Molecular pathways driving disease-specific alterations of intestinal epithelial cells

Abstract

Due to the fact that chronic inflammation as well as tumorigenesis in the gut is crucially impacted by the fate of intestinal epithelial cells, our article provides a comprehensive overview of the composition, function, regulation and homeostasis of the gut epithelium. In particular, we focus on those aspects which were found to be altered in the context of inflammatory bowel diseases or colorectal cancer and also discuss potential molecular targets for a disease-specific therapeutic intervention.

Introduction

The gastrointestinal tract forms the largest contact surface between the human body and the environment (more than 300 m²). This large size is indeed a physiological adaption to fulfil absorption/nutrition functions. In this context, the equilibrium between oral tolerance to food antigens and microflora and protective immune responses against pathogens and harmful substances is crucial for the maintenance of intestinal homeostasis. Accordingly, intestinal mucosa represents a complex system in which immune and non-immune cells work together in a tightly regulated manner. Within this system, intestinal epithelium seals the intestinal mucosa towards the lumen and builds up a primary physical and immune barrier [1]. Although cellular transport within intestinal epithelia cells (IECs) is required for absorption and nutrition, the tightness of intestinal epithelium has to be maintained to avoid invasion of luminal components which could otherwise activate immune cells in the subepithelial compartment and generate local immune responses. Consequently, epithelial barrier function plays a relevant role in intestinal homeostasis and impacts on the pathogenesis of gut disorders, such as inflammatory bowel diseases (IBD) and colorectal cancer (CRC).

Crohn’s disease (CD) and ulcerative colitis (UC) represent the two main forms of IBD and show an increasing incidence and high prevalence in Europe and North America [2]. Patients affected by this chronic relapsing disease develop clinical symptoms like diarrhea, abdominal pain, rectal bleeding, fever, weight loss, fatigue and a considerable loss of life quality. Despite intense scientific efforts in the field of gastroenterology and immunology, the exact etiology of IBD remains undefined. IBD represents a multifactorial disease and the characteristic breakdown of intestinal homeostasis is considered to arise from a complex interaction between immunological and environmental factors in a genetically predisposed individual [3, 4]. The complex interplay between dysregulated mucosal immune cells, bacterial flora and impaired epithelial barrier function in the inflamed gut makes it highly challenging to define molecular targets for optimized IBD therapy. While current therapeutic strategies mainly focus on the control of overwhelming immune cell activation, during the last years a growing attention has also been paid to mucosal healing and maintenance of the epithelial barrier integrity [5]. Several studies were able to describe promising approaches with a beneficial effect on epithelial permeability (vitamin D, AT1001, heparanoid compounds), mucus production or integrity of the epithelial monolayer (butyrate, TFF3, anti-TNF antibodies) [6–8]. Taking into account the increased colon cancer risk of IBD patients with a prolonged course of disease [9], restoration and maintenance of epithelial homeostasis becomes particularly important as inflammation-associated epithelial stress is strongly suggested to be a key driver of IBD-associated colorectal cancer [10]. Beside duration of disease, the risk of IBD patients for cancer development depends on the extent of disease at diagnosis, disease severity and efficacy of IBD treatment [11]. In general, intestinal cancer disease (including sporadic colon cancer and inflammation-associated colon cancer) represents one of the major causes of death in developed countries, and its incidence is growing over the last years. It is estimated that 1.4 million new cases of colorectal cancer and almost 694,000 disease-related deaths occurred for instance in 2012 [12]. According to the Fearon and Vogelstein’s model [13], colorectal tumorigenesis shows three key features: tumor development is initiated by clonal expansion of a small number of cells (monoclonal); a regulated sequence of common key genetic alterations provides tumor cells with growth and survival advantages over normal gut epithelial cells; the accumulation of mutations within a tumor determines clinical and histopathological tumor manifestations [14, 15]. In addition to well-described common genetic alterations, other low-frequency candidate mutations are known to contribute relevantly to the heterogeneity of colorectal cancer [16, 17]. Overall, the heterogeneity of involved tumor suppressors, oncogenes and low-frequency somatic mutations makes it highly sophisticating to therapeutically control intestinal tumor growth or metastasis [16–18]. Accordingly, an extensive scientific effort is currently ongoing to further improve our understanding of molecular events and signaling cascades involved in control of epithelial homeostasis in the gut, as well as to carefully describe the interplay between intestinal epithelial cells (IECs) and their surrounding environment.

Due to the fact that chronic inflammation as well as tumorigenesis in the gut is crucially impacted by the fate of intestinal epithelial cells, the following article will try to provide a comprehensive overview of the composition, function, regulation and homeostasis of the gut epithelium. In particular, we will focus on those aspects which were found to be altered in the context of inflammatory bowel diseases or colorectal cancer and will also discuss potential molecular targets for a disease-specific therapeutic intervention (Fig. 1).

Fig. 1.

Intestinal epithelium: structure and key molecular events

Cellular composition and turnover of intestinal epithelium

The intestinal epithelium consists of a monolayer of polarized columnar epithelial cells. Cell composition of this epithelial layer is not homogenous, but contains undifferentiated stem cells, partially differentiated progenitors (transient-amplifying cells) and up to six subtypes of fully differentiated IECs. Mature and differentiated IECs can be functionally categorized in two lineages: enterocytes or absorptive IECs and secretory IECs [19]. Secretory IECs include Goblet cells, Paneth cells, enteroendocrine cells and Tuft cells [20, 21]. Finally, M cells represent specialized IECs which don’t fit in one of the before mentioned categories. M cells are located at the luminal surface of intestinal lymphoid structures and participate in the mucosal uptake of luminal antigens [22]. Beyond cell composition, the complex 3D structure of the intestinal surface significantly contributes to the uniqueness of gut epithelium. The epithelium in small intestine is folded and builds up finger-like structures and invaginations, denoted villus and crypts [23]. The described surface formation results in a large contact area between epithelium and luminal contents and thus allows highly efficient absorption of water and nutrients [24]. However, large intestine lacks villus structures and therefore provides optimized conditions for peristalsis and transport of faecal materials [25].

In adults, the intestinal epithelium represents the most robustly self-renewing tissue. Only 4–5 days are required for its complete renewal by a continuous cell migration from the crypt bottom to the villus tip. Stem cells located at the crypt base proliferate permanently and give rise to the transit-amplifying cells, partially differentiated cells with proliferative activity (progenitors). Due to the pressure of these upcoming cells at the crypt bottom, IECs move upwards along the villus, where the majority of differentiated and specialized IECs are located. At the surface of the villus tip, aged IECs will be released from the monolayer (epithelial cell shedding).

Under physiological conditions, the dynamic interaction between stem cell differentiation, cell migration from the crypt bottom to the villus tip, cell proliferation, cell death and shedding of aged cells is tightly regulated in order to enable a permanent renewal of gut epithelium without loss of barrier function and tissue homeostasis [26]. In the following paragraphs, we will focus on specific aspects of epithelial turnover in the gut and on their impact on intestinal pathology.

Stem cell proliferation and differentiation

Self-renewal of the intestinal epithelium is enabled by a permanent proliferation of stem cells located at the crypt, leading to the replenishment of the intestinal epithelium [27, 28].

