The Stimulation of Macrophages with TLR Ligands Supports Increased IL-19 Expression in Inflammatory Bowel Disease Patients and in Colitis Models

The incidence of inflammatory bowel disease (IBD) has increased worldwide during the past five decades, posing a major health burden, especially in industrialized countries (1). IBD consists of two major disorders: ulcerative colitis (UC) and Crohn’s disease (CD) (2, 3). Although the exact etiology of IBD is still unclear, studies have indicated that an inappropriate innate and adaptive immune response to the intestinal microbiota is essential for the development of IBD (4). 16S rRNA sequencing has revealed an altered composition of the gut microbiota, called dysbiosis, which is characterized by a decrease in biodiversity, a low stability, a decrease in bacteria from the Firmicutes phylum, and an increased number of Enterobacteriaceae in IBD patients (5–7).

The development of colitis in most animal models, including spontaneous colitis in genetically modified animals, such as IL-10–deficient animals, requires the gut microbiota (8). Other than IL-10 itself, the IL-10 family includes IL-19, IL-20, IL-22, IL-24, and IL-26, as a result of their sequence homologies to IL-10 (9, 10). IL-19, IL-20, and IL-24 form the IL-20 subfamily within the IL-10 family, because these cytokines signal through the heterodimeric IL-20Rs, resulting in activation of the STAT3 signaling pathway (11). IL-20Rβ can pair with IL-20Rα, forming the type I IL-20R to allow signaling of IL-19, IL-20, and IL-24, or it can pair with IL-22Rα to form the type II IL-20R, which only signals from IL-20 and IL-24 (12).

IL-19 is involved in the development of autoimmune diseases, such as psoriasis (13, 14), whereas the data for IBD are inconsistent. Intestinal expression levels of IL-20Rα and IL-20Rβ are low compared with IL-22Rα (15, 16), but studies have suggested some role for IL-19 in the development of IBD. Elevated IL-19, IL-20, IL-24, and IL-26 expression in biopsies and peripheral blood has been reported in patients with active IBD (17–20). Changes in the methylation status of the IL19 promoter and gene locus have been associated with a severe disease state in CD (21), and single-nucleotide polymorphisms in the IL19 gene reduce the susceptibility to UC (22). Furthermore, in silico analysis of disease networks in CD patients demonstrated that the cytokines IL-19, IL-20, and IL-24 are interconnected with the core proteins IL12B/IL23R/JAK2/STAT3, for which single-nucleotide polymorphisms associated with IBD have been reported (23). On a cellular level, IL-19 has been shown to suppress LPS-induced TNF-α production by monocytes isolated from CD patients (24). In mice, increased expression of IL-6, IL-12, and TNF-α has been observed in bone marrow–derived macrophages (BMDMs) from IL-19–deficient animals compared with wild-type (WT) animals (25). Moreover, signaling via IL-20Rβ has inhibited IL-1β expression by keratinocytes and promoted disease in a model of cutaneous Staphylococcus aureus infection (26).

However, a proinflammatory effect of IL-19 was described in other contexts. For example, stimulation of monocytes with IL-19 may induce the production of TNF-α, IL-6, and reactive oxygen species, as well as induce monocyte apoptosis (27). Synovial fibroblasts have been reported to produce TNF-α and IL-6 after stimulation with IL-19 (28). In line with this, IL-19 stimulation has also been associated with the induction of the cytokines IL-1β, IL-6, and IL-8 and the chemokines CCL5 and CXCL9 in lung epithelial cells, as well as with neutrophil chemotaxis (29).

In this study, we analyzed the expression and the function of the cytokines IL-19, IL-20, and IL-24 in IBD patients and in a mouse model of colitis. To identify the cells that express IL-19, a reporter mouse line was constructed in which IL-19 expression can be identified by the red fluorescent protein tandem dimer tomato (tdTomato). Our findings suggest that microbial-derived products are required to induce the expression of IL-19 by macrophages. Colitis was attenuated in IL-19–deficient animals and was associated with reduced numbers of IL-6–producing monocytes and macrophages in the colonic lamina propria (cLP).

C57BL/6, Il19-tdTomato, and Rag^−/− mice were kept under specific pathogen–free (SPF) conditions in the animal colonies of the Department of Biomedicine, University of Basel. Il19-tdTomato–transgenic mice have been constructed as described below. In some cases, SPF C57BL/6 mice were purchased from Harlan Laboratories and kept in the animal colonies of the Department of Clinical Research, University of Bern. Germ-free (GF) Cx3cr1-GFP (B6.129P-Cx3cr1^tm1Litt/J) × Rag^−/− mice were kept at the clean mouse facility at the University of Bern. Mice with a simplified oligo-mouse microbiota with 12 bacterial species (Oligo-MM¹²) were maintained at the clean mouse facility (30, 31) at the University of Bern after colonizing of GF C57BL/6 mice with Acutalibacter muris sp. nov. KB18 (DSM 26090), Flavonifractor plautii YL31 (DSM26117), Clostridium clostridioforme YL32 (DSM 26114), Blautia coccoides YL58 (DSM 26115), Clostridium innocuum I46 (DSM26113), Lactobacillus reuteri I49 (DSM 32035), Enterococcus faecalis KB1 (DSM 32036), Bacteroides caecimuris sp. nov. I48 (DSM 26085), Muribaculum intestinale sp. nov. YL27 (DSM 28989), Bifidobacterium longum subsp. animalis YL2 (DSM 26074), Trichomonas muris sp. nov. YL45 (DSM 26109), and Akkermansia muciniphila YL44 (DSM 26127). Oligo-MM¹² is also known as stable defined moderately diverse microbiota mouse 2. To colonize GF animals with an SPF flora, GF animals were cohoused with SPF animals. All mouse experiments were performed in accordance with the Swiss Federal and Cantonal regulations (animal protocol number 2816 [canton of Basel-Stadt] and number BE132/14 [canton of Bern]).

To study the function of IL-19, Il19-tdTomato–transgenic animals were generated that express the red fluorescent protein tdTomato under the control of the Il19 promoter. In mice, Il19 is located on chromosome 1 (spanning 8.6 kb) with two transcripts comprising six or five exons. Both transcripts code for a protein with a size of 176 aa. The transcript with five exons is the dominant transcript. The start codon ATG is located on the first exon (five-exon transcript) at position 38. To keep the gene expression regulatory elements intact, a 100 aa truncated protein with only 79 aa from IL-19 is expressed in the transgenic animals. The reporter gene tdTomato is expressed on the Il19 transcript, but with an internal ribosomal entry site to translate into a nonfusion complete tdTomato protein. Using bacterial artificial chromosome–containing recombination, the pDTA–Il19-A-B-C targeting vector was generated and identified by PCR. After G418 selection, 400 clones were picked and selected by Southern blots using probe (WT 10.5 kb, mutant 7.5 kb [XbaI]) and probe (WT 10.0 kb, mutant 6.4 kb [EcoRI]). Positive clones were identified and microinjected into mouse blastocysts acquired from C57BL/6 mice. Blastocysts were implanted into the uterus of females to receive chimeras and backcrossed to receive the F1 generation. Transgenic animals were characterized by PCR. Primer sequences are given in Table I. To amplify the WT and mutated gene locus, separate PCR reactions were run. The following PCR conditions (35 cycles) were used: initial denaturation at 95°C for 5 min, denaturation at 95°C for 30 s, annealing at 62°C for 30 s, extension at 72°C for 60 s, and final extension at 72°C for 10 min.

To measure the concentration of IL-19, IL-20, and IL-24 in sera and mRNA expression levels in colonic biopsies of the same patient in this retrospective observational study, 72 patients with IBD (31 CD and 41 UC patients) were recruited by participating gastroenterologists in private practice, regional hospitals, and tertiary centers in the Swiss Inflammatory Bowel Disease Cohort Study (project 2015-05), which is a national cohort started in 2006 [described in detail in Ref. (32)]. Inclusion criterion was the diagnosis of CD or UC established ≥4 mo before, and inclusion was confirmed by radiological, endoscopic, or surgical assessment. Patients were excluded if they suffered from another form of colitis, were not regularly followed up for CD or UC, were not a permanent resident of Switzerland, or if they did not sign the informed consent form. Indications for ileocolonoscopy in patients with active CD or UC was worsening of symptoms (flare of the disease) or clinical suspicion of complication, and only patients with moderate to severe disease activity in endoscopic findings were included. Biopsies were taken from segments that appeared macroscopically inflamed. Indication for ileocolonoscopy in patients with quiescent CD or UC was colorectal cancer screening, and only patients with quiescent disease in endoscopic findings were included. For CD patients, biopsies were sampled from the ileum of active disease (n = 7), the ileum of quiescent disease (n = 7), the colon of active disease (n = 7), and the colon of quiescent disease (n = 9). Peripheral blood was taken, and serum was separated by centrifugation and stored at −80°C. Biopsies were taken during ileocolonoscopy, placed in RNAlater stabilization solution (Invitrogen), stored at −20°C, and processed later for RNA isolation. Detailed patient data are given in Table II.

Cohoused (for 3 wk), weight-matched female (8–12 wk of age) mice were given 1.5–2.0% dextran sodium sulfate (DSS; m.w. 36,000–50,000; MP Biomedicals) in drinking water for 5 d. To induce recovery from acute colitis, the DSS-containing water was exchanged with normal drinking water. Mice were controlled for the occurrence of clinical signs of colitis (see below), and weight loss was determined.

Female Rag^−/− mice (8 wk of age) were injected i.p. with 200 μg of anti-CD40 FGK4.5 (Bio X Cell) to induce CD40 signaling. Mice were monitored for the occurrence of clinical signs of colitis, and weight loss was determined.

Clinical signs of colitis were scored as follows: rectal bleeding: 0 - absent, 1 - bleeding; rectal prolapse: 0 - nil, 1 - prolapse - mice euthanized; stool consistency: 0 - normal, 1 - loose stools, 2 - diarrhea; position: 0 - normal movement, 1 - reluctance to move, 2 - hunched position; appearance of the fur: 0 - normal appearance, 1 - ruffled fur, 2 - spiky fur; and weight loss: 0 – no loss, 1 - body weight loss 0–5%, 2 - body weight loss >5–10%, 3 - body weight loss >10–15%, 4 - body weight loss >15%. The animals were monitored once a day, if the total score was ≥4, the animals were monitored twice a day. When a total clinical score ≥6 occurred, when an individual animal lost >15% body weight, when gross bleeding was observed, or when rectal prolapse appeared, the animal was euthanized.

LPS (2 μg/g body weight) purified from Escherichia coli O111:B4 (Sigma-Aldrich) was injected i.p. into age-matched male mice (9 wk of age).

After preparation of the femurs and tibias of hind legs, the muscle and connective tissues were carefully removed, and bones were cut open at the epiphysis. A syringe with a 25-gauge needle was inserted into the ends of the bones to flush out the bone marrow with complete RPMI 1640 medium (Life Technologies). The bone marrow cells were collected and passed through a 70-μm cell strainer to remove bone spicules and cell clumps.

Segments of the colon were washed with PBS to remove debris and mucus. The intestinal epithelium was removed by incubation in 5 mM EDTA in Ca²⁺/Mg²⁺-free PBS at 37°C under gentle shaking for 10 min for a total of three incubations. Tubes were vortexed for 30 s after each incubation, and the tissue pieces were transferred into fresh EDTA/PBS. In some experiments, the supernatants were collected, and the epithelial cells were pelleted and used for further analysis. The remaining tissue was washed in PBS to remove residual EDTA. Denuded tissues were cut as small as possible and digested with 0.5 mg/ml Collagenase type VIII (Sigma-Aldrich) and 10 U/ml DNase (Roche) in RPMI 1640 for 25–35 min at 37°C in a water bath with continuous shaking (200 rpm). Every 5 min, the tubes were vortexed manually for 30 s. Supernatants were collected and passed through a 70-μm cell strainer, and cLP cells were pelleted. The cells were counted and processed further. When gene-expression analysis was performed in isolated cLP cells, a Percoll gradient was performed to purify the leukocytes. Leukocytes were collected from the interface of a 30/100% Percoll gradient (density 1.13 g/ml; GE Healthcare), pelleted, and used for further analysis. When colonic macrophages and monocytes were used for cell cultures, macrophages and monocytes were sorted on a FACSAria cell sorter.

Lymph nodes and spleens were collected from euthanized mice, placed on a 70-μm cell strainer, and mashed with the plunger of a syringe into Falcon tubes. RBCs were lysed with ammonium-chloride-potassium buffer (150 mM NH₄Cl, 10 mM KHCO₃, 0.1 mM 0.5 M EDTA), and cells were washed, pelleted, counted, and used for flow cytometry or cell culture.

