Supported by research grants from INSERM Foundation Princesse Grace de Monaco, and by a BREMICI grant from Schering-Plough. J. Verdier was supported by the Association François Aupetit.
Compartmentalized expression of Th1 and Th17 cytokines in pediatric inflammatory bowel diseases†
Article first published online: 12 OCT 2011
Copyright © 2011 Crohn's & Colitis Foundation of America, Inc.
Inflammatory Bowel Diseases
Volume 18, Issue 7, pages 1260–1266, July 2012
How to Cite
Verdier, J., Begue, B., Cerf-Bensussan, N. and Ruemmele, F.M. (2012), Compartmentalized expression of Th1 and Th17 cytokines in pediatric inflammatory bowel diseases. Inflamm Bowel Dis, 18: 1260–1266. doi: 10.1002/ibd.21905
- Issue published online: 11 JUN 2012
- Article first published online: 12 OCT 2011
- Manuscript Accepted: 29 AUG 2011
- Manuscript Received: 23 AUG 2011
- Crohn's disease;
- ulcerative colitis;
Interleukin (IL)-23, IL-17A, IL-17F, and interferon-gamma (IFN-γ) are important mediators of inflammatory colitis and are potential therapeutic targets in inflammatory bowel disease (IBD). Their expression profile in the different parts of normal noninflammatory intestine is unclear and their changes during pathology have not yet been addressed in pediatric IBD patients.
We quantified the transcriptional expression of IL-23, IL-12, IL-17A, IL-17F, IL-6, and IL-10 in healthy, noninflammatory duodenum, ileum, and colon and in inflamed and noninflamed biopsies of pediatric patients with Crohn's disease (CD) and ulcerative colitis (UC).
In healthy tissue, expression of IL-17A is highest in the ileum, with IFN-γ expression lowest in the colon. Compared to healthy sections, CD patients displayed increased IL-12p35 and IFN-γ levels in noninflamed ileum and colon, respectively. Modifications of cytokine expression between noninflamed and inflamed tissues was characterized by increased IL-17A in UC colon, IFN-γ in CD colon, and IL-17A, IFN-γ and IL-6 in CD ileum. Elevated IL-17A levels were positively correlated with IFN-γ in both inflammatory CD and UC but IL-17A and IFN-γ were correlated with IL-23p19 in CD ileum only.
The expression of Th1 and Th17 cytokines varies along the intestine, indicating local specific regulation mechanisms. However, the cytokine expression patterns in the same tissue depends on the pathology, with a Th1 or a Th17 profile in the colon of CD and UC patients, respectively, and a Th1/Th17 profile in the ileum of CD patients. This indicates overlapping but distinct immune mechanisms driving intestinal inflammation in these two pathologies. (Inflamm Bowel Dis 2011)
Inflammatory bowel diseases (IBDs) comprise several chronic inflammatory disorders affecting the intestinal mucosa, with ulcerative colitis (UC) and Crohn's disease (CD) as distinct disease entities. CD usually affects the ileum and can affect the colon, whereas UC affects only the colon.1 Traditionally, CD was considered a classical Th1-mediated disorder, with increased local expression of interleukin (IL)-12 (p35/p40) and interferon-gamma (IFN-γ),2 opposed to UC, which was considered an atypical Th2 immune response.3 However, this simplistic dichotomy for IBD was challenged by the discovery of IL-23 (p19/p40), a heterodimeric cytokine composed of a p19 subunit and a common p40 subunit shared with IL-12. Indeed, similar to other experimental models of autoimmune diseases or chronic inflammation, murine models of colitis appeared to be dependent on IL-23 rather than on IL-12,4–7 and genetic studies in humans indicate an association of the IL-23R variant Arg381Gln with CD and UC.8, 9 IL-23 is an important factor for the development of Th17 cells, which are implicated in chronic inflammatory responses, and promotes the emergence of a combined Th1/Th17 phenotype during experimental colitis.10, 11 However, the respective role of the Th17 hallmark cytokines IL-17A and IL-17F and their relative importance compared to the Th1 cytokine IFN-γ in intestinal inflammation remains unclear.
