The Role of Human Defensins in Gastrointestinal Diseases

Jost Langhorst; Kyung-Eun Choi


Expert Rev Clin Immunol. 2011;7(6):779-787. 

In This Article

Crohn's Disease

In CD, the interest in antibacterial peptides as an innate mechanism of defense against bacteria has been intensified by the association between NOD2 mutations and decreased α-defensin production from Paneth cells in ileal CD.[19,28,48] Based on patient and confirmative mouse model data, Wehkamp and colleagues proposed a concept of slow bacterial invasion due to compromised defense by mucosal peptide antibiotic including defensins as a leading factor in the pathogenesis of CD.[4]

For CD patients, reduced α-defensin levels are seen in patients with ileal disease and reduced β-catenin levels are seen in those with colonic involvement.[2] Lower levels of induction of DEFB4 (HBD-2) mRNA have been reported in CD than in UC.[12] An association of reduced DEFB4(HBD-2) genomic copy number with colonic CD was connoted and supported by the reported reduction of mucosal DEFB4 mRNA expression in patients with a gene copy number of <four copies compared with four copies.[36] In addition to mucosal HBD expression, reduced gene copy numbers for HBD-1 and HBD-2 were consequently reasoned to predispose to the development of CD.[36]

In contrast to these findings alternate studies support the idea that HBD2 is not affected by NOD2 mutation status.[11,35] Reduced defensin expression was thus claimed to be a consequence of the mucosal surface destruction arising with inflammation[2] rather than being a key factor of the disease emergence. In addition, Aldhous et al.[49] recently queried a methodological blemish with Wehkamp's biopsies showing that HBD-2 mRNA is lower in CD compared with UC and induced with inflammatory influences solely in the latter.[12] Those biopsies were taken randomly from colonic mucosa at different areas, but DEFB4 mRNA is differentially expressed according to location within the inflamed colon. Aldhous et al. argued that the differential HBD-2 mRNA expression is upregulated in both CD and UC, if the biopsy location is controlled.[49] Furthermore, DEFB4 is highly inducible (up to several hundred-fold) in response to a range of proinflammatory mediators in a range of cell types.[36,50] However, Simms et al.[35] did not stratify according to NOD2 genotype even though only the frameshift mutation is affected.[18] This methodological aspect qualifies the drawn conclusions.

The dysregulated defensin response in IBD patients certainly requires further investigation to elucidate whether it is due to changes in the local bacterial flora and/or a sign of inflammation, or rather a resultant side effect of the primary IBD disease with a dysregulated immunity as part of its pathogenesis.[29,51,52] Initial evidence that regulation evolves in long standing disease is derived from data that HBD expression is less impaired in pediatric than adult CD.[34]

CD of the Ileum

As the location of inflammation commonly remains stable in CD and HD are differentially expressed in the gut,[4] it may be interesting to discuss CD of the ileum and CD of the colon separately. It has been hypothesized that disease locations in IBD are due to location-specific defects in host innate immunity.[3] The terminal ileum is highly susceptible for inflammation in CD, and a higher concentration of bacteria, large amount of lymphoid tissue and higher numbers of Paneth cells are seen as relevant contributing features.[2] Ileal CD is associated with reduced levels of HD-5 and HD-6:[35,53–54] in a case–control study, ileal CD patients had a significant decrease in HD-5 mRNA compared with non-IBD controls,[35] with HD-6 expression approximately fivefold lower than that of controls. This relative lack of AMPs naturally results in diminished antibacterial killing by the mucosa.[4]

Wehkamp and colleagues further proposed an association between the intracellular receptor NOD2,[55] which is expressed in intestinal epithelial cells (particularly in Paneth cells)[56,57] and the mechanisms for regulating α-defensin expression.[28] Intriguingly, approximately a third of patients with CD have a mutation in the NOD2 gene, which is associated with ileal but not colonic CD.[58] In the study by Wehkamp et al., CD patients with NOD2 mutations had a decreased ileal HD-5 mRNA expression in affected tissue compared with those with wild-type NOD2, while colonic HD-5 and HD-6 were only upregulated in inflamed tissue only in wild-type NOD2 patients.[28] Furthermore, the degree of inflammation did not influence the amount of defensins expressed, but was related to NOD2 genotype. A loss-of-function mutation of the CARD15 gene encoding for NOD2 was claimed to be responsible for the reduced HD5 and HD6 expression including lower antimicrobial activity at the mucosa. However, as Ramansundara recently contended, these data also demonstrated that individuals without this NOD2 mutation also have a defensin defiency corroborating the existence of other explaining factors involved.[2] In addition, healthy individuals carry the risk alleles and thus the mutation also occurs in unaffected individuals.[39] Support for the notion that the precise mechanisms responsible for the induction of those defensins remain insufficiently understood comes from a similar study to the Wehkamp trial with a larger sample size that revealed no association between HD-5 and -6 expression and NOD2 genotype.[35] However, this study has methodological limitations (as discussed previously).

