Defective Paneth Cell Mediated Host Defense in Pediatric Ileal Crohn s Disease

Transcription

1 452 ORIGINAL CONTRIBUTIONS nature publishing group Defective Paneth Cell Mediated Host Defee in Pediatric Ileal Crohn s Disease G ø r i Per minow, M D 1, 6, Ju l i a B e is ne r, Ph D 2, 6, Mau re e n Ko sl ow sk i, M S c 2, L ars Gust av Lyck and e r, M D 3, E du ard St ange, M D, Ph D 4, Morten H. Vatn, MD, PhD5 and Jan Weh k amp, M D 2, 4 OBJECTIVES: METHODS: Adult ileal Crohn s disease (CD) is characterized by a specific decrease in ileal Paneth cell -defei. In addition to NOD2, we previously identified a disturbance of the Wnt-signaling tracription factor TCF-4 as a major mechanism for this deficiency. The aim of this study was to evaluate human -defein-5 (HD-5) and TCF-4 in an independent cohort of pediatric CD patients. Expression levels of HD-5 and TCF-4 mrna were quantified by real-time PCR in biopsies from newly diagnosed untreated pediatric CD patients ( < 18 years, n = 36) and age-matched symptomatic non-inflammatory bowel disease controls with a histologically normal gut ( n = 29). To assess the influence of current inflammation, mucosal interleukin-8 (IL-8) and fecal calprotectin levels were determined. RESULTS: Small intestinal HD-5 and TCF-4 mrna were significantly reduced in pediatric ileal CD (L1 + L3) ( P =.22 and P =.5, respectively) and were significantly correlated ( r =.499; P =.1). In ileal but not colonic CD, TCF-4 was also reduced in the colon ( P =.5). Importantly, both HD-5 and TCF-4 were independent of inflammation, as measured by IL-8 or fecal calprotectin. In contrast to the small intestine, colonic Paneth cell HD-5 mrna was significantly elevated in colonic CD (L2) ( P =.26) and was correlated with fecal calprotectin levels ( r =.481; P =.2). CONCLUSIONS: In this study, we describe a specific decrease in HD-5 and TCF-4 mrna levels in children with ileal CD. In the small intestine, this decrease was independent of current inflammation, whereas inflammation seems to induce Paneth cell metaplasia in the colon. Our data extend the hypothesis of an important role of antimicrobial host defee in pediatric CD patients. Am J Gastroenterol 21; 15: ; doi:1.138/ajg ; published online 1 November 29 INTRODUCTION Inflammatory bowel disease (IBD), a chronic inflammation of the intestine, is commonly classified into ulcerative colitis and Crohn s disease (CD) on the basis of clinical features and histopathology. The pathogenesis of the disease is not fully understood, but during recent years there has been increasing evidence that a breakdown of the mucosal barrier function leads to increased immune activation and chronic inflammation in genetically predisposed individuals (1). IBD is common in most industrialized countries and childhood IBD accounts for nearly 3 % of total cases (2,3). The phenotype in early-oet IBD differs from adult-oet disease and is characterized by a more aggressive disease. The disease location exhibits a more dominant colonic involvement, with fewer young patients exhibiting an isolated purely ileal location. Stratifying adult patients according to phenotypic distinctio, e.g., location, has allowed the identification of distinct mechanisms in pathogenesis, but pediatric patients have not yet been studied (4). 1 Department of Pediatrics, Akershus University Hospital and Oslo University Hospital, Oslo, Norway ; 2 Dr Margarete Fischer-Bosch-Ititute of Clinical Pharmacology and Universitiy of T ü bingen, Stuttgart, Germany ; 3 Department of Pathology, Akershus University Hospital, Oslo, Norway ; 4 Department of Internal Medicine I, Robert-Bosch-Hospital Stuttgart, Stuttgart, Germany ; 5 Medical Department, Rikshospitalet, University of Oslo and Epigen, Akershus University Hospital, Oslo, Norway ; 6 These authors contributed equally to this work. Correspondence: Jan Wehkamp, MD, Dr. Margarete Fischer-Bosch-Ititute of Clinical Pharmacology, Auerbachstr. 112, 7376, Stuttgart, Germany. jan.wehkamp@ikp-stuttgart.de Received 17 June 29; accepted 3 September 29 The American Journal of GASTROENTEROLOGY VOLUME 15 FEBRUARY 21

2 TCF-4 and HD-5 in Pediatric Crohn s Disease 453 Different lines of evidence in animals and adult patients support the hypothesis that the antimicrobial barrier of the mucosa has a key role in the pathogenesis of CD. In patients with CD, a disturbed antimicrobial peptide shield has been suggested to result in the mucosal invasion of intestinal bacteria, as well as in alteratio of the luminal bacterial composition, leading to secondary activation of immune cells and mucosal inflammation (5,6). The two major types of CD, ileal and colonic localization, are both associated with a specific and distinct defect in defei, a group of antimicrobial peptides exhibiting broad microbicidal activity agait Gram-negative and Gram-positive bacteria, fungi, viruses, and protozoa (7,8). Defei regulate the composition and number of colonizing microbes and protect the epithelial cell barrier from bacterial invasion. Colonic CD is associated with an impaired induction of mucosal β -defei, partly because of a low copy number of the β -defein gene cluster (9,1). In contrast, pure ileal and ileal-colonic CD (Vienna Classification L1 and L3, respectively) is characterized by a specific decrease in small intestinal Paneth cell human α -defein (HD)-5 and -6 (11,12). Paneth cells are secretory epithelial cells at the base of crypts and are the major source of antimicrobial peptides in the small intestine (13,14). In case of disease, metaplastic Paneth cells can also be found in other parts of the intestinal tract, including the stomach, esophagus, and the colon (15 17). Their differentiation and maturation in the intestine is, to a large extent, mediated and tightly controlled by effector molecules of the Wnt-signaling pathway (18). We recently identified TCF-4 as a new factor in the pathogenesis of ileal CD, with adult patients with ileal CD showing a reduced of this Wnt-signaling tracription factor, which directly controls Paneth cell defein (19). Besides a specific reduction of TCF-4 tracription factor and binding, small intestinal involvement of CD is associated with a genetic polymorphism in the putative TCF- 4 promoter region (2). The purpose of this investigation was to study the of α -defein HD-5 and Wnt-signaling tracription factor TCF-4 in treatment-naive pediatric patients. In addition, we aimed to study the role of inflammation for small intestinal and colonic Paneth cell HD-5. population based in the catchment area, with no leakage of patients to surrounding health regio. A total of 39 pediatric patients with CD were included in the study, as well as 38 non-ibd controls. During endoscopy, several biopsies were taken from different segments for histological evaluation. The diagnosis of IBD was confirmed by histopathological verification of an expert pathologist (Lars Gustav Lyckander) according to the Porto criteria (21). Children with no evidence of IBD or mucosal inflammation, which was determined by either endoscopy or histopathological examination, served as a non-ibd control group. The final diagnoses of non- IBD children were recurrent abdominal pain or irritable bowel syndrome. Biopsies from the terminal ileum and colon ascende were obtained from 36 patients with CD and from 29 non- IBD controls. Biopsies were taken from noninflamed mucosa, if possible, and also from inflamed mucosa of IBD patients; altogether 15 small intestinal biopsies and 4 colonic biopsies were macroscopically inflamed. Following the Vienna classification, three subgroups of CD were defined: ileal disease only (L1), colonic disease only (L2), and disease (L3). Pediatric patients with CD were also classified according to disease behavior. In all, 3 pediatric CD patients had a notricturing and nonpenetrating disease type (B1) and 6 had a stricturing behavior (B2). Histological grading of routine biopsies was performed by a pathologist and the degree of inflammation was determined differentiating between normal mucosa (score ), no active inflammation (score 1), mild active inflammation (score 2), moderate active inflammation (score 3), and severe active inflammation staining (score 4). Stool samples were taken and the possibility of infectious diseases was excluded. Fecal calprotectin was determined as previously described (22). Characteristics of the pediatric study population are summarized in Table 1. Tissue preparation and histological evaluation Colonoscopy was performed and mucosal biopsies were harvested from the terminal ileum and from six colonic segments for histopathological diagnosis and evaluated by an expert pathologist. Additional study biopsies from the ileum and ascending colon were collected, directly snap-frozen in liquid nitrogen, and stored at 7 C until further preparation. METHODS Patients Coecutive undiagnosed and untreated pediatric patients ( < 18 years old) were screened ( n = 1) prospectively between May 25 and December 27. The screening was part of a larger study in Inflammatory Bowel South-Eastern Norway (IBSEN-II), investigating genetic, immunological, and environmental factors that contribute to IBD etiology, in which adults were also included. Pediatric patients were admitted to five different hospitals in the region after invitation to all general practitioners in the area to refer coecutive patients with symptoms suspicious of IBD. The pediatric cohort is a Isolation of total RNA and real-time PCR Frozen biopsies were disrupted mechanically and total RNA was isolated using TRIzol reagent according to the supplier s protocol. RNA was reverse tracribed with Superscript II reverse tracriptase (Invitrogen, Carlsbad, CA) into cdna, according to the standard protocol of the manufacturer. cdna derived from 1 ng of total RNA served as a template for real-time PCR reaction. The levels of HD-5, TCF-4, interleukin-8 (IL-8), and GAPDH were quantified by real-time PCR using a fluorescence detection monitor (Light- Cycler; Roche Diag nostics, Mannheim, Germany) as previously described (19). All levels were normalized with respect to the of GAPDH. 