Signaling the Junctions in Gut Epithelium

See allHide authors and affiliations

Science's STKE  29 Mar 2005:
Vol. 2005, Issue 277, pp. pe13
DOI: 10.1126/stke.2772005pe13


This Perspective summarizes recent developments in our understanding of the signaling pathways involved in the regulation of epithelial cell adhesion in the gut. The role of phosphatidylinositol 3-kinase signaling in the modulation of adherens junctions, and the connections between tight junctions and nuclear transcription factors, are discussed. The effect of gastrins on adherens and tight junctions is presented as an example of the regulation of adhesion by growth factors. The consequences of dysregulation of adherens junctions and tight junctions for human pathology are also considered.

Our understanding of the molecular signaling involved in the regulation of epithelial cell adhesion in the gut has increased significantly in the past 5 years. We summarize some exciting new results concerning the signaling pathways coupled to adherens junctions (AJs) and tight junctions (TJs) in the gut.

Architecture of the Gut Epithelium

The gut epithelium is a highly organized structure that is exquisitely adapted to the controlled absorption of dietary components from the gastrointestinal lumen. This function is maintained over a human lifetime of 70 or more years by tight regulation of the critical balance between proliferation, differentiation, migration, and death of the epithelial cells. In the small intestinal crypt, for example, stem cells differentiate into four major mature cell types (absorptive, enteroendocrine, goblet, and Paneth cells), and the past decade has seen major advances in our understanding of the regulation of differentiation and cell migration to appropriate positions within the crypt or along the associated fingerlike projections (villi) (1). A complex interplay between multiple signaling events within epithelial cells, under the control of extracellular signals stemming either from the immediate neighborhood (such as cell-cell or cell-matrix contacts) or from further away (such as growth factors or hormones), maintains the specific organization of this tissue, without breaching the integrity of the barrier separating luminal fluids from internal organs.

AJs and TJs

Specialized structures such as AJs and TJs play a major role in the formation of intracellular contacts and the establishment of cell polarity in vivo (2). AJs constitute the strongest physical link between neighboring cells (3) and are established through calcium-dependent homophilic binding of E-cadherin molecules, which are in turn connected to the actin cytoskeleton through proteins such as α- and β-catenin. TJs are formed by interactions between several families of transmembrane proteins, including occludin, claudins, and junctional adhesion molecules (JAMs). These proteins display organ-selective expression in epithelial and endothelial tissues and are connected to the actin cytoskeleton through a network of submembrane or "cytoplasmic plaque" proteins. Functional TJs form a tightly controlled barrier to diffusion across the epithelium and act as trafficking platforms that are involved in sorting apical and basolateral membrane proteins (4). Both AJs and TJs play an active role in the polarization of epithelial cells in vitro and in vivo (24), although recent results indicate that intestinal cells have the capacity to polarize in the absence of AJs or TJs upon activation of the serine-threonine kinase LKB1 (5).

The establishment and stability of both AJs and TJs are tightly regulated by growth factors and cytokines such as hepatocyte growth factor, gastrin, and interferon-γ (6, 7). This regulation is only partially understood but seems essential for development, morphogenesis, and wound healing in the gut; for the epithelium-mesenchyme transition; and for the modulation of paracellular permeability in various epithelia. AJs are subject to regulation by multiple overlapping signaling pathways (3). In turn, AJs and TJs themselves initiate or modulate a series of signaling pathways that regulate gene expression (2, 3, 8).

Signaling to and from AJs: The Complex Role of PI3K/Akt

Recruitment and activation of phosphatidylinositol 3-kinase (PI3K) is an essential step in the regulation of epithelial cell differentiation and polarity, and several recent publications have shed new light on the complex role played by this enzyme in AJ-associated signaling. Formation of E-cadherin–mediated cell-cell contacts between intestinal cells (the Caco-2/15 clone of the human colorectal carcinoma cell line Caco-2) was reported to lead to the recruitment of the p85 regulatory subunit of PI3K to AJs (Fig. 1) by binding to the human homolog of the Drosophila tumor suppressor Disc-large. Recruitment induced activation of PI3K and of the protein kinases Akt and p38 MAPK (mitogen-activated protein kinase) (9) and caused a stabilization of AJs, thereby promoting epithelial cell polarization and differentiation (10). The same pathway also caused inhibition of the kinases MEK [a mitogen-activated protein kinase kinase (MAPKK)] and ERK (extracellular signal–regulated kinase, a MAPK), leading to reduced Caco-2/15 cell proliferation (11). However, these results are in apparent conflict with other reports that inhibition of PI3K induces differentiation in the parental Caco-2 cell line (12) and in other epithelial cell types (13), and that bone morphogenic protein (BMP) increases amounts of nuclear β-catenin and activates Wnt signaling through PI3K in the mouse intestine (14). Studies on human tissues also argue in favor of a negative role for the PI3K signaling pathway in epithelial cell differentiation. Indeed, PI3K signaling is frequently activated in carcinomas, through amplification of PI3K itself or through inactivating mutations in the gene encoding the phosphatase PTEN (a tumor-suppressor protein that dephosphorylates phosphatidylinositol 3,4,5-trisphosphate and antagonizes the PI3K signaling pathway) (15).

Fig. 1.

Representative signaling pathways during the formation of enterocyte junctions. (A) In isolated cells, submembrane proteins from AJs and TJs are not associated with cell-cell contacts. Cytoplasmic β-catenin (β-cat) is mostly targeted to the APC–glycogen synthase kinase 3β–Axin complex for degradation but, upon activation of the Wnt pathway or in tumor cells with an APC mutation, β-catenin can activate TCF4-mediated transcription in the nucleus (green circle). Because amounts of ZO-1 are very low, the Y-box transcription factor ZONAB is free to interact with Cdk4 and to repress transcription of target genes. (B) Formation of AJs [yellow boxes in (B) and (C)]. When E-cadherin–mediated contacts are engaged, several proteins, including β-catenin and PI3K, are recruited to AJs. Activation of PI3K in turn leads to activation of the Akt and p38 MAPK protein kinases, which results in the transcription of genes associated with cell differentiation, such as the intestine-specific homeobox gene cdx2. Amounts of ZO-1 increase gradually, but ZONAB still represses transcription. (C) Formation of TJs (red box). The polarized state is stabilized by the segregation of apical and basolateral membranes. The expression of proteins associated with polarization, such as the brush border enzymes intestinal alkaline phosphatase (IAP) and sucrase-isomaltase (Suc), is increased. ZO-1 is targeted to the cytoplasmic plaque of TJs and sequesters ZONAB away from the nucleus, and β-catenin is similarly located at AJs. Some PI3K is located in the brush border membrane, presumably within lipid rafts, but can be mobilized to participate in junction dissociation by external signals from gastrins or other growth factors. The surprising finding that PI3K can promote not only cell polarization (A→B→C), but also TJ and AJ dissociation (C→B→A), is probably due to differentiation state–dependent changes in expression or subcellular compartmentation of the isoforms of downstream targets such as Akt.

There are several explanations for these discrepancies. First and perhaps most important, the Akt isoform(s) that mediates the downstream effects of PI3K may depend on the differentiation state of the cell (16). Such selectivity could be achieved by changes in expression or by isoform-specific subcellular compartmentation. For example, localization of Akt2 within lipid rafts has been described in differentiated Caco-2 cells and in HT-29 colon cancer cells, as well as in the enterocyte brush border membrane (17). Second, either or both the p85 regulatory subunit and the p110 catalytic subunit of PI3K may be differently expressed or compartmentalized within differentiating epithelial cells. The subcellular localization of the p110 catalytic subunit of PI3K during cell differentiation has not been considered (16, 17). Third, other downstream targets of PI3K such as Rac1 may also be involved in the modulation of intestinal cell polarity, as recently described in breast cancer cells (13). The dependence on cellular context is exemplified by the observation that the activity of Akt2 in enterocytes in vivo is decreased by the presence of PTEN within the same rafts (17), whereas in differentiated Caco2/15 cells, Akt2 activity is low and is PI3K-independent (16). Further studies should focus on specifying the precise compartmentation of AJ-associated proteins within epithelial cells along the intestinal crypt–villus axis, in order to fully understand the fine detail of PI3K involvement in the regulation of AJs and epithelial cell differentiation in the gut.

TJ-Coupled Signaling

Multiple signaling proteins have been identified as components of TJs. However, the kinetics of their interaction with integral TJ proteins and the exact nature of their role in junctional biogenesis, stability, or dissociation have not always been clearly established (8, 18). Recently, the TJ field has been spurred on by fascinating results that identify the signaling role of several cytoplasmic plaque proteins. Under various conditions, these proteins can shuttle between TJs and the cytoplasm or nucleus. Once in the nucleus, TJ proteins, like the zonula occludens proteins ZO-1 and ZO-2, both of which are members of the membrane-associated guanylate kinase family that includes the Drosophila tumor suppressor Disc-large, can regulate gene expression through their interaction with transcriptional regulators such as ZONAB or the scaffold attachment factor–B (SAF-B) (2, 19). Other TJ proteins such as symplekin alter mRNA stability in the nucleus and cytoplasm (20, 21). Of particular interest has been the identification of a physical interaction between ZO-1 and the Y-box transcriptional repressor ZONAB, which sequesters ZONAB away from the nucleus and hence reverses the inhibition of transcription by ZONAB (Fig. 1). ZONAB itself modulates the activity of the cyclin-dependent kinase 4 (Cdk4)–cyclin D1 complex, regulates the ErbB2 gene promoter, and controls cell proliferation (2). This regulatory process is very finely tuned, because maximal cell density is controlled even after epithelial cells have reached confluence and expressed functional TJs.

Although most of the currently available data indicate that signaling to and from AJs and TJs is highly selective, a degree of complexity is added by the existence of direct cross-talk between both types of junction. Thus, experimental disruption of TJs perturbs AJ stability and triggers the release of β-catenin from its E-cadherin binding site into the cytoplasm, activation by β-catenin of transcription factors of the T cell factor (TCF) family, and a subsequent increase in proliferation and induction of an epithelium-to-mesenchyme transition (22). These results were mostly obtained in the widely studied canine kidney cell line MDCK, but gastrointestinal AJ and TJ may well demonstrate a similar degree of interaction. Furthermore, transcriptional regulation of AJ and TJ protein expression seems to be correlated. For example, expression of both E-cadherin and claudin is repressed by the zinc-finger transcription factor Snail (23), and the claudin 1 and claudin 2 promoters are activated by β-catenin–TCF-4 (8).

Extracellular Regulation of AJs and TJs

AJ and TJ signaling pathways in the gut are also tightly controlled by various exogenous signals. For example, the complex interplay between Wnt and Hedgehog signaling is essential to achieve organized cell migration and differentiation in the embryonic and adult intestine (24, 25). A major advance in this area has been the targeted deletion of the gene encoding the adenomatous polyposis coli (APC) tumor-suppressor protein, loss of which increases both nuclear β-catenin and cell proliferation and abrogates migration along the crypt-villus axis (26). However, only one paper to date has provided direct evidence that connects the loss of APC function with modification of cell-cell adhesion (27). Multiple cytokines and growth factors, including interferon-γ and hepatocyte growth factor, have also been shown to induce temporary modulation of AJs and TJs (7). In addition, neurotransmitters are likely to affect these processes, as evidenced by the major functional alterations of the intestinal mucosa after traumatic brain injury in rats (28).

Recently, several groups have highlighted the relevance of peptides encoded by the gastrin gene in the regulation of AJ- and TJ-coupled signaling in the gut. Thus, amidated gastrin activates pathways critical for the migration of gastric parietal cells (29) and for the differentiation of pancreatic cells (30) in vivo. Dissociation of AJs upon gastrin stimulation of intestinal IEC-6 cells involved the sequential activation of the protein kinase JAK2 and of PI3K (31). Furthermore, partially processed gastrin gene products such as glycine-extended gastrin (Ggly) or progastrin induce dissociation and migration of intestinal epithelial cells (6, 32). In the case of progastrin, the signaling pathways leading to AJ and TJ dissociation activate PI3K and the Src tyrosine kinase, respectively (6). In view of the promotional effects of cytoplasmic β-catenin and ZO-1 on cell proliferation (2, 33), the dissociation of these proteins from their respective junctions is likely to facilitate the growth-promoting effect of progastrin on intestinal cells in vitro, and on tumors in vivo. Indeed, a high proportion of patients with colorectal carcinoma overexpress progastrin or Ggly in both tumor and plasma (34), and hypergastrinemia promotes adenoma progression in ApcMin mice, which are heterozygous for a mutation in the APC gene (35). The recent demonstration that the gastrin gene is itself a target of β-catenin–TCF-4 signaling (36, 37) emphasizes the relevance of progastrin as a promoter of colorectal cancer development, mediated in part by disruption of AJs and TJs.

Relevance to Human Disease

In view of the critical role of AJs in epithelial cell differentiation, it is hardly surprising that various genetic alterations of molecules involved in AJ structure and signaling have been implicated in gastrointestinal cancer. Thus, E-cadherin mutations are frequently detected in familial gastric cancers (38), whereas mutations in any of several genes modulating the level of cytoplasmic β-catenin seem to participate in the onset of colorectal cancer (39). A frequent target is the APC gene, mutation of which results in multiple intestinal polyps. The development of polyps is also a consequence of mutation of the human PTEN gene in Cowden disease, presumably because Akt is no longer maintained in an inactive state (17). In contrast, the relatively benign hamartomatous polyps observed in Peutz-Jeughers syndrome result from mutation of the serine-threonine kinase LKB1, activation of which normally recruits several AJ and TJ proteins to the actin ring (5).

TJ proteins are the target for several bacterial toxins and viruses (8). For example, translocation of the CagA virulence factor of Helicobacter pylori into cells permits association with the TJ proteins ZO-1 and JAM-1 and results in the ectopic assembly of TJs at sites of bacterial attachment (40). In contrast, the Clostridium perfringens enterotoxin binds to claudin 3 and claudin 4 and causes their dissociation from TJs and concomitant degradation (41). The VP8 surface protein of rotavirus, which is a major cause of infant diarrhea, can inhibit the development of TJs and reduce transepithelial resistance (42). Partial cytoplasmic delocalization of TJ proteins may also occur when the integrity of the epithelium is compromised; for example, when the mucosa suffers physical damage or when amounts of inflammatory cytokines increase locally (7). Disruption of TJ protein expression or localization (or both) is also correlated with tumor differentation and poor prognosis of gastric and colon cancer (43, 44). Partial delocalization of TJ proteins would be expected to induce leakage via the paracellular route, which in turn would permit luminal growth factors such as epidermal growth factor (EGF) to access their receptors located on the basolateral membrane and promote cell proliferation (45). This process could provide an explanation for the nonsteroidal anti-inflammatory drug–sensitive increase in EGF receptor signaling activity documented in ApcMin mice (46). Finally, more drastic TJ disruption would lead to a loss of cell polarization and consequent disorganization of cell architecture and signaling.


Cell-cell junctions are central to the complex information network that maintains tissue organization. They receive multiple signals that guide them though the processes of establishing contact with neighboring cells and installing the appropriate scaffolding in order to initiate cytoskeletal connections. They also behave as sensors that send back to the nucleus signals about both their tissue location and their immediate environment. In order to better understand the constant stream of new data generated in this exciting field, it will be critical to consider the compartmentation of the specific signaling molecules within each cell type. The development of new tools such as real-time fluorescence monitoring at the cellular level is therefore an essential step toward understanding AJ and TJ signaling in the gut. Although this area provides a difficult challenge for the future, the results should have an important impact on the development of new strategies for treating inflammation and cancer.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
View Abstract

Navigate This Article