acquire features of cDC1s and macrophages to orchestrate immunity to respiratory virus infection.
      CITE-seq analysis of DDC-fed liver DCs confirmed that intrahepatic cDC2s were expanded and acquired monocyte features in response to DDC (Fig. 5). Further analysis revealed that most of the cDC2s, which expanded in the liver, featured very similar gene expression profiles as the recently described proinflammatory Tbet- cDC2B subpopulation (Fig. 6).
      • Brown C.C.
      • Gudjonson H.
      • Pritykin Y.
      • Deep D.
      • Lavallée V.P.
      • Mendoza A.
      • et al.
      Transcriptional basis of mouse and human dendritic cell heterogeneity.
      We detected profound transcriptional changes of this Tbet- cDC2B subpopulation in response to DDC-induced biliary injury, indicating their activation and inflammatory maturation (Fig. 6).
      Among the upregulated inflammatory genes in cDC2Bs were the chemokine genes Cxcl2 and Ccl2. Ccl2 is a key chemokine inducing tissue infiltration of monocytes and macrophages.
      • Deshmane S.L.
      • Kremlev S.
      • Amini S.
      • Sawaya B.E.
      Monocyte chemoattractant protein-1 (MCP-1): an overview.
      Cxcl2 is a chemokine that facilitates tissue infiltration by neutrophils.
      • De Filippo K.
      • Dudeck A.
      • Hasenberg M.
      • Nye E.
      • van Rooijen N.
      • Hartmann K.
      • et al.
      Mast cell and macrophage chemokines CXCL1/CXCL2 control the early stage of neutrophil recruitment during tissue inflammation.
      Thus, inflammatory cDC2s contributed to biliary inflammation and recruitment of innate immune cells. Moreover, Cxcl2 also promotes the activation of the NLRP3 inflammasome.
      • Boro M.
      • Balaji K.N.
      CXCL1 and CXCL2 regulate NLRP3 inflammasome activation via G-protein-coupled receptor CXCR2.
      Inflammasome activation is, in addition, also promoted by the differentially induced molecules Ctsb,
      • Amaral E.P.
      • Riteau N.
      • Moayeri M.
      • Maier N.
      • Mayer-Barber K.D.
      • Pereira R.M.
      • et al.
      Lysosomal cathepsin release is required for NLRP3-Inflammasome activation by mycobacterium tuberculosis in infected macrophages.
      ,
      • Bruchard M.
      • Mignot G.
      • Derangère V.
      • Chalmin F.
      • Chevriaux A.
      • Végran F.
      • et al.
      Chemotherapy-triggered cathepsin B release in myeloid-derived suppressor cells activates the Nlrp3 inflammasome and promotes tumor growth.
      Lgals3,
      • Arsenijevic A.
      • Milovanovic J.
      • Stojanovic B.
      • Djordjevic D.
      • Stanojevic I.
      • Jankovic N.
      • et al.
      Gal-3 deficiency suppresses novosphyngobium aromaticivorans inflammasome activation and IL-17 driven autoimmune cholangitis in mice.
      ,
      • Tian J.
      • Yang G.
      • Chen H.Y.
      • Hsu D.K.
      • Tomilov A.
      • Olson K.A.
      • et al.
      Galectin-3 regulates inflammasome activation in cholestatic liver injury.
      and Clec4n.
      • Chang T.H.
      • Huang J.H.
      • Lin H.C.
      • Chen W.Y.
      • Lee Y.H.
      • Hsu L.C.
      • et al.
      Dectin-2 is a primary receptor for NLRP3 inflammasome activation in dendritic cell response to Histoplasma capsulatum.
      These findings are in accordance with previous descriptions of inflammasome activation in PSC models.
      • Liao L.
      • Schneider K.M.
      • Galvez E.J.C.
      • Frissen M.
      • Marschall H.U.
      • Su H.
      • et al.
      Intestinal dysbiosis augments liver disease progression via NLRP3 in a murine model of primary sclerosing cholangitis.
      ,
      • Maroni L.
      • Agostinelli L.
      • Saccomanno S.
      • Pinto C.
      • Giordano D.M.
      • Rychlicki C.
      • et al.
      Nlrp3 activation induces Il-18 synthesis and affects the epithelial barrier function in reactive cholangiocytes.
      Il1b is one of the most potent inflammatory mediators,
      • Dinarello C.A.
      A clinical perspective of IL-1β as the gatekeeper of inflammation.
      which is activated by the inflammasome,
      • Rathinam V.A.
      • Fitzgerald K.A.
      Inflammasome complexes: emerging mechanisms and effector functions.
      but can also be activated independent of inflammasomes by TNF-producing CD4 T cells.
      • Jain A.
      • Irizarry-Caro R.A.
      • McDaniel M.M.
      • Chawla A.S.
      • Carroll K.R.
      • Overcast G.R.
      • et al.
      T cells instruct myeloid cells to produce inflammasome-independent IL-1β and cause autoimmunity.
      Il1b is also known to induce Th17 responses, notably to C. albicans.
      • Zielinski C.E.
      • Mele F.
      • Aschenbrenner D.
      • Jarrossay D.
      • Ronchi F.
      • Gattorno M.
      • et al.
      Pathogen-induced human TH17 cells produce IFN-γ or IL-10 and are regulated by IL-1β.
      Note that Th17 cells and increased reactivity to C. albicans have been described in patients with PSC.
      • Katt J.
      • Schwinge D.
      • Schoknecht T.
      • Quaas A.
      • Sobottka I.
      • Burandt E.
      • et al.
      Increased T helper type 17 response to pathogen stimulation in patients with primary sclerosing cholangitis.
      Moreover, Th17 responses are promoted by AIF-expressing DCs,
      • Elizondo D.M.
      • Andargie T.E.
      • Yang D.
      • Kacsinta A.D.
      • Lipscomb M.W.
      Inhibition of allograft inflammatory factor-1 in dendritic cells restrains CD4+ T cell effector responses and induces CD25+Foxp3+ T regulatory subsets.
      and by DCs that express Spp1, also known as osteopontin.
      • Murugaiyan G.
      • Mittal A.
      • Weiner H.L.