21 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY

3 454 Perminow et al. Table 1. Characteristics of the pediatric patient cohort Non-IBD Number of patients Gender Female Male 2 11 Median age in years (range) 13.8 ( ) 12.2 ( ) Disease location L1 3 (8.3 % ) L2 9 (25 % ) L3 24 (66.7 % ) Disease behavior B1 3 (83.3 % ) B2 6 (16.7 % ) B3 ( % ) Fecal calprotectin in mg / kg (range) 1,278 (13 8,625) 3 (1 1,26) B1, non-stricturing, non-penetrating behavior; B2, structuring behavior; B3, penetrating behaviour; IBD, inflammatory bowel disease; L1, ileal disease only; L2, colonic disease only; L3, disease. DNA isolation and genotyping We obtained DNA from blood samples from 74 cases, 37 non- IBD and 37 CD patients. DNA was isolated with standard procedures as described (2). Matrix-Assisted-Laser-Desorption/ Ionization Time of Flight-Mass-Spectrometry (MALDI-TOF- MS)-based genotyping of DNA samples was carried out using a Mass ARRAY Compact System from Sequenom (San Diego, CA). Primers were designed using reference sequence NT 359 as described (2). Statistical analysis To calculate statistically significant differences between groups, samples were analyzed by unpaired Student s t-test for values with Gaussian distribution and by Mann Whitney test for values without normal distribution. Gaussian distribution was determined using the Kolmogorov Smirnov test. Spearman s rank correlation coefficient was calculated for correlatio between samples. Results are presented as mean ± s.e. Values of P >.5 were coidered to be statistically significant. Data were analyzed using Graphpad Prism version 4 (Graphpad Software, San Diego, CA) and SPSS software (SPSS, Chicago, IL). Ethical coideratio The study was carried out in accordance with the Declaration of Helsinki. All procedures, including the collected patient material, were approved by the Committee for Medical Research Ethics (Health Region East, Oslo, Norway). Written informed coent was obtained from all patients and from an additional parent. RESULTS To investigate the of α -defein HD-5 and TCF- 4 in pediatric CD patients, we studied a Norwegian cohort of untreated patients. Demographic data of the study population are presented in Table 1. A total of 36 coecutive pediatric patients with CD were included in the study. Age-matched controls were pediatric symptomatic non-ibd patients with a histologically normal gut ( n = 29). The median age of patients at disease oet was 13 years, with a wide range of years. The median and ranges of fecal calprotectin concentratio in CD patients and in the control group are shown in Table 1. Pediatric patients with CD had a high fecal calprotection level with a median concentration of 1278 mg / kg, whereas non-ibd controls showed a median concentration of 3 mg / kg. Fecal calprotectin was significantly increased in pediatric patients with CD compared with non-ibd controls ( Figure 1a, P.1) and was correlated with ileal mucosal IL-8 ( Figure 1b ; r =.556, P <.1), reflecting intestinal mucosa inflammation in these children. In contrast, in the colon, mucosal IL-8 levels did not correlate with fecal calprotectin ( Figure 1c ). HD-5 and TCF-4 is specifically reduced in pediatric patients with ileal CD We measured Paneth cell α -defein HD-5 and Wnt-signaling TCF-4 mrna in ileal biopsies from pediatric patients. Three subgroups were defined following the Vienna classification: ileal disease only (L1), colonic disease only (L2), and disease (L3). Coistent with the findings in adults, the HD-5 in small intestinal biopsies was found to be significantly reduced in CD patients with ileal involvement (L1 + L3, n = 2) compared with that in controls ( n = 27) (P =.22, Figure 2a ). In pediatric patients with pure colonic CD (L2), the ileal HD-5 was unchanged. We next examined mrna of the Wnt-signaling tracription factor TCF-4 in ileal biopsies, which was also significantly reduced in CD patients with ileal and localization (L1 + L3, n = 2) compared with non-ibd controls ( P =.5, Figure 2b ). In contrast, the TCF-4 mrna