Another possibly relevant factor in ileal CD is epithelial stem cell differentiation meditated by Wnt (wingless type) signaling.[8] Within the different mechanisms of Wnt signaling the 'canonical' pathway – in contrast to 'noncanonical' Wnt without β-catenin stabilization or β-catenin-mediated gene transactivation – is the best understood.[59] In the canonical pathway, β-catenin transduces the Wnt signal to the nucleus and Wnt signaling is attained from intestinal subepithelial myofibroblasts.[3] Progenitor cells are then transduced into secretory Paneth cells,[3] intracellular β-catenin is released and translocated into the nucleus, where it interacts with T-cell factor 4 (Tcf4).[59] The resulting complex activates target gene by binding to DNA and regulating their transcription, including antibacterial defensin genes.[3] In the majority of patients in a study of ileal CD, a Paneth cell deficiency was mediated by reduced Wnt signaling transcription-factor TCF-4,[21] which suggests a disturbed Paneth cell differentiation in this condition.[4] Intriguingly, the level of TCF-4 mRNA was highly correlated with both HD-5 and HD-6 mRNA within specimens of patients with ileal CD.[1] Irrespective of the degree of inflammation, the level of TCF-4 mRNA was reduced in ileal CD patients, whereas it was not reduced in colonic or UC.[1]

Recently, an association of ileal CD with the potassium intermediate/small conductance calcium-activated channel KCNN4 which is located in the IBD linkage region on chromosome 19q13 was drawn by Simms and colleagues[60] adding a new possible mechanism in the pathogenesis of this condition. The KCNN4 protein has an important role in T lymphocyte signaling,[3] is associated with Paneth cell secretion and shows lower mRNA levels in NOD2-mutated patients.[60] The relationship between NOD2 status, TCF-4 mRNA, KCNN4 protein, level of inflammation and α-defensin levels will thus have to be further elucidated to lead to a conclusion about genetically or acquired immunological mechanisms and deficiencies in the pathogenesis of ileal CD.

CD of the Colon

In contrast to ileal CD, in colonic CD an impaired induction of the epithelial β-defensins including HBD-2 has been affirmed.[36] There is little HBD-2 expression by the epithelium of uninflamed areas, but a significantly enhanced level in inflamed areas.[32] Bacteria might take advantage of the impairment formed by defensin deficiency,[4] partly caused by a reduction in β-defensin gene copy number on chromosome 8.[36] Furthermore, CD of the colon is associated with lessened functional antimicrobial activity against commensal gut microbiota compared with UC patients.[48] NOD2 receptors may contribute to host defense by recognizing intracellular patterns and, in case of need, triggering AMP expression.[11] Since changes to intracellular transcription by NF-κB and also NOD2 mutations are commonly seen in CD patients,[2] this might be a responsible mechanism for the defective HD expression in those patients.

There are three clusters of β-defensin genes, two on chromosome 20 and one on 8p23.1. The genes encoding the four β-defensins (HBD 1–4) expressed in the gut (genes DEFB1, DEFB4, DEFB103 and DEFB104) are clustered at the latter. While genomic copy number for HBD-1 is stable, different individuals have different numbers of the same DEFB4 gene in their genomes. This β-defensin cluster has a genomic copy number of between two and 12 per diploid genome, with the majority of individuals possessing between three and five copies of this cluster. Since the high as well as the low extreme gene copy numbers are not prevalent in the population, the question of whether this might be linked to the relative β-defensin defiency in CD has been raised. If this was the case, copy number of DEFB4 (HBD-2) could be used as a susceptibility factor for CD and predictor for disease emergence.

Fellerman and colleagues provided support for this assumption.[6] They found lower mean DEFB4 copy number in CD compared with healthy controls as well as a positive correlation between DEFB4 copy number and expression of DEFB4 mRNA in mucosal biopsies.[6] Interestingly, this genetic association was solely significant in patients suffering colonic CD,[6] which might reflect differences in defensin levels between ileum and colon.

Alternate studies, however, suggest that elevated DEFB4 copy number is associated with both ileal and colonic CD and might therefore be a risk factor for CD irrespective of intestinal location.[61] Bentley et al. found no evidence to support a role for DEFB4 as a risk factor for bowel resection in patients with colonic CD.[61] In ex vivo culture, no association of DEFB4 copy number with expression level in patients or controls was detected.[49] Fellerman's and Bentley's data are controversially discussed.[50] Hollox recently pointed out that Bentley's data are not yet replicated, but are based on a respectable sample size, which is in contrast to Fellerman's studies.[50] He argued that the differences in results might be due to inadequate sample sizes of the Fellerman group, referring to high false-positive rates also with high replication rates.[50] Further association and expression studies in large phenotypically well-defined cohorts with as reliable as possible methods are needed before we can come to a final conclusion.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: