SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Background  Crohn’s Disease (CD), a chronic intestinal inflammation, is currently treated primarily by therapeutics which are directed against inflammatory responses. Recent findings though suggest a central role of the innate immune barrier in the pathophysiology. Important factors providing this barrier are antimicrobial peptides like the α- and β-defensins. Little is known about in vivo effects of common drugs on their expression.

Aim  To analyse the influence of corticosteroids, azathioprine and aminosalicylate treatment on ileal and colonic antimicrobial peptides in active CD and also assess the role of inflammation.

Methods  We measured the expression of antimicrobial peptides and pro-inflammatory cytokines in 75 patients with active CD.

Results  Ileal and colonic α- and β-defensins as well as LL37 remained unaffected by corticosteroids, azathioprine or aminosalicylate treatment. Additionally, we did not observe a negative coherency between Paneth cell α-defensins and any measured cytokines. HBD2 and LL37 unlike HBD1 levels were linked to inflammatory cytokines and increased in highly inflamed samples.

Conclusions  Current oral drug treatment seems to have no major effect on the expression of antimicrobial peptides. In contrast to HBD2 and LL37, ileal levels of HD5 and HD6 and colonic HBD1 level are independent of current inflammation. Innovative drugs should aim to strengthen protective innate immunity.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Crohn’s disease (CD) is characterized by a chronic and severe inflammation of the intestine with diarrhoea, deep ulcerations and – in some cases – multiple fistulas.1 Both genetic and environmental factors were shown to contribute to disease risk.2 The leadoff but aged pathogenic disease concept was based on the idea of a ‘dysregulation’ in adaptive immunity.3, 4 Disease symptoms caused by an ‘over-reactive’ immune response at the mucosa may be successfully dampened by corticosteroids, thiopurines, methotrexate or tumour necrosis factor (TNF)-antibodies.5 Therefore, immunoregulatory agents remain the mainstay of conservative therapy in CD. However, the achievement of both induction and long-term maintenance of remission are still unsatisfactory.6

More than 10 years ago, it became evident that activated T-cells as well as antibodies in CD targeted bacterial flora.7–9 Inflammatory affected host tissue would therefore mostly count for collateral damage, rather then being the prime target of misdirected adaptive immune cells. On the basis of consecutive research, we developed an alternative but not mutually exclusive hypothesis, focusing on impaired barrier function with a secondary inflammatory response.10 One of its basic principles assumes that a healthy gut epithelium keeps the normal flora at a safe distance. This is achieved through a protective mucus layer containing defensins, cathelicidins and other antimicrobials as a first line of defence.11 The barrier thereby not only balances the commensal – host relationship at the mucosa, but also provides an effective protection against ingested pathogens.12 This barrier results in the observation that infections, stimulation of the immune system and subsequent inflammation are the exception rather then the rule in the healthy gastrointestinal tract. Crohn’s disease, however, was found to be associated with mucosal adherent bacteria.13–15 One likely but not exclusive explanation for this characteristic are localization-specific deficiencies in different antimicrobial peptides at the mucosal surface.16 In ileal CD patients, diminished Paneth cell α-defensin expression leads to reduced antibacterial killing activity of the mucosa.17 These α-defensins, the human alpha-defensin-5 (HD5) and human alpha-defensin-6 (HD6), are the most abundant Paneth cell products and the prominent antimicrobial peptides in the ileal mucosa.18, 19 Mechanisms underlying their decrease are represented by a genetic background with variations in several genes: NOD2,20, 21 ATG16L122 and TCF4.23 Those genes, all linked with a small intestinal involvement of CD are important players in Paneth cell biology and pinpoint to a primary and paramount role of this cell in the disease process.