Inserted between Paneth cells, crypt base columnar stem cells (CBCs; Lgr5+ cells) [29] give rise to rapidly dividing daughter cells, which will in turn differentiate into IECs of the secretory or absorptive lineage. Even ex vivo, Lgr5+ cells are able to generate organoids mimicking the architecture of the intestinal epithelium and containing all subtypes of differentiated IECs [30]. Multipotent Lgr5+ CBCs expand permanently and undergo a rapid random chromosome segregation, which increases the risks of genomic errors [31–33]. An alternative stem cell population is located in position +4 from the crypt, namely quiescent DNA label-retaining cells (LRCs) [34–36]. Bmi1+ LRCs are mitotic-inactive cells, whose multipotency has been demonstrated in vitro [30, 37] and in vivo [38]. LRCs can be activated by radiation and might play a role in recovery after injury [37]. However, their capacity to function as stem cells is still a matter of controversial discussions. There exists evidence which argues against an unidirectional and hierarchical pathway for differentiation of IECs, but rather supports the idea of a more complex plasticity and interconversion between these two multipotent cell populations depending on certain circumstances [39, 40]. Thus, Lgr5+ would work as main stem cells under physiological conditions and their function might be complemented or overtaken by LRCs in stress situations, such as tissue injury.

In general, stem cell differentiation is directed by various cell-extrinsic factors present in the intestinal stem cell niche located in the crypt area, which carefully control the balanced generation of different populations from identical precursors [41]. The majority of these factors are provided by neighboring cells located in close proximity to stem cells in the crypts. For instance, intestinal subepithelial myofibroblasts (ISEMFs) promote stem cell renewal and differentiation via secreted mediators (e.g. Wnt, TGF-β, growth factors and matrix metalloproteinases), but also by cell–cell interactions [42]. In addition, Paneth cells (or an analogous cell population in the colon) [43] are physically associated to Lgr5+ cells at the crypt bottom and produce essential niche signals for the proliferation and survival of these progenitors, such as EGF, TGF-α, Wnt3 and the Notch ligand Dll4 [44]. Interestingly, Paneth cells can also promote the differentiation of Bmi1+ LRCs [45]. Among these stem cell controlling factors, Wnt proteins fulfill a leading position [46, 47] and therefore the Wnt/β-catenin pathway and its impact on gut physiology will be discussed in a separate subchapter of this article.

The development of colorectal cancer is unavoidably linked to dysregulated proliferation and differentiation of stem cells in the crypt. Indeed, ISEMFs, which represent important regulators of stem cell differentiation, were found to be involved in the development of cancer [48]. However, there is evidence which supports the idea that any kind of epithelial cell is potentially able to drive neoplasia [49, 50], based on the recently proposed cellular plasticity processes [51]. Thus, as a consequence of key genetic alterations or external stimuli, differentiated somatic cells might be able to undergo reprogramming into induced pluripotent stem cells (iPSCs) and acquire the ability to proliferate in a stem cell-like manner [47]. Despite these observations, cancer stem cells represent the main drivers of colorectal cancer development and metastasis [52]. Accordingly, expression of stem cell markers like Lgr5, DCLKL1, CD133, CD44 and CD24 is increased in highly proliferating fractions of colorectal cancer [53]. One step further, the plasticity of stem cells allows the epithelial-mesenchymal transition (EMT), which crucially contributes to the growth and spread of primary tumors and represents a key process for the metastatic behavior of colorectal cancer [51]. Upregulation of EMT gene expression within human colorectal tumors could clearly be shown by transcriptomic analysis [54]. Furthermore, mesenchymal-epithelial transition (MET) is required for restoring the proliferative activity of metastatic tumor cells, which are located far away from the primary tumor site [51]. Summing up, the balance between EMT/MET is critical for the development and metastatic behavior of CRC.

Survival, apoptosis and necroptosis

In the intestinal epithelium, programmed cell death is a critical process which allows the elimination of non-functional (aged/damaged) cells and permits maintenance of epithelial integrity and consequently gut homeostasis [55]. Dysregulation of cell death within IECs was shown to be associated with severe gut pathology: increased cell death leads to breakdown of the epithelium barrier function, invasion of luminal content and subsequent inflammation, while decreased cell death contributes to the development of cancer [55–57].

Historically, cell death was differentiated into apoptosis and necrosis, or programmed and unregulated injury-like cell death, respectively [58]. This simplistic view has changed dramatically in the last 10 years, since new forms of cell death have been discovered and in particular the process of necroptosis has been defined [59]. Indeed, several studies demonstrated that these new players are crucial for the homeostasis of intestinal epithelium and development of gut pathologies, such as IBD and colorectal cancer [60, 61].

Described for the first time in 1970, apoptosis is defined as a programmed cell death induced by various stimuli and finally ending up in caspase activation. Apoptosis can occur via an intrinsic or an extrinsic pathway, depending on the triggering stimulus. Classically proposed as the antithesis of programmed cell death, necrosis is activated in response to external stress (infections, toxins, trauma, etc.) and results in the breakdown of the cell membrane and release of the intracellular content to the extracellular milieu. The immune system will be activated upon necrosis due to the production of pro-inflammatory cytokines by neighboring cells sensing this released material. On the interphase between apoptosis and necrosis, necroptosis was first described in 2003 as a new regulated necrotic-linked cell death pathway [62–65]. Although necroptosis is genetically programmed, regulated and activated by receptor ligation, necroptotic cells share the morphology of cells undergoing necrosis. Experiments in caspase-8 or FADD deficient mice nicely demonstrated that necroptotic cell death is negatively regulated by caspases and strictly depends on RIPK [66, 67]. Thus, inactivation of caspases inhibits apoptosis but simultaneously activates necroptosis, which is supposed to play a crucial role in maintenance of tissue homeostasis. Under physiological conditions, epithelial cell death activation in the gut can be observed in single IECs at the villus tip as well as at the crypt bottom. Those apoptotic IECs, which occur at the villus tip in small intestine or at the surface of the crypt in the colon, usually represent aged cells and are designated for shedding into the lumen. This specific process of programmed cell death, which is initiated by the loss of cell/matrix contact and involves caspase-3 activation, is called anoikis [68]. Although an improved understanding of the molecular link between activation of caspase-3 and subsequent cytoskeleton rearrangement in shedding IECs could be achieved recently, several aspects of this tightly regulated process are still a matter of discussion. Moreover, regulation of cell death in immature IECs at the crypt bottom remains even less defined. In general, epithelial apoptosis seems to be dispensable for the maintenance of gut architecture [69–74], but there exists strong evidence proving that dysregulated or excessive apoptosis within intestinal epithelium is associated with inflammation due to breakdown of the epithelial barrier function. The majority of genes associated with excessive IEC apoptosis belong to the NFκB pathway, such as NEMO [75], RELA [76], TAK1 [77], IKK1 [78] and IKK2 [79]. Another study showed the relevance of the transcription factor XBP1 for epithelial cell death, gut homeostasis and pathology in CD and UC patients [80, 81]. Finally, abrogation of STAT3 within IECs resulted in an increased rate of epithelial apoptosis and enhanced susceptibility to experimental colitis [10, 82]. All in all, these studies show the crucial role of apoptosis for the maintenance of epithelial homeostasis.

Apoptosis and necrosis have both been linked to IBD pathogenesis. A marked increase of epithelial cell death can be observed under chronic inflammatory conditions in the gut of IBD patients. UC patients show accumulation of apoptotic bodies in colon [83] and CD patients and individuals suffering from enteric infections are characterized by an increased number of apoptotic IECs [84, 85]. However, this excessive occurrence of programmed cell death in IECs seemed to be limited to inflamed gut areas [86] and therefore more likely represents a secondary phenomenon of inflammation than a causative effect underlying the inflammatory process. In contrast, there exists certain evidence for a causative role of non-apoptotic cell death in IECs in patients suffering from chronic intestinal inflammation. Increased numbers of dying cells in intestinal epithelium showing morphologic features of necrosis could be observed both in inflamed and uninflamed areas of the gut from CD patients and even in CD patients without any clinical sign of active disease [87, 88]. Interestingly, specific inhibition of apoptosis in IECs by inactivation of caspase-8 in mouse epithelium did not alter epithelial architecture but nevertheless went along with induction of inflammation in the terminal ileum [60, 89]. Several studies showed that the mechanism behind this phenomenon involves the induction of RIP3-dependent caspase-3-independent necroptotic death of Paneth cells. Although the location of Paneth cells is mainly restricted to the small intestine in healthy individuals, recent observational studies have shown the presence and/or metaplasia of Paneth cells in distal colon from IBD patients [90, 91]. Interestingly, the presence of Paneth cells directly correlated with the duration of the disease in UC patients, supporting the association between Paneth Cells and inflammation [90]. However, the lack of Paneth cells in healthy colon might indicate that the function of this cell type is not relevant for the initiation of colonic inflammation, but rather represents a consequence of the inflammatory milieu. Therefore, the potential impact of the described pathogenic mechanism and in particular of Paneth cell death on the development of colonic inflammation needs to be defined in future studies. However, a recent study described an increased expression of markers for necroptosis in the gut of pediatric IBD patients, while there were decreased levels of markers for apoptosis [92]. Furthermore, both isoforms of the master anti-apoptotic regulator cFLIP were found to be upregulated in gut tissue of CD and UC patients, implicating the inactivation of caspase-8 and activation of necroptosis in IBD [93].

Thus, an imbalance between apoptosis and necroptosis in intestinal epithelium can be suggested as key pathogenic factors in IBD.

Cell shedding

After differentiation and migration along the crypt-villus axis, aged IECs reach the villus tip, from where they are shed into the lumen. This continuous replacement of older and potentially damaged IECs by upcoming new cells enables the intestinal epithelium to maintain a constant barrier function and homeostasis of the tissue.

Considering epithelial cell extrusion as the final step in the process of epithelial turnover in the gut, it is important to keep in mind the relationship between cell death and shedding. For years, it was believed that activation of caspases occurs shortly before the completion of the shedding event (homeostatic caspase-dependent cell shedding) [94]. Importantly, in most of the underlying studies experimental cell shedding was initiated by administration of well-known inducers of cell death, such as LPS or TNF-α [95, 96]. More recently, it could be shown that cell shedding can also occur in a caspase-independent way, which was then defined as physiological cell shedding. In this process the activation of caspases occurs only secondary to the detachment of IECs from the monolayer and might represent a consequence of cytoskeleton rearrangement [97–99]. In conclusion, current data postulate the co-existence of a homeostatic caspase-independent cell shedding process important for physiological epithelial turnover and a caspase-dependent pathological cell shedding process, which often results in barrier dysfunction and inflammation.

In the process of pathological cell shedding, such as TNF-induced IEC extrusion, caspase activation initiates the following sequence of events: tight junction protein and actin redistribution; actin-myosin, microtubules and dynamin-dependent cell extrusion and closing of the epithelial gap by actin, ROCK, MLCK and dynamin [95]. Although numerous studies described mechanism involved in pathological or apoptosis-mediated cell extrusion [96, 100, 101], little is known about the recently identified process of physiological cell shedding. The latter is independent of caspase activation and occurs in epithelium of healthy tissues in order to maintain the balance between growth and cell death rates. Regarding the mechanism underlying physiological cell shedding, we can only assume that it is also dependent on cytoskeletal rearrangement and in particular involves the cytoskeletal protein ROCK, acto-myosin complexes and redistribution of tight junctions.

Under physiological conditions, the cell shedding event itself compromises the epithelial barrier function temporarily [102]. The immediate recovery of epithelial integrity is achieved by early redistribution of actin and tight junction proteins along the lateral membranes of shedding cells, a coordinated microfilament interaction and finally the sealing of the space left by the shedding cell by adjacent cells [95]. Under inflammatory conditions in IBD patients, the frequency of IEC shedding is markedly increased and subsequently the epithelial layer loses its capacity to maintain an effective barrier function via cytoskeletal rearrangement [103].

Cytoskeleton and tight junctions in gut epithelium

On a subcellular level, maintenance of epithelial cell shape, anchoring of cells to the basal membrane, homeostatic renewal of the intestinal epithelium and constant barrier function of the epithelial monolayer strongly depend on the complex and tightly regulated cytoskeleton network in IECs [26, 98, 104]. Cell membranes from individual epithelial cells act as an impermeable barrier, where solute passage is only possible via specific transporters. However, the space between neighboring cells (paracellular space) represents a vulnerable point within this physical barrier. Several intercellular junctions are therefore needed in order to maintain epithelial integrity: tight junctions, adherens junctions, desmosomes and gap junctions [104]. Tight junctions consist of transmembrane proteins, like for instance claudins, junctional adhesion molecules (JAM) and occludins, which intimately associate with cytoplasmic actin and myosin networks via several adaptor molecules or plaque proteins, such as ZO-1 or TJP1 [105, 106]. Actin cytoskeleton or intermediate filaments guarantee the mechanical strength of these intercellular junctions. Contraction of a perijunctional actomyosin ring further regulates the permeability of intestinal epithelium in a myosin light-chain kinase (MLCK) dependent manner [107]. In addition, the described intercellular junctions efficiently block the intramembranal diffusion of membrane components and thereby enable the apical-basolateral polarity of IECs [108]. Beside tight junctions, adherens junctions (AJs, e.g. E-cadherin, nectins) are assumed to form dynamic connection between epithelial cells, participating in actin polymerization [109, 110]. Finally, desmosomes (desmoglein, desmocolin) bind to keratin intermediate filaments and provide additional structural strength [111]. Tight junctions are known to seal the paracellular pathway, whereas adherens junctions and desmosomes rather provide the mechanical adhesive strength and thereby maintain cellular contacts and allow tight junction assembly. As a functionally unique intercellular junction, gap junctions (connexins, pannexins) act as direct communication sites (connexons) between neighboring cells and allow the exchange of intracellular material (ions, small molecules) [112].

As mentioned before, intercellular junctions are physically and functionally linked to the cytoskeleton of IECs [113, 114] and this association seems to be bidirectional [26]. Therefore, epithelial integrity and barrier function depends not only on the regulation of tight junctions or single components within the system of intercellular junctions, but rather on a dynamic and complex interaction between several of these components. Breakdown of epithelial integrity has been observed after disruption of intercellular junctions and cytoskeleton rearrangement, e.g. in the context of infection or inflammation [115–117]. Another important molecular player in this cellular network are small GTPases [118, 119], which will be discussed later in this article.

Loss of epithelial integrity and increased tight junction permeability is well known to be associated with development of gut inflammatory disorders, such as IBD [120–122]. In CD patients, a significant correlation between increased intestinal permeability and the presence of disease markers or clinical relapse could be shown [123, 124]. It still remains unclear whether this disease-associated loss of epithelial barrier function represents a cause or a consequence of intestinal inflammation. Performed studies in tight junction-deficient mice implicated that increased intestinal permeability due to the lack of single tight junction proteins can be efficiently compensated and so far only claudin-15 deficiency resulted in a pathological intestinal phenotype [125–128]. Rather than intrinsic dysregulation of specific cytoskeletal components, a potential impact of extrinsic factors on the interplay between intercellular junctions and cytoskeleton might underlie the loss of epithelial integrity in IBD patients, like for instance the pro-inflammatory cytokines IL-6 [129], IL-13 [130] and TNF [131, 132]. Indeed, release of TNF in inflamed mucosa triggers an increased expression and activity of myosin light chain kinase activity in IBD [133, 134]. Subsequently, an increased phosphorylation of myosin light-chain-II and the resulting overwhelming contraction of the perijunctional ring might explain the breakdown of epithelial barrier function in those patients [107]. However, in vivo confocal laser endomicroscopy in IBD patients during remission phase was able to demonstrate that loss of epithelial barrier function often precedes the clinical relapse of disease [101, 103, 135]. The latter argues for a causative role of disrupted permeability in IBD pathogenesis. Furthermore, even healthy relatives of Crohn’s disease patients showed an increased intestinal permeability [136–138].

Role of intestinal epithelial cells in immune surveillance

The described barrier function of the intestinal epithelium in the interphase between the gut lumen and the lamina propria of the mucosa also implicates a permanent exposition of IECs to extrinsic microbial as well as immune cell derived stimuli. Accordingly, there are multiple interactions between IECs, gut microbiota and lamina propria immune cells, which crucially impact on IEC function and intestinal homeostasis [139, 140].

After the initiation of postnatal enteral feeding, the gut is colonized by a large amount of microorganisms (microbiota or microflora). The majority of bacteria in the gut accumulate in the ileum and large intestine and a number of 10¹¹ bacteria per gram of tissue is estimated [141, 142]. The mutually beneficial symbiotic relationship between the microbiota and the host supports host nutrition, contributes to the maturation of the host immune system and prevents intestinal mucosa from colonization with enteric pathogens. On the other hand, bacteria from the microbiota obtain a niche and adequate nutrients for their survival. However, bacteria colonization also involves a permanent risk for activation of local immune responses. Therefore, the intestinal epithelium fulfils a lifelong crucial role in immune surveillance, impairing the invasion of luminal content and thereby maintaining tissue homeostasis. Several studies have demonstrated the association between dysbiosis or imbalance of the gut microbiota and IBD [143–145]. Although several bacterial species, such as Mycobacterium paratuberculosis [146, 147], Salmonella [148] or Campylobacter [147], have been discussed as potential causative triggers of IBD development, there is a broad consensus that mainly bacteria from the physiological gut flora are responsible for the initiation of overwhelming intestinal immune responses in IBD. Indeed, experiments with germ-free mice supported the hypothesis that participation of normal microflora components in genetically predisposed individuals contributes to the development of chronic intestinal inflammation [149, 150].

Mucus and antimicrobial peptide production

As an additional physical and chemical outer barrier between luminal microorganisms and the epithelium, highly glycosylated mucins secreted by goblet cells give rise to a gel-like film covering the apical epithelial surface. The production of mucins by Goblet cells contributes to the equilibrium between intestinal epithelium and commensal flora. Being the most abundant component of this mucus layer, MUC2 protein represents an indispensable factor for maintenance of intestinal homeostasis in the colon. Accordingly, MUC2 deficient mice spontaneously develop colitis and are highly predisposed to inflammation-associated tumor disease [151, 152]. Besides MUC2, additional components, such as trefoil factors (TFF) or resistin-like molecules (RELM), significantly impact on the barrier function of the mucus layer. For instance, TFF3 contributes to the integrity of the mucus layer, promotes epithelial repair and resistance to apoptosis [7], while RELMβ drives mucin secretion, regulates inflammation and protects against parasite infection [153, 154].

The protective effect of the mucus layer is further supported by its ability to retain antimicrobial peptides (AMP) which are able to target and kill luminal bacteria [155]. AMP represent endogenous antibiotics which are able to recognize essential and highly conserved bacterial patterns and thereby control the composition of the gut microflora. The majority of AMP, such as defensins, cathelicidins and lysozyme, are secreted by Paneth cells in the crypt base of the small intestine, although certain AMP (e.g. RegIIIγ and β-defensins) are produced by absorptive enterocytes in small and large intestine [156]. Accumulation of secreted AMP within the mucus layer results in the formation of an efficient innate barrier which relevantly limits the direct contact between gut microbiota and the gut epithelium [21, 157]. Data from animal and human studies demonstrated that an impaired antimicrobial defense promotes the development of chronic inflammatory diseases such as IBD [158]. For instance, lack of cathelecidin in mice lead to increased severity of DSS-induced colitis [159]. In humans, CD patient with an ileal disease manifestation were found to be characterized by an decreased expression of α-defensins in intestinal epithelium and a subsequently diminished epithelial barrier function [158, 160]. Interestingly, decreased expression levels of α-defensins in CD patients turned out to be independent on inflammation [160, 161] and might therefore represent a very early step within the pathogenesis of this disease. Regarding the underlying mechanisms which are responsible for the observed pathologic lack of α-defensins, a compromised Paneth cell function and the inability of monocytes to provide adequate amounts of WNT ligands for Paneth cell stimulation was described in CD patients [160, 161]. Interestingly, in UC patients the expression of α-defensins is upregulated [162]. However, the antimicrobial defense might still be compromised due to the observed decreased thickness of the mucus layer in those patients [163]. Concerning CRC, the potential impact of AMP on tumor development has not been studied extensively. However, some studies claim that the ability of AMP to orchestrate adaptive immune responses might play a role in tumorigenesis [164, 165], but this has not been investigated in the context of CRC. Therefore, further research is needed in order to improve our understanding of this interesting aspect.

Pattern recognition receptors

Beyond a simple physical barrier, the intestinal epithelium fulfills an important innate immune function by recognizing microbes via pattern recognition receptors (PRRs) [139]. Belonging to the PRR family, Toll-like receptors (TLRs), NOD-like receptors (NLRs) and RIG-like receptors (RLRs) expressed on the surface of IECs act as frontline sensors of microbes and are able to initiate antimicrobial as well as immunoregulatory responses. TLR binding of respective ligands results in dimerization of the TLR, recruitment of specific adaptor proteins and activation of subsequent signaling pathways. For instance, Mal or (TRAM)/TRIF represent important TLR associated adaptor proteins, which are linked to downstream NFκB or IRF signaling, respectively [166, 167]. Microbiota-driven TLR signaling impacts on IEC proliferation and differentiation, IgA production, maintenance of epithelial tight junctions and release of antimicrobial peptides and thus represents an important control element in the convoluted coexistence of gut microflora and host organism [168, 169].

The crucial involvement of TLRs in IEC biology implicates that TLR signaling also impacts on intestinal pathology. Indeed, TLR expression was found to be upregulated in polarized intestinal epithelial cells in the gut of IBD patients (107). This phenomenon is most probably driven by an increased presence of proinflammatory cytokines, such as IFNγ [170], TNF [171] or IL-13 [172]. As TLRs on the apical surface of IECs are assumed to be tolerant against luminal stimuli, breakdown of epithelial integrity under inflammatory conditions might further trigger TLR signaling by allowing basolateral stimulation [173–175]. Interestingly, some TLRs show a disease-specific alteration of their intracellular distribution pattern: TLR3 expression is shifted to the basolateral surface in colon epithelial cells of ulcerative colitis patients [176] and TLR4 accumulates at the apical side in Crohn’s disease patients [176]. In addition to their altered expression and subcellular localization, the functional relevance of TLRs in gut epithelium was also found to be modified under inflammatory conditions. Although TLR4 and the TLR adaptor protein MYD88 are dispensable for bacterial-regulated proliferation of IECs under physiological conditions, this turned out to be not true in the context of experimental colitis in dextran sulfate sodium (DSS) exposed mice. Decreased IEC proliferation and increased IEC apoptosis could be detected in the inflamed colon of MYD88 and TLR4 deficient mice. Regarding the underlying mechanism, TLR4 signaling was suggested to contribute to IEC proliferation by inducing a COX-2/PGE2 dependent expression of epithelial growth factors (EGFR) [177] and by supporting the recruitment of mesenchymal stromal cells (fibroblast, macrophages) [178]. Although activation of other TLRs, such as TLR2, TLR3 or TLR9, was also described to be beneficial in the context of colitis, these effects were probably not mediated by a direct initiation of IEC intrinsic signaling cascades [179–181]. Finally, there exist intracellular negative regulators of TLR function which are responsible for the tolerance of epithelial TLRs against physiological gut microbiota. Altered function of these regulatory molecules goes along with breakdown of intestinal homeostasis and subsequent development of inflammation [182, 183]. Interestingly, patients suffering from IBD show a defective function of some of these TLR negative regulators, such as TOLLIP [184].