level was unchanged in ileal biopsies from patients with colonic CD (L2) compared with controls ( Figure 2b ). TCF-4 mrna was highly correlated with HD- 5 mrna in all pediatric patients ( r =.499, P =.1; Figure 2c ). To assess the influence of current inflammation, of the proinflammatory cytokine IL-8 was determined in the mucosa, as well as in fecal calprotectin, a marker for intestinal inflammation. Importantly, both HD-5 and TCF-4 levels in pediatric CD patients did not show any correlation with inflammation markers, or with mucosal IL-8 or fecal calprotectin levels ( Figures 3a and b ). Furthermore, HD-5 and TCF-4 levels were also independent of the histological degree of inflammation (data not shown). The American Journal of GASTROENTEROLOGY VOLUME 15 FEBRUARY 21

4 TCF-4 and HD-5 in Pediatric Crohn s Disease 455 a Fecal calprotectin protein P.1 n=26 n=25 n=9 controls ileal + colonic b Relative IL-8 mrna c Colonic biopsies.6 r =.556 r =.186 P<.1.5 2,5 5, 7,5 1, Relative IL-8 mrna ,5 5, 7,5 1, Figure 1. Fecal calprotectin level and infl ammatory marker interleukin-8 (IL-8) in study samples. Patients were divided into three groups: non-inflammatory bowel disease (IBD) controls, pediatric Crohn s disease (CD) patients with ileal and presentation (Vienna, L1 + L3), and pediatric CD patients with colonic presentation (L2) ( a ). Data represent mean ± s.e. Correlation analysis of the fecal calprotectin level and mucosal IL-8 mrna in the small intestine ( b, left panel) and in the colon ( c, right panel). Spearman s rank correlation coeffi cient ( r ) and statistical signifi cance ( P ) were calculated using GraphPad Prism version 4. (GraphPad Software). Values of P <.5 were coidered to be statistically signifi cant. Paneth cell α -defein HD-5 in the colonic mucosa In contrast to the ileum, there was no significant difference in HD-5 mrna in the colonic biopsies from pediatric CD patients with ileal involvement (L1 + L3, n = 18). These patients showed slightly higher HD-5 mrna levels compared with non-ibd controls ( n = 19) (Figure 4a ). In contrast, in pure colonic CD (L2), the Paneth cell HD-5 mrna was significantly upregulated in colonic biopsies ( P =.26; Figure 4a ). The increase in colonic HD-5 in pediatric CD patients correlated with the fecal calprotectin level ( r =.481, P =.2; Figure 4b ) but not with local histological inflammatory scores (data not shown), suggesting that the general level of intestinal inflammation might trigger colonic Paneth cell metaplasia. TCF-4 in colonic mucosa In line with previous observatio, patients with pure colonic localization of CD (L2) did not show a significant difference in the mrna of TCF-4 ( Figure 4c ). In contrast, pediatric CD patients who had a small intestinal involvement (L1 + L3) and a significantly reduced of TCF-4 in the small intestine also showed a reduced TCF-4 in the colonic mucosa ( P =.5; Figure 4c ). Thus the decrease was not tissue specific, arguing for a primary importance of this tracription factor in case of this location-based phenotype in children. In line with an inflammatory, i.e., Paneth cell metaplasia-triggered in the colon, TCF-4 and HD-5 were correlated in the small intestine, but not in the colonic mucosa (data not shown). Genotyping for rs In this limited patient population, rs showed no significant association with CD (2). Pediatric CD patients had a T-allele frequency of % compared with non-ibd controls with 2.27 %. The distribution rates of genotypes among pediatric CD patients were 59.5 % CC, 37.8 % CT, and 2.7 % TT, and those among non-ibd controls were 62.2 % CC, 35.1 % CT, and 2.7 % TT. Owing to the small number of pediatric patients with ileal CD, we did not further analyze the data with respect to disease location. Because of the relatively small sample size, the data may not allow to draw conclusio about the impact of promoter variant rs allele on TCF-4 in pediatric patients, and additional studies in larger cohorts are necessary to explore the effects. DISCUSSION In this study, we assessed the profile of Paneth cell-mediated host defee in a Norwegian pediatric population. In line with previous studies in adult patients, we found a significant decrease in Paneth cell defein HD-5 21 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY

5 456 Perminow et al. a 1 P =.22 b.15 P =.5 Relative HD-5/GAPDH n=27 n=2 n=6 controls ileal + colonic Relative TCF4/GAPDH.1.5. n=29 n=2 n=6 controls ileal + colonic c Relative HD-5 mrna r =.499 P = Relative TCF-4 mrna Figure 2. Expression of Paneth cell human α -defein-5 (HD-5) and Wnt-signaling TCF-4 in the ileal mucosa. HD-5 ( a ) and Wnt TCF-4 ( b ) mrna were quantifi ed by LightCycler real-time reverse tracriptase PCR in a collection of ileal intestinal biopsies from non-infl ammatory bowel disease (IBD) controls, pediatric Crohn s disease (CD) patients with ileal and presentation (Vienna, L1 + L3), and pediatric CD patients with isolated colonic presentation (L2) at diagnosis. Expression of both genes was normalized to the corresponding level of GAPDH. Data represent the mean values ± s.e. ( c ) Correlation analysis of HD-5 and TCF-4 in all pediatric patients. Spearman s rank correlation coeffi cient ( r ) and statistical signifi - cance ( P ) were calculated using GraphPad Prism version 4. (GraphPad Software). Values of P <.5 were coidered to be statistically significant. in case of small intestinal involvement of CD. In addition, these children were characterized by a reduction of TCF-4, a tracription factor that directly binds to the promoter region of Paneth cell defei. The levels of HD-5 and TCF-4 in the small intestine showed a clear correlation. In contrast, there was no correlation with inflammation markers fecal calprotectin and mucosal IL-8, which was also confirmed at the histological level. In the colon, Paneth cell defein HD-5 was increased in case of colonic involvement because of metaplastic Paneth cells. At this site and in contrast to the small intestine, the colonic Paneth cell HD-5 was correlated with fecal calprotectin, suggesting that Paneth cell metaplasia is an additional protection mechanism in intestinal inflammation. Colonic TCF-4 was unchanged in case of pure colonic disease, but also significantly diminished in case of small intestinal involvement. There is increasing interest in investigating the differences between adult and pediatric patients, as childhood disease seems to be characterized by special clinical and pathogenetic aspects. It is especially noted that it is the youngest children ( < 8 years) who more often show a colonic phenotype at the oet of disease (23 25). The differences in phenotypes strongly suggest that the mechanisms of disease pathogenesis might be different in children. It is noteworthy that the clinical phenotype is the only clinical parameter that seems to be stable over time and does not change with disease progression (such as stenosis or fistulation) (26). We observed a decrease in Paneth cell defein HD-5 in young children with ileal CD at oet. This suggests that the defect in defei might be a key factor that actually causes the disease, probably by leading to a reduced mucosal barrier function toward commeal bacteria. Another recent investigation based on young patients from Italy already described a reduced Paneth cell defein in children with CD not differentiating between disease location (27). The cohort presented here is from Norway, mainly younger patients from Oslo, and was characterized by an HD-5 decrease in patients with small intestinal involvement. The effect seems to be less pronounced than in adults, which might also be partially attributed to the relatively low frequency of pure ileal disease in childhood oet of CD. Besides the confirmation of defein deficiency, this is the first description of impairment in Wnt / TCF-4 in pediatric CD patients. In contrast to adult patients and of special note, the decrease in TCF-4 was phenotype dependent but tissue spanning: it was found in the small intestine and in the colon, but only in case of small intestinal involvement. This observation argues for a primary role of Wnt tracription factor TCF-4, especially in children with early oet of CD. The American Journal of GASTROENTEROLOGY VOLUME 15 FEBRUARY 21

6 TCF-4 and HD-5 in Pediatric Crohn s Disease 457 a Relative HD-5 mrna Fecal calprotectin r =.219 2,5 5, 7,5 1, Relative HD-5 mrna Mucosal IL-8 r = Relative IL-8 mrna b Relative TCF-4 mrna r =.283 2,5 5, 7,5 1, Relative TCF-4 mrna.2 r = Relative IL-8 mrna Figure 3. Human α -defein-5 (HD-5) and TCF-4 levels in ileal biopsies are not correlated with fecal calprotectin and mucosal interleukin-8 (IL-8). Correlation analysis of HD-5 ( a ) and TCF-4 ( b ) mrna with either fecal calprotectin (left panel) or mucosal IL-8 (right panel) in pediatric Crohn s disease (CD) patients was performed. Spearman s rank correlation coeffi cient ( r ) and statistical signifi cance ( P ) were calculated using GraphPad Prism version 4. (GraphPad Software). The value of P <.5 was coidered to be statistically signifi cant. An important issue is the question whether the observed changes are primary or possibly secondary to inflammation. We tried to assess this question by measuring inflammatory markers in the tissue, as well as in the stool. Calprotectin is abundant in neutrophil granulocytes, monocytes, and macrophages and is measureable in the blood and other body fluids, as well as in stool