In contrast, colonic CD is characterized by an attenuated induction of inducible β-defensins.24–26 This defect in inducing these protective host factors is also explained by genetics, respectively by low gene copy numbers of human-beta-defensin-2 (HBD2).27 HBD2 has been shown to be induced not only by different pathogens28, 29 but also specifically by probiotic bacteria.30–32 In this process, probiotics are probably killed by the antimicrobial peptide they induce, but leave the host with an enhanced protective barrier.30 A defect in β-defensin induction might partly explain why probiotic treatment seems to have promising effects in pouchitis and the maintenance of remission in ulcerative colitis, but no benefit in CD.33

In the present study, we assessed the impact of standard medications in a translational disease patient study on the expression of various important defensins. The findings were also related to the degree of inflammation as measured by cytokine expression.

Patients

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

A total of 75 Crohn’s disease patients were included in the study and their current treatment was documented. Patients underwent diagnostic colonoscopy because of a clinical relapse. The study was approved by the ethics committee of the University of Tübingen. After informed consent, biopsies were collected from macroscopically inflamed Crohn’s disease tissue of the small (n = 53) and large intestine (n = 44) and immediately shock-frozen into liquid nitrogen.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

RNA isolation and cDNA synthesis were performed as described before.18 Levels of mRNA expression from human alpha-defensin-5 HD5 (DEFA5), human alpha-defensin-6 HD6 (DEFA6), human beta-defensin-1 HBD1 (DEFB1), human beta-defensin-2 HBD2 (DEFB4), human beta-defensin-3 HBD3 (DEFB103A), cathelicidin antimicrobial peptide LL37 (CAMP), interleukin-1β (IL1B), IL6, IL8, IL10, intercellular adhesion molecule 1 (ICAM1) and TNF (TNF superfamily, member 2, TNFα) were quantified by real-time RT-PCR using the LightCycler 2.0 and 480 (Roche Diagnostics GmbH, Roche Applied Science, Mannheim, Germany). Primers for HD5, HD6, IL-8, TNFα;21 HBD1, HBD2,34 HBD3,27 LL3735 and IL1B36 were previously published. Primer IL6 sense: 5′ AAT CAT CAC TGG TCT TTT GGAG 3′; antisense: 5′ GCA TTT GTG GTT GGG TCA 3′; Primer IL10 sense: 5′ AAG CCT GAC CAC GCT TTC TA 3′; antisense: 5′ ATG AAG TGG TTG GGG AAT GA 3′; Primer ICAM1 sense: 5′ TAG CAG CCG CAG TCA TAA 3′; antisense: 5′ CGG GAT AGG TTC AGG GAG 3′. Absolute quantification was determined by external standards.18

We stratified the mucosal samples according to ileal or sigmoid biopsy site as well as the concurrent use of corticosteroids, azathioprine and 5-Asa. Patient samples were combined into groups with low (<1000 copies/10 ng total RNA) or high (1000–1 000 000 copies/10 ng total RNA) IL8 expression, indicating moderate-to-severe inflammation. The Mann–Whitney U-test was performed for comparison of grouped data. P-values of <0.05 were considered statistically significant.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

Patient characteristics

The patient characteristics are given in Table 1. The age distribution was typical for this disease as was the balanced gender ratio. Patients with ileal biopsies all had localization L1 or L3 and those with sigmoid biopsies exclusively belonged to Vienna categories L2 or L3.37 Disease behaviour distribution in both groups was similar. The number of patients on or off steroids was rather balanced in this cohort, whereas the numbers of patients on azathioprine or 5-ASA were smaller. Very few patients additionally received the nutritional supplement zinc (three in the ileal and five in the colonic group), two patients with ileal biopsies taken and five with colonic biopsies received antibiotics and in both groups, one patient was additionally treated with anti-TNF antibodies.

Table 1.   Characteristics of patient cohort
 IleumSigmoid
  1. s.d., standard deviation; m, male; f, female; A, age of onset; L, localization; B, behaviour; according to Vienna classification.37 Data shown in absolutes numbers.

