The immunobiology of colitis and cholangitis in interleukin-23p19 and interleukin-17a deleted dominant negative form of transforming growth factor beta receptor type ii mice


  • Potential conflict of interest: Nothing to report.

  • Financial support was provided by the National Institutes of Health (grant no.: DK090019).


Dominant negative form of transforming growth factor beta receptor type II (dnTGFβRII) mice, expressing a dominant negative form of TGFβ receptor II under control of the CD4 promoter, develop autoimmune colitis and cholangitis. Deficiency in interleukin (IL)-12p40 lead to a marked diminution of inflammation in both the colon and the liver. To distinguish whether IL-12p40 mediates protection by the IL-12 or IL-23 pathways, we generated an IL-23p19−/− dnTGFβRII strain deficient in IL-23, but not in IL-12; mice were longitudinally followed for changes in the natural history of disease and immune responses. Interestingly, IL-23p19−/− mice demonstrate dramatic improvement in their colitis, but no changes in biliary pathology; mice also manifest reduced T-helper (Th)17 cell populations and unchanged IFN-γ levels. We submit that the IL-12/Th1 pathway is essential for biliary disease pathogenesis, whereas the IL-23/Th17 pathway mediates colitis. To further assess the mechanism of the IL-23-mediated protection from colitis, we generated an IL-17A−/− dnTGFβRII strain deficient in IL-17, a major effector cytokine produced by IL-23-dependent Th17 cells. Deletion of the IL-17A gene did not affect the severity of either cholangitis or colitis, suggesting that the IL-23/Th17 pathway contributes to colon disease in an IL-17-independent manner. These results affirm that the IL-12/Th1 pathway is critical to biliary pathology in dnTGFβRII mice, whereas colitis is caused by a direct effect of IL-23. (HEPATOLOGY 2012)

Murine strains with a deficiency in specific cytokine pathways are important tools for investigating the mechanism of immunopathogenesis of autoimmunity. Mice transgenic for directed expression of a dominant negative form of transforming growth factor beta receptor type II (dnTGFβRII), under the control of the CD4 promoter lacking the CD8 silencer, spontaneously develop an inflammatory bowel disease (IBD).1 In addition, dnTGFβRII mice develop an autoimmune biliary ductular disease with strong similarity to human primary biliary cirrhosis (PBC), an organ-specific autoimmune disease characterized by destruction of intrahepatic small bile duct biliary epithelial cells.2, 3 Deletion of interleukin (IL)-12p40 in dnTGFβRII mice, which results in deficiency of both IL-12 and IL-23, leads to marked diminution of inflammation in both the liver and the colon.4 In efforts to distinguish between the roles of the cytokine pathways mediated by IL-12 and IL-23 in the pathogenesis of liver and colon diseases in dnTGFβRII mice, we generated two new mutant strains of dnTGFβRII mice: an IL-23p19−/− strain, which is deficient in IL-23, but not other members of the IL-12 family, and an IL-17A−/− strain, which is deficient in IL-17, a major effector cytokine produced by IL-23-dependent T-heleper (Th)17 cells.5 The results of our study demonstrate that though deletion of IL-23p19 eliminates colitis, but not cholangitis, the deletion of IL-17A had no significant effect on either cholangitis or colitis. Therefore, the IL-12/Th1, but not the IL-23/Th17, pathway is important for autoimmune cholangitis. Our data also suggest that the IL-23/Th17 pathway contributes to colon disease in an IL-17-independent manner.


Ab, antibody; AMAs, antimitochondrial autoantibodies; ANA, antinuclear antibody; ANOVA, analysis of variance; BSA, bovine serum albumin; CXCL2, chemokine (C-X-C motif) ligand 2; dnTGFβRII, dominant negative form of transforming growth factor beta receptor type II; ELISA, enzyme-linked immunosorbent assay; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; H&E, hematoxylin and eosin; IBD, inflammatory bowel disease; IFN-γ, interferon-gamma; Ig, immunoglobulin; IL, interleukin; mAb, monoclonal antibody; MIP-2, macrophage inflammatory protein-2; MPO, myeloperoxidase; MNCs, mononuclear cells; mRNA, messenger RNA; PBC, primary biliary cirrhosis; PBS, phosphate-buffered saline; PCR, polymerase chain reaction; PDC-E2, pyruvate dehydrogenase E2 complex; Th, T-heleper cells; TNF-α, tumor necrosis factor alpha.

Materials and Methods


The dnTGFβRII colony on a B6 background (B6.Cg-Tg(Cd4-TGFBR2)16Flv/J) was maintained at the University of California at Davis animal facility (Davis, CA).3 B6 (IL-17A−/−) mice and B6 (IL-23p19−/−) mice were generous gifts from Dr. Yoichiro Iwakura (University of Tokyo, Tokyo, Japan) and Dr. Frederic J. de Sauvage (Genetech, South San Francisco, CA), respectively.6 IL-23p19−/− dnTGFβRII mice were generated as previously described.3, 4 Briefly, male dnTGFβRII mice were mated with female IL-23p19−/− mice to obtain IL-23p19+/− dnTGFβRII mice, which were subsequently back-crossed with female IL-23p19−/− mice to obtain IL-23p19−/− dnTGFβRII mice. Parental dnTGFβRII and the derived IL-23p19−/− dnTGFβRII mice were genotyped at 3-4 weeks of age to confirm the dnTGFβRII and IL-23p19−/− genes in their genomic DNA.3 IL-17A−/− dnTGFβRII mice were similarly generated. All mice were fed sterile rodent Helicobacter Medicated Dosing System (three-drug combination) diets (Bio-Serv, Frenchtown, NJ) and maintained in individually ventilated cages under specific pathogen-free conditions. Sulfatrim (Hi-Tech Pharmacal, Amityville, NY) was delivered through drinking water. At 24 weeks of age, animals were sacrificed to collect sera, spleen, liver, and colon tissues. Experimental protocols were approved by the University of California Animal Care and Use Committee.


The liver and colon from sacrificed mice were fixed in 4% paraformaldehyde, embedded in paraffin, cut into 4-μm sections, deparaffinized, stained with hematoxylin and eosin (H&E), and evaluated using a light microscopy.4 Liver histopathology was graded as follows: 0, no inflammation (or bile duct damage); 1, mild inflammation (or bile duct damage); 2, moderate inflammation (or bile duct damage); and 3, severe inflammation (or bile duct damage).4 Colon histopathology was graded as follows: 0, no significant changes; 1, minimal scattered mucosal inflammatory cell infiltrates, with or without minimal epithelial hyperplasia; 2, mild scattered to diffuse inflammatory cell infiltrates, sometimes extending into the submucosa and associated with erosions, with mild to moderate epithelial hyperplasia and mild to moderate mucin depletion from goblet cells; 3, moderate inflammatory cell infiltrates that were sometimes transmural, with moderate to severe epithelial hyperplasia and mucin depletion; and 4, marked inflammatory cell infiltrates that were often transmural and associated with crypt abscesses and occasional ulceration, with marked epithelial hyperplasia, mucin depletion, and loss of intestinal glands.7


To monitor neutrophil infiltration, sections of colon were stained for myeloperoxidase (MPO), as previously described.8 Colon sections were blocked using 20% (v/v) normal swine serum in Tris-buffered saline for 30 minutes and stained for MPO using a rabbit anti-MPO antibody (Ab) (0398; Dako, Carpinteria, CA), followed by staining with biotin-labeled antirabbit Ab (E0413; Dako). Avidin-biotin peroxidase (K0377; Dako) and histogreen (LINARIS Biologische Produkte GmbH, Mannheim, Germany) were used for color development.

Flow Cytometry.

Liver- and colon-infiltrating mononuclear cells (MNCs) were isolated as previously described.9, 10 Cells were resuspended in staining buffer (0.2% bovine serum albumin [BSA], 0.04% ethylene diamine tetraacetic acid, and 0.05% sodium azide in phosphate-buffered saline [PBS]), divided into 25-μL aliquots, and incubated with antimouse FcR-blocking reagent (eBioscience San Diego, CA) for 15 minutes at 4°C. Cells were washed and stained for 30 minutes at 4°C with cocktails containing combinations of fluorochrome-conjugated monoclonal antibody (mAb) for the cell-surface markers, CD4, CD8a, CD44, CD69, natural killer 1.1 (BioLegend, San Diego, CA), and T-cell receptor beta (eBioscience). To determine T-cell activation, mAbs for CD44 and CD69 were used.1, 11 Immunoglobulin (Ig)G isotype Abs with matching conjugates were used in parallel as negative controls. Cells were then washed with PBS containing 0.2% BSA. A FACScan flow cytometer (BD Immunocytometry Systems, San Jose, CA), upgraded for the detection of five colors by Cytek Development (Fremont, CA), was used to acquire data, which were analyzed with Cellquest PRO software (BD Immunocytometry Systems).

Serum Igs, Antimitochondrial Antibodies, and Antinuclear Antibodies (Gp210/Sp100).

Levels of serum IgG, IgM, and IgA were determined using a murine IgG, IgM, and IgA enzyme-linked immunosorbent assay (ELISA) quantitation kit (Bethyl Laboratories, Montgomery, TX). Serum antimitochondrial antibodies (AMAs) were detected using an ELISA assay based on recombinant murine pyruvate dehydrogenase E2 complex (PDC-E2), as previously described.12 Immunoreactivity was determined by measuring the optical density at 450 nm after incubation with 100 μL of tetramethylbenzidine substrate (BD Biosciences, San Jose, CA) for 30 minutes. Serum antinuclear antibodies (ANAs) (Gp210/Sp100) were measured by QUANTA Lite Gp210/Sp100 (INOVA Diagnostics, Inc., San Diego, CA).

Cytokine Analysis.

For analysis of cytokines secreted from cultured CD4 T cells, CD4 T cells were isolated from spleen MNCs with CD4 (L3T4) MicroBeads (Miltenyi Biotec Inc., Auburn, CA). Aliquots of 2.0 × 105 CD4 T cells were cultured in 96-well round-bottomed plates in 200 μL of RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum (Gibco-Invitrogen Corp., Grand Island, NY), 100 μg/mL of streptomycin, 100 U/mL of penicillin, and 0.5 μg/mL each of anti-CD3 (BioLegend) and anti-CD28 (BioLegend). Cultures were incubated for 72 hours at 37°C in a humidified 5% CO2 incubator, then centrifuged to collect supernatants.

For analysis of cytokine levels in tissue, total protein was extracted from 30 mg of frozen liver or colon tissues by homogenization in T-Per Tissue Protein Extraction buffer (Thermo, Rockford, IL) containing a protease inhibitor cocktail (Roche, Indianapolis, IN). The homogenized tissue suspension was centrifuged at 12,000×g for 20 minutes at 4°C, and the supernatant was stored at −80°C until use. The total protein concentration of each sample was measured using the BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA).

Levels of IL-17A, tumor necrosis factor alpha (TNF-α), IL-6, IL-10, IL-4, IL-2, and interferon-gamma (IFN-γ) in sera, cell-culture supernatant, and tissue lysates were measured with a cytokine bead array assay using the Mouse Th1/Th2/Th17 Cytokine Kit (BD Biosciences). Levels of IL-22 and macrophage inflammatory protein-2/chemokine (C-X-C motif) ligand 2 (MIP-2/CXCL2) were measured using the Quantikine mouse Mouse/Rat IL-22 Immunoassay kit and mouse CXCL2/MIP-2 kit (R&D Systems, Minneapolis, MN).

For measuring levels of cytokine gene messenger RNA (mRNA), total RNA was extracted from frozen colon tissues using the RNeasy Plus Mini Kit (QIAGEN, Venlo, The Netherlands), and complementary DNA was synthesized by Superscript III reverse transcriptase (Invitrogen), according to the manufacturer's protocols. The real-time polymerase chain reaction (PCR) system (ViiATM 7; Applied Biosystems, Foster City, CA) was used for quantitative PCR. The primers used were as follows: 5'-TCCAGAAGGCCCTCAGA CTA-3' (forward) and 5'-AGCATCTTCTCGACCCT GAA-3' (reverse) for mouse IL-17A, 5'-TAGCCAA GACTGTGATTGCGG-3' (forward) and 5'-AGAC ATCTCCTCCCATCAGCAG-3' (reverse) for mouse IFN-γ, 5'-AAGCCTGTAGCCCACGTCGTA-3' (forward) and 5'-AGGTACAACCCATCGGCTGG- 3' (reverse) for mouse TNF-α, 5'-TCCATCCAGTTGCC TTCTTG-3' (forward) and 5'-TTCCACGATTTCCC AGAGAAC-3' (reverse) for mouse IL-6, 5'- CATGGC CTTCCGTGTTCCTA-3' (forward) and 5'-CCTGC TTCACCACCTTCTTGAT-3' (reverse) for mouse glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Amplification was performed for 40 cycles in a total volume of 16 μL, and products were detected using SYBR Green. The relative expression level of each target gene was determined by normalizing its mRNA level to the internal control gene GAPDH.

Statistical Analysis.

Mann-Whitney's two-tailed unpaired, one-way analysis of variance (ANOVA) followed by Bonferroni's multiple-comparisons test, or Fisher's exact test were used for different analyses, as appropriate. P values <0.05 were considered statistically significant.


Depletion of IL-23p19 Ameliorated Colitis in dnTGFβRII Mice.

Because 5-month-old dnTGFβRII mice develop IBD, we examined IL-23p19−/− dnTGFβRII mice for colitis at 24 weeks of age. Colonic hyperplasia, crypt abscesses, and epithelial ulcers were readily observed in dnTGFβRII mice, but not in IL-23p19−/− mice (Fig. 1A). Colon weight and thickness, which correlates with severity of colitis, were significantly decreased in IL-23p19−/− dnTGFβRII mice, compared to age-matched dnTGFβRII mice (Fig. 1B). Colonic infiltration of total MNCs, as well as total and activated CD4 T cells, was significantly decreased in IL-23p19−/− mice, compared to dnTGFβRII mice, whereas no differences were observed in the levels of infiltrating CD8 T-cell populations (Fig. 2). MPO+ cells appeared to accumulate around the ulcer region in dnTGFβRII mice, whereas only a few of these cells were observed in the colon mucosal layer of IL-23p19−/− dnTGFβRII (Fig. 1A). In addition, a relatively higher incidence of dysplasia was observed in dnTGFβRII mice than IL-23p19−/− mice (Fig. 1A,C).

Figure 1.

Colitis is improved in IL-23p19−/− dnTGFβRII mice, compared to parental dnTGFβRII mice. (A) Representative histological staining of colon sections. Colonic hyperplasia, crypt abscess, and dysplasia were frequently observed in dnTGFβRII mice (wt), but not IL-23p19−/− dnTGFβRII mice (p19−/−). MPO+ cells accumulated around the ulcer region in dnTGFβRII mice, whereas few foci of MPO+ cells were noted within the mucosal layer in IL-23p19−/− dnTGFβRII mice. (B) Colon weight and colon wall thickness. (C) Colitis score and incidence rate of dysplasia. **P < 0.01; ***P < 0.001, determined using Mann-Whitney's two-tailed unpaired test.

Figure 2.

Numbers of total MNCs, CD4 T cells, and CD8 T cells in colon tissues from IL-23p19−/− mice (n = 6) and parental dnTGFβRII mice (n = 6), determined by flow cytometry. *P < 0.05, determined using Mann-Whitney's two-tailed unpaired test.

Depletion of IL-23p19 Did Not Suppress Autoimmune Cholangitis in dnTGFβRII Mice.

We next compared liver histology in IL-23p19−/− dnTGFβRII mice and dnTGFβRII mice at 24 weeks of age. There was no significant difference in the levels of inflammatory portal lymphoid cell infiltration and bile duct damage between the two mouse strains (Fig. 3A,B). In addition, the numbers of intrahepatic T cells, including the total CD8 T-cell population and activated CD8 T cells (defined by CD69+ and CD44+ phenotypes,1, 11 known to be pathogenic in liver disease of dnTGFβRII mice,13 did not differ significantly between the two mouse strains (Fig. 3C). These results indicate that the deficiency in IL-23p19 did not protect dnTGFβRII mice from developing liver disease.

Figure 3.

Cholangitis in livers of IL-23p19−/− mice and dnTGFβRII mice. (A) H&E-stained liver sections. (B) Liver portal inflammation and bile duct damage scores in IL-23p19−/− mice (n = 13) and parental dnTGFβRII mice (n = 13). (C) Numbers of total MNCs, CD8 T cells, and CD44+CD8 T cells in livers of dnTGFβRII mice (n = 7) and IL-23p19−/− dnTGFβRII mice (n = 7). No significant difference was found in any of these comparisons between the two strains (Mann-Whitney's two-tailed unpaired test).

Serum Levels of Ig, AMA, and ANA in IL-23p19−/− dnTGFβRII Mice.

To address whether IL-23 has a role in autoantibody induction, serum levels of AMA and ANA as well as those for total IgG, IgM, and IgA were measured by ELISA. Levels of IgG in IL-23p19−/− dnTGFβRII mice was higher than in healthy B6 mice, but were comparable with those of dnTGFβRII mice (Fig. 4). In contrast, levels of IgM and IgA in IL-23p19−/− mice were significantly higher than that of dnTGFβRII mice. In IL-23p19−/− dnTGFβRII mice, levels of AMA and anti-SP100 ANA were significantly higher than those of dnTGFβRII mice (P < 0.05), whereas levels of anti-GP210 ANA were significantly lower than that of dnTGFβRII mice (P < 0.001), but still significantly higher than that of B6 mice (P < 0.001) (Fig. 4). These data indicate that in this model of autoimmunity, IL-23 is, in general, not critical for autoantibody production, but has opposite effects on levels of different autoantibodies.

Figure 4.

Serum levels of AMA (PDC-E2), ANA (GP210 and SP100), and total Ig in dnTGFβRII mice (n = 11) and IL-23p19−/− dnTGFβRII mice (n = 11), as well as healthy B6 mice (n = 8) served as the negative control. Horizontal bars represent median values. *P < 0.05; **P < 0.01; ***P < 0.001, by one-way ANOVA followed by Bonferroni's multiple-comparisons test.

Effect of IL-23p19 Depletion on Serum Cytokine Levels.

We measured a panel of proinflammatory cytokines in sera from IL-23p19−/− and the parental dnTGFβRII mice. Sera from IL-23p19−/− mice, as compared with dnTGFβRII sera, contained significantly lower levels of most of the cytokines tested, which included IL-17A, TNF-α, IL-6, IL-22, IL-10, IL-4, IL-2, and MIP-2/CXCL2 (Table 1). The only exception was IFN-γ, which was not reduced in IL-23p19−/− mice, indicating that deletion of IL-23p19 does not affect the differentiation of Th1 cells.

Table 1. Serum Levels of Inflammatory Cytokines
CytokineConcentration (pg/mL)P Value
  • Serum samples from IL-23p19−/− dnTGFβRII mice (n = 10) and dnTGFβRII mice (n = 10) were tested. Data are presented as an average ± SD for each group.

  • *

    P < 0.05;

  • **

    P < 0.01;

  • ***

    P < 0.001 by Mann-Whitney's two-tailed unpaired test.

IL-17A34.4 ± 8.08.1 ± 3.1**
IFN-γ37.7 ± 8.943.5 ± 19.7 
TNF-α61.6 ± 7.328.0 ± 6.6**
IL-673.3 ± 11.512.9 ± 2.3***
MIP-263.4 ± 11.032.4 ± 3.7*
IL-227.6 ± 1.42.5 ± 1.2*
IL-10151.2 ± 35.542.5 ± 16.2*
IL-414.9 ± 3.05.3 ± 1.3**
IL-211.5 ± 3.12.7 ± 0.8*

Decreased Th17 Cell Population in IL-23p19−/− dnTGFβRII Mice.

To determine whether deletion of the IL-23p19 gene influences the generation of cytokine-based Th1, Th2, and Th17 cell populations, CD4 T cells isolated from spleens of IL-23p19−/− and the parental dnTGFβRII mice were cultured with anti-CD3/CD28 Ab for 3 days, and levels of secreted IFN-γ, IL-4, and IL-17A were measured in supernatant fluid. Levels of secreted IL-17A were significantly reduced in IL-23p19−/− mice, compared to dnTGFβRII mice (22.4 ± 3.6 pg/mL in IL-23p19−/− mice versus 4.7 ± 0.9 pg/mL in dnTGFβRII mice; P < 0.01); levels of IL-4 and IFN-γ were not significantly different between the two strains. These data suggest that deletion of IL-23p19 reduced the population of Th17 cells, but not Th1 or Th2 cells, in the spleen.

Deletion of IL-23p19 Reduced IFN-γ, TNF-α, and IL-6 in Colon but Not in Liver.

We next compared levels of select inflammatory cytokines in colon and liver tissues from IL-23p19−/− and parental dnTGFβRII mice. Altough deletion of IL-23p19 resulted in a significant decrease in levels of IFN-γ, TNF-α, and IL-6 in the colon, there was no detectable change of these cytokines in the liver (Fig. 5A). In contrast, levels of IL-17A were increased in the colo,n but decreased in the liver, tissues from IL-23p19−/− mice. The different cytokine levels in the colon of these two mouse strains are in agreement with their mRNA levels in the colon (Fig. 5B).

Figure 5.

Expression of cytokines in liver and colon tissues. (A) Cytokines in whole protein lysates of liver and colon tissues from dnTGFβRII mice (n = 12) and IL-23p19−/− mice (n = 12). (B) Relative levels of cytokine gene mRNA in colon tissues of dnTGFβRII mice (n = 6) and IL-23p19−/− mice (n = 6). *P < 0.05; **P < 0.01; ***P < 0.001, determined using Mann-Whitney's two-tailed unpaired test.

Depletion of IL-17A Did Not Affect the Severity of Autoimmune Cholangitis or Colitis in dnTGFβRII Mice.

To determine whether IL-17 was critical for the pathogenesis of autoimmune liver or colon diseases in dnTGFβRII mice, we generated IL-17A−/− dnTGFβRII mice. Histological examination of liver and colon sections detected no significant differences in either levels of lymphoid cell infiltration and/or tissue damage in both liver and colon tissues between IL-17A−/− mice and paternal dnTGFβRII mice (Fig. 6A,B). These results indicate that IL-17A is not critical for the spontaneous development of either autoimmune cholangitis or colitis in dnTGFβRII mice.

Figure 6.

Histological evidence of cholangitis and colitis in IL-17A−/− dnTGFβRII mice and parental dnTGFβRII mice. (A) Representative H&E-stained liver and colon sections. Upper panels demonstrate lymphoid cell infiltration within the portal tract area around a damaged bile duct. Lower panels demonstrate diffused lymphoid cell infiltration from the mucosal layer into the muscular layer. (B) Scoring of liver portal inflammation, bile duct damage, and colitis in IL-17A−/− dnTGFβRII mice (IL-17A−/−), compared to parental dnTGFβRII mice (wt). No significant differences were detected in any of the comparisons (Mann-Whitney's two-tailed unpaired test).


Previously, we observed that deletion of IL-12p40 protected dnTGFβRII mice from the spontaneous development of autoimmune disease in both the liver and the colon.4 Because IL-12p40 is a subunit shared by IL-12 and IL-23, deletion of this subunit disrupts both the IL-12/Th1 pathway and IL-23/Th17 pathway. The first aim of this study was to examine the role of IL-23 in liver and colon diseases of the dnTGFβRII mouse model by deleting p19 of the IL-23 heterodimer, which is unique to this cytokine in the IL-12 family. Although IL-12 is required for the development of IFN-γ-producing Th1 cells, IL-23 induces the differentiation of naïve CD4 T cells into a highly pathogenic helper T-cell population, termed Th17, that produces IL-17A, IL-17F, IL-6, and TNF, but not IFN-γ or IL-4.5 Several previous studies have suggested a potential link between IL-17 and PBC.14-16 Therefore, the second strategy we used in the current study was to delete the gene encoding IL-17A in efforts to examine whether this cytokine contributes to autoimmune pathogenesis in dnTGFβRII mice.

Results from these studies demonstrate that disrupting the IL-23/Th17 pathway by deleting IL-23p19 abolished colitis, but had no detectable effect on the severity of cholangitis in dnTGFβRII mice, indicating that the IL-23/Th17 axis is involved in the pathogenesis of autoimmune colitis, but not in the cholangitis of this mouse model. However, deletion of the IL-17 gene from dnTGFβRII did not affect either colitis or cholangitis, indicating that IL-17 is not a key factor in the pathogenic IL-23/Th17 axis in the spontaneous development of colon disease of the dnTGFβRII mouse strain. Of note, deletion of IL-23 resulted in increased levels of AMA and anti-SP100, but decreased levels of anti-GP210. The mechanism for these differential effects of IL-23 should be addressed in future studies.

The autoimmune cholangitis that developed in the IL-23p19−/− mice was associated with an intact IL-12/Th1 pathway, as indicated by the high levels of IFN-γ in this mouse strain. In contrast, cholangitis did not develop in IL-12p40−/− mice that lack the IL-12/Th1 pathway.4 Taken together, these results confirm that IL-12/Th1 immunity is necessary and sufficient for the development of cholangitis in dnTGFβRII mice. We have recently reported that adoptive transfer of CD8 T cells from dnTGFβRII into B6/Rag1−/− mice led to liver histopathology similar to that in donor mice. In contrast, adoptive transfer of CD4 T cells predominantly induced IBD in recipient mice.13 These data demonstrated that in dnTGFβRII mice, CD8 T cells are the major pathogenic effector of cholangitis, whereas CD4 T cells are involved in IBD. This is in agreement with our current finding that whereas comparable levels of CD8 T cells are present in liver tissues of IL-23p19−/− and dnTGFβRII mice, both develop cholangitis, and that protection against colitis in IL-23p19−/− mice was associated with reduced numbers of total and activated CD4 T cells, but not CD8 T cells, in the colon. These findings further support the organ-specific pathogenic role of CD4 and CD8 T cells in dnTGFβRII mice. dnTGFβRII mice were completely protected from autoimmune diseases in both the liver and colon only when the IL-12/Th1 pathway was eliminated by deletion of IL-12p40.4

It has been previously shown that the IL-23/Th17 pathway plays a key role in T-cell-mediated IBD and other autoimmune diseases in murine models that either involved cytokine gene knockouts or Ab treatment in mice.17-26 Our current study in dnTGFβRII mice showed that deletion of the IL-23p19 gene resulted in a marked reduction of the Th17 population in the spleen, which is associated with the prevention of colitis. However, deletion of the IL-17 gene did not prevent colitis, suggesting that the pathogenic effect in the colon of dnTGFβRII mice was not mediated by IL-17. Actually, levels of IL-17 cytokine and mRNA in the colon of IL-23p19−/− mice was even higher than those in dnTGFβRII mice, despite the fact that colitis was present in the latter, but not in the former. It is important to note that in addition to the synthesis of IL-17, Th17 cells are also a major source for a number of other cytokines, including IL-6.5

One of the most prominent features in the cytokine profile of IL-23p19−/− mice is the significant decrease in the levels of IL-6 in both serum and colon (Table 1; Fig. 6). This is in agreement with the previously reported role of IL-23-dependent IL-6 in the development of colon inflammation, as shown in other animal models of IBD.19, 27-30 It was recently observed that IL-6 levels were elevated in active IBD patients at diagnosis and during therapy.28 It has also been suggested that IL-23 might directly activate a subset of macrophages and dendritic cells expressing the IL-23 receptor, resulting in the production of inflammatory mediators, such as TNF-α, IL-6, and IL-1.25 Of note, using our dnTGFβRII mice model, we recently reported that depletion of IL-6 significantly improved colitis, but exacerbated autoimmune cholangitis in the liver.31 These studies indicate that the role of IL-6 in the pathogenesis of organ-specific autoimmune diseases is also different between the liver and colon. These data should become a major consideration in the use of anticytokine therapy in the treatment of organ-specific autoimmune diseases. We note recent data from our laboratory on the therapeutic manipulation of this and similar models of autoimmune cholangitis.12

It has been known for some time that individuals with IBD have a 10- to 40-fold increased risk of developing colorectal cancer, compared with the general population. This is in agreement with the fact that colitis-associated cancer frequently develops from persistently inflamed mucosa and progresses through dysplasia to adenocarcinoma, following an “inflammation-dysplasia-carcinoma sequence” that contrasts the “adenoma-carcinoma sequence” of sporadic colorectal cancer. Therefore, effective anti-inflammatory treatment, such as infliximab therapy, could reduce the development of colorectal dysplasia and cancer in IBD.32-34 Although colonic dysplasia was frequently observed in dnTGFβRII mice (Fig. 1A), deletion of IL-23p19 reduced the incidence of dysplasia (Fig. 1C), suggesting that immunotherapies aimed at blocking the IL-23 pathway26 could prevent IBD-related colon cancer.

In summary, our studies demonstrate that deletion of IL-23p19 improved colitis and reduced the rate of colonic dysplasia, but had no effect on cholangitis, in dnTGFβRII mice. These findings confirm that in this mouse model, the IL-12/Th1 pathway is critical to biliary pathology, whereas colitis is caused by a direct effect of IL-23. This study demonstrates that disruption of a pathway with a global effect, such as transforming growth factor beta signaling in CD4 T cells, leads to pathogenesis in different sites with distinct immune mechanisms. Therefore, care needs to be taken before the institution of immunotherapeutic strategies for organ-specific autoimmune diseases, which should be tailored to address different targets in each disease.


The authors thank Katsunori Yoshida, Thomas P. Kenny, Hajime Tanaka, and Chen-yen Yang for their technical support in this experiment. The author also thank Ms. Nikki Phipps for her support in preparing this article.