samples. Fecal calprotectin has been shown to be a valid and easily accessible (stool) inflammatory marker. In healthy children, a cutoff value of < 5 mg / kg has been suggested (22). We found a median fecal calprotectin level in our non-ibd controls of 3 mg / kg. In contrast to adult patients, the recruitment of healthy controls is limited if not impossible, which might bias some of the observatio. All children controls were clinically symptomatic, even though they showed no sig of mucosal inflammation (as assessed by a gastrointestinal pathologist). Thus it might be speculated that some of these controls are future IBD patients. This might partly explain why some of the changes in defein are less pronounced in children than in adult patients, even though they were statistically significant ( Figure 2a ). As expected and supporting the inflammatory diagnosis, fecal calprotectin levels in CD specime (median 1278 mg / kg) were significantly increased compared with those in controls ( P =.1, Figure 1a ). To assess mucosal inflammation, we also measured inflammatory cytokine IL-8 in the biopsies. It has been shown that IL-8 correlates with the grade of histological inflammation as assessed by a blinded pathologist (12). Supporting the clinical value of both markers, calprotectin in the stool, as well as mucosal IL-8, showed a good correlation in the small intestine ( Figure 1 ). In contrast, there was no correlation with either small intestinal HD-5 or TCF-4, arguing that the grade of inflammation does not account for the decrease in this patient population. This is in line with other studies in CD patients (12), as well as in those with non-cd-ileitis such as ulcerative colitis pouchitis and graft-vs.-host ileitis, both independent and dependent on NOD2 status (12,28,29). In contrast to the small intestine, of Paneth cell defei in the colon is rare and limited to Paneth cell metaplasia (3). The recruitment of metaplastic (non-small intestinal) Paneth cells has been described at different sites in case of intestinal inflammation such as esophagitis, gastritis, and IBDcolitis, with no difference between CD and ulcerative colitis (31). We report here that, similar to adults, the Paneth cell HD-5 is significantly elevated in pediatric CD, with isolated colonic presentation at oet of diagnosis demotrating the presence of metaplastic Paneth cells. As expected, colonic HD-5 in non-ibd controls is very low because of the few number of Paneth cells physiologically located at this site. In contrast to the small intestine, colonic HD-5 in pediatric CD patients was significantly correlated with fecal calprotectin, whereas there was no dependency on local histological inflammation, suggesting that the level of inflammation throughout the intestinal tract might contribute to this effect rather than local inflammatory respoes. In summary, these data show that colonic Paneth cell metaplasia is dependent on clinical 21 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY

7 458 Perminow et al. a Relative HD-5/GAPDH Colonic biopsies P =.26 b Relative HD-5 mrna. n=19 n=18 n=7 2,5 5, 7,5 1, controls ileal + colonic Colonic biopsies r =.481 P =.2 c Relative TCF4/GAPDH Colonic biopsies P =.5 n=19 n=18 n=7 controls ileal + colonic Figure 4. Colonic of Paneth cell human α -defein-5 (HD-5) and TCF-4. HD-5 ( a ) and TCF-4 ( c ) mrna levels were quantified by LightCycler real-time reverse tracriptase PCR in colonic biopsies from non-inflammatory bowel disease (IBD) controls, pediatric Crohn s disease (CD) patients with ileal and presentation (Vienna, L1 + L3), and pediatric CD patients with isolated colonic presentation (L2) at diagnosis. Expression was normalized to the corresponding level of GAPDH. Data represent the mean values ± s.e. ( b ) Correlation analysis of HD-5 and fecal calprotectin levels in pediatric CD patients. Spearman s rank correlation coeffi cient ( r ) and statistical signifi cance ( P ) were calculated using GraphPad Prism version 4. (GraphPad Software). Values of P <.5 were coidered to be statistically signifi cant. phenotype, as well as on the extent of mucosal inflammation as assessed by fecal calprotectin. These findings suggest that Paneth cell metaplasia is an important additional protection system in children with colonic disease. In conclusion, this study confirms that pediatric patients with ileal and disease exhibit a significantly lower of HD-5 and TCF-4 in the small intestinal mucosa. Our results suggest a similar α-defein deficiency in childhood CD with affection of the small intestine as previously reported in adults ( Figure 2 ) (9). Besides NOD2, ATGL1, and possibly XBP1, TCF-4 (TCF7L2) is another genetic link of small intestinal CD and Paneth cell antimicrobial host defee (2,32 36). This is the first study describing an influence of the Wnt-signaling pathway in children and it is possible that other upstream factors in this pathway may account for the observed effects. In contrast, Crohn s colitis is typically characterized by an impaired induction of β-defei HBD2 and HBD3 (37,38), which is associated with fewer gene copy numbers in the β -defein gene locus on chromosome 8 (1). This observation has also been confirmed and extended in a European pediatric population (39). A better understanding of the exact molecular mechanisms underlying ileal and colonic CD both in children and adults may give rise to new therapeutic and preventive strategies focusing on stimulating protective innate immune factors. ACKNOWLEDGMENTS We thank the members of the IBSEN-II study group for the recruitment of patients and specime, Jorunn Bratlie at Oslo University Hospital, Rikshospitalet for expert technical assistance, and Gunn Seim Ekeland at AHUS for assisting with the administrative work. We thank K. Siegel and J. Bader for excellent technical assistance. CONFLICT OF INTEREST Guarantor of the article: Jan Wehkamp, MD. Specific author contributio : Study design: Gøri Perminow, Julia Beisner, Morten H. Vatn, Eduard F. Stange, and Jan Wehkamp; performance of the experiments: G ø ri Perminow, Julia Beisner, and Lars Gustav Lyckander; data analysis: Julia Beisner, G ø ri Perminow, Maureen Koslowski, and Jan Wehkamp; manuscript writing: Julia Beisner, G ø ri Perminow, Morten H. Vatn, Eduard F. Stange, and Jan Wehkamp. Financial support : The study was supported by a Norwegian Helse Ø st research grant, as well as by the Robert Bosch Foundation (Stuttgart, Germany), the Deutsche Forschungsgemeichaft (DFG), and the DCCV (German patient organization of CD and ulcerative colitis). Jan Wehkamp is an Emmy Noether scholar of the Deutsche Forschungsgemeichaft (WE 436 / 1-1). Potential competing interest : None. The American Journal of GASTROENTEROLOGY VOLUME 15 FEBRUARY 21

8 TCF-4 and HD-5 in Pediatric Crohn s Disease 459 Study Highlights WHAT IS CURRENT KNOWLEDGE 3 Paneth cell α -defei are specifically reduced in the small intestine in adult ileal Crohn s disease. 3 In adults, in addition to NOD2, a major mechanism for this reduction is mediated by the Wnt-signaling tracription factor TCF-4 (TCF7L2), which regulates the differentiation from epithelial stem cells to Paneth cells. 3 Paneth cell metaplasia occurs in the adult IBD colon, which results in an increase in the colonic of Paneth cell defei. WHAT IS NEW HERE 3 In pediatric, treatment-naive patients with ileal Crohn s disease, the decrease in Paneth cell small intestinal HD-5 seems to be mediated by Wnt signaling / TCF-4, similar to adults. 3 This provides evidence that Wnt signaling and likely stem cell differentiation to Paneth cells have primary roles in regulating innate antimicrobial host defee in early disease pathogenesis. 3 Colonic Paneth cell metaplasia in children with Crohn s disease seems to be a protective factor triggered by inflammation. REFERENCES 1. S ch re ib e r S, R os e t iel P, A lbre cht M et al. Genetics of, an archetypal inflammatory barrier disease. Nat Rev Genet 25 ;6 : Russel MG. Changes in the incidence of inflammatory bowel disease: what does it mean? Eur J Intern Med 2 ;11 : Papp a H M, S em r i n G, Wa l ke r T R et al. Pediatric inflammatory bowel disease. Curr Opin Gastroenterol 24 ;2 : Weh k amp J, Kosl ow sk i M, Wang G et al. Barrier dysfunction due to distinct defein deficiencies in small intestinal and colonic. Mucosal Immunol 28 ;1 (Suppl 1) : S Wehkamp J, Schmid M, Stange EF. Defei and other antimicrobial peptides in inflammatory bowel disease. Curr Opin Gastroenterol 27 ;23 : S ar tor R B. Therapeutic manipulation of the enteric microflora in inflammatory bowel diseases: antibiotics, probiotics, and prebiotics. Gastroenterology 24 ;126 : Huttner KM, Bevi CL. Antimicrobial peptides as mediators of epithelial host defee. Pediatr Res 1999 ;45 : Harder J, Glaser R, Schroder JM. Human antimicrobial protei effectors of innate immunity. J Endotoxin Res 27 ;13 : Weh k amp J, Fel l e r man n K, Her rl i nge r K R et al. Human beta-defein 2 but not beta-defein 1 is expressed preferentially in colonic mucosa of inflammatory bowel disease. Eur J Gastroenterol Hepatol 22 ;14 : Fel l e r man n K, St ange DE, S cha e ffeler E et al. A chromosome 8 gene-cluster polymorphism with low human beta-defein 2 gene copy number predisposes to of the colon. Am J Hum Genet 26 ;79 : Wehkamp J, Harder J, Weichenthal M et al. NOD2 (CARD15) mutatio in are associated with diminished mucosal a-defein. Gut 24 ;53 : Weh k amp J, S a l z man N H, Por te