      Increased osteopontin expression in dendritic cells amplifies IL-17 production by CD4+ T cells in experimental autoimmune encephalomyelitis and in multiple sclerosis.
      ,
      • Kourepini E.
      • Aggelakopoulou M.
      • Alissafi T.
      • Paschalidis N.
      • Simoes D.C.
      • Panoutsakopoulou V.
      Osteopontin expression by CD103- dendritic cells drives intestinal inflammation.
      Both, AIF and Spp1 were upregulated by inflammatory cDC2s in response to DDC feeding, indicating that inflammatory cDC2s promote Th17 responses. Furthermore, these cells upregulated Clec4n, also known as dectin-2, which is a major pattern recognition receptor for alpha-mannans in the cell wall of C. albicans promoting Th17 responses.
      • Saijo S.
      • Ikeda S.
      • Yamabe K.
      • Kakuta S.
      • Ishigame H.
      • Akitsu A.
      • et al.
      Dectin-2 recognition of alpha-mannans and induction of Th17 cell differentiation is essential for host defense against Candida albicans.
      Thus, expansion and inflammatory activation of cDC2s in response to cholangiocyte damage might possibly be linked to increased immune reactivity to C. albicans and Th17 responses found in patients with PSC. This notion was confirmed by the finding that liver DCs from DDC-fed mice are superior inducers of Th17 cells, compared to control liver DCs (Fig. 2D). Other pattern recognition receptors that promote phagocytosis of microbes were also upregulated by hepatic cDC2s in response to DDC, including C1qb, C1qc, Clec4e (mincle), Clec4d (dectin-3), Clec4a1 (Dcir), CD14, Fcgr3, CD68, C3, Fcgr1, CD300lb.
      Another molecule that was overexpressed in inflammatory cDC2s was Sat1, which is a spermidine-deactivating enzyme; spermidine is involved in maintaining a tolerogenic DC phenotype.
      • Mondanelli G.
      • Bianchi R.
      • Pallotta M.T.
      • Orabona C.
      • Albini E.
      • Iacono A.
      • et al.
      A relay pathway between arginine and tryptophan metabolism confers immunosuppressive properties on dendritic cells.
      ,
      • Li G.
      • Ding H.
      • Yu X.
      • Meng Y.
      • Li J.
      • Guo Q.
      • et al.
      Spermidine suppresses inflammatory DC function by activating the FOXO3 pathway and counteracts autoimmunity.
      Thus, Sat1 can function as a metabolic switch regulating the maturation of tolerogenic DCs into inflammatory DCs. Accordingly, inflammatory cDC2s overexpressed S100a4, which is a proinflammatory molecule that bypasses tolerance to endotoxin and other damage-associated molecules;
      • Neidhart M.
      • Pajak A.
      • Laskari K.
      • Riksen N.P.
      • Joosten L.A.B.
      • Netea M.G.
      • et al.
      Oligomeric S100A4 is associated with monocyte innate immune memory and bypass of tolerance to subsequent stimulation with lipopolysaccharides.
      moreover, S100a4 is profibrotic and activates hepatic stellate cells.
      • Chen L.
      • Li J.
      • Zhang J.
      • Dai C.
      • Liu X.
      • Wang J.
      • et al.
      S100A4 promotes liver fibrosis via activation of hepatic stellate cells.
      Inflammatory maturation of cDC2s was also reflected by upregulation of the antigen processing and presentation machinery, indicated by upregulated expression of genes like Ctss,
      • Mahiddine K.
      • Hassel C.
      • Murat C.
      • Girard M.
      • Guerder S.
      Tissue-specific factors differentially regulate the expression of antigen-processing enzymes during dendritic cell ontogeny.
      Gabarap, also known as Atg8,
      • Romao S.
      • Gasser N.
      • Becker A.C.
      • Guhl B.
      • Bajagic M.
      • Vanoaica D.
      • et al.
      Autophagy proteins stabilize pathogen-containing phagosomes for prolonged MHC II antigen processing.
      ,
      • Schmid D.
      • Pypaert M.
      • Münz C.
      Antigen-loading compartments for major histocompatibility complex class II molecules continuously receive input from autophagosomes.
      and Plin2.
      • Bougnères L.
      • Helft J.
      • Tiwari S.
      • Vargas P.
      • Chang B.H.
      • Chan L.
      • et al.
      A role for lipid bodies in the cross-presentation of phagocytosed antigens by MHC class I in dendritic cells.
      Taken together, our study revealed that inflammatory Tbet- cDC2Bs accumulated in livers as an early response to biliary injury; yet hepatic cDC2 accumulation persisted in chronic cholangitis both in mice and humans. The intrahepatically accumulated cDC2s matured into inflammatory cells, expressing molecules that augment antigen presentation and promote inflammation. Thus, a sizeable fraction of inflammatory Tbet- cDC2s seems to stay in the liver and function as local organisers of cholangitis and local antigen-presenting cells, maintaining the pathogenic Th17 response associated with cholangitis. Therefore, Tbet- type 2 conventional DCs and their inflammatory mediators are potential therapeutic targets for the treatment of PSC. However, it is not entirely clear how well the findings in the mouse models translate into human cholangitis. Particularly the mechanisms behind the unexpected sparing of cDC2s in the CD11c.DOG model are not entirely clear yet. Therefore, although associated with an intriguing increase in ALT, the findings obtained with this model should be taken with a pinch of salt. However, the single-cell and in vitro data clearly confirmed a proinflammatory and Th17-promoting role of cDC2s in mice. Therefore, our findings should be followed by further translational studies investigating DCs in early PSC lesions in comparison to other inflammatory liver diseases.