Lymphopenic mice reconstituted with naive T cells develop a colitis that can be inhibited by IL-17A and IL-6 blocking antibodies,4 and Th17 cells are more colitogenic than Th1 cells in transfer models of flagellin-specific T cells.7, 12 IL-17A or IL-17F are likely to have an effect in colitis since mice deficient for the IL-17RA subunit or for its downstream adaptor Act1, which mediate the effects of both IL-17A and IL-17F,13, 14 are protected in the trinitrobenzenesulfonic acid (TNBS) and in the dextran sulfate sodium (DSS) models of colitis.15, 16 However, mice injected with blocking antibodies against IL-17A or mice deficient for IL-17A have worsened colitis severity, possibly due to uncontrolled expression of IFN-γ.17, 18 Alternatively, other reports suggest that blocking either or both cytokines can reduce intestinal inflammation.19, 20
Although several studies addressed the expression patterns of these cytokines in human IBD, their relative expression in different tissues remains unclear. IL-23, IL-17A, and IFN-γ expression in colonic tissue were increased in both active CD and UC,21–23 and IL-23 expression was correlated with IL-17 expression in UC and IFN-γ in CD.22 In the ileum of CD patients, T cells can coexpress IL-17A and IFN-γ24; however, mRNA expression of IL-17A and IL-23p19 were only mildly increased and were not compared to expression of IFN-γ.25, 26 IL-17F has been reported to be weakly expressed in UC and CD, irrespective of the location,26 or to be more strongly expressed in UC patients but to increase with inflammation only in colonic biopsies from CD patients.27
Under the hypothesis of IL-17A as a therapeutic target for IBD patients, it is particularly important to understand and precisely define the expression patterns of IL-17A, IL-17F, and IFN-γ. There is good evidence indicating that the inflammatory profile in IBD patients changes with time (disease duration) and most of the aforementioned studies were performed in adult patients with a longstanding disease; thus, we chose to address this question in pediatric IBD with disease duration of less than 1 year. Pediatric disease onset reflects a more severe and often more aggressive disease pattern28 without numerous confounding factors, complicating analyses in adult patients, making it a privileged access for the study of pathogenetic mechanisms.
MATERIALS AND METHODS
Patients and Samples
Intestinal biopsies were obtained during routine endoscopic procedures after having obtained informed consent from the patients and parents (this study was approved by our local Ethics Committee, CPP II, 05/09/11). A total of 42 IBD patients (CD n = 30, UC n = 12) and 29 noninflammatory controls were included in this study. The clinical details of the patients and controls are shown in Table 1.
|Healthy Controls||Ulcerative Colitis||Crohn's Disease|
|Number of patients||29||12||30|
|Number of samples||29||21||54|
|Mean age (SD) (years)||12 (4)||14 (3)||12 (3)|
|Mean disease duration (SD) (months)||–/–||15 (7)||16 (8)|
Real-time Polymerase Chain Reaction (PCR) Analysis of Tissue
Biopsies were immediately stored in RNALater until homogenization with a FastPrep (MP Biomedicals, Illkirsh, France). Total RNA were extracted with the RNEasyPlus kit (Qiagen, Courtaboeuf, France) following the manufacturer's recommendations. Quality of RNA was assessed by electrophoresis and RNA concentrations were determined by measuring the absorbance at 260 nm in a spectrophotometer. One μg of RNA was reverse-transcribed in the presence of 500 ng oligo(dT) and 250 ng oligonucleotides with 1 μL reverse-transcriptase M-MLV, 1 μL RNaseOUT (40U/μL), 2 μL DTT 10 mM, and 1 μL dNTP 10 mM in 4 μL of 5× buffer (Invitrogen, Courtaboeuf, France). Samples were conserved at −20°C until analysis.