Age (median ± s.d.)36 ± 1337 ± 15
Gender ratio (m/f)25/2813/31
Vienna A1/A247/634/10
Vienna L1/L2/L325/0/280/19/25
Vienna B1/B2/B3/unknown19/17/15/221/8/13/2
Corticosteroids (on/off)23/3023/21
Azathioprine (on/off)10/439/35
Aminosalicylate (on/off)14/3915/29

Ileal defensin expression

The major defensin expressed in the terminal ileum was HD5, followed by HD6 and at a hundredfold lower level, by HBD1 and HBD2, whereas the cathelecidin LL37 exhibited the lowest mRNA level. None of the studied therapies significantly affected the expression of both α- and β-defensins, as well as LL37 (Figure 1 and Table 2). As expected, for ileal HD5 and HD6 expression, a significant (< 0.0001) and strong (Spearman r = 0.9486) correlation has been found (Figure 2). Furthermore, supporting the hypothesis of an inflammation independent and primary role of Paneth cell α-defensins in ileal CD, no negative coherency between either HD5 or HD6 mRNA level and inflammatory cytokine expression was observed (Table 3, Figure 2). Grouping the samples according to IL8 expression levels which correlates with the histological grade of inflammation,38 we again found no difference in Paneth cell α-defensin expression. The expression level of all measured inflammatory cyctokines was highly matching and on a weaker base also correlated with the anti-inflammatory IL10 (Table 3). Ileal expression of HBD3 was minimal (Table 2). Not surprisingly, for HBD2, we found a significant increase in the high inflammation group (data not shown), which could also be observed for LL37 mRNA level (Figure 5). MRNA level of both antimicrobial peptides correlated with the expression of inflammatory cytokines (Table 3).

image

Figure 1.  Current treatment and the expression of ileal defensins and LL37 in active Crohn’s Disease. In Crohn’s disease patients, the ileal expression of antimicrobial Paneth cell α-defensins is not significantly affected by the studied therapies (corticosteroids, azathioprine and aminosalicylates) (upper panel). Similarly, the constitutively expressed β-defensin HBD1 as well as LL37 seems unchanged (lower panel).

Download figure to PowerPoint

Table 2.   Expression levels of antimicrobial peptides and cytokines: mRNA transcripts/10 ng RNA in human ileal tissue
Products SteroidsAzathioprine5-ASA
Medication+++
n 302343103914
  1. Shown are the means ± the standard errors.

HBD1726.7 (±166.4)762.1 (±136.8)1585 (±426.2)1129 (±232.6)1076 (±463.1)1104 (±253.5)1163 (±345.6)
HBD21766 (±975.7)1966 (±875.9)912.4 (±254.5)1745 (±620.7)494.3 (±265.1)1304 (±517.4)2080 (±1309)
HBD31558 (±1234)1077 (±826)187.8 (±135.7)849.8 (±579.9)7.64 (±3.323)909.5 (±639)81.81 (±74.58)
HD51 403 000 (±442 410)2 240 000 (±532 497)1 933 000 (±406 767)1 938 000 (±378 721)2 831 000 (±857 263)1 997 000 (±347 582)2 412 000 (±910 334)
HD6440 552 (±106 481)610 644 (±118 496)761 554 (±201 771)629 730 (±119 640)875 667 (±274 286)659 779 (±118 727)721 689 (±258 723)
LL3755.35 (±21.47)54.23 (±14.78)51.24 (±11.67)49.99 (±10.96)65.59 (±21.14)52.86 (±12.35)53.15 (±13.49)
IL836 741 (±33 357)27 005 (±22 215)10 128 (±4830)23 664 (±15 629)2556 (±1010)24 734 (±17 232)5606 (±2370)
IL64961 (±3452)3317 (±2202)1295 (±926.8)2967 (±1604)169.5 (±46.87)3272 (±1811)353.9 (±171.1)
IL101535 (±441.0)1267 (±313.5)844.7 (±181.6)1189 (±235.7)632 (±136.5)1217 (±251.4)713.5 (±212)
IL1B mean (±S.E.)67 464 (±34 513)53 789 (±23 517)16 454 (±6066)44 725 (±16 730)6892 (±3066)44 401 (±18 218)18 605 (±10274)
TNFα mean (±S.E.)1764 (±672.7)1447 (±444.1)609.7 (±110.7)1171 (±318.1)644.5 (±160.7)1192 (±340.7)737.5 (±268.2)
ICAM15869 (±2688)5322 (±1841)3314 (±818.3)4971 (±1371)2710 (±678.3)5006 (±1438)2903 (±1125)
image