However, the capacity of epithelial TLRs to sense injury in the intestine and to limit the extent of damage by inhibition of IEC apoptosis, dampening of proinflammatory pathways and induction of IEC proliferation also has a negative side. Studies performed in experimental models of colorectal cancer clearly demonstrated that intensive TLR signaling is able to drive tumorigenesis and promote cancer development [185]. Interestingly, a protective effect in cancer development was described for some NLRs. This beneficial effect of NLRs was mediated by different mechanisms, such as regulation of proliferation and cell death of transformed IECs or IL-18-dependent tissue repair [183, 186, 187].

Lympho-epithelial interactions

IEC function and epithelial integrity is relevantly regulated by lamina propria immune cells and surrounding stromal cells [139]. Thus, a number of extrinsic factors, like pro-inflammatory cytokines or pro-apoptotic stimuli, are well known to impact on barrier function of the epithelial layer in the gut. Vice versa, IECs themselves produce soluble compounds and express specific surface proteins that in turn regulate the function of immune cells in the epithelial environment in order to generate tolerance, limit inflammation and support adequate immune responses. IECs constitutively express major histocompatibility complex (MHC) class I and II molecules and a diverse set of non-classical MHC class I molecules on their surface [188]. Thereby, IECs are able to act as non-professional antigen-presenting cells and to process and present luminal antigens to antigen-specific lymphocytes within the lamina propria [189]. Although IECs lack classical co-stimulatory molecules, the expression of non-classical MHC class I molecules and molecules of the B7 family [190] enables them to fulfill a costimulatory function on immune cells. Interestingly, the expression of epithelial MHC-II in the gut is upregulated in response to pro-inflammatory signals [191, 192] and intestinal inflammation in IBD patients significantly impacts on MHC-I and II compartments in intestinal epithelial cells [193]. A recently performed in vivo study showed that abrogation of the inflammation-induced MHCII expression on IECs in the context of experimental colitis resulted in a more severe course of disease and in an accumulation of CD4⁺ Th1 cells in the gut, implicating a relevant involvement of MHC-II expressing IECs in T cell-dependent mechanisms of intestinal mucosal tolerance [194].

Among those IEC derived factors which relevantly impact on T cell fate, tumour necrosis factor (TNF) represents one of the most intensively studied and best characterized players in intestinal homeostasis.

Members of the tumour necrosis factor superfamily (TNFSF; including 19 ligands and 29 receptors) display a broad spectrum of activities within intestinal mucosa and impact both on intestinal epithelium and immune cells. TNF proteins initiate signaling pathways related to cell survival, proliferation, differentiation and apoptosis and trigger production of inflammatory mediators. In brief, the ligand binds to its corresponding receptor, leading to the recruitment of adaptor proteins and initiation of subsequent signaling pathways [195]. The overall consequence of TNF signaling in intestinal mucosa strongly depends on the cellular context and on involved signaling pathways. Despite the well-established role of TNF members in T cell-mediated immune responses and autoimmune diseases [196, 197], TNF signaling also represent a key event in the maintenance of epithelial integrity [198] with relevant impact on IBD pathogenesis.

Tumour necrosis factor-alpha (TNF-α, or TNFSF2) represents the best known member of the TNF family. There exist two forms of TNF-α, a transmembrane and a soluble protein. While transmembrane TNF-α preferentially binds to TNF receptor 2 (TNFR2), soluble TNF-α predominantly interacts with TNFR1. TNF-α expression occurs in various cell types and is upregulated by various proinflammatory factors [199]. TNF gene was identified as a susceptible IBD locus [200–202] and expression of TNF-α was found to be elevated in IBD patients in several tissue compartments: TNF-α mRNA in colonic tissue [203, 204]; TNF-α protein in serum [205–207], stool [208], intestinal lamina propria and submucosa [209–212]. However, there also exist few studies which were not able to confirm increased TNF-α levels in IBD [213, 214]. Expression levels of TNFR are also elevated in serum [215], intestinal epithelial cells [216] and CD4+ cells [217] of IBD patients. Furthermore, regarding functional aspects of TNF-α signaling in the gut, experimental overexpression of TNF-α in mice resulted in the development of intestinal pathology resembling Crohn’s disease [218]. Finally, the highly successful use of anti-TNF-α agents in the treatment of IBD patients provides an unequivocal proof of the crucial functional role of this pathway in the pathogenesis of CD and UC. In addition to its beneficial effects on immune cells, therapeutic blockade of TNF-α is assumed to impact on rearrangement and expression of tight junction proteins and to protect epithelial barrier function [219–221].

In particular under inflammatory conditions, IECs are able to express TNFR1,TNFR2 and the pro-inflammatory cytokine TNF-α [216, 222, 223], which is also produced by invading immune cells and stroma cells in the gut mucosa. Accordingly, there is a well-established role of TNF-α in epithelial barrier regulation in the gut. As a first line of defense, TNF significantly contributes to a balanced mucus production by favoring IEC differentiation into the secretory lineage [224] and inducing MUC2 mRNA expression [225], but also by induction of cell death in Goblet cells [8]. Regarding intercellular junctions and epithelial cytoskeleton, TNF-α is able to modulate the expression of several tight junction proteins [171, 226] and, as already mentioned, triggers expression and activation of myosin light-chain kinase [107, 133, 227, 228]. Redistribution of tight junction and adherens junction proteins in the presence of TNF-α results in the induction of IEC shedding and thereby further impacts on epithelial integrity [95, 101, 198]. Interestingly, several studies implicated an association between hyperpermeability of intestinal epithelium and regulation of TNF sheddase enzymes, which catalyze shedding of the membrane-bound TNF-α protein from the cell surface [132, 229]. Finally, TNF-α represents an important regulator of IEC survival. On the one hand, TNF signaling is able to directly activate cell death within IECs [230–232] or induce the expression of the pro-apoptotic protein p53 [231] and thereby contributes to barrier defects. On the other hand, TNF-α protects from intestinal apoptosis and promotes wound healing in a ErbB- and Wnt-dependent pathway [233–235]. Thus, the overall consequence of TNF-α signaling in the intestinal epithelium is context-dependent and is determined by the selected intracellular signaling cascade which is triggered by TNFR ligation. Beside TNF-α, other TNF family members were found to be associated with dysfunction of epithelial barrier function in the context of IBD. These are for instance TL1A [236, 237], FasL [238–240], LIGHT [241], TRAIL [242] or TWEAK [131].

Although TNF might be considered as the main extrinsic regulator of IEC function, it is not the only immune cell derived factor involved in epithelial homeostasis in the gut. In particular, IL-6 and IL-22 should be taken into account in this context. IL-22 is produced predominantly by T cells and natural killer cells, while the IL-22 receptor is expressed solely on non-hematopoietic cells including intestinal epithelial cells [243]. In Crohn’s disease patients, lamina propria T cells and macrophages could be identified as major cellular source of intensively expressed IL-6 [244]. Most of IL-6 and IL-22 initiated effects on IECs are mediated via activation of STAT-3 and finally result in induction of IEC proliferation and inhibition of apoptosis. Due to the key importance of the STAT/JAK pathway in the intestinal epithelium of IBD patients, this pathway will be described in a separate subchapter in this article.

Master molecular regulators and signaling pathways in intestinal epithelium

As described before, stability, function and maintenance of the intestinal epithelial monolayer strictly depends on a balanced interplay between IEC differentiation, proliferation, survival and extrusion as well as on the capacity of IECs to respond to multiple epithelial-extrinsic triggers, like for instance components of gut microflora or immune cell derived mediators. The fate of IECs within this complex setting is tightly regulated by an intracellular network of key molecular factors organized in different signaling pathways. Selected candidates of this network, their disease-specific alteration and their applicability as potential therapeutic targets will be discussed in the following paragraphs.

NFκB signaling

The transcription factor NFκB controls the expression of a variety of genes critically involved in IEC homeostasis [245]. The NFκB/Rel family consists of 5 peptides (p50, p52, RelA, RelB, c-Rel) which form homo or heterodimers [246]. Inactive dimers are localized in the cytosol, where they are retained by an inhibitor protein (IκB). Ubiquitination and proteosomal degradation of IkB finally enables nuclear translocation and subsequent activation of NFκB [247]. NFκB target genes can be classified in four main groups: inflammation related genes, genes involved in cell cycle, anti-apoptotic mediators and negative regulators of NFκB signaling [248, 249].

Numerous studies clearly demonstrated a crucial involvement of NFκB in the pathogenesis of IBD. Increased levels of activated NFκB could be detected in the inflamed gut in different mouse models of IBD and, moreover, specific blockade of NF-kB signaling was able to ameliorate severity of experimental colitis [250–252]. An increased NFκB activation could also be shown in ex vivo stimulated IECs, strongly implicating a functional relevance of this pathway for gut epithelium [253, 254]. In the context of human disease, macrophages and IECs in the intestinal mucosa of IBD patients are characterized by an increased expression and activation of NFκB [250, 255]. With regard to immune cells, activation of NFκB leads to induction of proinflamamtory cytokines like IL-1, IL-2, IL-6 or TNF-α, which in turn contribute to the perpetuation of inflammation. Although NFκB is involved in the regulation of cytokines in gut epithelium as well, in the last ten years a growing attention has been paid to the protective effects of NFκB on epithelial barrier function. Two parallel studies in genetically modified mice strains demonstrated that inhibition of NFκB within IECs results in extensive intestinal inflammation [75, 256]. Interestingly, lack of NFκB activation led to increased apoptosis in IECs and thereby caused a disruption of barrier function and uncontrolled bacterial translocation. The observed protective role of NFκB signaling could further be confirmed by subsequent studies which focused on the role of other components of the NFκB pathway for maintenance of epithelial homeostasis in the gut. In particular, these studies were able to identify a key role of RelA [76], TAK1 [77], IKK1 and IKK2 [78, 79] in IEC biology. Overall, NFκB signaling represents an attractive therapeutic target in the context of IBD, although the opposite quality of NFκB mediated effects in IECs (epithelial protective) and in immune cells (pro-inflammatory) should be taken into account, which makes it necessary to target NFκB signaling in a cell type specific manner.

STAT-3

Signal transducer and activators of transcription (STAT) proteins represent critical components within cytokine induced signaling pathways. Among the 7 members of the STAT family, STAT3 is a pleiotropic mediator that can be activated by a broad panel of cytokines, among which IL-6, IL-11 and IL-22 have been extensively characterized. In general, interaction of the respective cytokine receptor (e.g. gp130) with its ligand results in receptor dimerization and activation of JAK by transphosphorylation. JAK thereupon catalyzes tyrosine phosphorylation of the receptor, which enables binding of STAT proteins. STAT is then activated by a double phosphorylation. First, JAK-mediated phosphorylation allows STAT dimerization, nuclear translocation and DNA binding. In addition, a maximal transcriptional activity of STAT is achieved by MAPK-mediated STAT phosphorylation. IEC relevant target genes of STAT-3 mainly include proteins involved in cell survival and proliferation [10, 257].

In the context of IBD, STAT-3 gene represents a well-known susceptible locus for CD and UC [258, 259], and increased IL-6 levels could be detected in serum and mucosa of affected patients [260]. The consequence of STAT-3 modulation for gut homeostasis depends on the cytokine which triggers STAT-3 activation and on the involved cell type [261–264]. Focusing on the intestinal epithelium, STAT-3 is a decisive factor in the transition between chronic colitis and colitis associated cancer (CAC). Accordingly, increased levels of STAT-3 activation in gut epithelium could be detected in IECs from ulcerative colitis patients with and without diagnosed dysplasia or colon cancer [265]. Conditional knockout mice carrying an IEC specific STAT-3 deficiency are characterized by a disturbed IL-22-induced wound healing and, subsequently, are highly susceptible to development of colitis [82]. Due to the well described impact of IL-22 triggered STAT-3 activation on the expression of mucins and anti-microbial peptides [266], STAT-3 signaling is also crucially involved in the epithelial defense against toxic and infectious agents [266, 267]. Despite the essential role of permanent epithelial turnover for maintenance of epithelial barrier function and wound healing, uncontrolled and overwhelming IEC proliferation represents a hallmark of intestinal cancer development, which might also be driven by activated STAT-3 proteins. While transient activation of STAT-3 is tightly regulated under physiological conditions, cancer cells are characterized by a persistent STAT-3 activation resulting in unlimited promotion of cell growth and inhibition of apoptosis. Accordingly, direct or indirect inhibition of STAT-3 signaling in IECs went along with a reduced intestinal tumor growth in the experimental azoxymethane/DSS model of colitis-associated cancer [10, 257]. Mechanistically, this observation might be explained by an increased induction of apoptosis in colorectal cancer cells in the absence of STAT-3 signaling [268] or by the recently postulated capacity of epithelial STAT-3 to promote gut homing of CD8 T cells and inhibit intestinal accumulation of regulatory T cells [269]. Moreover, the STAT-3 triggering cytokine IL-22 could be identified as a potent inducer of IEC hyperproliferation and driver of colorectal cancer progression in a bacteria-induced model [270]. Interestingly, an increased risk for the development of colon cancer was found to be associated with IL-22 gene polymorphisms [271].

Wnt/β-catenin pathway

Beyond its role during embryogenesis and in maintenance of tissue homeostasis, Wnt pathway is essential for epithelial replenishment and repair of epithelial injury in the gut. Under physiological conditions, the function and growth of intestinal stem cells is tightly regulated by multiple factors. Within this complex interplay, inhibition of the Wnt pathway results in crypt loss and a marked inhibition of epithelial proliferation [272]. Accordingly, the Wnt/β-catenin pathway was found to be altered in the vast majority of colorectal tumors [273].

Briefly, Wnt pathway activation is driven by intracellular accumulation of its main effector β-catenin. In the absence of Wnt, phosphorylated cytosolic β-catenin is included in a protein complex (destruction complex) which facilitates its ubiquitination and proteosomal degradation [274, 275]. Beside others, Axin and Adenomatous polyposis coli (APC) represent functionally important components of this destruction complex. Binding of the Wnt ligand to its respective receptor results in the release of β-catenin from the complex [276]. Subsequently, β-catenin accumulates in the cytosol and translocates into the nucleus, where it forms an active transcriptional complex (TCF/LEF) leading to the expression of Wnt target genes [277, 278]. The list of important Wnt/β-catenin target genes include regulators of cell migration (e.g. EPH), inducers of proliferative signals (e.g. c-myc, cyclin D1) and stem cell and cancer cell markers (e.g. Lgr5, Bmi1) [277, 278]. Regarding the ISC marker Lrg5, it could be shown that binding of R-spondin to Lrg5 is able to promote Wnt signaling [29, 279, 280]. In addition to the before described β-catenin dependent canonical Wnt signaling pathways, Wnt can alternatively affect cell fate via β-catenin independent noncanonical pathways. For instance, noncanonical signaling pathways initiated by Wnt-2 are able to induce a pro-invasive phenotype in cancer cells. Interestingly, recently performed whole-exome sequencing analyses of IBD-associated colon tumors implicated a predominant involvement of noncanonical Wnt pathways in colitis-associated tumorigenesis [281].