      Abbreviations

      ADT, antibody-derived tags; AIH, autoimmune hepatitis; ALT, alanine aminotransferase; cDC1, conventional type 1 dendritic cell; cDC2, conventional type 2 dendritic cell; DCs, dendritic cells; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; DTR, diphtheria toxin receptor; moDCs, monocyte-derived DCs; PBC, primary biliary cholangitis; pDC, plasmacytoid DC; PSC, primary sclerosing cholangitis; Th, T helper; UMAP, uniform manifold approximation and projection.

      Financial support

      This study was supported by the DFG - Deutsche Forschungsgemeinschaft ( KFO306 ) (to JH), the DFG NGS Competence Centre for Genomic Analysis, Kiel and the DFG 2167 Cluster of Excellence Precision Medicine in Chronic Inflammation, RTF III (to PR).

      Authors’ contributions

      Experimental work, analysis and interpretation of data: Anna-Lena Müller, Christian Casar, Max Preti, Daria Krzikalla, Cornelia Gottwick, Pia Averhoff, Philip Rosenstiel. Study concept and design: Antonella Carambia, Johannes Herkel. Drafting of the manuscript: Anna-Lena Müller, Antonella Carambia, Johannes Herkel. Critical revision of the manuscript: all authors. Technical or material support and resources: Mathias Gelderblom, Marcus Altfeld, Ansgar W. Lohse, Silja Steinmann, Marcial Sebode, Jenny Krause, Dorothee Schwinge, Christoph Schramm. Funding acquisition: Johannes Herkel. All authors read and approved the manuscript.

      Data availability statement

      The single-cell data has been reposited in in the European Bioinformatics Institute (EBI) ArrayExpress database under accession number E-MTAB-11821.

      Conflict of interest

      The authors declare no conflicts of interest that pertain to this work.
      Please refer to the accompanying ICMJE disclosure forms for further details.

      Supplementary data

      The following are the supplementary data to this article:

      References

        • Dyson J.K.
        • Beuers U.
        • Jones D.E.J.
        • Lohse A.W.
        • Hudson M.
        Primary sclerosing cholangitis.
        Lancet. 2018; 391: 2547-2559
        • Lazaridis K.N.
        • LaRusso N.F.
        Primary sclerosing cholangitis.
        N Engl J Med. 2016; 375: 1161-1170
        • Azad A.I.
        • Krishnan A.
        • Troop L.
        • Li Y.
        • Katsumi T.
        • Pavelko K.
        • et al.
        Targeted apoptosis of ductular reactive cells reduces hepatic fibrosis in a mouse model of cholestasis.
        Hepatology. 2020; 72: 1013-1028
        • Chen Y.Y.
        • Arndtz K.
        • Webb G.
        • Corrigan M.
        • Akiror S.
        • Liaskou E.
        • et al.
        Intrahepatic macrophage populations in the pathophysiology of primary sclerosing cholangitis.
        JHEP Rep. 2019; 1: 369-376
        • Guicciardi M.E.
        • Trussoni C.E.
        • Krishnan A.
        • Bronk S.F.
        • Lorenzo Pisarello M.J.
        • O'Hara S.P.
        • et al.
        Macrophages contribute to the pathogenesis of sclerosing cholangitis in mice.
        J Hepatol. 2018; 69: 676-686
        • Reuveni D.
        • Gore Y.
        • Leung P.S.C.
        • Lichter Y.
        • Moshkovits I.
        • Kaminitz A.
        • et al.
        The critical role of chemokine (C-C Motif) receptor 2-positive monocytes in autoimmune cholangitis.
        Front Immunol. 2018; 9: 1852
        • Katt J.
        • Schwinge D.
        • Schoknecht T.
        • Quaas A.
        • Sobottka I.
        • Burandt E.
        • et al.
        Increased T helper type 17 response to pathogen stimulation in patients with primary sclerosing cholangitis.
        Hepatology. 2013; 58: 1084-1093
        • Oo Y.H.
        • Banz V.
        • Kavanagh D.
        • Liaskou E.
        • Withers D.R.
        • Humphreys E.
        • et al.
        CXCR3-dependent recruitment and CCR6-mediated positioning of Th-17 cells in the inflamed liver.
        J Hepatol. 2012; 57: 1044-1051
        • Nakamoto N.
        • Sasaki N.
        • Aoki R.
        • Miyamoto K.
        • Suda W.
        • Teratani T.
        • et al.
        Gut pathobionts underlie intestinal barrier dysfunction and liver T helper 17 cell immune response in primary sclerosing cholangitis.
        Nat Microbiol. 2019; 4: 492-503
        • Jeffery H.C.
        • Hunter S.
        • Humphreys E.H.
        • Bhogal R.
        • Wawman R.E.
        • Birtwistle J.
        • et al.
        Bidirectional cross-talk between biliary epithelium and Th17 cells promotes local Th17 expansion and bile duct proliferation in biliary liver diseases.
        J Immunol. 2019; 203: 1151-1159
        • Kunzmann L.K.
        • Schoknecht T.
        • Poch T.
        • Henze L.
        • Stein S.
        • Kriz M.
        • et al.
        Monocytes as potential mediators of pathogen-induced Th17 differentiation in patients with primary sclerosing cholangitis (PSC).
        Hepatology. 2020; 72 (Epub ahead of print. PMID: 31975436): 1310-1326
        • Weston C.J.
        • Zimmermann H.W.
        • Adams D.H.
        The role of myeloid-derived cells in the progression of liver disease.
        Front Immunol. 2019; 10: 893
        • Bassler K.
        • Schulte-Schrepping J.
        • Warnat-Herresthal S.
        • Aschenbrenner A.C.
        • Schultze J.L.
        The myeloid cell compartment-cell by cell.
        Annu Rev Immunol. 2019; 37: 269-293
        • Guilliams M.
        • Ginhoux F.
        • Jakubzick C.
        • Naik S.H.
        • Onai N.
        • Schraml B.U.
        • et al.
        Dendritic cells, monocytes and macrophages: a unified nomenclature based on ontogeny.
        Nat Rev Immunol. 2014; 14: 571-578
        • Durai V.
        • Murphy K.M.
        Functions of murine dendritic cells.
        Immunity. 2016; 45: 719-736
        • Deczkowska A.
        • David E.
        • Ramadori P.
        • Pfister D.
        • Safran M.
        • Li B.
        • et al.
        XCR1+ type 1 conventional dendritic cells drive liver pathology in non-alcoholic steatohepatitis.
        Nat Med. 2021; 27: 1043-1054
        • Reizis B.
        Plasmacytoid dendritic cells: development, regulation, and function.
        Immunity. 2019; 50: 37-50
        • Langlet C.
        • Tamoutounour S.
        • Henri S.
        • Luche H.
        • Ardouin L.
        • Grégoire C.
        • et al.
        CD64 expression distinguishes monocyte-derived and conventional dendritic cells and reveals their distinct role during intramuscular immunization.
        J Immunol. 2012; 188: 1751-1760
        • Bosteels C.
        • Neyt K.
        • Vanheerswynghels M.
        • van Helden M.J.
        • Sichien D.
        • Debeuf N.
        • et al.
        Inflammatory Type 2 cDCs acquire features of cDC1s and macrophages to orchestrate immunity to respiratory virus infection.
        Immunity. 2020; 52: 1039-1056.e9
        • Brown C.C.
        • Gudjonson H.
        • Pritykin Y.
        • Deep D.
        • Lavallée V.P.
        • Mendoza A.
        • et al.
        Transcriptional basis of mouse and human dendritic cell heterogeneity.
        Cell. 2019; 179: 846-863.e24
        • Fickert P.
        • Stöger U.
        • Fuchsbichler A.
        • Moustafa T.
        • Marschall H.U.
        • Weiglein A.H.
        • et al.
        A new xenobiotic-induced mouse model of sclerosing cholangitis and biliary fibrosis.
        Am J Pathol. 2007; 171: 525-536
        • Fickert P.
        • Fuchsbichler A.
        • Wagner M.
        • Zollner G.
        • Kaser A.
        • Tilg H.
        • et al.
        Regurgitation of bile acids from leaky bile ducts causes sclerosing cholangitis in Mdr2 (Abcb4) knockout mice.
        Gastroenterology. 2004; 127: 261-274
        • Guilliams M.
        • Dutertre C.A.
        • Scott C.L.
        • McGovern N.
        • Sichien D.
        • Chakarov S.
        • et al.
        Unsupervised high-dimensional analysis aligns dendritic cells across tissues and species.
        Immunity. 2016; 45: 669-684
        • Hildner K.
        • Edelson B.T.
        • Purtha W.E.
        • Diamond M.
        • Matsushita H.
        • Kohyama M.
        • et al.
        Batf3 deficiency reveals a critical role for CD8alpha+ dendritic cells in cytotoxic T cell immunity.
        Science. 2008; 322: 1097-1100
        • Hochweller K.
        • Striegler J.
        • Hämmerling G.J.
        • Garbi N.
        A novel CD11c.DTR transgenic mouse for depletion of dendritic cells reveals their requirement for homeostatic proliferation of natural killer cells.
        Eur J Immunol. 2008; 38: 2776-2783
        • Stuart T.
        • Butler A.
        • Hoffman P.
        • Hafemeister C.
        • Papalexi E.
        • Mauck 3rd, W.M.
        • et al.
        Comprehensive integration of single-cell data.
        Cell. 2019; 177: 1888-1902.e21
        • Hahne F.
        • LeMeur N.
        • Brinkman R.R.
        • Ellis B.
        • Haaland P.
        • Sarkar D.
        • et al.
        flowCore: a Bioconductor package for high throughput flow cytometry.
        BMC Bioinformatics. 2009; 10: 106
        • Liao Y.
        • Wang J.
        • Jaehnig E.J.
        • Shi Z.
        • Zhang B.
        WebGestalt 2019: gene set analysis toolkit with revamped UIs and APIs.
        Nucleic Acids Res. 2019; 47: W199-W205
        • Heier E.C.
        • Meier A.