Quantitative gene expression of IFN-γ, IL-6, IL-10, IL-12, IL-17A, IL-17F, IL-23, and the housekeeping gene GAPDH was performed by real-time PCR using commercially available TaqMan gene assays (Applied Biosystems, Courtaboeuf, France) according to established standard protocols. Forty cycles of amplification were performed with denaturation at 95°C for 15 seconds and annealing/extension at 60°C for 1 minute using an ABI PRISM 7300. Results were normalized to the housekeeping gene coding glyceraldehyde 3-phosphate dehydrogenase (GAPDH) using the 2-ΔΔCt method and multiplied by 106 for clarity of the figures.
Comparative statistical analyses were performed using GraphPad Prism (GraphPad Software, San Diego, CA). Tissue genes expressions were compared between patient groups using a non parametric Mann–Whitney test and relationships between parameters were analyzed by using a Spearman's rank correlation coefficient test. Values were considered significant at P < 0.05.
In a first step, we aimed to determine basal expression levels of IL-12p35, IL-23p19, and of the cytokines IL-17A, IL-17F, and IFN-γ in normal noninflammatory intestinal tissue (Fig. 1). There was a clear difference in IL-17A expression in biopsies from control individuals, with significantly higher levels in the ileum compared to duodenum (P = 0.003) or colon (P = 0.0482), while no tissue-specific difference was observed for IL-17F. IL-23p19 expression was slightly higher in ileal tissue than in the duodeum (P = 0.0745) or the colon (P = 0.0575), but this did not reach statistical significance, and IL-12p35 expression was significantly higher in the duodenum than in the ileum (P = 0,0451). A marked shift in mRNA levels was observed for IFN-γ between colonic tissue and duodenum (P = 0,0062) and ileum (P = 0,0038), while no differences were observed for IL-6 or IL-10 (data not shown).
Compared to controls, in UC patients the expression of cytokines in noninflammatory colon or ileum was not different but inflamed colon had increased expression of IL-23p19 (P = 0.0044), IFN-γ (P = 0.0418), IL-17A (P = 0.0004), and IL-17F (P = 0.0080) (Fig. 2 and data not shown). By contrast, in CD patients noninflammatory colon expressed higher levels of IFN-γ (P = 0,0021) and inflammatory colon was associated with a sustained expression of IFN-γ (P = 0.0002) and increased expression of IL-17A (P = 0.0144), but neither IL-23p19 nor IL-17F. In the ileum of CD patients, higher expression of IL-12p35 was observed in the absence of inflammation (P = 0.0398), whereas inflammatory ileum expressed significantly higher levels of each one of the measured cytokines, except IL-23p19 (IL-23p19, P = 0.0944; IL-12p35, P = 0.0056; IL-17A, P = 0.0059; IFN-γ, P = 0.0029; IL-17F, P = 0.0232; IL-6, P = 0.0241).
When comparing noninflamed with inflamed colon of IBD patients (− vs. +; Fig. 2), only IFN-γ was increased in CD patients (P = 0.0330), and only IL-17A was increased in UC patients (P = 0.0070). With regard to inflammatory lesions in the ileum of CD patients, only IL-17A, IFN-γ, and IL-6 were upregulated (P = 0.0102, P = 0.0013, and P = 0.0361 respectively) in comparison to noninflammatory CD ileum. Increased expression of the latter cytokines in the inflamed ileum of CD patients was confirmed upon stratification of CD patients according to macroscopic and histological disease severity. We observed an increased expression of IL-6 as well as IL-17A and IFN-γ with the degree of inflammation but not of IL-17F, and expression of IL-12p35 or IL-23p19 remained unchanged (data not shown).
To further define the inflammatory profile comparing UC and CD tissues, we performed correlation analyses of cytokine expression profiles in the intestinal tissues (Table 2). A positive correlation between IL-17A, IL-17F, and IFN-γ expression was observed in the inflamed colon and ileum of both UC and CD patients, but neither in controls nor in noninflamed IBD tissue. A positive correlation between IL-23p19 expression and IL-17A or IFN-γ expression was only observed in inflamed CD ileum, whereas a positive correlation between IL-12p35 and IFN-γ expression was observed in both noninflamed and inflamed ileum.
|Ulcerative Colitis||Crohn's Disease|
|IL-17A||IL-23p19||ns||ns||ns||P = 0.0031 r = 0.6404|
|IFNγ||ns||P = 0.0007 r = 0.9333||ns||P = 0.0019 r = 0.6649|
|IL-17F||ns||P = 0.0368 r = 0.7619||P = 0.0426 r = 0.6182||P = 0.0025 r = 0.7000|
|IFNγ||IL-23p19||ns||ns||ns||P = 0.0021 r = 0.6596|
|IL-12p35||ns||ns||P < 0.0001 r = 0.9580||P = 0.0047 r = 0.6193|
|IL-17F||ns||P = 0.0458 r = 0.7381||ns||P < 0.0001 r = 0.8588|
IBDs are yet incurable chronic inflammatory diseases mediated by dysregulated cytokine expression, as attested by clinical efficiency of immunosuppressive drugs and of blocking antibodies against tumor necrosis factor alpha (TNF-α) in CD.29 Genome-wide association studies indicate that IL23R, IL12B, JAK2, and STAT3 are associated with susceptibility to CD and UC,30 thus pointing out the importance of IL-23 signaling pathways and the Th17 response in intestinal inflammation.
Despite their importance in IBD, only a limited number of studies have addressed expression of Th1- and Th17-related cytokines in the different parts of healthy intestine. In contrast to one of these studies reporting only a few differences between ileum and colon,26 we observed that children without any intestinal inflammation had significantly higher levels of IFN-γ and IL-17A mRNA expression in the ileum compared to colon. This difference between small and large bowel probably reflects the presence of IFN-γ- and IL-17A-producing cells in the small intestine, as reported in mice,31, 32 resulting from ileum-restricted interactions with specific bacteria.33, 34 In addition, although not significant, the increased IL-23p19 mRNA expression in the ileum is consistent with a study showing preferential expression of IL-23 in the distal ileum of mice.35 Unexpectedly, although IL-17A and IL-17F are considered to be concomitantly regulated,36, 37 expression of IL-17F was not increased in the ileum. Strong TCR stimulation can preferentially promote expression of IL-17A rather than IL-17F38; however, this mechanism is unlikely to occur in the ileum, where IL-17A expression driven by microbial colonization is largely TCR-independent.39 In our study, expression of the proinflammatory cytokine IL-6 and the regulatory cytokine IL-10 were also comparable in the different segments of the gut, suggesting that basal expression of these cytokines is probably less influenced by the local environment. Given the observed differences of cytokine expression between the ileum and the colon of healthy control children, we compared noninflamed and inflamed tissue of IBD patients with tissue-matched controls.
The expression profile of cytokines in noninflammatory tissues of healthy controls and in UC patients was comparable. By contrast, noninflamed tissue of CD patients was prone to express IL-12p35 in the ileum and IFN-γ in the colon. Yet p35 might not only account for IL-12 expression in the ileum, since p35 is also expressed together with EBI3 to form IL-35, a presumably antiinflammatory cytokine preferentially expressed at higher levels in the small intestine than in large bowel of mice.40, 41 However, higher levels of IFN-γ in noninflamed colon of CD patients, but not UC patients, suggests that the intestinal colon of children with CD is maintained in a Th1-skewed response. Accordingly, a positive correlation was found between IFN-γ and IL-12p35 expression in noninflamed and inflamed ileum of CD patients. Interestingly, compared to controls and regardless of the tissue, only IFN-γ and IL-17A expression were increased in inflamed biopsies of both UC and CD patients. A positive correlation between expression of IFN-γ and IL-17A was also observed in both diseases. However, the cytokine shift between noninflamed and inflamed tissue delineates different tissue- and disease-dependent cytokine patterns. Inflammatory colon of UC expressed higher levels of IL-17A only, whereas in CD patients inflammatory colon expressed higher levels of IFN-γ only and inflammatory ileum was associated with an increased expression of IL-17A, IFN-γ, and IL-6. Therefore, these cytokine patterns reflect a Th1 profile in colonic CD and a Th17 profile in UC, whereas the inflamed ileum of CD patients had a cytokine expression profile more evocative of a combined Th1/Th17 profile.
