Innate immune dysfunction in inflammatory bowel disease


Michael Gersemann MD, Internal Medicine I, Robert Bosch Hospital, Auerbachstr. 110, D-70376 Stuttgart, Germany.
(fax: +49-711-81013793; e-mail:


Abstract.  Gersemann M, Wehkamp J, Stange EF (Robert Bosch Hospital, Stuttgart; Dr. Margarete Fischer-Bosch Institute of Clinical Pharmacology, Stuttgart and University of Tübingen, Tübingen; Germany). Innate immune dysfunction in inflammatory bowel disease (Review). J Intern Med 2012; 271: 421–428.

The pathogenetic mechanisms that cause the two types of inflammatory bowel disease (IBD), Crohn’s disease (CD) and ulcerative colitis (UC), are still under investigation. Nevertheless, there is broad agreement that luminal microbes are of particular relevance in the development of these conditions. In recent years, increasing evidence has shown that defects in the innate immunity are at the centre of both types of IBD. The innate intestinal barrier is provided by the epithelium which secretes antimicrobial peptides (so-called defensins) that are retained in the mucus layer. In ileal CD, the alpha-defensins are lacking owing to several Paneth cell defects. In colonic CD, the expression of beta-defensins is inadequate. This may be related to downregulation of the transcription factor peroxisome proliferator-activated receptor-gamma and in some cohorts is associated with a reduced HBD2 gene copy number. In UC, the mucus layer, which protects the host from the enormous amounts of luminal microbes, is defective. This is accompanied by an insufficient differentiation from intestinal stem cells towards goblet cells. All these disturbances in the gut barrier shift the balance from epithelial defence towards bacterial offence. The current treatment for CD and UC is based on suppression of this secondary inflammatory process. In future, patients may benefit from new therapeutic approaches stimulating the protective innate immune system.


The human intestine is constantly being exposed to an enormous number of microorganisms [1]. However, in most cases, infection is resisted because of a highly developed immune system. The innate immune system, also known as the nonspecific immune system, is the first line of control against luminal organisms. This evolutionarily old defence system, which is found in all plants and animals, is characterized by an antigen-independent response against microorganisms without the development of an immunological memory. Amongst other functions, the innate immune system preserves a defence wall between the luminally located microbes and the gut epithelium, thereby protecting the host against infections.

Crohn’s disease (CD) and ulcerative colitis (UC), the two major forms of inflammatory bowel disease (IBD), are characterized by chronic inflammation of the distal ileum and/or colon [2]. Both disease locations have an abundance of pathogenic and nonpathogenic bacteria [1]. There is general consensus that these microbes are relevant in the development and maintenance of CD and UC [3]. In this review, we focus on the challenge of the innate intestinal barrier by luminal microbes, the contribution of this challenge to the pathogenesis of IBD and its relevance in novel therapeutic strategies.

Intestinal bacteria and their role in IBD

In the human gastrointestinal tract, the distal ileum and the colon, the two typical locations of CD and UC, contain 107–108 and 1011–1012 microorganisms per gram of luminal content, respectively [1]. These microbes are both advantageous and disadvantageous to the gastrointestinal epithelium. On the one hand, they help to degrade complex carbohydrates to enhance absorption; on the other hand, they may adversely affect the host by causing inflammation. This was demonstrated by the findings that surgical diversion of the faecal stream improved CD inflammation distally [4] and that inflammation increased following the exposure of the terminal ileum to the luminal contents [5]. In addition, patients with IBD have antibodies against several microbes and microbial antigens, such as anti-Saccharomyces cerevisiae antibodies [6], showing that the interaction between these microbes and the host is relevant in IBD pathogenesis. Similarly, T cells may also be directed against microbial and not against self epitopes [7] suggesting that both types of IBD are not autoimmune diseases. Therefore, it is not surprising that treatment for IBD with antibiotics may improve the clinical outcome of patients with these conditions [8].

The composition of the microbial flora is altered in IBD towards fewer anti-inflammatory and a greater number of proinflammatory bacteria [9–11]; for example, the number of protective Bacteroides fragilis is reduced and that of Enterobacteriaceae, especially Escherichia coli with enhanced virulence, is increased in patients with CD [10, 12, 13]. Moreover, a higher number of mucus-, mucosal- and intraepithelial-associated bacteria were found in patients with IBD as compared to healthy individuals [14]. Mucolytic bacteria, such as Ruminococcus gnavus and Ruminococcus torques, were also described in colonic IBD [15]. Current evidence suggests that these alterations in the microbial flora are secondary to the defects in innate immunity described in the following sections [16, 17].

Pathogen-associated molecular patterns (PAMPs) are microbial motifs that enable the innate immune system to recognize bacteria. PAMPs are common and highly conserved microbial structures, such as lipopolysaccharides, peptidoglycanes, flagellin and lipoproteins [18]. The host recognizes PAMPs with the help of pattern recognition receptors, such as the toll-like membrane receptors (TLRs) and the intracellular nucleotide-binding oligomerization domain (NOD) receptors. Amongst the NOD family, NOD2 is crucially involved in IBD pathogenesis. It is expressed in the epithelium and senses muramyl dipeptide (MDP), which is a constituent of Gram-positive and Gram-negative bacteria [19]. Specific mutations of the NOD2 gene (Arg702Trp, Gly908Arg and leu1007fsinsC) are linked to an increased susceptibility to ileal CD [20, 21]. The risk of developing ileal CD is increased two to fourfold and 20- to 40-fold, respectively, for heterozygous and homozygous carriers of these NOD2 mutations [22]. In patients with NOD2 mutations, the activation of NF-κB in response to MDP is defective, enabling bacteria to trigger inflammation [23, 24]. The human TLR family consists of 10 members [25]. The TLRs detect microbiota- and damage-associated molecular patterns and are, therefore, involved in the maintenance of the commensal flora and mucosal homoeostasis [25]. In the healthy intestine, TLRs are expressed in small amounts not only by epithelial cells, but also by monocytes, macrophages and dendritic cells [26]. In IBD, their expression profile is altered [25, 27]; for example, TLR3 is downregulated in active CD but not in UC and TLR5 is upregulated in both forms of IBD [25, 27]. Essentially, these receptors provide a danger signal which, amongst other effects, stimulates the formation of alpha- and beta-defensins [28–30].

The intestinal barrier

The intestinal epithelium plays a pivotal role in gut barrier function. This monolayer consists of four different epithelial cell types, columnar cells, Paneth cells, goblet cells and neuroendocrine cells, all derived from multipotent intestinal stem cells [31].

The columnar cells are the most abundant cells in the intestinal epithelium. They have microscopic, finger-like, cellular membrane protrusions, termed microvilli, which are located on the luminal side of the cells and are covered by the glycocalyx. This sticky monolayer acts as a physical barrier preventing the luminal microbes from invading the mucosa [32]. In addition, columnar cells produce several antimicrobial peptides, which are known as defensins [33]. Defensins are small cationic peptides with a broad spectrum of microbicidal activity against Gram-negative and Gram-positive bacteria, fungi, viruses and protozoa [34]. They are capable of forming pores in the microbial membranes leading to disruption of the bacteria [35]. Owing to the position of three characteristic intramolecular disulphide bonds, defensins can be classified into two major subgroups: alpha-defensins and beta-defensins. Whereas the beta-defensins HBD1–3 are mainly produced by columnar cells [33], the alpha-defensins HD5 and HD6 originate from Paneth cells [36]. The main feature of these cells, located deep within the small intestinal crypts, is to keep the crypts sterile by secretion of these antimicrobial peptides [37]. Goblet cells also secrete defensins [38] as well as mucins, which are structural molecules of the protective luminal mucus layer [39]. This negatively charged biochemical and physical barrier lines the entire gastrointestinal tract to prevent pathogens from coming too close to the epithelium [39]. Thus, the entire intestinal epithelium, covered by a shield of cationic defensins which are fixed in the negatively charged mucus layer [40], forms the innate barrier of the gut. In healthy individuals, this defensive wall keeps the luminal microorganisms at sufficient distance from the epithelium (Fig. 1) [32]. In IBD, defects in this intestinal shield enable the luminal microorganisms to attack the epithelium.

Figure 1.

Intact innate intestinal barrier in healthy individuals. In the healthy intestinal mucosa, the positively charged antimicrobial peptides are secreted to a sufficient extent and fixed to the intact negatively charged mucus layer, thereby preventing the luminal microbes from coming too close to the epithelium.

Barrier dysfunction in ileal CD

Paneth cells, located between the stem cells at the base of each intestinal crypt of Lieberkühn [41], produce great quantities of the alpha-defensins HD5 and HD6 [42]. In healthy individuals, these broad spectrum antimicrobials sufficiently protect the epithelium against bacterial attack once secreted into the crypt lumen [43]. In ileal CD, the production of alpha-defensins is decreased, in contrast to colonic CD and UC (Table 1) [30, 42]. Notably, this decreased production is independent of inflammation and more pronounced in patients with NOD2 mutations [30, 42].

Table 1. Antimicrobials in inflammatory bowel disease (IBD)
AntimicrobialIntestinal locationCellular locationMolecular mass (kDa)Changes in IBD
  1. CD, Crohn’s disease; UC, ulcerative colitis.

HD5IleumPaneth cells3.5–4.5Reduced in ileal CD (but not in colonic IBD)
HD6IleumPaneth cells3.5–4.5Reduced in ileal CD (but not in colonic IBD)
HBD1ColonEpithelial and inflammatory cells3.5–4.5Reduced in colonic CD (but not in UC)
HBD2ColonEpithelial and inflammatory cells3.5–4.5Attenuated induction in colonic CD (but not in UC)
HBD3ColonEpithelial and inflammatory cells3.5–4.5Attenuated induction in colonic CD (but not in UC)
ElafinColonEpithelial and inflammatory cells9.8Attenuated induction in colonic CD (but not in UC)
SLPIColonEpithelial and inflammatory cells12Attenuated induction in colonic CD (but not in UC)
LL37ColonEpithelial and inflammatory cells18Attenuated induction in colonic CD (but not in UC)

The lack of alpha-defensins in ileal CD was confirmed by two other research groups [44, 45]. One group linked this finding to inflammation and loss of Paneth cells [44], whereas others reported no change in the number of Paneth cells in ileal CD [42, 46]. The relevance of NOD2 in mediating defensin synthesis was verified in the NOD2-knockout mouse [47]. Alpha-defensin deficiency results in a weakened antimicrobial shield as shown by reduced antibacterial activity in mucosal extracts from patients with ileal CD [42].

Paneth cells originate from intestinal stem cells under the control of TCF4, a Wnt signalling transcription factor [48]. Mouse studies showed that TCF4 regulates cryptidins, the mouse homologues of the human alpha-defensins HD5 and HD6 [48]. Similar to the alpha-defensins, TCF4 expression was found to be reduced in ileal CD but not in colonic CD or UC [49]. Again this reduction was independent of inflammation and – in contrast to HD5 and HD6 – also independent of the NOD2 genotype [49]. A clear correlation between TCF4 and alpha-defensin expression implies that HD5 and HD6 are controlled by TCF4 not only in mice, but also in humans [49]. Moreover, the promotor of these peptides possesses a high-affinity binding site for TCF4, and genetic variants in the putative promoter region of the WNT transcription factor TCF4 were found to be associated with ileal CD [50]. This finding suggests a primary role of TCF4 in the pathogenesis of small intestinal CD. Additionally, the functional relevance of this reduced TCF4 expression was confirmed in a murine TCF4-knockout model, in which reduced TCF4 expression in heterozygous mice caused a decrease in both alpha-defensins and the bactericidal activity of the mucosa [49].

In summary, a TCF4-mediated defect in Paneth cell differentiation is linked to a specific absence of the alpha-defensins, especially in patients with NOD2 mutations. Consequently, the deficient mucosal barrier allows the luminal microbes to invade the mucosa and trigger a secondary inflammatory disease (Fig. 2) [51, 52]. TCF4 and other Paneth cell-associated molecules (NOD2, ATG16L1, XBP1, KCNN4 and HD5) are genetically defective in ileal CD. Therefore, it is likely that the lack of alpha-defensins is a primary factor in small intestinal CD pathogenesis [53].

Figure 2.

Defect in the innate intestinal barrier in Crohn’s disease. Alpha-defensins and beta-defensins are insufficiently expressed in ileal Crohn’s disease (CD) and in colonic CD, respectively. Therefore, the antimicrobial mucus shield is ineffective leading to secondary inflammation because of invasion of the luminal microorganisms.

Barrier dysfunction in colonic CD

Compared with the small intestine, the normal colonic alpha-defensin expression is very low and is almost equally elevated in patients with CD and UC [30, 54]. This enhanced expression is apparently linked to the formation of metaplastic Paneth cells in the colon in IBD.

By contrast, the beta-defensins HBD1–3 are secreted by columnar and goblet cells [33, 38] and are more important in the colon. HBD1, one of the most prominent antimicrobials in the human colon, is a constitutively expressed antimicrobial peptide. However, the results of a recent study showed that HBD1 expression is significantly reduced in patients with colonic CD independent of inflammation (Table 1) [55]. It is interesting that HBD1 expression seems to be maintained by the peroxisome proliferator-activated receptor gamma (PPAR-gamma) acting as an antimicrobial factor [55]. In line with this, the colonic mucosa of PPAR-gamma mutant mice showed a reduced killing capacity against Candida albicans, B. fragilis, Enterococcus faecalis and E. coli as compared to wild-type littermates [55].

Another enigma was that HBD1, although being expressed in large amounts in various tissues that are in contact with bacteria, has only a marginal antibiotic activity against microbes under aerobic conditions. However, the partial pressure of oxygen in the human colon is low and, in addition, the luminal microbes are responsible for a decreased redox potential owing to their metabolism [56]. Recently, it was shown that the reduction of HBD1 disulphide bridges led to an enormous increase in the antimicrobial activity of this beta-defensin against C. albicans, Bifidobacterium and Lactobacillus species [56]. Thus, HBD1 seems to be a potent antibiotic agent in the human intestine in vivo.

In contrast to HBD1, the expressions of HBD2 and HBD3 are augmented in inflamed IBD [57–59]. However, this induction was less intense in colonic CD than in UC (Table 1) [57, 59]. Other antimicrobial peptides, such as elafin, SLPI and LL37, were also induced to a lesser extent in CD compared with UC mucosa (Table 1) [60, 61]. Accordingly, the bactericidal capacity was decreased in biopsy extracts from patients with CD as compared to those with UC and healthy individuals [62]. In humans, the beta-defensins HBD2 and HBD3 are neighbouring genes on chromosome 8p23.1. This beta-defensin gene cluster is known to be highly polymorphic [63]. There is evidence that the reduced induction of HBD2 and HBD3 is associated with a reduced gene copy number of the HBD2 gene in CD compared with controls [64]. However, this link was not confirmed in other cohorts [65, 66].

In conclusion, a reduced expression of HBD1 (possibly linked to dysfunction of the PPAR-gamma gene) and a reduced induction of HBD2 and HBD3 (possibly linked to HBD2 gene copy numbers) as well as other peptide antimicrobials in colonic CD causes a weakening of the mucosal defence. Therefore, luminal organisms are able to overcome the innate barrier leading to inflammation (Fig. 2) [67].

Barrier dysfunction in UC

Beta-defensins and other defensin-like peptides are sufficiently induced in UC under inflammatory conditions, and moreover, the antimicrobial activity of the inflamed UC mucosa is significantly higher compared with the inflamed CD mucosa [61, 62, 68]. Therefore, it seems that defensin synthesis and activity are not disturbed in UC. Instead, the colonic mucus layer covering the whole gastrointestinal tract is impaired. This physical and chemical barrier against luminal microbes is altered or even deficient in UC [69–71]. The thickness of the mucus layer in the healthy colon varies between 100 and 300 μm, increasing from the ascending colon to the rectum [70, 72, 73]. In UC, this mucus layer is thinner, more variable and partly denuded [69, 70].

Mucins that form the mucus layer are secreted by goblet cells, which migrate upwards the crypt during differentiation. Differentiation from the intestinal stem cell towards the goblet cell is regulated by several transcription factors, such as the basic helix–loop–helix transcription factor Hath1 and the zinc-finger transcription factor KLF4 [74, 75]. It was recently found that these two transcription factors are significantly augmented in inflamed CD, in contrast to UC and independent of the degree of inflammation [76]. However, a decrease in Hath1 in inflamed UC, but not in CD, has also been described [77]. Nevertheless, both these findings are consistent with a higher expression of Hath1 in patients with inflamed CD compared to those with inflamed UC. This fits well with the observation that the number of mature goblet cells, located in the upper part of the crypts, is decreased in inflamed UC as compared to controls [76]. Accordingly, mucin synthesis is decreased in inflamed UC [76, 78–80].

Overall, defective goblet cell differentiation might explain the deficient mucus layer in UC. The sufficiently secreted defensins are not retained by the defective mucus layer, thus allowing bacteria to pass through the epithelium and induce inflammation (Fig. 3) [51].

Figure 3.

Defect in the innate intestinal barrier in ulcerative colitis (UC). In UC, the mucus layer is reduced and biochemically altered. Therefore, although sufficient defensins are secreted, they cannot be fixed to the mucus. Luminal microbes can penetrate the mucosa and cause secondary inflammation.

Treatment perspectives

Inflammatory bowel disease is currently treated with steroids, azathioprine, methotrexate, tacrolimus and antitumour necrosis factor-alpha antibodies. These agents suppress the response of the host against bacteria that have penetrated the epithelium, thus reducing secondary inflammation. They are all accompanied by several potentially serious side effects including opportunistic infections [81].

There is increasing evidence for treatment concepts that augment the innate immune barrier to prevent microbials from penetrating the epithelium; for example, probiotic E. coli Nissle is as effective as 5-aminosalicylic acid in maintaining remission in UC [82]. Interestingly, E. coli Nissle was shown to induce defensins as a potential mechanism of action [83]. Moreover, the orally administered antibiotic agent rifaximine effectively induces CD remission [84] and metronidazol has been shown to reduce the risk of postoperative relapse in CD [85]. Therefore, it would be interesting to investigate oral defensins as treatment for CD, particularly in maintaining remission. However, it will take several years to complete the relevant clinical studies. Meanwhile, Trichuris suis ova are further tested as a preliminary study showed that they improve CD [86]. It is possible that these T. suis ova modulate the innate immune response once attached to the epithelium. Another promising innate immune modulator is phosphatidylcholine, a mucus component that establishes a protective hydrophobic surface in healthy individuals [87]. The phosphatidylcholine content is decreased by about 70% in UC [87]; therefore, treatment with orally administered phosphatidylcholine could be of benefit in these patients [88–90].

In summary, there are several treatment strategies that reinforce the innate barrier which are currently being tested in IBD. It remains to be seen whether these drugs have equivalent or even better efficacy than the present conventional therapy, hopefully with fewer side effects.


Despite many remaining unanswered questions, the understanding of IBD has increased in recent years. For instance, the commensal intraluminal flora is essential in the development and maintenance of IBD, and moreover, the battle against these microbes is lost in IBD because of a defective innate barrier: the production of alpha-defensins is reduced in ileal CD, that of beta-defensins is decreased in colonic CD, and the mucus layer is deficient in UC. All these defects of the barrier shield cause inflammation by bacterial invasion into the mucosa, thus activating the adaptive immune system. The current standard treatment for IBD is based on the suppression of these secondary inflammatory processes. In future, we should focus on causal treatment strategies towards stimulation of the antimicrobial shield in CD and towards stabilization of the mucus layer in UC.

Conflict of interest statement

No conflict of interest was declared.