        • Julich-Haertel H.
        • Djudjaj S.
        • Rau M.
        • Tschernig T.
        • et al.
        Murine CD103+ dendritic cells protect against steatosis progression towards steatohepatitis.
        J Hepatol. 2017; 66: 1241-1250
        • Deshmane S.L.
        • Kremlev S.
        • Amini S.
        • Sawaya B.E.
        Monocyte chemoattractant protein-1 (MCP-1): an overview.
        J Interferon Cytokine Res. 2009; 29: 313-326
        • De Filippo K.
        • Dudeck A.
        • Hasenberg M.
        • Nye E.
        • van Rooijen N.
        • Hartmann K.
        • et al.
        Mast cell and macrophage chemokines CXCL1/CXCL2 control the early stage of neutrophil recruitment during tissue inflammation.
        Blood. 2013; 121: 4930-4937
        • Boro M.
        • Balaji K.N.
        CXCL1 and CXCL2 regulate NLRP3 inflammasome activation via G-protein-coupled receptor CXCR2.
        J Immunol. 2017; 199: 1660-1671
        • Amaral E.P.
        • Riteau N.
        • Moayeri M.
        • Maier N.
        • Mayer-Barber K.D.
        • Pereira R.M.
        • et al.
        Lysosomal cathepsin release is required for NLRP3-Inflammasome activation by mycobacterium tuberculosis in infected macrophages.
        Front Immunol. 2018; 9: 1427
        • Bruchard M.
        • Mignot G.
        • Derangère V.
        • Chalmin F.
        • Chevriaux A.
        • Végran F.
        • et al.
        Chemotherapy-triggered cathepsin B release in myeloid-derived suppressor cells activates the Nlrp3 inflammasome and promotes tumor growth.
        Nat Med. 2013; 19: 57-64
        • Arsenijevic A.
        • Milovanovic J.
        • Stojanovic B.
        • Djordjevic D.
        • Stanojevic I.
        • Jankovic N.
        • et al.
        Gal-3 deficiency suppresses novosphyngobium aromaticivorans inflammasome activation and IL-17 driven autoimmune cholangitis in mice.
        Front Immunol. 2019; 10: 1309
        • Tian J.
        • Yang G.
        • Chen H.Y.
        • Hsu D.K.
        • Tomilov A.
        • Olson K.A.
        • et al.
        Galectin-3 regulates inflammasome activation in cholestatic liver injury.
        FASEB J. 2016; 30: 4202-4213
        • Chang T.H.
        • Huang J.H.
        • Lin H.C.
        • Chen W.Y.
        • Lee Y.H.
        • Hsu L.C.
        • et al.
        Dectin-2 is a primary receptor for NLRP3 inflammasome activation in dendritic cell response to Histoplasma capsulatum.
        PLoS Pathog. 2017; 13e1006485
        • Liao L.
        • Schneider K.M.
        • Galvez E.J.C.
        • Frissen M.
        • Marschall H.U.
        • Su H.
        • et al.
        Intestinal dysbiosis augments liver disease progression via NLRP3 in a murine model of primary sclerosing cholangitis.
        Gut. 2019; 68: 1477-1492
        • Maroni L.
        • Agostinelli L.
        • Saccomanno S.
        • Pinto C.
        • Giordano D.M.
        • Rychlicki C.
        • et al.
        Nlrp3 activation induces Il-18 synthesis and affects the epithelial barrier function in reactive cholangiocytes.
        Am J Pathol. 2017; 187: 366-376
        • Dinarello C.A.
        A clinical perspective of IL-1β as the gatekeeper of inflammation.
        Eur J Immunol. 2011; 41: 1203-1217
        • Rathinam V.A.
        • Fitzgerald K.A.
        Inflammasome complexes: emerging mechanisms and effector functions.
        Cell. 2016; 165: 792-800
        • Jain A.
        • Irizarry-Caro R.A.
        • McDaniel M.M.
        • Chawla A.S.
        • Carroll K.R.
        • Overcast G.R.
        • et al.
        T cells instruct myeloid cells to produce inflammasome-independent IL-1β and cause autoimmunity.
        Nat Immunol. 2020; 21: 65-74
        • Zielinski C.E.
        • Mele F.
        • Aschenbrenner D.
        • Jarrossay D.
        • Ronchi F.
        • Gattorno M.
        • et al.
        Pathogen-induced human TH17 cells produce IFN-γ or IL-10 and are regulated by IL-1β.
        Nature. 2012; 484: 514-518
        • Elizondo D.M.
        • Andargie T.E.
        • Yang D.
        • Kacsinta A.D.
        • Lipscomb M.W.
        Inhibition of allograft inflammatory factor-1 in dendritic cells restrains CD4+ T cell effector responses and induces CD25+Foxp3+ T regulatory subsets.
        Front Immunol. 2017; 8: 1502
        • Murugaiyan G.
        • Mittal A.
        • Weiner H.L.
        Increased osteopontin expression in dendritic cells amplifies IL-17 production by CD4+ T cells in experimental autoimmune encephalomyelitis and in multiple sclerosis.
        J Immunol. 2008; 181: 7480-7488
        • Kourepini E.
        • Aggelakopoulou M.
        • Alissafi T.
        • Paschalidis N.
        • Simoes D.C.
        • Panoutsakopoulou V.
        Osteopontin expression by CD103- dendritic cells drives intestinal inflammation.
        Proc Natl Acad Sci U S A. 2014; 111: E856-E865
        • Saijo S.
        • Ikeda S.
        • Yamabe K.
        • Kakuta S.
        • Ishigame H.
        • Akitsu A.
        • et al.
        Dectin-2 recognition of alpha-mannans and induction of Th17 cell differentiation is essential for host defense against Candida albicans.
        Immunity. 2010; 32: 681-691
        • Mondanelli G.
        • Bianchi R.
        • Pallotta M.T.
        • Orabona C.
        • Albini E.
        • Iacono A.
        • et al.
        A relay pathway between arginine and tryptophan metabolism confers immunosuppressive properties on dendritic cells.
        Immunity. 2017; 46: 233-244
        • Li G.
        • Ding H.
        • Yu X.
        • Meng Y.
        • Li J.
        • Guo Q.
        • et al.
        Spermidine suppresses inflammatory DC function by activating the FOXO3 pathway and counteracts autoimmunity.
        iScience. 2020; 23100807
        • Neidhart M.
        • Pajak A.
        • Laskari K.
        • Riksen N.P.
        • Joosten L.A.B.
        • Netea M.G.
        • et al.
        Oligomeric S100A4 is associated with monocyte innate immune memory and bypass of tolerance to subsequent stimulation with lipopolysaccharides.
        Front Immunol. 2019; 10: 791
        • Chen L.
        • Li J.
        • Zhang J.
        • Dai C.
        • Liu X.
        • Wang J.
        • et al.
        S100A4 promotes liver fibrosis via activation of hepatic stellate cells.
        J Hepatol. 2015; 62: 156-164
        • Mahiddine K.
        • Hassel C.
        • Murat C.
        • Girard M.
        • Guerder S.
        Tissue-specific factors differentially regulate the expression of antigen-processing enzymes during dendritic cell ontogeny.
        Front Immunol. 2020; 11: 453
        • Romao S.
        • Gasser N.
        • Becker A.C.
        • Guhl B.
        • Bajagic M.
        • Vanoaica D.
        • et al.
        Autophagy proteins stabilize pathogen-containing phagosomes for prolonged MHC II antigen processing.
        J Cell Biol. 2013; 203: 757-766
        • Schmid D.
        • Pypaert M.
        • Münz C.
        Antigen-loading compartments for major histocompatibility complex class II molecules continuously receive input from autophagosomes.
        Immunity. 2007; 26: 79-92
        • Bougnères L.
        • Helft J.
        • Tiwari S.
        • Vargas P.
        • Chang B.H.
        • Chan L.
        • et al.
        A role for lipid bodies in the cross-presentation of phagocytosed antigens by MHC class I in dendritic cells.
        Immunity. 2009; 31 (PMID: 19699172; PMCID: PMC2803012): 232-244