IL-23 can promote the emergence of a double-positive phenotype with a concomitant expression of IFN-γ and IL-17, which is observed both in experimental models of colitis and in CD.11, 24 Surprisingly, IL-23p19 expression was positively correlated with expression of IL-17A and IFN-γ only in the inflamed ileum of CD patients, where these two cytokines are increased. We also observed an elevated expression of IL-6 in the inflamed ileum of CD patients. IL-6 can induce the differentiation of Th17 cells,42 and its expression is associated with colitis in several models.4, 19 Despite IL-17F expression being correlated with expression of IL-17A and IFN-γ in both diseases, this cytokine is not further increased with inflammation, suggesting that its association with the disease is less pronounced than the one observed for IL-17A and IFN-γ.
The exact mechanisms of IFN-γ or IL-17A alone or simultaneously impacting inflammatory reactions in different tissues are poorly understood. IL-17A can act in synergy with IFN-γ43 but whether this occurs in the intestinal tissue is unclear. It has been reported that IL-17A could also restrain expression of IFN-γ.18 It is therefore possible that concomitant expression of IFN-γ and IL-17 in the ileum of CD patients mediates unique inflammatory effects.
In patients with IBD, clinical trials based on targeting the p40 subunit, shared by IL-12 and IL-23, proved encouraging44 and is efficient in patients unresponsive to anti-TNF-α therapies.45 The involvement of IL-23, rather than IL-12, in different inflammatory pathologies, together with the association of IL-23 with Th17 responses, gave rise to developing therapies targeting IL-17A directly.46 Recently, a first clinical trial indicated beneficial effects of anti-IL-17A treatments in several Th17-associated diseases.47 However, therapies based on the inhibition of the Th17 pathway did not always fulfill the expected results,48 and a recent study presented at the 6th Congress of ECCO49 indicated that treatment of patients with CD by anti-IL-17A blocking antibodies were inefficient or even worsened the disease course. Inefficiency of anti-IL-17A treatment would then contrast with the relative efficiency of anti-IFN-γ in CD,50 and if IL-17A acts as a protective brake against inflammatory effects mediated by IFN-γ, it might be interesting to determine if increased levels of IFN-γ are detected in patients receiving anti-IL-17 antibodies. Our results, although not covering the variety of cytokines that likely acts together in the gut, indicate that IL-17A and IFN-γ have distinct expression patterns depending on the localization of the lesions and the disease. Further studies might therefore indicate if these distinct patterns have an impact on the physiopathology of IBD in these patients.
We thank all patients for their participation in this study. The authors thank the staff of endoscopic exploration for providing intestinal samples and the U989 laboratory for stimulating discussions.
- 27Role of the novel Th17 cytokine IL-17F in inflammatory bowel disease (IBD): upregulated colonic IL-17F expression in active Crohn's disease and analysis of the IL17F p.His161Arg polymorphism in IBD. Inflamm Bowel Dis. 2008. 14: 437–445., , , et al.
- 49Inhibition of IL-17A by secukinumab is ineffective for Crohn's disease (CD). Journal of Crohn's and Colitis. 2011. 5: S1–S192., , , et al.