Tissue samples were fixed in 4% paraformaldehyde and embedded in paraffin blocks. Five-micrometer sections were stained with H&E. Histological features of colonic inflammation were evaluated using a previously published scoring system ranging from 0 (normal tissue) to 4 (maximal severity of inflammation) as follows: 0, no evidence of inflammation; 1, low level of inflammation with scattered infiltrating mononuclear cells with one or two foci; 2, moderate inflammation with multiple foci; 3, high level of inflammation with increased vascular density and marked wall thickening; and 4, maximal severity of inflammation with transmural leukocyte infiltration and loss of goblet cells (33). Histological scores were assessed in a blinded fashion by two independent investigators, and the mean histological score was determined for each animal.

Cryopreserved biopsies embedded in Tissue-Tek O.C.T. compound (Sakura) were acquired from the Basel IBD cohort. Immunohistochemistry (IHC) was performed on 7-μm sections using rabbit anti-human IL-20Rα mAb 024 (Sino Biological) and anti-human IL-20Rβ polyclonal Ab (product code ab124332; Abcam). Primary Ab binding was detected with a goat anti-rabbit IgG secondary Ab (Thermo Fisher). Ab binding was revealed with standard AEC Solution (Thermo Fisher), and sections were counterstained with hematoxylin.

To further determine the cells that express IL-20Rα, 6-μm cryosections from the Basel IBD cohort were fixed with 4% formaldehyde and stained with mouse anti-human CD64 mAb 10.1 (BioLegend) and rabbit anti-human IL-20Rα mAb 024 (Sino Biological). Primary Ab binding was detected with goat anti-mouse IgG1-A555 (Thermo Fisher) and with a goat anti-rabbit IgG secondary Ab (Thermo Fisher). Sections were counterstained with Hoechst 33342 (Thermo Fisher) and imaged with a Nikon A1R Nala confocal microscope.

Combined in situ RNA hybridization (ISH) for Il19 and IHC for F4/80 was used to detect colocalization of Il19 expression with macrophages. ISH was performed using an RNAscope 2.5 HD Assay-BROWN Kit (Advanced Cell Diagnostics) as per the manufacturer’s protocol with minor changes. In brief, paraffin-embedded formalin-fixed tissue from intestine was cut into 5-μm-thick sections. Thereafter, sections were deparaffinized, and endogenous peroxidase was blocked using H₂O₂, followed by RNAscope Target Retrieval at 95°C for 6 min. Finally, slides were incubated with RNAscope Protease Plus for 10 min at 40°C. A custom-designed Il19 probe (gene ID: 329244) was added to the tissue, and the signal was developed according to the manufacturer’s protocol. Slides were washed with 1× PBS for 5 min to be prepared for IHC after ISH. IHC for macrophages was performed using rat anti-mouse anti-F4/80 clone Cl:A3-1 (Bio-Rad). In brief, endogenous peroxidase was blocked using BLOXALL (Vector Laboratories) for 10 min at room temperature. Thereafter, IHC was performed using an ImmPRESS HRP goat anti-rat IgG (Peroxidase) Polymer Detection Kit (Vector Laboratories), following the manufacturer’s protocol. Slides were blocked using normal goat serum for 30 min and incubated with primary Ab (1:100) overnight at 4°C. Subsequently, slides were washed with PBS for 5 min, incubated with ImmPress Reagent for 30 min, and washed with 1× PBS for 5 min. HIGHDEF yellow IHC chromogen (Enzo Life Sciences) was used to reveal signal by incubating slides with chromogen solution for 30 s. Slides were washed with dH₂O and counterstained with 50% Hematoxylin Solution, Gill No. 1 (Sigma-Aldrich), washed, dried, and mounted using VectaMount Permanent Mounting Medium. The images were acquired using an Olympus AX70 microscope and processed using Cell^F software (both from Olympus).

Murine bone marrow cells were cultured in 6- or 24-well plates in RPMI 1640 medium (Life Technologies) containing 10% FCS and supplemented with 0.05 mM 2-ME, 100 U/ml penicillin, and 100 μg/ml streptomycin. To generate macrophages, 20 ng/ml M-CSF (BioLegend) was added; to generate macrophages and dendritic cells (DCs), 10 ng/ml GM-CSF (BioLegend) was added, and to receive DCs, 100 ng/ml human FLT3 ligand (FLT3L; kindly provided by R. Tussiwand) was added. After 7 d (M-CSF and GM-CSF) or 12 d (FLT3L) of culture, cells were stimulated with the TLR ligands LPS from E. coli O111:B4 (Sigma-Aldrich), flagellin from Salmonella typhimurium (InvivoGen), Pam2CSK4 (InvivoGen), low m.w. polyinosinic:polycytidylic acid [poly (I:C); InvivoGen], and CpG ODN 1668 (5′-TCCATGACGTTCCTGAATAAT-3′; Microsynth), as specified in the figure legends.

Purified colonic macrophages and monocytes were cultured in 96-well plates in RPMI 1640 GlutaMAX medium (Life Technologies) containing 10% FCS and supplemented with 0.05 mM 2-ME, 200 U/ml penicillin, and 200 μg/ml streptomycin. Cells were stimulated with 100 ng/ml LPS for 24 h, supernatants were collected, and cytokine concentrations were analyzed with a custom Mouse ProcartaPlex Multiplex assay system (Affymetrix).

