An Emerging Function of E-Cadherin: Gene Transcriptional Regulation
Recently, reports have described other functions of E-cadherin beyond its role in mediating AJs. In particular, E-cadherin was found to regulate gene expression (Du et al. 2014). Sasaki et al. first reported that a loss of E-cadherin in invasive breast cancer cells resulted in an increase of Bcl-2 expression, contributing to chemotherapy resistance in tumor cells. Wang et al. found an inverse correlation between E-cadherin and EGFR expression in tissue specimens of head and neck sarcoma, further demonstrating that EGFR signaling activation inhibited the expression of E-cadherin, and knockdown of E-cadherin resulted in the elevation of EGFR transcription. Strumane et al. found an inverse correlation between E-cadherin and human Nanos1 expression in various cell lines and showed that the re-expression of E-cadherin in a human breast cancer cell line decreased hNanos1 expression. The first systematic study on the role of E-cadherin in global gene transcription was performed by Onder et al. in 2008 (Onder et al. 2008). They inhibited the function of E-cadherin through either siRNA-mediated knockdown or expression of a truncated form of E-cadherin (cytoplasmic region) in human mammary epithelial cells and found that the expression of many genes had been altered significantly as a result. Indeed, they showed that it was the loss of E-cadherin itself and not the loss of cell-cell contacts or the subsequent activation of β-catenin that contributed mostly to this alteration. This was the first time to show that a loss of E-cadherin resulted in the transcriptional elevation of Twist and ZEB1, two well-known transcriptional repressors of E-cadherin. Recently, Francesca et al. compared global transcript expression in E-cadherin null (E-cad-/-) embryonic stem (ES) cells and E-cadherin wide-type ES cells, showing that E-cadherin depletion led to the altered expression of 2,265 genes. Notably, they did not detect an elevation of β-catenin activity after E-cadherin depletion in their model. However, they did observe an enhancement of FGF signaling activity due to the increase of FGF5 transcription in E-cad-/- ES cells.
These results implied that E-cadherin was a novel regulator of gene transcription, even though the molecular mechanisms involved had not yet been fully detailed. Analysis of the available E-cadherin data, particularly in regards to its regulation of cell signaling pathways, may help to shed some light on this issue (Du et al. 2014).
E-Cadherin Regulates the Wnt/β-Catenin Signaling Pathway
Over the past few decades, some reporters have surmised that a loss of E-cadherin may elevate the activity of β-catenin, having evaluated its activity through luciferase reporter systems and determination of TCF/β-catenin target gene expression. However, it was commonly accepted that E-cadherin loss alone was not sufficient to activate β-catenin signaling, requiring instead the presence of other effectors, such as Wnt and FGFR signaling activity (Kuphal and Behrens 2006). With the combined use of time-lapse microscopy and image analysis, the cadherin-bound pool of β-catenin was shown to accumulate at the perinuclear endocytic recycling compartment (ERC) upon AJ dissociation and then translocate into the nucleus upon Wnt signaling pathway activation, which suggests that the ERC may be a site of residence for β-catenin following its liberation from the membrane cadherin complex and prior to entering the nucleus. In most cases, restoration or overexpression of E-cadherin inhibited β-catenin activity by sequestering cytoplasmic β-catenin (Orsulic et al. 1999). Notably, overexpression of the cytoplasmic region of E-cadherin was sufficient to achieve this response (Orsulic et al. 1999). Additionally, a recent research demonstrated that the aberrant nuclear localization of E-cadherin is a potent inhibitor of Wnt/β-catenin signaling pathway in cancer stem cells (Su et al. 2015). These results imply that E-cadherin may be a negative regulator of the Wnt/β-catenin signaling pathway. However, Sara et al. recently showed that the ability to bind E-cadherin was necessary for β-catenin’s transcriptional activity, and E-cadherin was required for augmented activation of the Wnt/β-catenin pathway in vivo, which suggests that E-cadherin could be a positive regulator of the Wnt/β-catenin pathway in certain models. In yet another study, the Wnt/β-catenin pathway seemed to regulate E-cadherin expression. The E-cadherin gene promoter contains TCF/β-catenin binding sites, and Wnt signaling activation represses the expression of E-cadherin in a TCF/β-catenin-dependent manner, which suggests that a feedback circuit may exist between E-cadherin and Wnt/β-catenin signaling (Jamora et al. 2003).
E-Cadherin Regulates RTK Signaling Pathway
Growth factors, such as EGF, FGF, TGF, and HGF, are known to promote cell proliferation and prevent apoptosis through binding to their receptors in the cell membrane, inducing dimerization of the receptors and concomitant activation of the intracellular tyrosine kinase domains. The activated receptor tyrosine kinases can then phosphorylate their substrates, resulting in the activation of multiple downstream signaling pathways, including MAPK, PI3K/AKT, and STAT signaling pathways.
As early as 1994, the E-cadherin-β-catenin complex was shown to interact with Erb-B2, a member of the EGF receptor family of receptor tyrosine kinases, in the cancer cell membrane (Ochiai et al. 1994). Soon after, several other groups demonstrated that E-cadherin could bind the EGF receptor (EGFR). In one study, interaction of the extracellular domain of E-cadherin with EGFR was required for the transient activation of EGFR signaling in mammary cells (Fedor-Chaiken et al. 2003). In another, Pece et al. showed that E-cadherin interacted with EGFR and activated EGFR-mediated MAPK signaling in a ligand-independent manner. More recently, the extracellular domain of soluble E-cadherin was shown to interact with EGFR and activate EGFR-mediated PI3K/AKT and ERK1/2 signaling in breast cancer cells and squamous cell carcinoma (Du et al. 2014).
In addition to EGFR, E-cadherin is also shown to interact with FGFR. In MCF-7 breast cancer cells, treatment with FGF induced the endocytosis of E-cadherin and FGFR. The interaction of E-cadherin with FGFR was required for the nuclear translocation of FGFR and subsequent activation of FGF-induced MAPK signaling. Overexpression of E-cadherin blocked the endocytosis of both molecules, the nuclear translocation of FGFR, and the activation of FGFR-mediated MAPK signaling (Bryant et al. 2005). In Ewing tumor cells, under anchorage-independent growth conditions, E-cadherin was upregulated and correlated with the formation of multicellular spheroids and the suppression of anoikis. The mechanism study showed that E-cadherin activated the Erb-B4 receptor tyrosine kinase coupled with the activation of PI3K/AKT signaling.
E-cadherin may also directly regulate PI3K activity. Indeed, PI3K was recruited to the site of cell-cell contact by the ligation of homophilic E-cadherin, resulting in the activation of PI3K signaling (Kovacs et al. 2002). Recently, the p85 subunit of PI3K was shown to be directly targeted by the E-cadherin complex and activated in ovarian cancer cells (De Santis et al. 2009).
E-Cadherin Regulates the GTPase Signaling Pathway
GTPases are molecular switches that control multiple processes in eukaryotic cells while cycling between a GTP-bound active state and a GDP-bound inactive state. GTPases consist of five major groups: Rho, Ras, Rab, Ran, and Arf. Rho GTPases are primarily known for regulating the actin cytoskeleton and cell polarity (De Santis et al. 2009). Recently, Rho GTPases were also found to regulate gene transcription. For example, in mid-G1 phase of the cell cycle, Rho GTPases inhibited the expression of cyclin/CDK inhibitor P21 but induced the expression of cyclin D1 through promoting the sustained activation of MAPK signaling (Olson et al. 1998). E-cadherin was found to regulate the activity of Rho, Rac, and Cdc42, the three most well-characterized members of the Rho GTPases, implying that E-cadherin may regulate transcription through regulating GTPase signaling activity.
Rac activation was observed as an early-immediate response of E-cadherin adhesion formation (Nakagawa et al. 2001), and PI3K seemed to play a critical role in the E-cadherin-mediated activation of Rac (Coniglio et al. 2001). Furthermore, inhibition of PIK3 activity prevented the E-cadherin-mediated activation of Rac (Nakagawa et al. 2001). As mentioned earlier, E-cadherin recruits and activates PI3K at sites of cell-cell contact (Kovacs et al. 2002). Guanine nucleotide exchange factors (GEFs), which activate Rho GTPases by promoting the exchange of GDP for GTP, were found to recognize activated PI3K through their pleckstrin homology (PH) domains (Hawkins et al. 1995). These results suggest that PI3K may be an upstream activator of Rac in E-cadherin-mediated cell signaling.
E-Cadherin Regulates the NF-κB Signaling Pathway
E-Cadherin Regulates the HIPPO Signaling Pathway
It has been revealed that E-cadherin also regulates the HIPPO signaling pathway by inhibiting the activity of Yap, one of the key effectors of this pathway. The status of Hippo pathway usually varied according to cell density. At the edge of a colony of cells grown in culture, cells display a greater proportion of nuclear Yap but less mature E-cadherin-mediated cell-cell contacts, with higher proliferative and migratory ability, when compared with cells toward the center of the colony. This observation implies that Hippo signaling activity is sensitive to the E-cadherin-mediated junction status of cells. A recent research also found that maintenance of quiescence required E-cadherin extracellular engagement, which facilitates the Yap1 nuclear exclusion (Benham-Pyle et al. 2015). Kim et al. found that Yap is predominantly located in the nuclear in E-cadherin-deficient cancer cell line, whereas restored the expression of E-cadherin, Yap moved to the cytoplasm and lost its transcriptional activity. Gumbiner et al. recently also reported that hemophilic ligation of E-cadherins between cells in trans prevent Yap’s nuclear location. The mechanisms about E-cadherin mediating its effects upon Yap is not clear; even some researchers suggested that it may be associated with α-catenin (McCrea et al. 2015).
E-Cadherin Proteolytic Fragments as Nuclear Signaling
E-cadherin can be proteolysis by numerous enzymes. The ectodomain of E-cadherin can be cleaved by ADAMs and MMPs and then binds with EGFP and activates MAPK signaling pathway. This has been discussed in the “E-Cadherin Regulates the RTK Signaling Pathways” section. The E-cadherin cytodomaincleavage occurs via presenilins or caspases, and the cytodomain fragment mediates nuclear signaling (McCrea et al. 2015). For example, the cytodomain fragment of E-cadherin generated by presenilin enters the nucleus to bind CBP (Creb-binding protein) and degrade CBP. This in turn attenuates the transactivation ability of CREB (cAMP response element-binding protein), a transcription factor and partner of CBP. Additionally, a fragment of E-cadherin arising as a consequence of presenilin 1 activity is reported to enhance β-catenin nuclear activity (McCrea et al. 2015). Interestingly, p120, one of the components of cadherin complex, appears to enhance the nuclear entry of an E-cadherin cytodomain fragment, with the fragment enhancing p120-facilitated relief of Kaiso-mediated target-gene repression (McCrea et al. 2015).
Mediation of Cross-Talk Between Signaling Pathways by E-Cadherin
The role of p120 in the regulation of Rho GTPase was extensively reviewed previously. In summary, p120 was found to directly interact with and regulate Rho GTPase and indirectly modulate Rho activity through interacting with and regulating Rho GEFs. Additionally, p120 was able to promote or suppress the activation of Rho GTPases in different situations. For example, p120 dominantly inhibited Rho activity but consistently activated Rac and Cdc42 (Anastasiadis and Reynolds 2001). Furthermore, GTPase regulation could occur either at the site of E-cadherin-mediated cell-cell contacts or in the cytoplasm. When associated with E-cadherin, p120 modulated local GTPases and affected cytoskeletal structures; once dissociated from E-cadherin, p120 could diffuse into the cytoplasm and activate GTPases, thereby affecting the expression of genes involved in a variety of cellular processes, including cell-cycle regulation (Anastasiadis and Reynolds 2001).
Alternatively, liberated p120 could enter the nucleus to regulate gene transcription directly. Like β-catenin, p120 has an Arm-repeat domain, and proteins with this domain may have dual localization at cell-cell junctions and in the nucleus. In the nucleus, p120 was reported to interact with the zinc finger transcriptional repressor Glis2 and induce its C-terminal cleavage, though the mechanism of action for this process is unknown (Hosking et al. 2007). Nuclear p120 was also shown to interact with the BTB/POZ transcriptional repressor Kaiso, inhibiting Kaiso transcriptional activity (Kim et al. 2004). Kaiso is an inhibitor of the Wnt signaling pathway, directly inhibiting the transcription of Wnt11 (Kim et al. 2004) and the expression of Wnt signaling targets, such as c-Myc, cyclin D1, and matrilysin (MMP-7), through competitive binding of TCF/LEF with β-catenin (Spring et al. 2005). The inhibitory role of Kaiso on Wnt signaling and Wnt signaling targets can be attenuated by p120, suggesting that p120 may play a positive role in activation of the Wnt signaling pathway (Spring et al. 2005). Interestingly, it was demonstrated that Wnt signaling activation stabilized p120, which in turn promoted Kaiso sequestration or removal from the nucleus and elevated Wnt signaling (Park et al. 2006). These data therefore suggest the existence of a possible positive feedback circuit between p120 and Wnt signaling activity.
Over the past decade, E-cadherin has been reported to function as a gene transcriptional regulator, but further studies are needed to more clearly define its likely numerous modes of action in this process. The well-known associations of E-cadherin-mediated AJs with multiple signaling pathways leaves little room for doubt that altering E-cadherin would also affect gene transcription through impacting cell signaling. This hypothesis provides a model that signals originating from E-cadherin relay ultimately to the nucleus by molecules that play a central role in the associated signaling pathways. Given the complexity of interaction between E-cadherin-mediated AJs and cell signaling and the existence of cross-talk among different pathways, however, the discrete contribution of each pathway to E-cadherin-mediated gene transcriptional modulation is currently difficult to ascertain. The combined knockdown of E-cadherin and relevant pathway-related molecules may be a useful strategy for tackling this issue.
Nuclear translocation of E-cadherin has also been observed in numerous cancer cell lines and tissues (Du et al. 2014), which raises the following questions: What is the function of nuclear E-cadherin, and does it directly regulate gene transcription? Ferber et al. reported that a cleaved cytoplasmic domain of E-cadherin could enter the nucleus, form a complex with DNA via p120, and regulate gene transcription (Ferber et al. 2008). This data implies the possibility that E-cadherin itself may function in the nucleus as a novel transcriptional regulator, which is definitely an interesting topic and deserving of further systematic study.
- Tong Z, Chakraborty S, Sung B, Koolwal P, Kaur S, Aggarwal BB, et al. Epidermal growth factor down-regulates the expression of neutrophil gelatinase-associated lipocalin (NGAL) through E-cadherin in pancreatic cancer cells. Cancer. 2011;117:2408–18. doi: 10.1002/cncr.25803.PubMedCrossRefGoogle Scholar