Figure 2.  Inflammation severity and the expression of ileal defensins in active Crohn’s disease. The mRNA levels of Paneth cell α-defensins remained unchanged comparing moderate to severely inflamed patient samples (left upper panel). The expression of both defensins strongly correlated with each other (right upper pane), but was independent of the expression of inflammatory cytokines (right lower panel). Latter exhibited matching mRNA level (left lower panel).

Download figure to PowerPoint

Table 3.   Correlations of ileal mRNA expression levels
Samples ileumIL8IL10Il-1bIl-6TNFαICAM1
  1. Shown are the correlation scores and the respective P-values. We tested for correlation of HD5 and HD6 as well as LL37 mRNA level with different cytokines. We additionally performed tests to determine the strength of correlations between IL8 and different other measured cytokines.

HD5
 Spearman r0.19900.0167−0.0040.02750.06300.1809
 P-value0.15320.90580.97690.84480.65400.1949
HD6
 Spearman r0.15890.02360.0590−0.00920.10930.1901
 P-value0.25570.86690.67640.94790.43590.1728
LL37
 Spearman r0.45150.45190.47670.65440.59880.5799
 P-value0.00070.00070.0003<0.0001<0.0001<0.0001
IL8
 Spearman r 0.44870.82550.80220.62040.7844
 P-value 0.0008<0.0001<0.0001<0.0001<0.0001
image

Figure 5.  Distribution of ileal and colonic antimicrobial peptides in active Crohn’s disease. (a) upper panel: comparison of mRNA levels of defensins and LL37 in ileal active Crohn’s disease; lower panel: Inflammation severity and the expression of ileal LL-37 in active Crohn’s disease. The Paneth cell α-defensins HD5 and HD6 are clearly the dominating members of the measured antimicrobial factors. HBD2 seems to be more prominently expressed than HBD1 in the studied active CD samples. According to mRNA levels, LL37 was significantly highly expressed in the severely inflamed tissue samples (lower panel). (b) upper panel: comparison of mRNA levels of defensins and LL37 in colonic active Crohn’s disease; lower panel: Inflammation severity and the expression of colonic LL-37 in active Crohn’s disease. Comparable to the results in the ileum, HBD2 seems to have a more prominent role in the studied active CD samples than HBD1. Also matching to the ileal situation, mRNA levels of LL37 were significantly higher in the severely inflamed tissue samples (lower panel).

Download figure to PowerPoint

Colonic defensin expression

The major β-defensins expressed in colonic mucosa were HBD1 and the inducible HBD2, followed by HBD3 and at the lowest level, LL37 (Figures 3 and 5, Table 4). The two inducible HBDs exhibited matching expression patterns (Spearman r = 0.7484; P < 0.0001). HBD2 mRNA levels also correlated with those of the measured inflammatory cytokines (Table 5, Figure 4). The strongest effect was seen with IL1B (Spearman r = 0.5838; P < 0.0001). HBD1 however, exhibited mRNA levels independent of HBD2 and inflammatory cytokines (Table 5, Figure 4). Consistent with these results, we found an increase in HBD2 in the high inflammation group, whereas HBD1 remained unchanged (Figure 4). Similar to HBD2, LL37 showed an increased expression after stratifying the samples according to inflammation severity and analogous mRNA levels in comparison with cytokine expression (Figure 5 and Table 5). There were no significant differences in the treatment groups comparing HBD2, HBD3 and LL37 expression levels in the colonic mucosa. We saw a HBD1 decrease in the 5-ASA ‘on’ compared to the ‘off’ group, which did not hold up after P-value adjustment according to multiple testing (Figure 3). In conclusion, none of the treatments seemed to have any paramount effect on antimicrobial peptide expression.

image

Figure 3.  Current treatment and the expression of colonic defensins and LL37 in active Crohn’s disease. Possible differences in the expression of inducible β-defensins as well as LL37 were not statistically significant.

Download figure to PowerPoint

Table 4.   Expression levels of antimicrobial peptides and cytokines: mRNA transcripts/10 ng RNA in human colonic tissue
Products SteroidsAzathioprine5-ASA
Medication+++
n2121233592915
  1. Shown are the means ± standard errors.

HBD12043 (±448.0)2112 (±373.5)2403 (±416.8)2285 (±281.3)2182 (±859.4)2715 (±369.8)1392 (±301.5)
HBD21673 (±753.3)2470 (±1049)7603 (±5723)5812 (±3766)2592 (±2269)7006 (±4564)1570 (±742.2)
HBD3535.7 (±355.7)325 (±207)187 (±84.26)297 (±130.3)48.1 (±24.92)290.7 (±151.7)178.1 (±119.5)
LL3768.47 (±17.65)79.54 (±13.55)88.39 (±15.0)87.13 (±10.65)70.66 (±28.43)91.19 (±12.75)69.31 (±15.83)
IL826 711 (±20 346)10 043 (±3089)25 238 (±11 688)19 831 (±3861)10 808 (±3861)22 795 (±9351)8688 (±3582)
IL6754.2 (±435.0)966.1 (±292.6)3340 (±1393)2436 (±929.3)1316 (±774.7)2046 (±940.5)2519 (±1308)
IL101429 (±548.4)1073 (±299.1)1131 (±209.2)1249 (±207.8)460.2 (±127.7)1319 (±255.5)686.9 (±115.1)
IL1B13432 (±5274)14405 (±3826)17324 (±5933)17647 (±4368)9258 (±3428)17712 (±4876)12488 (±4619)
TNFα584.7 (±137.5)688.0 (±111.6)736.7 (±159.7)731.9 (±114.7)642 (±184.2)714.7 (±125.3)711.1 (±160.7)
ICAM13620 (±1501)4553 (±1044)6096 (±1625)5634 (±1194)4291 (±1236)5745 (±1393)4613 (±1048)
Table 5.   Correlations of colonic mRNA expression levels
Samples colonIL8IL10IL1BIL6TNFαICAM1
  1. Shown are the correlation scores and the respective P-values. We tested for correlation of HBD1 and HBD2 as well as LL37 mRNA level with different cytokines. We additionally performed tests to determine the strength of correlations between IL8 and different other measured cytokines.

HBD1
 Spearman r0.16320.12090.15630.11510.21350.2307
 P-value0.28980.44560.31090.45690.16410.1320
HBD2
 Spearman r0.51850.29650.58380.43050.44180.5386
 P-value0.00030.0536<0.00010.0040.00340.0002
IL8
 Spearman r 0.30220.84690.90050.68810.8513
 P-value 0.0436<0.0001<0.0001<0.0001<0.0001
image

Figure 4.  Inflammation severity and the expression of colonic defensins in active Crohn’s disease. The constitutively expressed beta defensin HBD1 exhibited unchanged mRNA level after stratifying according to inflammation severity. Differently, the inducible HBD2 was significantly higher expressed in the severely inflamed tissue samples (left upper panel) correlating with inflammatory cytokines, e.g. IL1B (lower panel left) and HBD3 (upper panel right).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

The present study is the first which systematically assessed the effect of standard treatments in Crohn’s disease on the expression of important barrier peptides in affected patient mucosa. It should be noted that all patients had an indication for the endoscopy because of a clinical relapse. As a consequence, none of the patients were in clinical remission and, indeed, none in endoscopic remission. We stratified the samples according to the severity of inflammation which was determined by IL8 mRNA levels.17 We found a strong link between IL8 and all other proinflammatory cytokines in either tissue, supporting its beneficial use as a marker.

Patients with ileal localization of CD did not show significant differences in mRNA expression of antimicrobial peptides under standard medication including corticosteroids, azathioprine or aminosalicylates. The lack of an effect on α-defensins was not surprising. These peptides are constitutively expressed and appear to be under the control of different factors including the intracellular receptor nucleotide binding oligomerization domain NOD2. The detailed mechanisms behind the link between Paneth cell α-defensins and NOD2 are not clearly understood, but supported by investigations in NOD2 knockout mice39 as well as by translational studies in European34, 40 and US patients.41, 42 In addition, Paneth cells are adversely affected by mutations in the autophagy gene ATG16L122 and potentially by polymorphisms in the TCF-4 promotor.23 The mRNA expression of TCF-4, a key effector of the Wnt pathway is, similarly to HD5 and HD6, specifically reduced in ileal CD. Besides directly controlling the expression of these Paneth cell α-defensins TCF-4 plays an important role in stem cell proliferation and Paneth cell maturation.43, 44 In agreement with previous studies,34, 41 the degree of inflammation in the mucosa did not affect HD5 and HD6 mRNA. Antithetically, in a recent Australian study, Simms et al.45 suggested inflammation dependence for the decrease in ileal CD. They proposed tissue damage as the cause for the diminished levels in inflamed ileal CD mucosa. Based on their results, differences according to inflammation severity would be expected in our samples. We did not find any decrease in HD5 or HD6 in the high inflammation group compared to moderate inflammation. We also did not see a negative correlation between HD5 or HD6 mRNA levels and inflammatory cytokine expression. A reason for the unequal interpretations might be differences in study design. We only used biopsies or surgical specimens separately, whereas the Australian study mixed resection specimens and biopsies disproportionately in the different groups. In support of this idea, the Paneth cell product PLA2G2A was significantly lower in the normalized control group suggesting a possible bias because of normalization effects.46 In summary, these data presented here further support that the decrease in Paneth cell α-defensins is seen independent of inflammation and not because of epithelial loss at the base of the crypts where Paneth cells are located.

Like in the small intestine, the different treatments did not seem to have great effects on defensin expression in the colon. HBD2 and HBD3 can be induced by proinflammatory cytokines, such as IL1B, mostly through NFκB- and AP1-dependent pathways,47 by bacteria by means of similar pathways31 and also by the cytokines IL-23 and IL-17.48 NFκB has a key role in the transcriptional regulation of a proinflammatory gene programme in intestinal epithelial cells (e.g. IL-8, ICAM-1, inducible NO synthase, COX-2, and growth-related oncogene alpha). The inflammatory target genes are thereby much more affected by changes in NFκB than HBD2. The latter must be additionally regulated by factors, which have not yet been identified.34 We observed a significant, but relatively weaker correlation between colonic HBD2 and the different cytokines, supporting the partly common but also fractionally distinct regulation. Nothing is known so far about the transcriptional regulation of HBD1.

Considering the effects of corticosteroid treatment, the situation in the colon is quite complex and the conclusion based on this study design must be drawn carefully. Glucocorticoids are known to inhibit the transcription factors NFκB and AP1.49 These are, as mentioned above, involved in the bacterial or cytokine-dependent induction of HBD2 and HBD331, 47, whereas HBD1 is very likely independent of them. Interestingly, in a study by Witthoft et al.50in vitro experiments showed an upregulation of HBD2 in Caco-2 cells following corticosteroid treatment. Similarly, in an in vitro study using a human epithelial lung cancer cell line, another group reported upregulation of HBD2 after dexamethason treatment.51 These studies are compatible with the, by trend, marginal induction seen in the corticosteroid treated group and support a partly inflammatory independent regulation of HBD2.

The nucleotide analogue Azathioprine is metabolized to mercaptopurine through reduction by glutathione.52 It is relatively lymphocyte-specific because these cells lack a salvage pathway for purine synthesis.53 Nonetheless, as a purine-analogue blocking the synthesis of inosinic acid, it generally affects cells in the S-phase. As the intestinal epithelium undergoes cell renewal at extraordinary rates,54, 55 an impact on the proliferating cell compartments might not be unlikely. We observed a trend for decreased expression of inducible HBDs in the azathioprine group, which did not achieve statistical significance, but potential effects can so far not completely be excluded.

It is well known that 5-ASA can modulate various inflammatory pathways (e.g. production of inflammatory cytokines, activity of inducible nitric oxide synthase, activation of NFkB).56–58 More recently, it has been shown that 5-ASA can interfere with biological pathways that control growth and survival of colorectal cancer cells as well as colorectal mucosal cells in patients with sporadic large bowel polyps.59, 60 If and how it has an influence on small intestinal mucosa or on healthy colonic mucosa is not clear. Mechanisms by which 5-ASA influences CRC proliferation are inhibition of both the Wnt/β-catenin pathway and epidermal growth factor receptor signalling pathways, as well as the activation of peroxisome proliferator-activated receptor-γ.61 We see no difference in ileal defensin expression following 5-ASA treatment and no significant changes in the colonic expression of inducible β-defensins. However, we could observe lower mRNA levels of HBD1 in the colon, comparing 5-ASA treated patients with those who did not receive this medication. The effect did not hold up after P-value adjustment.

Stratifying the patients according to IL8, we found that an increase in colonic HBD2 expression in severe inflammation conforms to the significant induction in the ileum in the respective group. The same was seen for LL37. This antimicrobial peptide is known to have an influence on innate immune responses, angiogenesis and wound healing62 and dysfunctions of LL37 emerge as a central factor in the pathogenesis of several cutaneous diseases.63 There are different reports available on the influence of inflammation and inflammatory cytokines on LL37 expression, with some rebutting a role,64 but others supporting the idea.65, 66 Supporting an inflammatory controlled regulation, we found a correlation of LL-37 with inflammatory cytokines and an increase with more inflammation.

In conclusion, currently used standard treatments including corticosteroids, azathioprine and 5-ASA did not significantly influence the overall α-defensin expression in the terminal ileum as well as colonic β-defensin expression and LL37 mRNA level. A possible limitation of the study is that most patients received multiple treatments and thus the results are based on a cross-sectional design as well as lacking information on protein level. Further limitations might be that the study was conducted in a clinical patient – rather than in an experimental animal setting, which limits the possibility of drawing ultimate conclusions. The patients were often transferred to our tertiary care centre by practising physicians and therefore sometimes received therapies like 5-ASA, which are not necessarily evidence-based5

For future treatment of IBD, strategies might aim to strengthen protective innate immunity. In this context, the reported efficiency of Trichuris suis therapy in CD provokes the testable question of whether the stimulation of defensins by parasitic worms might be an explanation for their therapeutic effect. Probiotic bacteria like Escherichia coli bacteria and Lactobacilli, have been shown to induce antimicrobial peptides in vitro and in vivo.30, 32, 34 This might be an important mechanism to support the mucosa in preventing bacterial invasion. Probiotic bacteria are the first therapeutic agents for IBD of which we know that they bolster the production of antimicrobial peptides, but it is likely that others could have similar effects. In the future, further therapeutic strategies aimed at restoring the host–microbe balance at the intestinal mucosa should be identified and may prove superior to those that broadly suppress inflammation and adaptive immunity.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References

We thank the members of the Robert-Bosch-Hospital for recruitment of patients and for the specimens. We thank K. Siegel and J. Bader for excellent technical assistance. Declaration of personal interests: None. Declaration of funding interests: The study was supported by the Robert Bosch Foundation (Stuttgart, Germany), the Deutsche Forschungsgemeinschaft (DFG) and the DCCV (German patient organization of Crohn’s disease and ulcerative colitis). Jan Wehkamp is an Emmy Noether scholar of the Deutsche Forschungsgemeinschaft (WE 436/1-1).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients
  5. Methods
  6. Results
  7. Discussion
  8. Acknowledgements
  9. References