To isolate naive CD4 T cells for T cell–differentiation assays, single-cell suspensions were aseptically prepared from the spleen of C57BL/6 animals, washed, passed through a cell strainer (pore size 40 μm), and resuspended in PBS/2% FCS–supplemented 2 mM EDTA. Naive CD4 T cells were negatively sorted by magnetic cell isolation using a Naive CD4+ T Cell Isolation Kit, mouse (Miltenyi Biotec), according to the manufacturer’s instructions. Naive CD4 T cells (5 × 10⁴ cells per well) in IMDM containing 10% FCS and supplemented with 0.05 mM 2-ME, 100 U/ml penicillin, and 100 μg/ml streptomycin were placed in 96-well flat-bottom plates that had been coated with 5 μg/ml anti-CD3 145-2C11 (BioLegend), and 5 μg/ml soluble anti-CD28 37.51 (eBioscience) was added to all assays. Th1 cells were differentiated in the presence of 10 μg/ml anti–IL-4 11B11 (BioLegend), 20 ng/ml IL-2 (R&D Systems), and 20 ng/ml IL-12 (R&D Systems). Th2 cells were differentiated in the presence of 10 μg/ml anti–IFN-γ XMG1.2 (BioLegend) and 20 ng/ml IL-4 (BioLegend). Th17 cells were generated in the presence of 10 μg/ml anti–IL-4, 10 μg/ml anti–IFN-γ, 1 ng/ml human TGF-β (R&D Systems), 10 ng/ml IL-1β (R&D Systems), 50 ng/ml IL-6 (R&D Systems), and 20 ng/ml IL-23 (R&D Systems). One hundred nanograms per milliliter of IL-19 (R&D Systems) were added to the cultures. In some experiments, IL-19 was added after the differentiation of naive T cells to the respective Th effector cells. Cells were harvested after 4–5 d of culture and processed for flow cytometry.

After isolation, cells were washed in PBS, and subsequently stained with fixable viability dye eFluor 455UV (eBioscience) or Zombie Aqua fixable viability dye (BioLegend) and mAb 2.4G2 (BD Biosciences) directed against the FcγRIII/II CD16/CD32 (0.5 μg mAb per 10⁶ cells) to stain dead cells or to block nonspecific binding of Abs to Fc receptors, respectively, for 30 min at 4°C. Cells were washed in PBS/2% FBS supplemented with 0.1% w/v sodium azide and 10 mM EDTA, incubated with the relevant mAb for 20 min at 4°C, and washed again twice. When biotin-coupled Abs were used as primary Abs, cells were incubated with Streptavidin–eFluor 450 (eBioscience) for 20 min at 4°C. Data were acquired with an LSR II or a Fortessa flow cytometer (BD Biosciences) and analyzed using FlowJo software version 10.0.7r2 (TreeStar). In all experiments, forward scatter (FSC)-H versus FSC-A was used to gate on singlets, with dead cells excluded using the fluorescence-coupled fixable viability dye. When lineage exclusion was performed, CD3-, CD19-, and NK1.1-expressing cells were removed from further analysis.

The following mAbs were used for surface staining: allophycocyanin-Cy7–conjugated anti-CD11b M1/70, FITC-conjugated anti-CD11b M1/70 (both from BioLegend), allophycocyanin-conjugated anti-CD11c HL3 (BD Biosciences), allophycocyanin-Fire 750–conjugated anti-CD11c N418, Brilliant Violet 570–conjugated anti-CD19 6D5, Brilliant Violet 785–conjugated anti-CD19 6D5, Alexa Fluor 700–conjugated anti-CD19 6D5, biotin-conjugated anti-CD19 6D5, Alexa Fluor 700–conjugated anti-CD3 17A2 (all from BioLegend), allophycocyanin-Cy7–conjugated anti-CD4 GK1.5 (BD Biosciences), Brilliant Violet 510–conjugated anti-CD4 RM4-5 (BioLegend), allophycocyanin-conjugated anti-CD4 RM4-5 (BD Biosciences), PE-conjugated anti-CD44 IM7 (BioLegend), eVolve 655–conjugated anti-CD45 30-F11 (eBioscience), FITC-conjugated anti-CD62L MEL-14 (BD Biosciences), Brilliant Violet 711–conjugated anti-CD64 X54-5/7.1 (BioLegend), PE-Cy7–conjugated anti-CD64 X54-5/7.1 (BioLegend), PerCP-conjugated anti-CD8α 53-6.7 (BioLegend), allophycocyanin-conjugated anti-F4/80 BM8 (BioLegend), FITC-conjugated anti–Gr-1 RB6-8C5 (BD Biosciences), Pacific Blue–conjugated anti–I-A/I-E M5/114.15.2, Alexa Fluor 700–conjugated anti–I-A/I-E M5/114.15.2, PerCP-Cy5.5–conjugated anti-Ly6C HK1.4 (all from BioLegend), PE-Cy7–conjugated anti-NK1.1 PK136 (BD Biosciences), Alexa Fluor 700–conjugated anti-NK1.1 PK136, and allophycocyanin-conjugated anti-NKp46 29A1.4 (both from BioLegend).

Isolated cells (0.4–1 ×10⁶ per milliliter) from the cLP or cells from T cell cultures were stimulated for 4 h at 37°C with 50 ng/ml PMA and 750 ng/ml Ionomycin (both from Sigma-Aldrich) in the presence of 3 μg/ml Brefeldin A (eBioscience). Cells were harvested, washed, and surface stained. Surface stained cells were fixed and permeabilized (Fixation/Permeabilization Solution Kit; BD Biosciences). Permeabilized cells were incubated for 30 min at 4°C in the dark with the following Abs: FITC-conjugated anti–IFN-γ XMG1.2 (BD Biosciences), allophycocyanin-conjugated anti–IL-17 eBio17B7 (eBioscience), PE-conjugated IL-17A TC11-18H10 (BD Biosciences), PE-conjugated anti–IL-13 eBio13A (eBioscience), and allophycocyanin-conjugated anti–IL-6 MP5-20F3 (BioLegend). Stained cells were washed twice in permeabilization buffer and resuspended in PBS supplemented with 2% FBS and 0.1% w/v sodium azide. The number of cytokine-expressing cells was determined by flow cytometry.

Bone marrow cells were differentiated to macrophages in the presence of 20 ng/ml M-CSF. To exclude STAT3 signaling by M-CSF, macrophages were cultured in medium that did not contain M-CSF for 16–24 h before stimulation. For stimulation, macrophages were collected, adjusted to 5 × 10⁵ cells per 500 μl, and left to rest for 4 h before stimulation with 100 ng/ml recombinant murine (rm)IL-10 (BioLegend) or rmIL-19 (R&D Systems). At the indicated time points, cells were directly fixed with BD Phosflow Lyse/Fix Buffer (BD Biosciences), washed twice in PBS/2% FBS supplemented with 0.1% w/v sodium azide and 10 mM EDTA, and permeabilized with BD Phosflow Perm Buffer III (BD Biosciences). Cells were washed three times, and nonspecific binding of Abs to Fc receptors was blocked by preincubation of cells with mAb 2.4G2 (BD Biosciences) directed against the FcγRIII/II CD16/CD32 (0.5 mg mAb per 10⁶ cells). Cells were then incubated with 0.125 μg per test PE-conjugated anti-STAT3 (pY705) 4/P-STAT3 (BD Biosciences) or PE-conjugated mouse IgG2aκ isotype MOPC-173 (BioLegend) for 60 min at room temperature. Cells were washed twice before data acquisition with an LSR II flow cytometer (BD Biosciences).

RNA was extracted from the indicated cell populations and organs of mice or from human biopsies using TRIzol (Ambion) or TRI Reagent (Zymo Research). RNA isolated from the tissue of DSS-treated animals was cleaned with an RNeasy MinElute Cleanup Kit (QIAGEN) or isolated with a Direct-zol RNA MiniPrep Kit (Zymo Research) to ensure removal of residual DSS. RNA was extracted from FACS-sorted BMDMs using a Direct-zol MicroPrep Kit (Zymo Research). Genomic DNA was removed with provided in-column treatment or with recombinant DNase I (Invitrogen) and reverse transcribed into cDNA using Superscript RT III (Invitrogen) or MultiScribe MuLV (Invitrogen) reverse transcriptase, according to the manufacturer’s instructions. Quantitative real-time PCR (qPCR) was carried out in 384-well plates using gene-specific primers and either SsoFast EvaGreen Supermix (Bio-Rad) and run on a Bio-Rad CFX384 cycler or with a QuantiNova SYBR Green PCR kit (QIAGEN) and run on an ABI ViiA 7 cycler. All reactions were run in triplicates. Samples were normalized to the expression of Gapdh or Actb by calculating 2^(−deltaCt). QuantiTect Primer assays for human IL19, human IL20, human IL24, human GAPDH, mouse Il20, and mouse Il24 were purchased from QIAGEN. Table III shows the sequences of primers used for amplification of mouse Gapdh, Actb, Il1b, Il6, Il10, Il19, Il20ra, Il20rb, Il22ra1, and Tnfa.

Total RNA was quality checked on a Bioanalyzer instrument using the RNA 6000 Nano Chip (both from Agilent Technologies) and quantified by spectrophotometry using a NanoDrop ND-1000 (NanoDrop Technologies). Library preparation was performed with 220 ng of total RNA using a TruSeq Stranded mRNA LT Sample Prep Kit (Illumina). Libraries were quality checked with a Fragment Analyzer using a Standard Sensitivity NGS Fragment Analysis Kit (both from Advanced Analytical, Ames, IA) revealing excellent quality of libraries (average concentration was 126 ± 13 nmol/l, and average library size was 345 ± 7 bp). Samples were pooled to equal molarity. Each pool was quantified by PicoGreen Fluorometric measurement to be adjusted to 1.5 pM and was used for clustering on the NextSeq 500 instrument (Illumina).

Samples were sequenced by single-reads 81 bases using the NextSeq 500 High Output Kit 75-cycles (Illumina), and primary data analysis was performed with an Illumina RTA version 2.4.11 and bcl2fastq-2.18.0.12. Reads were aligned to the mouse genome (mm10) using the spliced read aligner STAR. Sequencing and mapping quality were assessed using the qQCReport function of the R package QuasR. Gene expression was quantified by the qCount function of QuasR using an exon-union model of mm10 RefSeq genes (downloaded from UCSC 2015-12-18). The R package edgeR was used for differential gene-expression analysis. An additive generalized linear model including the factors “group” (tdTomato-positive and tdTomato-negative) and “replicate” (1–4) was fitted to the raw counts (function glmfit), and differential expression between tdTomato-positive and -negative samples was evaluated by likelihood ratio tests (function glmLRT). The p values were adjusted by controlling the false-discovery rate.

The commercially available mouse IL-19 ELISA Ready-SET-Go! (eBioscience) was used. A Human IL-19 Quantikine ELISA Kit, a Human IL-20 Quantikine ELISA Kit, and a Human IL-24 DuoSet ELISA were purchased from R&D Systems. Extinction was measured at 450 nm on a microplate ELISA reader.

The data were analyzed with GraphPad Prism software (version 7.0b) and are presented as dot plots in which the median of each experimental group is presented in addition to the individual samples. Statistical significance was calculated using the Mann–Whitney U test for two groups or using the Kruskal–Wallis test followed by the Dunn correction test for multiple comparisons. When the data are presented as a time course, the arithmetic mean ± SD is shown. Statistical significance was calculated using two-way ANOVA with the Sidak correction. Outliers were identified with the Grubb test during the analysis of data acquired from samples from the Swiss IBD Cohort only. The p values are indicated as follows: *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001.

Recent reports have suggested that members of the IL-20 cytokine family may play a role in IBD (17), but the published data are scarce and, in part, contradicting. Therefore, we analyzed the concentrations of IL-19, IL-20, and IL-24 in sera from IBD patients and determined IL19, IL20, and IL24 expression in biopsies from patients with quiescent and active IBD. Active disease, for CD and UC, was defined as moderate to severe endoscopic disease activity, worsening of symptoms, or suspicion of complications, whereas remission was defined as quiescent endoscopic disease activity. We found that the expression of all three cytokines was significantly increased in biopsies from patients with active UC compared with quiescent UC, whereas the expression levels were not affected by the disease status in CD patients (Fig. 1a–c). The expression levels of IL19, IL20, and IL24 correlated with each other, irrespective of the disease activity, indicating that inflammation in patients with active disease leads to the expression of all three cytokines (data not shown). Serum levels of IL-19 (p = 0.03) and IL-20 (p = 0.01), but not IL-24, were elevated in IBD patients with active disease compared with patients with quiescent disease (Fig. 1d, 1f). It should be noted that sample age did not affect the measurements of the respective cytokine in sera of IBD patients and that the expression of IL19, IL20, and IL24 in biopsies did not correlate with serum levels (data not shown).

Biopsies from inflamed and noninflamed colon segments of UC patients were stained for IL-20Rα and IL-20Rβ (Fig. 1g, 1h). We found that mononuclear cells in the inflamed and noninflamed cLP expressed IL-20Rα and that IL-20Rα was also expressed by intestinal epithelial cells. cLP mononuclear cells and intestinal epithelial cells showed a positive IL-20Rβ staining, with a higher staining intensity observed in intestinal epithelial cells than in cLP mononuclear cells. There was no significant difference in IL-20Rα or IL-20Rβ expression between inflamed and noninflamed regions (data not shown). Given that IL19 expression is increased in patients with active UC and that the type I IL-20R is present in the cLP, we next assessed the expression of Il19 in a mouse model of colitis.