r E et al. Reduced Paneth cell alphadefei in ileal. PNAS 25 ;12 : G e orge M D, Weh k amp J, Kay s R J et al. In vivo gene profiling of human intestinal epithelial cells: analysis by laser microdissection of formalin fixed tissues. BMC Genomics 28 ;9 : B e v i C L. The Paneth cell and the innate immune respoe. Curr Opin Gastroenterol 24 ;2 : Cun l i ffe R N, R os e F R AJ, Ke y te J et al. Human defein 5 is stored in precursor form in normal Paneth cells and is expressed by some villous epithelial cells and by metaplastic Paneth cells in the colon in inflammatory bowel disease. Gut 21 ;48 : Wehkamp J, Schmidt K, Herrlinger KR et al. Defein pattern in chronic gastritis: HBD-2 is differentially expressed with respect to Helicobacter pylori status. J Clin Pathol 23 ;56 : She n B, Por ter E M, R e y no s o E et al. Human defein 5 in intestinal metaplasia of the upper gastrointestinal tract. J Clin Pathol 25 ;58 : v an E s J H, Jay P, Gre gor i eff A et al. Wnt signalling induces maturation of Paneth cells in intestinal crypts. Nat Cell Biol 25 ;7 : Wehkamp J, Wang G, Kubler I et al. The Paneth cell alpha-defein deficiency of ileal is linked to Wnt/Tcf-4. J Immunol 27 ;179 : Kosl ow sk i M J, Kubl er I, C hamai l l ard M et al. Genetic variants of Wnt tracription factor TCF-4 (TCF7L2) putative promoter region are associated with small intestinal. PLoS One 29 ;4 :e IBD Working Group of the European Society for Paediatric Gastroenterology, Hepatology and Nutrition. Inflammatory bowel disease in children and adolescents: recommendatio for diagnosis the Porto criteria. J Pediatr Gastroenterol Nutr 25 ; 41 : Fage rb e rg U L, L o of L, Mer z ou g R D et al. Fecal calprotectin levels in healthy children studied with an improved assay. J Pediatr Gastroenterol Nutr 23 ;37 : B ous varo s A, Antoni ol i DA, C ol l e tt i R B et al. D ifferentiating ulcerative colitis from in children and young adults: report of a working group of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition and the Crohn s and Colitis Foundation of America. J Pediatr Gastroenterol Nutr 27 ;44 : He y man M B, Kirs ch ne r B S, G ol d BD et al. Children with early-oet inflammatory bowel disease (IBD): analysis of a pediatric IBD coortium registry. J Pediatr 25 ; 146 : L e v i ne A, Kug at has an S, Anne s e V et al. Pediatric oet Crohn s colitis is characterized by genotype-dependent age-related susceptibility. Inflamm Bowel Dis 27 ;13 : L ou is E, C ol l ard A, O ge r A F et al. Behaviour of according to the Vienna classification: changing pattern over the course of the disease. Gut 21 ;49 : Negroni A, Stronati L, Pierdomenico M et al. Ac t iv at ion of NOD2 - mediated intestinal pathway in a pediatric population with. Infl amm Bowel Dis 29 ;15 : Kieh ne K, Br u n ke G, Weg ne r F et al. Defein in chronic pouchitis in patients with ulcerative colitis or familial adenomatous polyposis coli. World J Gastroenterol 26 ;12 : Fishbein T, Novitskiy G, Mishra L et al. NOD2-expressing bone marrow-derived cells appear to regulate epithelial innate immunity of the traplanted human small intestine. Gut 28 ;57 : Weh k amp J, C hu H, She n B et al. Paneth cell antimicrobial peptides: topographical distribution and quantification in human gastrointestinal tissues. FEBS Lett 26 ;58 : Weh k amp J, S chw i nd B, Her rl i nge r K R et al. Innate immunity and colonic inflammation: enhanced of epithelial alpha-defei. Dig Dis Sci 22 ;47 : Hugot J -P, C hamai l l ard C, Z ou a li H et al. Association of NOD2 leucine-rich repeat variants with susceptibility to. Nature 21 ;411 : Ogura Y, Bonen DK, Inohara N et al. A f r ame shi ft mutation in NOD2 associated with susceptibility to. Nature 21 ;411 : Hamp e J, Fran ke A, R os e t i el P et al. A genome-wide association scan of noynonymous SNPs identifies a susceptibility variant for in ATG16L1. Nat Genet 27 ;39 : Kas e r A, L e e A H, Fran ke A et al. XBP1 links ER stress to intestinal inflammation and confers genetic risk for human inflammatory bowel disease. Cell 28 ;134 : C adwel l K, L iu J Y, Brow n SL et al. A key role for autophagy and the autophagy gene Atg16l1 in mouse and human intestinal Paneth cells. Nature 28 ;456 : Wehkamp J, Harder J, Weichenthal M et al. Inducible and cotitutive beta-defei are differentially expressed in and ulcerative colitis. Inflamm Bowel Dis 23 ;9 : Fa h l g re n A, Ham marst rom S, D an i elss on A et al. beta-defein-3 and -4 in intestinal epithelial cells display increased mrna in ulcerative colitis. Clin Exp Immunol 24 ; 137 : Kap el N, B ena hme d N, Mora li A et al. Fecal beta-defein-2 in children with inflammatory bowel diseases. J Pediatr Gastroenterol Nutr 29 ;48 : by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY