Final Progress Report

Proposal No. IBD-0154R
Principal Investigator:  Giovanni Monteleone, M.D., Ph.D.
Applicant Organization:  University Tor Vergata of Rome (Italy)
Project Title:  Interleukin-21 triggers inflammatory signals in the gut.  Relevance for human inflammatory bowel diseases.
Period of Award:  July 1, 2005 – June 30, 2007

Summary of project aims

This project was aimed at investigating the role of interleukin (IL-21) in gut inflammation.

IL-21 biological effects are mediated through a class I cytokine receptor (IL-21R) that interacts with the common γ-chain receptor subunit. Although, IL-21R was originally detected in immune cells, we preliminarily showed that, in the gut, IL-21R is also expressed by non-immune cells, such as epithelial cells and fibroblasts. Specific aims of this project were to evaluate: a. factors/mechanisms involved in the control of IL-21R expression on epithelial cells and fibroblasts in inflammatory bowel disease (IBD); b. whether these cell types respond to IL-21 by making molecules that sustain the ongoing inflammation and promote mucosal degradation. In particular, our plan was to examine whether IL-21-stimulated epithelial cells secrete macrophage inflammatory protein 3 alpha (MIP-3alpha), a chemokine involved in the intestinal recruitment of memory T cells, to dissect the molecular mechanism by which IL-21 regulates the functional activity of gut epithelial cells, and to prove that blockade of endogenous IL-21 in ex vivo cultures of IBD mucosal explants reduces the synthesis of MIP-3alpha. We also planned to evaluate whether IL-21-treated fibroblasts produce matrix metalloproteinases (MMPs), a family of proteases that are supposed to cause tissue damage and remodeling in the gut. In IBD, mucosal T lymphocytes are resistant against apoptotic stimuli. Since, cytokines that signal through the common γ-chain receptor are known to regulate T cell death, we also wanted to examine whether IL-21 controls intestinal T cell survival. Another goal of this project was to examine whether up-regulation of IL-21 is a specific hallmark of IBD. Finally, we planned to bring our human observations to murine models of IBD and examine whether IL-21 is produced in excess in experimental models of IBD. 

Accomplishments towards meeting those aims, and list of significant results (positive or negative).

1. IL-21R-mediated signals enhance MIP-3alpha production in colonic epithelial cells.

By immunostaining and Western blotting, we showed that IL-21R was constitutively expressed by  colonic epithelial cells, and such an expression was more pronounced in samples from patients with Crohn’s disease (CD) and patients with ulcerative colitis (UC) in comparison to normal controls. Additionally, it was shown that colonic epithelial cells expressed the common γ-chain receptor.  IL-21R was also seen in several colonic cancer cell lines. We next performed experiments to examine whether colonic epithelial cells respond to exogenous IL-21. As primary intestinal epithelial cells rapidly undergo apoptosis in vitro, we used colonic cell lines. Initially, we showed that IL-21 induced the  phosphorylation of various proteins in colonic epithelial cells, but it regulated neither the growth nor the survival of such cells.

In recent years, several studies have shown that intestinal epithelial cells actively respond to inflammatory molecules by releasing increased amounts of both chemokines and cytokines, which are believed to sustain the ongoing mucosal inflammation. Therefore, we explored the possibility that IL-21 could regulate the synthesis of inflammatory mediators by colonic epithelial cells. To this end, we stimulated both DLD-1 and HT-29 cells with IL-21 and then analyzed the content of 120 different proteins into the culture supernatants by a human protein antibody array. Stimulation of cells with IL-21 enhanced the secretion of macrophage inflammatory protein 3 alpha (MIP-3alpha). These data were then confirmed through the analysis of MIP-3alpha by ELISA.

Previous studies have shown that MIP-3alpha is up-regulated on IBD epithelium, and it is supposed to contribute to the recruitment of T cells within the inflammatory microenvironment. Therefore we examined whether the IL-21-stimulated epithelial cell-derived MIP-3alpha was capable of enhancing the migration of peripheral blood lymphocytes using an in vitro chemotaxis assay.  Notably, the fraction of lymphocytes migrating in response to the stimulation with IL-21-treated cell culture supernatants (conditioned medium) was significantly higher than that induced by supernatants of untreated cells. FACS analysis showed that the conditioned medium enhanced the migration of both CD4+ and CD8+ T cells. To determine whether the lymphocyte migration in response to the stimulation with conditioned medium was mediated by MIP-3alpha, the conditioned medium was incubated with or without a neutralizing anti-MIP-3alpha or control isotype antibody prior to testing its effect on lymphocyte migration. Notably, the anti-MIP-3alpha but not control antibody significantly inhibited the conditioned medium-induced lymphocyte migration.

In subsequent experiments we examined whether the neutralization of endogenous IL-21 in organ cultures of IBD mucosal explants reduced MIP-3alpha production. For this purpose, mucosal samples were taken from patients with CD and  patients with UC and cultured with or without the initial addition of a neutralizing IL-21 or control antibody for 24 hours. At the end, organ culture supernatants were collected and analyzed for the content of MIP-3alpha by ELISA. The neutralizing IL-21 but not control antibody significantly decreased the production of MIP-3alpha. Overall these data suggest that IL-21 can regulate the production of MIP-3a in the gut, thus contributing to the recruitment of inflammatory cells.

2. Control of matrix metalloproteinase production in human intestinal fibroblasts by interleukin-21 (IL-21).

Fibroblasts were isolated from the inflamed colon of patients with CD, patients with UC, and from macroscopically and microscopically unaffected colonic areas of patients undergoing colectomy for colon cancer (normal controls). Total protein extracts were then prepared from fibroblasts and analyzed for IL-21R by Western blotting. We showed that IL-21R was constitutively expressed by gut fibroblasts regardless of whether cells were from IBD patients or controls. Additionally, IL-21R was seen in the fetal gut fibroblast cell line, CCD18CO. Semi-quantitative analysis of immunoreactive bands of IL-21R revealed no significant difference between IBD and normal controls. Moreover, expression of IL-21R and the common g-chain receptor in gut fibroblasts was confirmed by RT-PCR. In subsequent experiments, we examined whether IL-21R expression on gut fibroblasts is regulated by pro-inflammatory cytokines. To this end, normal intestinal fibroblasts were stimulated with IL-1beta or TNF-alpha, two cytokines which are produced in excess in IBD and known to regulate fibroblast activity. Both cytokines up-regulated IL-21R expression.

To examine whether IL-21 regulates the synthesis of MMPs by intestinal fibroblasts, serum-starved fibroblasts were treated with graded doses of IL-21, and MMPs were evaluated by Western blotting using specific monoclonal antibodies. Stimulation of intestinal fibroblasts with IL-21 resulted in enhanced secretion of MMP-1, -2, -3 and -9. IL-21 dose-dependently stimulated the secretion of both inactive and active isoforms. This effect was seen regardless of whether fibroblasts were isolated from normal or IBD patients, as well as in CCD18CO cells. The ability of IL-21 to enhance the synthesis of gelatinases (MMP-2 and MMP-9) was also confirmed by zymography.  Since matrix degradation results from an alteration in the balance between MMPs and their physiological inhibitors (TIMPs), we also analyzed the effect of IL-21 on TIMP-1 and TIMP-2. No significant change in the secretion of TIMP-1 and -2 was seen in IL-21-stimulated fibroblast cultures.

In the gut, TNF-alpha up-regulates MMP-1 and MMP-3 production by fibroblasts, thus leading to the degradation of ECM and basement membrane, and epithelial shedding.  The fact that TNF-alpha up-regulated IL-21R expression prompted us to examine whether IL-21 cooperates with TNF-alpha in inducing MMP production. Indeed, cells stimulated with IL-21 and TNF-alpha together produced more MMPs than either IL-21 or TNF-alpha alone. By contrast, secretion of TIMP-1 and -2 remained unchanged.

To investigate increased transcription as a potential mechanism for the IL-21-induced increased secretion of MMPs, MMP RNA expression was assessed by RT-PCR. MMP RNA transcripts were seen in untreated fibroblasts and were not increased by IL-21. This was seen regardless of whether normal or IBD cells were analyzed and at any time of cell passage. In contrast, TNF-alpha significantly enhanced MMP-1, MMP-3, and MMP-9 but not MMP-2 RNA. To determine whether IL-21 up-regulated the level of MMPs before secretion, cell-associated MMPs were assayed by Western blotting using total extracts. IL-21 did not alter the cellular content of any MMP, whereas TNF-alpha increased the expression of MMP-1, MMP-3 and MMP-9, but not MMP-2. Western blotting of total extracts revealed the appearance of single bands for each MMP consistent with the demonstration that MMPs are mostly cleaved after secretion.  Inhibitors of RNA and protein synthesis or secretion were then used to confirm our findings. Incubation of fibroblasts with the protein secretion inhibitor, brefeldin A, abrogated the effect of IL-21 on MMP production. In contrast, neither actinomycin D nor cycloheximide inhibited the IL-21-mediated MMP secretion. Taken together, these results suggest that IL-21 affects neither gene transcription nor de novo protein synthesis but it may enhance the secretion of pre-constituted or newly synthesized MMPs. This hypothesis is also supported by the demonstration that IL-21 induced a rapid secretion of MMPs which was evident as early as 2 hours after stimulation.

In a final set of experiments, we evaluated whether CD lamina propria mononuclear cells (LPMC)-derived IL-21 can modulate MMPs production. To this end,  intestinal fibroblasts isolated from CD mucosal specimens were cultured in the presence or absence of CD LPMC supernatants with or without the addition of a neutralizing human IL-21R/Fc fusion protein. Unstimulated fibroblasts released low levels of MMPs. However, the addition of CD LPMC supernatants to the fibroblast cultures resulted in enhanced synthesis of all MMPs. The addition of IL-21R/Fc to the fibroblast cultures decreased the production of MMPs induced by CD LPMC supernatants.

3.  IL-21 does not regulate the survival of IBD mucosal T cells

The marked accumulation of T cells in the inflamed mucosa of CD patients is believed to rely partly on the resistance of such cells against apoptotic stimuli. There is also evidence that such a phenomenon plays a key role in maintaining the mucosal inflammatory state. Indeed drugs that are able to increase T cell susceptibility to apoptotic stimuli promote the resolution of the ongoing tissue inflammatory process. The exact mechanism underlying the resistance of T cells against apoptosis during IBD is not yet known, but locally-produced cytokines seem to be involved.

We therefore examined whether IL-21 played a role in the resistance of mucosal IBD T cells against apoptotic stimuli. To address this issue, we first examined whether the addition of recombinant IL-21 to cultures of normal intestinal CD4+ lamina propria T lymphocytes (LPL) resulted in enhanced survival, and second  whether the neutralization of endogenous IL-21 in cultures of IBD LPL was sufficient to restore the susceptibility of LPL to apoptosis. Studies from these experiments showed that IL-21 is not involved in the control of intestinal LPL survival.

4. Activation of NF-κB by interleukin-21 results in enhanced synthesis of  gelatinases in gastric epithelial cells.

Helicobacter pylori (Hp) infection causes a T cell-associated chronic inflammation which can lead to gastric ulceration. Although the mechanism by which T cells promote epithelial  damage in Hp-infected subjects remains unknown,  studies in other systems suggest that cytokines produced by T cells and macrophages can enhance the production of MMPs by both epithelial and LPMC, thereby promoting mucosal injury. The demonstration that enhanced expression of IL-21 occurs in T cell-mediated human diseases, such as IBD, and that IL-21 regulates the production of MMPs in intestinal fibroblasts prompted us to examine whether IL-21 is also involved in Hp-associated gastritis.  To this end, we first examined IL-21 in gastric biopsies taken from patients with Hp-associated gastritis, patients with Hp-negative gastritis, and normal subjects. IL-21 was analyzed by Western blotting.  IL-21 was detected in all samples, but its expression was more pronounced in patients with Hp than in controls. Moreover, the content of IL-21 was higher in the mucosa of patients with Hp-negative gastritis than in normal controls. Additionally, enhanced IL-21 expression was seen in purified CD3+ LPL from Hp-infected patients, thus suggesting that the high IL-21 seen in biopsy samples of Hp-infected patients does not rely on the increased number of mucosal lymphocytes.  

Since in Hp-infected mucosa, epithelial cells are a major target of T cell-derived cytokines, we examined whether IL-21 regulated gastric epithelial cells activity. To this end, we initially showed that IL-21R was constitutively present in gastric epithelial cells, but its expression was increased in Hp-infected epithelium. In line with these data, IL-21R and the common-g chain were seen in gastric cancer epithelial cells lines (AGS and MKN28). We then tested the hypothesis that IL-21 could regulate MMP production by gastric epithelial cells. Serum-starved AGS and MKN28 cells were either left untreated or treated with graded doses of IL-21 and the cell-free supernatants were analyzed by Western blotting for MMPs. IL-21 dose-dependently enhanced the synthesis of gelatinases A and B (MMP-2, and MMP-9). In contrast, the secretion of MMP-1,  MMP-3, MMP-7, TIMP-1 and TIMP-2 remained unchanged.     

Analysis of molecular mechanisms underlying the IL-21 effect on MMP-2 and MMP-9 production revealed  that IL-21 did not significantly alter the activation of MAP kinases (i.e. p38, ERK1/2, and JNK), and treatment of AGS  with inhibitors of MAP kinases did not reduce the IL-21-induced synthesis of MMP-2 and MMP-9.  In contrast, a reduction of IkBα, an inhibitor of NF-kB activation, was seen in IL-21-stimulated AGS and MKN28 cells. Consistently, IL-21 enhanced NF-kB DNA  binding activity. To evaluate the role of NF-kB in the IL-21-induced MMP, AGS cells were treated with an inhibitor of NF-kB (TPCK) and then stimulated with IL-21. Inhibition of NF-kB  was associated with a reduction  in the IL-21-mediated MMP-2 and MMP-9 synthesis.

5. Up-regulation of IL-21 occurs in active celiac disease

The above findings indicate that up-regulation of IL-21 is not a specific hallmark of IBD. To confirm further this notion, we examined IL-21 in duodenal biopsies taken from patients with celiac disease and normal control. IL-21 RNA and protein were evaluated by real-time PCR and Western blotting respectively. Immunoreactivity for IL-21 protein was seen in all untreated celiac disease patients and 9/14 control samples. IL-21 expression quantitated by densitometry and normalized by beta-actin expression was significantly increased in biopsies taken from untreated celiac disease patients as compared with controls. In contrast, biopsies taken from treated celiac disease patients showed the same levels of IL-21 as controls. Moreover, IL-21 RNA expression was significantly higher in celiac disease patients than controls.

Gluten-specific immune response in celiac disease is associated with a marked induction of T-bet, a transcription factor that directs Th1 cell commitment. Since T-bet expression is transcriptionally regulated by signaling pathways that are inducible by IL-21, we next carried out ex vivo organ cultures and examined whether blockade of endogenous IL-21 reduced T-bet in duodenal biopsies of active celiac disease patients. Duodenal biopsies taken from untreated celiac disease patients were cultured with or without a neutralizing anti-IL-21 antibody or control antibody for 24 hours and then analyzed for T-bet by Western blotting. T-bet was clearly detectable in biopsies cultured with medium alone, was down-regulated by the anti-IL-21, but not control antibody. Treatment of biopsies from untreated celiac disease patients with the anti-IL-21 also resulted in a significant inhibition of T-bet RNA expression.

We also showed that the anti-IL-21-mediated down-regulation of T-bet was associated with a decrease in IFN-gamma expression.

6.  IL-21 is over-produced in experimental models of colitis

To assess whether IL-21 is produced in excess in experimental models of IBD, colitis was induced in SJL mice by either 2,4,6-trinitrobenzene sulfonic acid (TNBS) or oxazolone. These two models of IBD were selected because they show immuno-morphologic similarities with CD and UC respectively.  Four days after induction of colitis, tissues were removed from TNBS- or oxazolone-treated mice and controls and fixed in 10% neutral buffered formalin solution, embedded in paraffin, cut into tissue sections, and stained with Haematoxylin & Eosin. Stained sections were examined for evidence of colitis. Additional colonic samples were used for extracting LPMC or immediately frozen for subsequent analysis of IL-21. LPMC were cultured with or without the initial addition of anti-CD3 and CD28 antibodies in the presence or absence of a blocking IL-21R/Fc or control fusion protein for 48 hours. At the end, cells and cell-free supernatants were harvested and used for analysis of cytokines. Enhanced production of IL-21 was seen in cultures of LPMC isolated from the colon of  mice with  either TNBS or oxazolone colitis in comparison to controls. Moreover, neutralization of endogenous IL-21 by IL-21R/Fc significantly reduced the production of inflammatory cytokines by stimulated LPMC. In particular, blockade of IL-21 in cultures of LPMC isolated from mice with TNBS- or oxazolone-colitis inhibited the secretion of IFN-gamma or IL-13 respectively.  

We also performed preliminary experiments and tested the in vivo effect of IL-21R/Fc on the ongoing inflammation. To this end, the day after the induction of colitis, IL-21R/Fc (0,4 mg/mouse) was intraperitoneally administered to mice. IL-21R/Fc was not effective in inhibiting the ongoing colitis. However, we are not able to ascertain if the dose of IL-21R/Fc we selected for these studies was sufficient to neutralize the endogenous IL-21 in the gut.

Lay summary of the progress report:

T lymphocytes have been long presumed to play a decisive role in initiating and shaping pathologic responses in patients with Crohn’s disease (CD) and patients with ulcerative colitis (UC). There is also evidence that T lymphocytes actively interact with other cell types to promote tissue damage, and that cytokines are essential mediator of this cross-talk. This project was aimed at investigating the role of interleukin (IL-21), a T cell-derived cytokine which is produced in excess in inflammatory bowel disease (IBD), in gut inflammation. To this end, we first characterized cell targets of IL-21 in the inflamed intestine. We showed that non-immune cells, such as fibroblasts and epithelial cells, express receptors for IL-21 and are able to respond to IL-21 by producing molecules that are known to amplify the ongoing inflammation and to promote tissue damage. We also dissected the molecular mechanisms by which IL-21 controls the functional activity of gut fibroblasts and epithelial cells. Our studies also revealed that up-regulation of IL-21 is not a specific hallmark of IBD, as this cytokine is over-produced in the stomach of patients with Helicobacter pylori infection and in the intestine of patients with celiac disease. Additionally, we showed that IL-21 production is enhanced in mice with IBD-like colitis. Overall these observations indicate that IL-21 is an active and critical participant of the immune response that causes tissue damage in the gut and suggest that neutralization of IL-21 may be a novel therapeutic approach for IBD.