Abstract
Anti–tumor necrosis factor (TNF) therapy appears to be effective in the treatment of Crohn’s disease (CD), a chronic inflammatory disease of the gastrointestinal tract. However, the mechanisms involved are not completely understood.
Fifty-seven active CD patients were enrolled, and cytokine profiles in colonic biopsies of patients with active CD receiving anti-TNF monoclonal antibody (mAb) (infliximab [IFX]) treatment were determined using quantitative real-time polymerase chain reaction (qRT-PCR). Colonic biopsies of active CD patients and healthy donors were cultured with IFX in vitro, and cytokine profiles were measured by qRT-PCR. Peripheral blood (PB)–CD4+ T cells were stimulated with anti-CD3 and anti-CD28 mAbs in the presence of human immunoglobin (HIg), IFX, recombinant human TNF-α converting enzyme (rhTACE), and aryl hydrocarbon receptor (AhR) inhibitor (CH-223191), respectively, to determine interleukin (IL)-22 expression by CD4+ T cells. Caco2 cells were also utilized to study their potential role in modulating epithelial cell barrier repairs in vitro.
IFX therapy markedly upregulated IL-22 mRNA expression in the gut mucosa of CD patients. In vitro treatment with IFX greatly promoted CD CD4+ T cells to express IL-22, which was inhibited by rhTACE, indicating that reverse signaling through binding to membrane-bound TNF mediates anti-TNF-induced IL-22 expression of CD CD4+ T cells. However, blockade of AhR markedly inhibited anti-TNF-induced IL-22+CD4+ T (Th22) cell differentiation in CD patients. Moreover, treatment with IL-22 induced intestinal epithelial cell expression of tight junction proteins (eg, claudin1 and ZO-1) and facilitated transepithelial resistance, indicating that IL-22 protects intestinal mucosa from inflammation via maintenance of epithelial barrier integrity.
Our results uncover a novel mechanism whereby anti-TNF therapy upregulates IL-22 production in CD patients through promoting Th22 cell differentiation and contributes to intestinal epithelial barrier repairs.
INTRODUCTION
Crohn’s disease (CD) is a type of inflammatory bowel disease (IBD) characterized by chronic or progressive inflammatory conditions in the gastrointestinal tract and colonic mucosa. Although its etiology and pathogenesis remain unclear, it is believed that aberrant innate and adaptive mucosal immune response to luminal commensal microbiota contribute to the pathogenesis of IBD.1 Data from animal models and CD patients have demonstrated that intestinal mucosal inflammation is predominately mediated by type 1 helper T cells (Th1) and Th17 cells.2 Th17 cells have been found to be markedly infiltrated in the inflamed mucosa of active IBD patients, and the levels of Th17 cell-related cytokines (eg, interleukin [IL]-17A, IL-22, IL-23) are correlated with the disease activity of IBD.3 Our previous work has demonstrated that expression of IL-17A, IL-21, and IL-23 is significantly upregulated in inflamed areas of IBD patients and profoundly associated with the pathogenesis of IBD.4, 5 In contrast, IL-22 appears to play a protective role in the induction of intestine mucosal inflammation through promoting epithelial barrier repairs in a variety of experimental colitis models in mice.6, 7 IL-22+CD4+ T cells have recently been termed “Th22 cells”, characterized by producing high levels of IL-22 but neither IL-17A nor interferon (IFN)-γ.8 Th22 cells express low or undetectable levels of the Th17 cell–related transcription factor retinoic acid–related orphan receptor γt (RORγt). In addition, recent data have demonstrated that the type 3 innate lymphoid cells (ILC3), as well as natural killer cell subsets, produce IL-22 in the human intestine.9, 10 Moreover, IL-22-producing neutrophils have also been identified in intestinal mucosa of mice.11, 12 Aryl hydrocarbon receptor (AhR) is reported to be critical for IL-22 expression and functions as a regulator for either IL-22 expression or ILC3 and Th22 cell development.13 Recently, Th22 cells have been demonstrated to be associated with the pathogenesis of several autoimmune diseases such as Hashimoto’s thyroiditis, rheumatoid arthritis, and Behcet’s disease.14–16 However, the role of Th22 in the pathogenesis of IBD is still illusive.
Accumulating lines of evidence have established that tumor necrosis factor (TNF) is an essential cytokine for regulating proinflammatory responses and is associated with several immune-mediated diseases, including IBD. Secretion of TNF has been attributed largely to T cells and macrophages,17, 18 initially expressed as a transmembrane protein, and the membrane-bound TNF form transduces reverse signaling upon binding to anti-TNF antibody or soluble cognate receptors, such as TNFR1 and TNFR2.19 Previous work has demonstrated that the reverse signaling of membrane-bound TNF regulates the inflammatory response acting primarily locally20 and that the metalloproteinase tumor necrosis factor–α converting enzyme (TACE), also referred to as ADAM17, is able to cleave the extracellular domain of TNF to release soluble TNF.21 Importantly, TACE-dependent release of soluble TNF is reported to be involved in the pathology of TNF-mediated inflammatory diseases, including colitis.22, 23 It has been well established that targeted therapies directed against TNF can inhibit inflammation and support the wound healing of the intestinal mucosa. As a crucial member of anti-TNF monoclonal antibodies (mAbs), infliximab (IFX) binding soluble TNF and transmembrane TNF (tmTNF) specifically with high affinity has provided a vital strategy in the management of active CD patients. However, the mechanism in anti-TNF therapy in CD patients remains unclear, in addition to neutralizing TNF.
In the present study, we determined IL-22 expression in active CD patients during anti-TNF therapy and the potential roles in modulating intestinal mucosal immune responses. We observed that stimulation with anti-TNF mAb significantly enhanced IL-22 production of CD CD4+ T cells, which was significantly reversed by rhTACE treatment. Anti-TNF therapy facilitated Th22 cell differentiation by CD CD4+ T cells and produced high levels of IL-22. Blockade of AhR significantly reduced Th22 cell differentiation and IL-22 production by CD CD4+ T cells, suggesting that AhR is associated with anti-TNF-induced IL-22 production of CD4+ T cells. Importantly, treatment with IL-22 in vitro could alleviate epithelial cell injuries characterized by promoting tight junction protein expression and maintenance of epithelial integrity. Collectively, these data reveal a novel mechanism whereby anti-TNF therapy promotes Th22 cell differentiation through the induction of reverse signaling in an AhR-dependent manner.
METHODS
Anti-TNF Therapy in Active CD Patients
Fifty-seven patents with active CD were recruited and intravenously treated with anti-TNF mAb (IFX; Cilag AG, Schaffhausen, Switzerland) at a dose of 5 mg/kg at weeks 0, 2, and 6 according to the manufacturer’s instructions. We monitored all patients weekly during the follow-up period and collected intestinal biopsies at weeks 0 and 12 after initial administration.5 Endoscopic biopsies were collected from macroscopically inflamed and normal terminal ileum and colon. We assessed the efficacy of IFX at week 12 after the first infusion when patients visited for continuous treatment of IFX. Disease activity of CD was evaluated at weeks 0 and 12 according to the Crohn’s Disease Activity Index (CDAI), biochemical parameters such as serum C-reactive protein (CRP) and calprotecin, and endoscopic examination. We defined CD patients with CDAI scores of <150 points as being in clinical remission, and a decrease in the CDAI score ≥70 points in comparison with the baseline index as clinical response to IFX.4, 24 Endoscopic response to IFX therapy was evaluated according to the Simple Endoscopic Severity for CD (SES-CD) at week 12 after the first infusion, and endoscopic variables were also assessed as described elsewhere.25 Active CD was determined by SES-CD >2, and endoscopic remission was defined as SES-CD ≤2 with a disappearance of ulcers. To evaluate SES-CD, we selected 4 endoscopic variables as follows: ulcers, proportion of the surface covered by ulcers, proportion of the surface with any other lesions, and stenosis. Each variable ranged from 0 to 3 in each segment. This protocol was evaluated and approved by the Institutional Review Board for Clinical Research of our institute.
Reagents
Cell culture reagents including RPMI-1640 medium, fetal bovine serum (FBS), streptomycin and penicillin, 2-mercaptoethanol, and L-gentamycin were purchased from HyClone (Logan, UT, USA). Antihuman CD3 and antihuman CD28 mAbs were purchased from BD Biosciences (San Diego, CA, USA). Phorbol 12-myristate 13-acetate (PMA) and ionomycin were purchased from Sigma-Aldrich (St. Louis, MO, USA). CellTrace CFSE Cell Proliferation Kit and Live/Dead Fixable Dead Cell stain kit were purchased from Life Technologies (Carlsbad, CA, USA). Enzyme-linked immunosorbent assay (ELISA) kits for IL-22 were purchased from eBioscience. Rabbit polyclonal antibody against human IL-22 (IgG) was purchased from Abcam (Cambridge, MA, USA). The mAbs used for flow cytometry in this study were all purchased from Biolegend (San Diego, CA, USA). All primers were synthesized and purchased from ShengGong BioTeck (Shanghai, China).
Ex Vivo Colonic Tissue Culture
Colonic inflamed biopsies were collected from CD patients during endoscopic examination. After being washed with sterile PBS, biopsies (2 biopsies/well) were cultured in 1 mL of complete RPMI-1640 medium (10% FBS, 100 μg/mL streptomycin, and 100 U/mL penicillin) in the presence of IFX or HIg (50 μg/mL). The biopsies were collected after 24 hours of culture, followed by extraction of total RNA, and IL-22 mRNA levels were then measured by quantitative real-time polymerase chain reaction (qRT-PCR).
Immunofluorescence
In situ expression of tight junction protein ZO-1 in Caco2 cells was performed by immunofluorescence staining, as described previously.5 Briefly, paraffin-embedded intestinal mucosal tissue sections (5 μm) were washed, followed by incubation with 3% H2O2 in phosphate buffered saline with Tween-20 (PBS-T). After using PBS buffer supplemented with 5% donkey serum, 3% BSA, and 0.1% Triton-X-100 to block nonspecific proteins, the sections were incubated with primary rabbit anti-ZO-1 antibody (Abcam) at 4°C overnight. On the next day, the sections were incubated with FITC-conjugated donkey antirabbit IgG at room temperature for 1 hour. After being washed and mounted with cover slips, sections were observed with a confocal microscope (Zeiss LSM510; Jena, Germany). We also treated sections with PBS instead of primary antibody as a negative control. The slides were read blindly without any code to avoid observer bias.
Quantitative Real-Time Polymerase Chain Reaction
We extracted total RNA from freshly collected biopsies or cultured cells using the Trizol reagents according to the protocol obtained from the manufacturer. The quality and quantity of RNA of each sample were assessed through Nanodrop 2000 (Quawell; Waltham, MA, USA) with a A260/A280 ratio of >1.8 and <2.0 for samples. We synthesized cDNA from 400 ng of RNA using an all-in-1 reverse transcription (RT) reagent kit (abm; Richmond, BC, Canada). PCR was performed using a SYBR Green PCR kit (Takara; Dalian, China) in the ABI prism 7900 HT sequence detector (Applied Biosystems; Foster City, CA, USA). The 2-ΔΔCT method was used to determine the differences in each target gene expression relative to the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene.
Isolation and Culture of CD4+ T Cells
We collected heparin anticoagulated blood samples from CD patients and healthy donors to isolate peripheral blood mononuclear cells (PBMCs) over Ficoll-hypaque density gradient by centrifugation, as described previously.26 Antihuman CD4 particles (BD Biosciences; San Diego, CA, USA) were used to isolate CD4+ T cells from PBMC. After counting cells, freshly isolated CD4+ T cells were seeded in an anti-CD3 mAb-coated 24-well plate (5 × 105 cells/well) with anti-CD28 mAb in RPMI-1640 medium supplemented with 10% FBS, streptomycin (100 mg/mL) and penicillin (100 U/mL), sodium pyruvate (1 mM; Life Technologies), 4-(2-hydroxyethyl) pipera-zine-1-ethanesulfonic acid (HEPES; 10 mM; Amresco, Solon, OH, USA), and β-mercaptoethanol (50 μM; MP Biomedicals, Santa Ana, CA, USA). We stimulated CD4+ T cells under different conditions and then harvested cells for qRT-PCR and flow cytometry analysis. Moreover, supernatants were also harvested for ELISA.
Flow Cytometry
After staining with suitable mAbs according to the manufacture’s protocols, PB-CD4+ T cells were analyzed using flow cytometry. For intracellular staining of cytokines, CD4+ T cells were incubated for 5 hours with ionomycin (1 μg/mL), PMA (50 ng/mL; Sigma-Aldrich), and IL-23 (10 ng/mL; R&D Systems; Minneapolis, MN, USA) in RPMI-1640 supplemented with 10% FBS, streptomycin (100 mg/mL), and penicillin (100 U/mL). Brefeldin A (eBioscience) was added in the last 3 hours, with a final concentration of 10 μg/mL. Cells were fixed by fixation/permeabilization buffer (eBioscience) after 5 hours and stained for intracellular cytokines according to the manufacture’s protocol. We used BD FACSCanto II (BD Biosciences) to acquire results and FlowJo vX.0.7 (Tree Star; Ashland, OR, USA) to do flow cytometric analysis.
Enzyme-Linked Immunosorbent Assay
IL-22 in supernatants of CD4+ T cells cultured under different conditions was measured with the human IL-22 ELISA kit (eBioscience) according to the manufacturer’s instructions. Briefly, we coated a 96-well plate with capture antibody and incubated it at 4°C overnight. After 3 washes, we added blocking diluent and incubated the 96-well plate at room temperature for 1 hour, followed by washing again. Then we added samples and standards to the wells and incubated the 96-well plate for 2 hours. Wells were added to the detection antibody and incubated again for 1 hour after washing. After being aspirated and washed, wells were added with avidin-HRP and incubated for 30 minutes. H2SO4 was added to each well to incubate for 10 minutes to stop color reaction, and the wells were read at 450 nm in Epoch (BioTek; Winooski, VT, USA).
T-Cell Proliferation Assay
We labeled freshly purified human PB-CD4+ T cells with carboxyfluorescein diacetate succinimidyl ester (CFSE; 10 μM, Life Technologies). Cells were cultured in an anti-CD3 mAb-coated 24-well plate in the presence of anti-CD28 mAb in the culture medium of RPMI-1640 as described above. Cells were harvested 3 days later for intracellular staining of cytokines, as mentioned before, and analyzed by flow cytometry.
Transepithelial Electrical Resistance Measurement
Caco2 cells were first seeded in the inserts of the Transwell system (2 × 104 cells/well). We changed the medium every 1–2 days, and transepithelial electrical resistance (TEER) was assessed every day according to the manufacturer’s instructions. When the TEER value became stable, we added lipopolysaccharides (LPS), IL-22, and LPS plus IL-22, respectively, in the medium and assessed TEER as well. Results of 5 independent experiments were utilized to determine the relativity to the initial TEER value and expressed as mean ± standard error of the mean (SEM).
Statistical Analysis
Data were analyzed using Prism V.6.0 software (Graphpad software; San Diego, CA, USA). Statistic comparisons were applied to an unpaired 2-tailed Student t test for 2 groups and 1-way analysis of variance (ANOVA) for more than 2 groups. Significance of difference was presented as follows: *P < 0.05; **P < 0.01; ***P < 0.001.
RESULTS
Anti-TNF mAb Treatment Promotes IL-22 Production in CD Patients
Given that IL-22 plays a critical role in modulating epithelial barrier repair and protecting the intestines from inflammation, we sought to study whether anti-TNF therapy could upregulate IL-22 production in the intestinal mucosa of active CD patients. We treated active CD patients (n = 57) with IFX at a dose of 5 mg/kg at weeks 0, 2, and 6, as indicated previously,24 and measured IL-22 production in intestinal biopsies prior to and 12 weeks after treatment with IFX. All information from CD patients including clinical and demographic characteristics was shown in Table 1. According to the change of CDAI, 29 participants (50.9%) achieved clinical remission (CDAI < 150). Sixteen patients (28.1%) did not achieve clinical remission, but they reached the standard clinical response with a decrease of CDAI ≥70. Unfortunately, 12 patients (21.1%) had both CDAI ≥150 and a decrease of CDAI ≤70 and failed to respond to IFX therapy; some even got worse, with an increase of CDAI from baseline. Figure 1A–C shows that the levels of IL-22 expression were significantly higher in the intestinal mucosa of CD patients from both the clinical remission group and the response group after IFX induction therapy than those prior to IFX administration, whereas no differences in IL-22 expression were observed in the failure group. We then determined whether treatment with IFX in vitro could also promote intestinal mucosal production of IL-22. To this end, we collected normal and inflamed intestinal biopsies from healthy donors and CD patients who did not receive any medical treatment including IFX, respectively, and treated them with IFX in vitro for 24 hours. We analyzed the levels of IL-22 expression using qRT-PCR and found that IL-22 expression was also markedly increased in the IFX-treated group compared with controls (Fig. 1D; Supplementary Fig. 1A).
Demographic and Clinical Characteristics of the Study Population
| Characteristics | Remission Group | Response Group | Failure Group |
|---|---|---|---|
| No. patients | 29 | 16 | 12 |
| Sex (M/F) | 14/15 | 14/12 | 7/5 |
| Age, y | 29.1 ± 13.3 | 26.7 ± 9.8 | 30.2 ± 11.2 |
| Disease duration, mo | 16.5 ± 9.7 | 34.8 ± 16.5 | 51.4 ± 20.7 |
| Age at diagnosis (A1/A2/A3)a | 2/25/2 | 1/11/4 | 0/9/3 |
| Disease location (L1/L2/L3/L4)a | 7/12/10/0 | 3/7/6/0 | 1/3/8/0 |
| Behavior (B1/B2/B3)a | 15/7/7 | 8/5/3 | 7/3/2 |
| Perianal disease (yes/no)a | 3/26 | 4/12 | 2/10 |
| Biopsy locationb | 9/6/9/4/1 | 3/5/7/1/0 | 3/2/5/2/0 |
| Past IFX exposure (yes/no) | 4/25 | 3/13 | 1/11 |
| Concomitant azathioprine (yes/no) | 17/12 | 9/7 | 6/5 |
| CRP, mg/L | 10.4 ± 7.2 | 20.1 ± 17.5 | 35.8 ± 14.3 |
| CDAI | 65.3 ± 39.4 | 163.7 ± 15.6 | 193.8 ± 58.2 |
| SES-CD | 2.4 ± 6.3 | 4.1 ± 5.3 | 12.7 ± 9.4 |
| Characteristics | Remission Group | Response Group | Failure Group |
|---|---|---|---|
| No. patients | 29 | 16 | 12 |
| Sex (M/F) | 14/15 | 14/12 | 7/5 |
| Age, y | 29.1 ± 13.3 | 26.7 ± 9.8 | 30.2 ± 11.2 |
| Disease duration, mo | 16.5 ± 9.7 | 34.8 ± 16.5 | 51.4 ± 20.7 |
| Age at diagnosis (A1/A2/A3)a | 2/25/2 | 1/11/4 | 0/9/3 |
| Disease location (L1/L2/L3/L4)a | 7/12/10/0 | 3/7/6/0 | 1/3/8/0 |
| Behavior (B1/B2/B3)a | 15/7/7 | 8/5/3 | 7/3/2 |
| Perianal disease (yes/no)a | 3/26 | 4/12 | 2/10 |
| Biopsy locationb | 9/6/9/4/1 | 3/5/7/1/0 | 3/2/5/2/0 |
| Past IFX exposure (yes/no) | 4/25 | 3/13 | 1/11 |
| Concomitant azathioprine (yes/no) | 17/12 | 9/7 | 6/5 |
| CRP, mg/L | 10.4 ± 7.2 | 20.1 ± 17.5 | 35.8 ± 14.3 |
| CDAI | 65.3 ± 39.4 | 163.7 ± 15.6 | 193.8 ± 58.2 |
| SES-CD | 2.4 ± 6.3 | 4.1 ± 5.3 | 12.7 ± 9.4 |
aMontreal Classification.
bIleum/right colon/transverse colon/left colon/rectum.
Demographic and Clinical Characteristics of the Study Population
| Characteristics | Remission Group | Response Group | Failure Group |
|---|---|---|---|
| No. patients | 29 | 16 | 12 |
| Sex (M/F) | 14/15 | 14/12 | 7/5 |
| Age, y | 29.1 ± 13.3 | 26.7 ± 9.8 | 30.2 ± 11.2 |
| Disease duration, mo | 16.5 ± 9.7 | 34.8 ± 16.5 | 51.4 ± 20.7 |
| Age at diagnosis (A1/A2/A3)a | 2/25/2 | 1/11/4 | 0/9/3 |
| Disease location (L1/L2/L3/L4)a | 7/12/10/0 | 3/7/6/0 | 1/3/8/0 |
| Behavior (B1/B2/B3)a | 15/7/7 | 8/5/3 | 7/3/2 |
| Perianal disease (yes/no)a | 3/26 | 4/12 | 2/10 |
| Biopsy locationb | 9/6/9/4/1 | 3/5/7/1/0 | 3/2/5/2/0 |
| Past IFX exposure (yes/no) | 4/25 | 3/13 | 1/11 |
| Concomitant azathioprine (yes/no) | 17/12 | 9/7 | 6/5 |
| CRP, mg/L | 10.4 ± 7.2 | 20.1 ± 17.5 | 35.8 ± 14.3 |
| CDAI | 65.3 ± 39.4 | 163.7 ± 15.6 | 193.8 ± 58.2 |
| SES-CD | 2.4 ± 6.3 | 4.1 ± 5.3 | 12.7 ± 9.4 |
| Characteristics | Remission Group | Response Group | Failure Group |
|---|---|---|---|
| No. patients | 29 | 16 | 12 |
| Sex (M/F) | 14/15 | 14/12 | 7/5 |
| Age, y | 29.1 ± 13.3 | 26.7 ± 9.8 | 30.2 ± 11.2 |
| Disease duration, mo | 16.5 ± 9.7 | 34.8 ± 16.5 | 51.4 ± 20.7 |
| Age at diagnosis (A1/A2/A3)a | 2/25/2 | 1/11/4 | 0/9/3 |
| Disease location (L1/L2/L3/L4)a | 7/12/10/0 | 3/7/6/0 | 1/3/8/0 |
| Behavior (B1/B2/B3)a | 15/7/7 | 8/5/3 | 7/3/2 |
| Perianal disease (yes/no)a | 3/26 | 4/12 | 2/10 |
| Biopsy locationb | 9/6/9/4/1 | 3/5/7/1/0 | 3/2/5/2/0 |
| Past IFX exposure (yes/no) | 4/25 | 3/13 | 1/11 |
| Concomitant azathioprine (yes/no) | 17/12 | 9/7 | 6/5 |
| CRP, mg/L | 10.4 ± 7.2 | 20.1 ± 17.5 | 35.8 ± 14.3 |
| CDAI | 65.3 ± 39.4 | 163.7 ± 15.6 | 193.8 ± 58.2 |
| SES-CD | 2.4 ± 6.3 | 4.1 ± 5.3 | 12.7 ± 9.4 |
aMontreal Classification.
bIleum/right colon/transverse colon/left colon/rectum.
IFX enhances IL-22 expression in intestinal mucosa and CD4+ T cells of CD patients. Fifty-seven patients with active CD were receiving IFX treatment, as indicated previously. Intestinal mucosal biopsies were collected from 29 patients in clinical remission (A), 16 patients responsive to IFX therapy but not in clinical remission (B), and 12 patients nonresponsive to IFX therapy (C) before and 12 weeks after the first infusion, and IL-22 mRNA expression was measured by qRT-PCR and normalized to the GAPDH in each sample. ***P < 0.001, *P < 0.05 vs data before IFX treatment. D, Freshly obtained intestinal biopsies from CD patients (n = 13) were cultured with IFX or control human IgG (50 μg/mL) for 24 hours, and IL-22 expression was detected by qRT-PCR and normalized to the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in each sample. ***P < 0.001 vs controls. E, PB-CD4+ T cells (5 × 105/mL) from CD patients (n = 6) were stimulated with immobilized anti-CD3 (5 μg/mL) and anti-CD28 (2 μg/mL) mAbs in the presence of IFX or HIg (50 ng/mL) for 48 hours; qRT-PCR was used to detect cytokine expression. *P < 0.05, ***P < 0.001 vs data from the same group stimulated with anti-CD3 and anti-CD28 mAbs in the presence of HIg. F, PB-CD4+ T cells from active CD patients (A-CD; n = 6) or healthy volunteers (HC; n = 6) were stimulated with anti-CD3 (5 μg/mL) and anti-CD28 (2 μg/mL) mAbs in the presence of IFX or HIg (50 ng/mL) for 3 days, and the cells were harvested to analyze IL-22 and IL-17A expression by flow cytometry. Percentages of IL-22+CD4+ T cells are shown in the bar chart. ***P < 0.001. Abbreviation: ns, not significant.
Although ILC3 and IL-22+CD4+ T cells have been shown as major resources of IL-22 production in the intestines, CD4+ T cells, but not ILC3, produce high levels of TNF. To determine whether CD4+ T cells in CD patients contribute to increased IL-22 production upon IFX treatment, PB-CD4+ T cells isolated from 6 CD patients who did not receive any medical treatment were stimulated in vitro with immobilized antihuman CD3 and antihuman CD28 mAbs in the presence of IFX (50 ng/mL)5 or HIg (50 ng/mL) as controls for 48 hours. As shown in Figure 1E, IFX significantly promoted CD4+ T cells to produce IL-22 and IL-10. However, IL-21 production was decreased in the IFX-treated group, consistent with our previous work.5 We then investigated whether CD4+ T cells responded to IFX treatment, particularly in CD patients. PB-CD4+ T cells from CD patients and healthy volunteers were isolated and stimulated as described above. Interestingly, we found an increase of IL-22+CD4+ T cells in the IFX-treated group but not in medium alone or HIg-treated controls (Fig. 1F; Supplementary Fig. 1B), while we did not find a significant difference in IL-22+CD4+ T cells in healthy donors (Supplementary Fig. 1C). Taken together, our findings indicate that anti-TNF treatment promotes IL-22 production from CD CD4+ T cells.
Anti-TNF Treatment Induces Th22 Cell Differentiation
As we found that anti-TNF mAb promoted IL-22 production of CD CD4+ T cells, we next studied whether anti-TNF treatment could upregulate Th22 cell differentiation. We isolated PB-CD4+ T cells from patients with active CD and healthy volunteers and applied CFSE (10 μM) dilution to them under stimulation with anti-CD3/CD28 mAb in the presence of IFX in vitro. The frequencies of Th22, Th1, and Th17 cells in the proliferated part of CD4+ T cells were measured by flow cytometry. As shown in Figure 2A and B, Th22 cells (IL-22+IL-17−IFN-γ− CD4+ T cells) were markedly increased in proliferated CD CD4+ T cells treated with IFX, while there were no differences in Th1 (IFN-γ+IL-22−IL-17− CD4+ T cells) or Th17 cells (IL-17+IL-22−IFN-γ− CD4+ T cells) between IFX-treated patients and controls. The difference was statistically significant according to 5 independent experiments (Fig. 2C). In contrast, Th22 cells were only slightly increased in CD4+ T cells of healthy individuals treated with IFX compared with controls (Supplementary Fig. 2A, B), but the difference is not statistically significant (Supplementary Fig. 2C). Therefore, the results indicate that anti-TNF treatment promotes CD4+ T cells to differentiate into Th22 cells in CD patients rather than in healthy controls.
IFX treatment induces Th22 cell differentiation. Freshly isolated PB-CD4+ T cells (5 × 105/mL) from CD patients (n = 5) were labeled with CFSE (10 μM) and stimulated with anti-CD3 (5 μg/mL) and anti-CD28 (2 μg/mL) mAbs in the presence of HIg (A) or IFX (B) at the concentration of 50 ng/mL for 3 days. Cells were harvested, and intracellular staining with fluorochrome-conjugated antibodies was performed as indicated. Intracellular expression of IL-17A, IL-22, and IFN-γ was analyzed by flow cytometry. C, Percentages of IL-22+ IL-17− IFN-γ− CD4+ T cells are shown in the bar chart. ***P < 0.001.
TACE Downregulates Anti-TNF-Induced IL-22 Expression by CD CD4+ T Cells
Following post-translational modification, TNF anchors into the plasma membrane as transmembrane TNF (tmTNF). Upon binding to anti-TNF antibody or soluble cognate receptors, the membrane-bound TNF form transduces reverse signaling,19 which is involved in the immune regulation of the inflammatory response acting primarily locally.20 To determine whether anti-TNF mAb binding to tmTNF could initiate reverse signaling, leading to increased IL-22 production of CD4+ T cells in CD patients, we added TACE, a metalloproteinase that is able to strip tmTNF to soluble TNF, into CD4+ T-cell cultures. We isolated PB-CD4+ T cells from both active CD patients and healthy volunteers and stimulated in vitro with anti-CD3/CD28 mAbs in the presence of HIg, IFX, rhTACE, and IFX plus rhTACE, respectively. The dose of rhTACE was determined in previous work (Supplementary Fig. 3). As shown in Figure 3A and C, rhTACE markedly suppressed IL-22 production in IFX-treated CD CD4+ T cells. We collected the supernatants of CD CD4+ T cells treated as indicated above and analyzed the levels of IL-22 using ELISA. Consistently, the levels of IL-22 were found to be elevated in IFX-treated CD CD4+ T cells but decreased in rhTACE-treated groups (Fig. 3B). On the contrary, we did not find any effect on IL-22 production of CD4+ T cells from healthy donors (Fig. 3D–F). Collectively, our results indicate that reverse signaling by anti-TNF mAb binding to tmTNF promotes CD CD4+ T cells to produce IL-22, and that TACE appears to be a functional target in CD patients, particularly those who are nonresponsive to anti-TNF therapy.
rhTACE downregulates IFX-induced IL-22 production by CD4+ T cells of CD patients. Freshly isolated PB-CD4+ T cells (5 × 105/mL) from active CD patients (A-CD; n = 5) or healthy volunteers (HC; n = 5) were stimulated with anti-CD3 (5 μg/mL) and anti-CD28 (2 μg/mL) mAbs in the presence of HIg (50 ng/mL), IFX (50 ng/mL), rhTACE (100 ng/mL), and IFX plus rhTACE, respectively. After 3 days of culture, cells were harvested and analyzed for intracellular expression of IL-22 and IL-17A by flow cytometry. A and D, Percentages of IL-22+CD4+ T cells in A-CD (A) and HC (D) are shown in the bar chart. B and E, The supernatants were collected from A-CD (B) and HC (E) for the detection of IL-22 using ELISA. C and F, IL-22 transcription expression in CD4+ T cells from A-CD (C) and HC (F) was analyzed by qRT-PCR and normalized to GAPDH. Representative results from 5 independent experiments are shown. *P < 0.05, **P < 0.01, ***P < 0.001 vs data from the same group stimulated with anti-CD3 and anti-CD28 mAbs in the presence of HIg; ΔP < 0.001 vs data from the same group stimulated with anti-CD3 and anti-CD28 mAbs in the presence of IFX. +P < 0.05 vs data from the same group stimulated with anti-CD3 and anti-CD28 mAbs in the presence of IFX.
Anti-TNF Promotes IL-22 Production in an AhR-Dependent Manner
As anti-TNF treatment significantly promoted CD CD4+ T cells to express IL-22, we then measured the relative expression of the transcriptional factors specifically in these populations. As shown in Figure 4A, RORC expression was found to be significantly reduced in CD CD4+ T cells after IFX treatment, indicating an inhibitory role of anti-TNF mAb in regulating Th17 cell differentiation, as described previously,5 whereas there was no difference in T-bet and Foxp3 expression after IFX treatment (Fig. 4B, C). Interestingly, the expression of AhR was observed to be markedly increased in IFX-treated CD CD4+ T cells (Fig. 4D). To clarify whether AhR is responsible for IL-22 expression of anti-TNF-treated CD CD4+ T cells, we applied an AhR antagonist (ie, CH-223191; 20 ng/mL)27 in CD4+ T cells stimulated with anti-CD3/CD28 mAbs in the presence of IFX or HIg. Blockade of AhR significantly decreased the frequency of IL-22+CD4+ T cells in IFX-treated CD4+ T cells compared with controls (Fig. 4E, F).
IFX induces Th22 cell differentiation in an AhR-dependent manner. A–D, Freshly isolated PB-CD4+ T cells (5 × 105/mL) from CD patients (n = 5) were stimulated with anti-CD3 (5 μg/mL) and anti-CD28 (2 μg/mL) mAbs in the presence of IFX or HIg (50 ng/mL) for 3 days. These cells were then collected, and cytokine transcript expression was analyzed by qRT-PCR as indicated. An AhR antagonist, CH-223191 (20 ng/mL), was applied to the culture of CD4+ T cells as indicated. IL-22 expression was analyzed by qRT-PCR (E) and flow cytometry (F). Gene expression was normalized to GAPDH mRNA levels in each sample. Representative results from 5 independent experiments are shown. ***P < 0.001 vs data from the same group stimulated with anti-CD3 and anti-CD28 mAbs in the presence of HIg. ΔP < 0.001 vs data from the same group stimulated with anti-CD3 and anti-CD28 mAbs in the presence of IFX. #P < 0.001 vs data from the same group stimulated with anti-CD3 and anti-CD28 mAbs in the presence of CH-223191.
IL-22 Alleviates LPS-Induced Intestinal Epithelial Injuries
To investigate whether the increase of IL-22 production after anti-TNF therapy could contribute to intestinal epithelial tissue repair, we treated a human intestinal cell line, Caco2 cells, with IL-22 (100 ng/mL)28 in the presence or absence of LPS (100 μg/mL),29 which causes epithelial cell injuries. As shown in Figure 5A and B, LPS significantly reduced the expression of tight junction (TJ) proteins. We then employed IL-22 in the damaged Caco2 cell monolayer and tested TJ protein expression by qRT-PCR and immunoblotting. Caco2 cell monolayer transepithelial resistance was tested using the Transwell system. Figure 5B illustrates that IL-22 significantly upregulated the expression of ZO-1, Claudin1, and Clacudin 2 mRNA, which was further confirmed by immunoblotting (Supplementary Fig. 4A, B). Moreover, IL-22 treatment also enhanced Caco2 transepithelial resistance to LPS-induced damage (Fig. 5C).
IL-22 alleviates LPS-induced epithelial injuries in Caco2 cells. A, Caco2 cells were stimulated with LPS (100 μg/mL), IL-22 (100 ng/mL), and LPS plus IL-22, respectively, as indicated for 48 hours. Representative immunofluorescent images for detection of tight junction protein ZO-1 (green) in Caco2 cells are shown. B, Caco2 cells were stimulated with LPS, IL-22, and LPS plus IL-22, respectively, as indicated for 24 hours, and transpcription levels of tight junction proteins were measured by qRT-PCR and normalized to GAPDH mRNA levels in each sample. The data represent 3 independent experiments. *P < 0.05, **P < 0.01 vs data from Caco2 cells treated with medium alone. #P < 0.05 vs data from Caco2 cells treated with LPS. C, The Transwell system was employed in a Caco2 cell culture experiment. Transepithelial electrical resistance remained stable since day 20, and different stimulation was applied starting on day 23; then transepithelial electrical resistance was measured daily until day 30.
DISCUSSION
Apart from neutralizing TNF, the mechanisms whereby anti-TNF therapy mediates in modulating immune response in IBD are still not completely understood. Accumulating lines of evidence have suggested that anti-TNF therapy is involved in anti-inflammatory activities by inhibiting proinflammatory cytokines and promoting anti-inflammatory cytokines to facilitate wound healing in colitis.5, 30 In the present study, we did observe that the expression of IL-22 was significantly higher in inflamed mucosa in CD patients who achieved clinical remission after IFX therapy than before IFX treatment. However, we did not find any change of IL-22 expression in the failure group. Interestingly, recent studies have demonstrated that treatment of patients with medically refractory severe IBD secondary to IL-10R deficiency with anakinra, an IL-1 receptor antagonist, leads to marked clinical, endoscopic, and histological improvement and mucosal healing accompanied by increased numbers of IL-22+CD4+ T cells in the terminal ileum, thus confirming the significance of an increase of IL-22+CD4+ T cells in the treatment of IBD.31
Our current study indicates that anti-TNF therapy enhances IL-22 expression by CD CD4+ T cells. Consistent with our previous studies,4, 5 we further confirmed that anti-TNF therapy suppresses CD CD4+ T cells to express Th17 transcription factor RORC. Therefore, our data demonstrated that anti-inflammatory functions of anti-TNF therapy are likely related to its ability to induce CD4+ T cells to express IL-22 and inhibit proinflammatory T-cell subsets (ie, Th17 cells).
It is known that IFX can bind to both soluble TNF and tmTNF to block TNF-mediated signaling pathways. Reversing signaling of tmTNF has been demonstrated to confer resistance to LPS stimulation and inhibit proinflammatory cytokine production by macrophages.19, 20 Binding of anti-TNF mAb to tmTNF in CD4+ T cells results in simultaneous inhibition of signaling through TNF receptors and reverse signaling via tmTNF.32 Our present data showed that TACE downregulated anti-TNF-induced IL-22 production, indicating that tmTNF-mediated reverse signaling is at least partially involved in promotion of IL-22 production in CD CD4+ T cells in response to anti-TNF therapy. AhR is considered a pivotal transcription factor in driving IL-22 production in T cells, ILC3, and neutrophils.33 In our study, IFX treatment promoted expression of AhR in CD CD4+ T cells, and blockade of AhR remarkably decreased IL-22 expression by IFX-treated CD4+ T cells, indicating that AhR mediates IFX promotion of IL-22 production in CD CD4+ T cells. Our data thereby provide a potential approach of adjuvant therapy for anti-TNF in CD patients.
The protective properties of IL-22 in the intestines have been well established in animal models of colitis.6, 10, 11 Transgenic expression or exogenous administration of IL-22 has been shown to have a protective effect on regulating intestinal homeostasis and intestinal mucosal inflammation.6, 10, 11, 34, 35 In our study, increased IL-22 production after anti-TNF treatment has been observed to be responsive for epithelial cell repairs, which further sheds light on therapeutic effects in CD patients. Importantly, recent data have proven that the IL-9 level is decreased in patients with CD who had favorable therapeutic effects after 30 weeks of IFX therapy, thus providing preliminary evidence that the level of IL-9 at week 14 in CD patients could predict the therapeutic effects of IFX at week 30.36 In our study, we found that IL-22 level was elevated in CD patients who reached clinical remission at week 12, suggesting that IL-22, together with IL-9, may be a promising noninvasive biomarker as to predicting disease activity and IFX efficacy in these patients. However, the precise manner in which IL-22 contributes to IFX efficacy in CD patients warrants further investigation.
TACE has been identified as a membrane-anchored multidomain metalloproteinase responsible for cleavage of pro-TNF.37 TACE activity is increased in inflamed mucosa of IBD, which promotes T-cell proliferation.38, 39 Our study showed that the addition of rhTACE to CD CD4+ T cells markedly inhibited anti-TNF-induced IL-22 production. Increased TACE mRNA expression has been reported in various human tumors, which is considered to be one of the reasons that tumor patients who have high levels of TNF in sera fail to respond to anti-TNF therapy.40 Increased TACE expression has also been shown in inflamed colon tissue.38 As TACE is able to inhibit anti-TNF-induced CD4+ T-cell IL-22 production, it will be very interesting to determine whether different levels of TACE are involved in some CD patients who are not responsive to IFX. If this is the case, TACE could be used as a potential target for the treatment of IBD patients with nonresponse to anti-TNF therapy.
In summary, the data in this study describe that anti-TNF therapy is involved in enhancing CD CD4+ T cells to differentiate into Th22 cells and IL-22 production, possibly through reverse signaling induced by anti-TNF mAb binding to tmTNF in T cells in an AhR-dependent way. Importantly, we found that TACE plays a key role in the reverse signaling of CD4+ T cells; it may function as a potential target for IBD patients who are nonresponsive to anti-TNF therapy.
SUPPLEMENTARY DATA
Supplementary data are available at Inflammatory Bowel Diseases online.
ACKNOWLEDGMENTS
Patient consent: obtained.
Ethics approval: Institutional Review Board for Clinical Research of the Shanghai Tenth People’s Hospital of Tongji University.
Supported by: grants from the National Natural Science Foundation of China (816300017 and 81470822).
Conflicts of interest: The authors have no conflicts of interest to disclose.
REFERENCES
Author notes
Co-first authors.