The majority of colorectal tumors are classified as nonhypermutated microsatellite stable (MSS) tumors. The sequence of tumorigenic events in MSS tumors was nicely described by the Fearon and Vogelstein’s model [13], which indeed demonstrated the crucial role of Wnt activation in this context. Frameshift or non-sense mutations in the APC gene (tumor suppressor) initiate tumor development, while tumor growth and progression are determined by the accumulation of somatic mutations, affecting genes like KRAS, PI3 K, TGFB, p53 and/or SMAD4 [282, 283]. On a molecular level, loss of APC function impairs the β-catenin destructive protein complex, subsequently blocks β-catenin ubiquitination and results in an uncontrolled activation of the Wnt/β-catenin pathway. Accordingly, APC mutations, both in mouse and human, result in the development of multiple intestinal adenomas [284, 285]. On the other hand, 10–15 % of colorectal tumors are classified as hypermutated microsatellite instable (MSI) tumors, which arise from defects in DNA mismatch repair mechanisms. Among other oncogenes or tumor suppressor genes, proteins involved in Wnt pathway are often mutated in MSI CRC (e.g. APC, β-catenin or AXIN2). Moreover, an epigenetic silencing of several negative regulators of this master pathway could be observed in colorectal cancer [286, 287].

Another important link between Wnt pathway and intestinal epithelial homeostasis arises from the bidirectional regulation between mucin expression and Wnt activation. This is of particular interest because several mucins are known to be aberrantly expressed or downregulated in cancer [288]. In the gut, MUC5AC is exclusively expressed in tumor tissue [289], while expression of MUC1 [290], MUC4 [291, 292] or MUC2 [293, 294] turned out to be decreased in colorectal tumors. Via their cytoplasmic tail, mucins are able to interact with cytosolic β-catenin. Dependent on the specific mucin type and on the cellular context, this interaction might either promote nuclear translocation of β-catenin or result in an accumulation of membrane-bound β-catenin. Vice versa, several studies strongly implicated that β-catenin might directly impact on the transcription of various mucin genes [295]. Subsequently it was postulated that alterations of the mucin expression pattern in colorectal cancer might occur as a consequence of Wnt activation [296–298]. All in all, activation of Wnt pathway in IECs has to be considered as master regulator of colorectal cancer development.

Rho GTPases

Small GTPases of the Ras superfamily act as molecular switches transducing extracellular signals into the intracellular compartment. The more than 50 members of this protein family are characterized by a low molecular weight (between 20 and 29 kDa) and by the presence of GTP binding domains. Five different subfamilies can be distinguished (Ras, Rho, Rab, Sar/ARF and Ran), which participate in different biological processes, such as cell differentiation, cell proliferation, cell migration, cytoskeleton rearrangement and vesicular trafficking. Overall, the maintenance of epithelial barrier integrity relies on a GTPase-mediated fine-tuning of junctional and cytoskeletal dynamics [299]. The activation status of small GTPases is regulated by the transition from inactive GDP-bound into active GTP-bound state [300]. Guanosine exchange factors (GEF), guanosine activating triphosphatases (GAP) and guanosine dissociation inhibitors (GDI) are responsible for a balanced regulation of this transition [301]. GTP-bound proteins interact with effector proteins in order to activate downstream pathways. Activation of small GTPases as well as the capacity to initiate downstream signaling cascades crucially depends on their subcellular localization [302].

Rho-A (Ras homology family member A), together with Rac-1/2, and Cdc42 represent the best described members of the Rho family; although growing attention has also been paid to other family members, like Rho-B, Rho-C and Rho-H. Impaired small GTPase function in intestinal epithelium is associated with junctional and cytoskeletal dysfunction as it is often observed in the gut of IBD patients [303, 304]. A wide number of in vitro studies indeed defined a relevant role of Rho proteins for epithelial integrity, cytoskeleton regulation, cell morphology and IEC migration [305–310]. Despite these numerous publications supporting the role of Rho proteins in maintenance of cytoskeleton in epithelium, it still remained unclear whether Rho inhibition, activation or both would modify epithelial integrity and permeability [311]. Moreover, there exist only very few data about the complex in vivo regulation of epithelial Rho proteins so far.

In the context of gastrointestinal tract and inflammation, few in vivo studies tried to clarify the functional role of Rho GTPases within IECs. Genetic deletion of Cdc42 in mice caused a defect in IEC proliferation and differentiation and resulted in an intestinal phenotype which resembled human microvillus inclusion disease [312, 313]. Similarly, epithelial cell differentiation turned out to be affected after mutations or abrogation of Rac-1 expression and interfered with physiological development of the intestine [314, 315]. A potential role of Rho-A in IBD pathogenesis was first suggested in 2003 when it was observed that experimental colitis as well as intestinal inflammation in IBD patients goes along with increased Rho-A activation in intestinal mucosa. Interestingly, systemic pharmacological inhibition of Rho-A was able to interfere with NFκB signaling and to ameliorate experimental colitis [316]. Supporting the hypothesis that dysregulated Rho-A function, both activation or inhibition, might negatively impact on epithelial integrity, altered Rho-A function could be observed in the epithelium of IBD patients [317]. Inflamed areas in the gut of Crohn’s disease or ulcerative colitis patients depicted an altered subcellular localization of Rho-A within IECs and showed a marked cytosolic accumulation of the presumably inactive small GTPase. On a functional level, IEC-restricted lack of Rho-A resulted in the development of spontaneous colitis [317].

With regard to tumorigenesis, Rho-A expression is upregulated in different kinds of cancer and has therefore been discussed as a potential cancer biomarker [318]. Several studies showed that Rac-1 and Cdc42-mediated regulation of IEC differentiation impacts on cancer development. For instance, activation of these two proteins is required for a Wnt-dependent mechanism of tumor initiation [315, 319, 320]. However, most of these studies used experimental models of sporadic tumors and did not consider inflammation-associated tumor diseases like colitis-associated cancer. Interestingly, recent investigations based on whole-exome sequencing were able to identify a potential key involvement of the Rho/Rac pathway in the pathogenesis of colitis-associated cancer [281]. Although some genetic alterations in IBD-associated colon tumors turned out to be similar to sporadic tumors (e.g. Ras, p53, TGF, Wnt) [273], tumors in IBD patients showed less alterations in APC and Ras and could instead be characterized by new pathways and mutations. These pathways were mainly related to cell communication, cell–cell signaling and adhesion. Interestingly, 50 % of IBD-associated tumors possessed at least a single mutation in genes related to the family of Rho GTPases. Somehow, alterations of Rho/Rac signaling could be observed more frequently in UC patients than in CD patients.

All in all, we conclude that small GTPases represent important mediators of epithelial homeostasis in the gut and significantly impact on the development of colitis and colitis-associated colon cancer. Further research in this area is needed in order to achieve a full understanding of their complex pathogenic involvement and their applicability as potential therapeutic targets.

Prenylation

In order to allow their association with cellular membranes, small GTPases first have to undergo prenylation. Prenylation represents a post-translational process consisting of the attachment of hydrophobic isoprenoid residues to the C-terminal CAAX motif of target proteins and is of particular importance for the intracellular localization and function of small GTPases. The attached hydrophobic residue finally enables the GTPase to anchor to the lipophilic cell membrane and to achieve its full activation status [302, 321]. Indeed, GEF-mediated GTPase activation preferentially takes place when GTPases are associated with cellular membranes [302, 322]. Dependent on the type of GTPase, it is either a geranylgeranyl pyrophosphate residue (e.g. Rac, Rho, Cdc42, Rab) or a farnesyl pyrophosphate residue (e.g. Ras), which is covalently attached to the target GTPase [323, 324]. These enzymatic reactions are catalyzed by respective prenyltransferases, namely geranylgeranyltransferases (GGTase-I and GGT-ase-II) and farnesyltransferase.

The initial idea of a potential link between prenylation and inflammation and/or cancer development arose from the identification of K-Ras as proto-oncogene [325]. Based on this observation and on the dependency of Ras function on appropriate prenylation, inhibitors of prenylation have been investigated in preclinical and clinical studies as anticancer agents and showed promising results [326–329]. In the context of inflammation, it was described that therapeutic use of statins in patients with atherosclerosis resulted in anti-inflammatory effects which were mediated via inhibition of prenylation [330, 331]. Based on this observation, the therapeutic use of statins as potent inhibitors of isoprenoid precursor synthesis has also been discussed in the context of intestinal inflammation and other inflammatory disorders not primary related to lipid metabolism [332–334]. Moreover, nitrogen-containing bisphosphonates, which are known to interfere with prenylation by inhibition of isoprenoid synthesis, could be identified as effective anti-inflammatory agents in several experimental models of colitis [335, 336]. Taken together, these observations convincingly argued for a functional impact of prenylation on intestinal homeostasis, chronic gut inflammation and cancer development.

In line with this hypothesis, a recent study clearly identified the indispensable role of geranylgeranylation within IECs for maintenance of intestinal homeostasis. Mice lacking GGTase-Iβ expression in intestinal epithelium develop a severe intestinal disease with increased intestinal permeability and marked intestinal inflammation [317]. The mechanism behind this striking phenomenon involves the destruction of intestinal architecture due to dramatic cytoskeleton rearrangement within IECs. In the absence of adequate epithelial geranylgeranylation, Rho-A dysfunction might alter cytoskeleton arrangement leading to dysregulation of cell shedding, increased epithelial permeability and breakdown of barrier function. Interestingly, enteral application of a synthetic Rho activator was able to partly compensate the lack of epithelial geranylgeranylation and to ameliorate gut pathology [317]. Thus, geranylgeranylation of Rho-A represents a crucial epithelium-intrinsic mechanism to maintain intestinal barrier integrity and gut homeostasis. The clinical transferability of these findings was further supported by the observation that gut epithelium of IBD patients with active disease is characterized by decreased GGTase-I expression levels [317].

In summary, we can postulate that prenylation represents a novel relevant pathway for maintenance of gut homeostasis and epithelial integrity. Regarding involved target proteins of prenylation, a particular role of Rho-A in IEC biology could be identified. In line with this, we can hypothesize that other proteins belonging to Rho family, which are also regulated by prenylation, might be important in this context as well. Future studies are needed in order to further elucidate the molecular events which are affected by Rho-A and other targets of prenylation within intestinal epithelium.

Conclusions and future perspectives

The here reviewed aspects of IEC biology and function clearly indicate that the maintenance of epithelial integrity and homeostasis should be considered as a key target in the therapeutic and/or even preventive clinical management of IBD and CRC.

However, due to the vast number of different mechanisms regulating epithelial homeostasis in the intestine and the complexity of the structure and cellular composition of gut tissue, profound collaborative scientific efforts are required in order to allow a detailed understanding of the interplay between involved molecular pathways. During the last years, an enormous advance has been achieved in this field and enabled us to gain interesting new insights into the role of IECs in gut homeostasis and pathology. In the context of inflammation, special attention has been paid to Paneth cell biology and necroptotic cell death, mucus composition, as well as AMP and epithelial-microbiom interactions, while stem cell proliferation/differentiation, Wnt pathway and STAT-3 signaling represent important candidates in colorectal tumorigenesis. Moreover, the recently identified crucial role of prenylation and small GTPases in the maintenance of epithelial architecture introduced further potential new target structures for an optimized treatment or early diagnosis of IEC-derived gut diseases. In future studies, it will now be important to translate these findings into innovative therapeutic strategies and pave the way for a clinical applicability and, subsequently, an optimized therapy of IBD and CRC patients.

Abbreviations

AJs

Adherens junctions

AMPs

Antimicrobial peptides

APC

Adenomatous polyposis coli

Bmi

B cell-specific Moloney murine leukemia virus integration site

CAC

Colitis-associated cancer

CBCs

Columnar stem cells

CD

Crohn’s disease

cFLIP

Cellular FLICE-inhibitory protein

COX-2

Cyclooxygenase-2

CRC

Colorectal cancer

DCLKL

Doublecortin like kinase 1

DNA

Deoxyribonucleic acid

DSS

Dextran sodium sulfate

EGF

Epidermal growth factor

EMT

Epithelial-mesenchymal transition

FADD

Fas-associated protein with death domain

FasL

Fas ligand

GAP

Guanosine activating triphosphatases

GDI

Guanosine dissociation inhibitors

GDP

Guanosine diphosphate

GEF

Guanosine exchange factors

GGTase

Geranylgeranyltransferase

GTP

Guanosine triphosphate

IBD

Inflammatory bowel diseases

IECs

Intestinal epithelia cells

IFN

Interferon

IL

Interleukin

IKK

IκB kinase

iPSCs

Induced pluripotent stem cells

IRF

Interferon regulatory factors

ISC

Intestinal stem cell

ISEMFs

Intestinal subepithelial myofibroblasts

JAK

Janus kinase

JAM

Junctional adhesion molecules

Lgr

Leucine-rich repeat-containing G-protein-coupled receptor

LIGHT

TNF ligand superfamily member 14

LPS

Lipopolysaccharides

LRCs

Label-retaining cells

Mal

MyD88-adapter-like

MAPK

Mitogen-activated protein kinases

MET

Mesenchymal-epithelial transition

MHC

Major histocompatibility complex

MLCK

Myosin light-chain kinase

MSI

Microsatellite instable

MSS

Microsatellite stable

MUC2

Mucin 2

MYD88

Myeloid differentiation primary response gene 88

NFκB

Nuclear factor ‘kappa-light-chain-enhancer’ of activated B-cells

NLR

NOD-like receptor

PGE2

Prostaglandin E2

PI3K

Phosphatidylinositol-4,5-bisphosphate 3-kinase

PRR

Pattern recognition receptor

R

Receptor

RELM

Resistin-like proteins

Rho

Ras homology family member

RIPK

Receptor-interacting protein kinases

RLR

RIG-like receptor

ROCK

Rho associated kinase

STAT

Signal transducer and activators of transcription

Tak

Tat-associated kinase

TFF3

Trefoil factor 3

TGF

Transforming growth factor

Th

T helper cell

TJP1

Tight junction protein 1

TL1A

TNF-like ligand 1A

TLR

Toll-like receptor

TNF

Tumor Necrosis Factor

TOLLIP

TOLL interacting protein

TRAIL

Tumor necrosis factor related apoptosis inducing ligand

TRIF

TIR-domain-containing adapter-inducing interferon-β

TWEAK

TNF-related weak inducer of apoptosis

UC

Ulcerative colitis

Wnt

Wingless-type MMTV integration site family member

XBP1

X-box binding protein 1

ZO-1

Zona occuldens protein 1