We first analyzed the expression of Il19 and its receptor in unmanipulated SPF C57BL/6 mice. Although Il19 expression could not be detected in spleen, peripheral lymph nodes, bone marrow, Peyer’s patches, liver, kidneys, or the entire gastrointestinal tract, it was detectable in the mesenteric lymph nodes in six of eight animals and in the epidermis in one of eight animals (Supplemental Fig. 1a, 1b). Previous work has shown that IL-19 binds to the type I IL-20R, which is formed by IL-20Rα and IL-20Rβ (12). We found IL-20Rα to be expressed in the stomach, proximal and distal colon, bone marrow, and epidermis, but not in the small intestine of C57BL/6 mice, with the highest expression in the forestomach and epidermis. IL-20Rβ was expressed in the entire gastrointestinal tract and all tested organs, with the highest expression in the epidermis (Supplemental Fig. 1a–c). IL-22Rα, which, together with IL-20Rβ, forms the type II IL-20R, was expressed in the entire gastrointestinal tract and the epidermis but not in peripheral lymph nodes (Supplemental Fig. 1a–c).

Next, we sought to determine whether the expression of Il19, Il20, and Il24 is regulated in mice with DSS colitis. DSS (2%) was provided in the drinking water for 5 d and then replaced by regular water. The animals lost body weight and displayed clinical signs of colitis, as assessed by a clinical colitis score (Fig. 2a). After 8 and 12 d, cLP cells and epithelial cells were isolated. Il19 expression peaked in cLP cells, but not in intestinal epithelial cells, at day 8 and declined 12 d after the onset of colitis (Fig. 2b). In contrast to human disease (Fig. 1b, 1c), expression of the cytokines Il20 and Il24 could not be detected in animals with colitis (data not shown). Intestinal epithelial cells and cLP cells expressed Il20rb, with higher expression levels in the cLP (Fig. 2c). Il20ra expression by cLP cells, but not by intestinal epithelial cells, was increased during colitis (Fig. 2d), whereas intestinal epithelial cells primarily expressed Il22ra1 (Fig. 2e). In situ hybridization showed expression of Il19 by F4/80⁺ cLP macrophages, which increased during inflammation. Some intestinal epithelial cells, other cLP cells, and cells within the muscle layer also expressed Il19 (Fig. 2f). These results suggest that expression of Il19 is induced during colitis, Il20ra is expressed in the colon, and the expression pattern of Il20ra is different from that of Il22ra1.

Because it is well established that local and systemic immune responses are largely influenced by the commensal microbiota (34), we tested whether the expression of IL-20 family cytokines is changed by the colonization of GF animals with an SPF flora. Therefore, GF mice were cohoused with SPF mice for 1, 2, or 7 d. Although Il19, Il20, and Il24 were not detected in intestinal tissues, mesenteric lymph nodes, or epidermis at any point during the colonization (data not shown), the expression of Il20rb was transiently increased in the proximal parts of the small intestine (Supplemental Fig. 1d, 1e).

GF mice develop a severe disease when exposed to DSS (35); this is in contrast to most colitis models, including genetic models and the CD45Rb^high transfer colitis model (36, 37). In our hands, GF animals quickly succumb after exposure to DSS. To test whether Il19 expression during colitis depends, in part, on the presence of an intact microbiota, DSS was given to Oligo-MM¹² mice that are colonized with a consortium of 12 bacterial species that cover the five major phyla of prokaryotes present in the gut and that can be maintained with success over at least five generations (30, 31). Although the animals showed clinical signs of disease (Fig. 3a), Il19 was not expressed in the proximal or distal colon of diseased Oligo-MM¹² mice (Fig. 3b), which is in contrast to SPF mice with colitis (Fig. 3c). Low Il20ra expression was observed in the proximal, but not the distal, colon of Oligo-MM¹² mice with colitis. Il20rb was expressed in the proximal and distal colon of Oligo-MM¹² mice, but DSS colitis did not increase the expression of Il20ra or Il20rb in these mice (Fig. 3b). To further address whether the hygiene status of the animals influences the expression of IL-19, LPS was injected into GF, Oligo-MM¹², and SPF mice. The expression of IL-19 was reduced in GF and Oligo-MM¹² mice compared with SPF mice after injection of LPS (Fig. 3d). Because increased IL-19 expression was observed during colitis after the disruption of the epithelium with DSS, supporting the entrance of microbial-derived products into the host, we next investigated whether the direct activation of macrophages by CD40 ligation is able to induce the expression of IL-19. In this model, CD40 activation leads to colitis in Rag^−/− mice (38). Although injection of the activating anti-CD40 Ab FGK45 into Rag^−/− mice (200 μg per mouse i.p.) caused inflammation, as indicated by a rapid and severe weight loss (Fig. 3e, 3f), it did not lead to increased expression of Il19 in colonic tissue (Fig. 3g).

To further investigate the effect of microbial products on Il19 expression, we stimulated BMDMs and DCs with TLR ligands. M-CSF–differentiated nonstimulated macrophages expressed Il19.

Moreover, stimulation with the TLR4 ligand LPS, the TLR9 ligand CpG, and the TLR2/6 ligand Pam2CSK4, but not the TLR5 ligand flagellin or the TLR3 ligand poly (I:C), increased the expression of Il19 by M-CSF–differentiated macrophages (Fig. 4a). In contrast, the stimulation of DCs, which were purified from GM-CSF–treated bone marrow cell cultures by fluorescence assisted cell sorting (Supplemental Fig. 2a, 2b), did not increase Il19 expression (Fig. 4a). Il20rb was expressed by DCs and macrophages, and its expression was further upregulated by stimulation with LPS or Pam2CSK4, respectively (Fig. 4b). We did not detect expression of Il20ra by macrophages or DCs (data not shown). To further identify the cells that express Il19 and to study the effects of IL-19 on colitis, Il19-tdTomato reporter animals were constructed (Supplemental Fig. 2c). The reporter gene tdTomato is expressed on the Il19 transcript and contains an internal ribosomal entry site to translate into a nonfusion complete tdTomato protein. Flow cytometry of LPS-stimulated BMDMs from Il19-tdTomato mice confirmed tdTomato expression in stimulated macrophages, and the measurement of IL-19 in the culture supernatants by ELISA confirmed that WT animals, but not homozygous Il19-tdTomato mice, produce IL-19 (Supplemental Fig. 2d). The characterization of immune cell populations did not reveal major changes in Il19-tdTomato animals (data not shown). Hence, homozygous Il19-tdTomato mice express tdTomato but lack IL-19 expression. Bone marrow was then isolated from Il19-tdTomato animals and differentiated in the presence of M-CSF, FLT3L, or GM-CSF to generate macrophages, DCs, and mixed cultures, respectively (Fig. 4c). LPS-stimulated macrophages, but not GM-CSF–differentiated cells or FLT3L-differentiated DCs, expressed the fluorescent protein tdTomato, indicating that macrophages, but not DCs, produce IL-19 (Fig. 4c, 4d). We carried out RNA sequencing analysis of LPS-stimulated macrophages derived from heterozygous Il19-tdTomato mice to compare the transcriptome of tdTomato⁻ macrophages with tdTomato⁺ macrophages. RNA sequencing data have been uploaded to the Gene Expression Omnibus (accession number GSE94939) and can be viewed at the following link: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE94939. Unsupervised analysis revealed that 1711 transcripts were downregulated, including the cytochrome P4502S1 (CYP2S1), which is involved in the metabolism of polyunsaturated fatty acids (39), and 1254 transcripts were upregulated. However, no transcript was increased >4-fold in tdTomato⁺ macrophages. The genetic and protein interactions extracted from transcripts that were downregulated >4-fold in tdTomato⁺ cells included, for example, collagens and interactomes involved in cell adhesion and migration, cell motility, and cell–extracellular matrix interactions. RNA sequencing confirmed that Tlr5 expression by BMDMs is very low or not detectable (Supplemental Fig. 3). Il19 expression by BMDMs peaked 6 h after stimulation with LPS and declined within 24 h (Supplemental Fig. 2e). When C57BL/6 mice were challenged i.p. with LPS (2 μg/g body weight), increased expression of Il19 was observed in the colon of WT mice after 4 h, which declined at 8 h (Fig. 4e). tdTomato was expressed by CD64⁺CD11⁻ colonic macrophages isolated 8 h after i.p. injection of LPS into Il19-tdTomato mice (Fig. 4f). Furthermore, LPS stimulation of cLP cells isolated from Il19-tdTomato mice with colitis induced the expression of tdTomato by CD64⁺ macrophages, but not DCs, as determined by flow cytometry (Fig. 4g).

Whether IL-19 has autocrine effects on macrophages is unclear, because different studies give conflicting answers about its influence on the expression of cytokines by macrophages (40, 41). IL-19 signals via the type I IL-20R, formed by IL-20Rα and IL-20Rβ, and results in the activation of STAT3 (12). We did not detect Il20ra expression by BMDMs (data not shown), indicating that these cells do not express a functional receptor for IL-19. In agreement, stimulation of BMDMs with IL-19 did not lead to STAT3 phosphorylation, in contrast to stimulation with IL-10 (Fig. 5a, 5b). Nonstimulated and LPS-stimulated macrophages from C57BL/6 mice did not differ significantly with regard to Il1b, Il6, Il10, or Tnfa expression compared with macrophages from homozygous Il19-tdTomato mice (Fig. 5c–f). Stimulation of WT macrophages with IL-19, as well as treatment of macrophages with IL-19 before LPS stimulation, did not significantly influence the production of Il1b, Il6, Il10, or Tnfa (Fig. 5g–j). Macrophages and monocytes isolated from C57BL/6 and Il19-tdTomato mice with DSS colitis were stimulated with LPS for 24 h, supernatants were collected, and the concentration of IL-1β, IL-6, IL-10 and TNF-α was determined. LPS-stimulated macrophages from Il19-tdTomato mice produced more IL-10 compared with macrophages from WT mice. A major difference in IL-1β, IL-6, or TNF-α production between WT and Il19-tdTomato mice was not observed (Fig. 5k–n). Furthermore, we stained sections from IBD patients with anti-CD64 and anti–IL-20Rα to identify the cells that express IL-20Rα. Immunofluorescence revealed that some CD64⁺ macrophages expressed IL-20Rα (Fig. 6), indicating that colonic macrophages, but not BMDMs, express a functional type I IL-20R.

Because previous reports indicated that members of the IL-20 cytokine family might influence CD4 Th responses (42), naive T cells were differentiated into IFN-γ–secreting Th1 cells, IL-13–secreting Th2 cells, and IL-17A–secreting Th17 cells in the presence of IL-19 (Supplemental Fig. 4a–c). The addition of IL-19 to these cultures did not influence the differentiation of naive T cells into effector cells. When previously differentiated Th1, Th2, and Th17 cells were stimulated with IL-19, the cytokine stimulation did not alter the production of IFN-γ, IL-13, or IL-17A. respectively (Supplemental Fig. 4d–f). Furthermore, significant differences in B cell or T cell numbers or the production of IFN-γ, IL-13, or IL-17A were not observed between C57BL/6 WT and Il19-tdTomato mice with colitis (Supplemental Fig. 4g–i).

Because the expression of IL-19 was increased in mice with DSS colitis, we induced colitis in IL-19–deficient animals. DSS colitis in homozygous Il19-tdTomato mice was attenuated compared with cohoused C57BL/6 WT mice, as indicated by a reduced loss of body weight, increased colon length, and reduced histopathological signs of colitis (Fig. 7a–e). We determined the numbers of B cells, T cells, neutrophils, monocytes, and mononuclear cells in the cLP at day 7 after the onset of DSS treatment. Because there were no significant differences in the total numbers of B cells and T cells (Supplemental Fig. 4g–i), we further analyzed the numbers of mononuclear phagocytes in C57BL/6 and IL-19–deficient animals with colitis. After lineage exclusion (T, B, and NK cells) I-A/I-E⁺ cells were further categorized into macrophages and DCs based on CD11c and CD64 expression (43), whereas I-A/I-E⁻ cells were identified as neutrophils and monocytes based on Ly6C and Ly6G staining (Fig. 7f). It is evident that the reduced cellular infiltration observed in histological sections was due to reduced monocyte numbers in the cLP of colitic Il19-tdTomato mice compared with WT mice (Fig. 7g–i). Moreover, the total numbers of CD64⁺CD11c⁻, but not CD64⁺CD11c⁺ or CD64⁻CD11c⁺, mononuclear cells were reduced in the cLP of Il19-tdTomato mice (Fig. 7j–l).

We then analyzed the production of Il6, Il1b, Tnfa, and Il10 in IL-19–deficient mice with DSS colitis. In the inflamed colon, the percentage of IL-6–producing CD64⁺CD11c⁻ and CD64⁺CD11c⁺ macrophages, but not of CD64⁻CD11c⁺ DCs or monocytes, was higher in WT mice than in Il19-tdTomato animals (Fig. 8a, 8b). Additionally, the total numbers of IL-6–secreting CD64⁺CD11c⁻ and CD64⁺CD11c⁺ macrophages and of monocytes were reduced in Il19-tdTomato mice (Fig. 8c). Significant changes in Il6, Il1b, Tnfa or Il10 expression in total colon between C57BL/6 and Il19-tdTomato mice were not observed (Fig. 8d). In conclusion, DSS colitis is attenuated in IL-19–deficient animals associated with reduced numbers of IL-6–producing macrophages in the cLP.

Cytokines that are produced by macrophages in response to microbial-derived compounds in IBD patients may hinder the resolution of inflammation and facilitate tissue destruction (44). We found an increased expression of Il19, Il20, and Il24 in biopsies of patients with active UC compared with quiescent UC and an increased IL-19 expression in mice with DSS colitis. As a consequence, colitis was attenuated in IL-19–deficient animals.

Our findings indicate that a barrier breach, which allows the entry of bacteria or bacterial-derived products, is needed to induce Il19 expression during colitis. Such a barrier breach occurs in the DSS colitis model, in which the intestinal epithelium is damaged (45), and Il19 was expressed in cells of the cLP. In contrast, when we administered DSS to gnotobiotic animals with a simplified oligo-mouse microbiota, Il19 was not detectable in the colon. It has not been investigated whether Oligo-MM¹² consortium is unable to or whether the microbial pressure in mice with a simplified oligo-mouse microbiota is not sufficient to induce the expression of IL-19 during colitis. Reduced IL-19 expression was observed in GF and Oligo-MM¹² mice compared with SPF animals after LPS injection, suggesting that the production of IL-19 is impaired in GF and Oligo-MM¹² mice. Furthermore, the ligation of APCs with an activating anti-CD40 Ab alone was not sufficient to induce Il19 expression in vivo. In this model, colitis is induced by CD40 activation (38), in contrast to the DSS colitis model, in which disruption of the intestinal barrier and subsequent entrance of microbial-derived products initiates the development of colitis. To confirm that macrophages produce IL-19 after stimulation with microbial-derived compounds, BMDMs were stimulated with TLR ligands, and LPS was injected into mice. Stimulation of BMDMs with the TLR ligands LPS, CpG, and PAM2CSK4, but not poly (I:C), induced production of IL-19 in vitro. Because LPS signals in MyD88-dependent and TRIF-dependent pathways, and poly (I:C) signals only via TRIF, these results suggested that IL-19 production is dependent on MyD88. Injection of LPS into mice confirmed in vivo that IL-19 production by colonic macrophages is also induced by TLR ligands.

Whether microbial-driven IL-19 expression acts on macrophages in an autocrine manner is unclear: some groups have reported increased IL-6 and TNF-α expression by macrophages after stimulation with IL-19 (28), whereas other groups reported reduced IL-6 and TNF-α expression (40). We could not detect expression of IL-20Rα by BMDMs, and IL-19 treatment did not activate the STAT3 pathway and did not regulate Il6 and Tnfa expression, suggesting that these cells did not express a functional type I IL-20R. These results are in line with the findings of Kunz et al. (41), who analyzed the effects of IL-19 on human PBMCs and did not observe any IL-19 signaling in these cells.

In contrast to BMDMs, IL-19 may have direct effects on intestinal macrophages, because human colonic macrophages expressed IL-20Rα, LPS-stimulated macrophages from IL-19–deficient mice produced more IL-10 compared with WT mice, and IL-6 production by colonic macrophages freshly isolated from IL-19–deficient mice with colitis was reduced. In line with these data, different functional properties between BMDMs and intestinal macrophages have been described, such as reduced production of IL-1β and TNF-α by colonic macrophages (46, 47). As a consequence, colitis was attenuated in our study, which is in contrast to a previous study by Azuma et al. (25). Several explanations could account for the discrepancy between these studies. First, we used a novel Il19-tdTomato reporter mouse line instead of the IL-19–deficient mouse line generated by Azuma et al. (25). The Il19-tdTomato reporter mouse is on a C57BL/6 background, whereas the IL-19–deficient mouse line described by Azuma et al. is on a mixed 129/Sv × C57BL/6 background. The different genetic backgrounds of these mouse lines may explain, in part, the discrepancy between our study and the study by Azuma et al. (25). Second, different microbiotas in different animal facilities may have a substantial effect on the outcome of colitis experiments (48). Furthermore, it needs to be considered that the cytokines IL-20 and IL-24 belong to the same cytokine family as IL-19 (10). IL-19 signals through type I IL-20R, whereas IL-20 and IL-24 bind to the type I and type II IL-20Rs, with higher affinity to the type II IL-20R (49). The expression pattern of the type I IL-20R differs from the expression pattern of the type II IL-20R in colonic tissues. IL-19 may act in concert with IL-20 and IL-24, which could explain the different outcomes in different experimental settings.

In conclusion, the current study begins to reveal how IL-19 production is modulated in the intestine during colitis, because microbial-derived compounds induced the expression of IL-19 by macrophages. The deletion of individual members of the IL-20 cytokine family and individual receptor chains is needed to further dissect the importance of this cytokine family in IBD.

We thank the entire staff of the Swiss-IBD Cohort study for supporting this project (2015-05) and all gastroenterologists who enrolled patients into the Swiss-IBD Cohort. We acknowledge Philippe Demougin for help with RNA sequencing and Florian Geier for help with bioinformatics analysis of the RNA sequencing data. We thank Simone Keck for providing reagents and support for CD64 immunofluorescence staining and Biocytogen for generation of Il19-tdTomato mice. This work is part of the Ph.D. thesis of A.S. and part of the M.D. thesis of I.L.

The authors have no financial conflicts of interest.