Regulation of cell-cell adhesion
Content
Introduction to the regulation of cell-cell adhesion[Edit]
The assembly and disassembly of cell-cell junctions, such as adherens junctions (AJ), are regulated by a vast array of factors functioning in a highly complex network of molecular switches and mechanical cues. Although adhesion complexes are often transient, effective regulation of their assembly and disassembly is crucial for regular cell function. This is evidenced by the fact that destabilization of AJ dynamics is correlated to the onset of cancer cell metastasis where increased cell motility and invasion is commonly observed [1]. Indeed, one recent study attributed the oncosuppressive properties of the RhoGAP protein DLC1 (deleted in liver cancer 1) to a role in the stabilization of AJs [2].
With AJs essentially being a functional extension of the cytoskeleton, their formation and stability is heavily dependent on that of the actin filament network. Regulation of the adhesion complex is maintained by a combination of actin cytoskeleton dynamics, the influence of AJ scaffolding proteins, and the effect of posttranslational modification of AJ components. Each protein involved in the regulation of the adhesion complex must first be recruited to the complex, and in most cases, must be activated in order to carry out their function. Thus, regulation of the adhesion complex is a highly complicated system involving several independent molecular pathways [3]. Efforts to dissect each regulatory mechanism are therefore ongoing and many interactions and consequences are still unknown.
Actin Cytoskeleton Dynamics and AJ Regulation[Edit]
Given the function and stability of clustered cell-cell adhesions is tightly coupled to the formation and function of the actin cytoskeleton, it is not surprising that many of the components that regulate actin filament dynamics, also localize to adhesions sites, and have been implicated in their regulation. This has been highlighted for the Arp2/3 complex, formins and various nucleation promotion factors (NPFs), where, depending on the experimental conditions and cell types used, their presence was required for stability and function of the adhesion complex [4, 5, 6, 7, 3].
For example, down-regulation of Dia1 in human MCF7 cells by shRNA reportedly disrupted both E-cadherin localization to adhesion sites, and AJ integrity [4]. Likewise, it was found that knockdown of N-WASP diminished the integrity of zonula adherens [5]. This was manifested in the fragmentation of the apical actin ring structure and was attributed to the role of the WIP-family protein WIRE; which is normally recruited to the ZA by N-WASP where it interacts with E-cadherin, and regulates actin cytoskeleton formation at the junction via direct interactions with F-actin. Knocking-down, WIRE, as well as N-WASP, induced the same level of disruption to ZA integrity [5]. Similar results were found in MDCK cells where actin filament assembly at adhesion sites was diminished by inhibition of the Arp2/3 complex. In this case, the necessity for nucleators, and their associated NPFs, to localize to adhesion sites was attributed to a finding that indicated pre-existing filaments are unable to link to mature adhesion sites, and instead, new actin filaments must be polymerized [6]. As discussed in recent reviews, [7, 3] the requirement for nucleators and regulators of the actin cytoskeleton to localize to sites of cell-cell adhesion may reflect a system of synergistic regulation where the function of the adhesion complex contributes to the regulation of the actin cytoskeleton, and vice versa in a form of feedback [7, 3].
For example, down-regulation of Dia1 in human MCF7 cells by shRNA reportedly disrupted both E-cadherin localization to adhesion sites, and AJ integrity [4]. Likewise, it was found that knockdown of N-WASP diminished the integrity of zonula adherens [5]. This was manifested in the fragmentation of the apical actin ring structure and was attributed to the role of the WIP-family protein WIRE; which is normally recruited to the ZA by N-WASP where it interacts with E-cadherin, and regulates actin cytoskeleton formation at the junction via direct interactions with F-actin. Knocking-down, WIRE, as well as N-WASP, induced the same level of disruption to ZA integrity [5]. Similar results were found in MDCK cells where actin filament assembly at adhesion sites was diminished by inhibition of the Arp2/3 complex. In this case, the necessity for nucleators, and their associated NPFs, to localize to adhesion sites was attributed to a finding that indicated pre-existing filaments are unable to link to mature adhesion sites, and instead, new actin filaments must be polymerized [6]. As discussed in recent reviews, [7, 3] the requirement for nucleators and regulators of the actin cytoskeleton to localize to sites of cell-cell adhesion may reflect a system of synergistic regulation where the function of the adhesion complex contributes to the regulation of the actin cytoskeleton, and vice versa in a form of feedback [7, 3].
Scaffolding (adaptor) proteins as regulators of adhesion stability[Edit]
Scaffolding proteins are known to regulate the stability of adhesion complexes through their physical interactions with the core components. In some cases however they may also recruit additional regulatory proteins to the adhesion site. These proteins are often identified as binding partners to one or more of the key components of AJs such as E-cadherin, α-catenin and p120-catenin.
In some cases scaffolding proteins, such as cortactin (link to GT), may recruit regulators of actin filament dynamics like the Arp2/3 complex or N-WASP, which will regulate the function of the adhesion complex as described above. In other cases they will stabilize the core components and regulate their function.
One adaptor protein that is crucial to the stabilization of the AJ complex is p120-catenin. This protein prevents the endocytosis of the classical cadherins in a function that was recently attributed to its binding at the juxtamembrane position of the cytoplasmic tail of cadherin [8]. Three VE-cadherin residues in particular, (DEE 646-648), which are well conserved across the classical cadherins and lie in the p120-catenin binding site, were identified as producing an endocytic signal that is blocked upon binding of p120-catenin. Mutations within this motif not only impeded p120-catenin binding but also prevented endocytosis of cadherin [8]. Similar findings have shown that binding of the small GTPase RAP1 to afadin enhances p120-catenin binding to E-cadherin, preventing endocytosis of E-cadherin and stabilizing the E-cadherein–nectin complex [9].
Additionally, an earlier study highlighted that p120-catenin may influence signaling pathways that regulate cadherin complex stability and clustering. Binding of p120-catenin to cadherin was proposed to negatively regulate Rho GTPase signaling by preventing the interaction between p120-catenin and RhoA which would otherwise result in the inhibition of RhoA activity, dissolution of actin stress fibres and the mislocalization of ezrin to the cytoplasm [10]. The latter effect is directly correlated to AJ formation and stability, with Ezrin having been shown to regulate the transport of E-cadherin to the membrane in a process that involves Rac1 activation [11]. p120-catenin’s role in regulating the Rho GTPases was again highlighted in a more recent study that indicated ROCK1 and p190A RhoGAP, which are recruited to the adhesion complex by Rho A, interact transiently with p120-catenin. These interactions may control a variety of processes including cadherin clustering and stability [12].
In some cases scaffolding proteins, such as cortactin (link to GT), may recruit regulators of actin filament dynamics like the Arp2/3 complex or N-WASP, which will regulate the function of the adhesion complex as described above. In other cases they will stabilize the core components and regulate their function.
One adaptor protein that is crucial to the stabilization of the AJ complex is p120-catenin. This protein prevents the endocytosis of the classical cadherins in a function that was recently attributed to its binding at the juxtamembrane position of the cytoplasmic tail of cadherin [8]. Three VE-cadherin residues in particular, (DEE 646-648), which are well conserved across the classical cadherins and lie in the p120-catenin binding site, were identified as producing an endocytic signal that is blocked upon binding of p120-catenin. Mutations within this motif not only impeded p120-catenin binding but also prevented endocytosis of cadherin [8]. Similar findings have shown that binding of the small GTPase RAP1 to afadin enhances p120-catenin binding to E-cadherin, preventing endocytosis of E-cadherin and stabilizing the E-cadherein–nectin complex [9].
Additionally, an earlier study highlighted that p120-catenin may influence signaling pathways that regulate cadherin complex stability and clustering. Binding of p120-catenin to cadherin was proposed to negatively regulate Rho GTPase signaling by preventing the interaction between p120-catenin and RhoA which would otherwise result in the inhibition of RhoA activity, dissolution of actin stress fibres and the mislocalization of ezrin to the cytoplasm [10]. The latter effect is directly correlated to AJ formation and stability, with Ezrin having been shown to regulate the transport of E-cadherin to the membrane in a process that involves Rac1 activation [11]. p120-catenin’s role in regulating the Rho GTPases was again highlighted in a more recent study that indicated ROCK1 and p190A RhoGAP, which are recruited to the adhesion complex by Rho A, interact transiently with p120-catenin. These interactions may control a variety of processes including cadherin clustering and stability [12].
Posttranslational modifications to adhesion complex components[Edit]
Posttranslational modification of adhesion complex components is one of the most complicated aspects in the regulation of cell-cell adhesions. It involves an array of protein enzymes and despite a growing body of knowledge on the topic, many of the specific mechanisms, from recruitment of regulators, to the consequence each modification has on adhesion stability and function has yet to be described.
Phosphorylation events are the most prominent posttranslational modification to play a role in the regulation of cell-cell adhesions. These are mediated by phosphatases and kinases, of which at least 34 are known to localize to AJs alone [13]. In many cases the method by which phosphatases/kinases are recruited to the adhesion site has yet to be established, however, in most cases direct binding with either core components or scaffolding proteins is likely to be involved. Indeed this has been established in some specific cases; for example CSK and VE-PTP bind directly to VE-cadherin [14] , MET and PTPRF to β-catenin [15, 16] and ROCK1 to p120-catenin [12].
Addition or removal of a phosphate molecule may not only alter the targets binding affinity to other components, but may activate or deactivate enzymatic properties to produce either positive (stabilizing), or negative (destabilizing) consequences. Furthermore, the activity of kinases/phosphatases is often correlated to the action of the cell-cell adhesion complex, with mechanical cues from neighboring cells triggering the enzymatic activity of the protein. This was observed, for example, when the H-Ras/Raf/MEK/ERK signaling cascade was induced following the attachment of invasive breast cancer cells to endothelial cells in vitro. Here, tyrosine phosphorylation of VE-cadherin was induced, and subsequently, the dissociation of β-catenin from the complex observed [17].
Similarly, attachment of neutrophils and lymphocytes to endothelial cells activated by TNF-α lead to the rapid dissociation of the tyrosine phosphatase VE-PTP from VE-cadherin [14]. VE-PTP had previously been found to bind to, and reduce the phosphorylation of VE-cadherin, but not β-catenin [18], which effectively stabilized the adhesion complex. Leukocyte mediated dissociation of VE-PTP from VE-cadherin was associated with increased phosphorylation of VE-cadherin as well as β-catenin and plakoglobin. Importantly, it was the plakoglobin that was found to be essential in VE-PTP stabilization VE-cadherin [14] .
As mentioned, leukocyte mediated dissociation of VE-PTP was only observed in cells that had been activated by the inflammatory cytokine, TNF-α, highlighting that regulators themselves may be activated only under specific cellular conditions. This was again reflected when the tyrosine phosphatase SHP2 was found to dissociate from β-catenin as a result of thrombin stimulation [19]. Later studies noted that dissociation of SHP2 promoted catenin phosphorylation [20], confirming it to be a substrate of SHP2[21], and revealed, through silencing or inhibition of SHP2 its role in promoting the mobility of E-cadherin, and hence, the integrity of adhesion complexes [21]).
Similarly, posttranslational modification of scaffolding proteins such as p120-catenin may be equally influential to the stability of the adhesion complex. This was demonstrated when it was found that PKCa mediated phosphorylation of p120-catenin at Serine 879 which initiated dissociation of p120-catenin from the AJ, and subsequently promoted its disassembly [22] .
Phosphorylation events are the most prominent posttranslational modification to play a role in the regulation of cell-cell adhesions. These are mediated by phosphatases and kinases, of which at least 34 are known to localize to AJs alone [13]. In many cases the method by which phosphatases/kinases are recruited to the adhesion site has yet to be established, however, in most cases direct binding with either core components or scaffolding proteins is likely to be involved. Indeed this has been established in some specific cases; for example CSK and VE-PTP bind directly to VE-cadherin [14] , MET and PTPRF to β-catenin [15, 16] and ROCK1 to p120-catenin [12].
Addition or removal of a phosphate molecule may not only alter the targets binding affinity to other components, but may activate or deactivate enzymatic properties to produce either positive (stabilizing), or negative (destabilizing) consequences. Furthermore, the activity of kinases/phosphatases is often correlated to the action of the cell-cell adhesion complex, with mechanical cues from neighboring cells triggering the enzymatic activity of the protein. This was observed, for example, when the H-Ras/Raf/MEK/ERK signaling cascade was induced following the attachment of invasive breast cancer cells to endothelial cells in vitro. Here, tyrosine phosphorylation of VE-cadherin was induced, and subsequently, the dissociation of β-catenin from the complex observed [17].
Similarly, attachment of neutrophils and lymphocytes to endothelial cells activated by TNF-α lead to the rapid dissociation of the tyrosine phosphatase VE-PTP from VE-cadherin [14]. VE-PTP had previously been found to bind to, and reduce the phosphorylation of VE-cadherin, but not β-catenin [18], which effectively stabilized the adhesion complex. Leukocyte mediated dissociation of VE-PTP from VE-cadherin was associated with increased phosphorylation of VE-cadherin as well as β-catenin and plakoglobin. Importantly, it was the plakoglobin that was found to be essential in VE-PTP stabilization VE-cadherin [14] .
As mentioned, leukocyte mediated dissociation of VE-PTP was only observed in cells that had been activated by the inflammatory cytokine, TNF-α, highlighting that regulators themselves may be activated only under specific cellular conditions. This was again reflected when the tyrosine phosphatase SHP2 was found to dissociate from β-catenin as a result of thrombin stimulation [19]. Later studies noted that dissociation of SHP2 promoted catenin phosphorylation [20], confirming it to be a substrate of SHP2[21], and revealed, through silencing or inhibition of SHP2 its role in promoting the mobility of E-cadherin, and hence, the integrity of adhesion complexes [21]).
Similarly, posttranslational modification of scaffolding proteins such as p120-catenin may be equally influential to the stability of the adhesion complex. This was demonstrated when it was found that PKCa mediated phosphorylation of p120-catenin at Serine 879 which initiated dissociation of p120-catenin from the AJ, and subsequently promoted its disassembly [22] .
Alternative Mechanisms of Adherens Junction Regulation[Edit]
Alternative mechanisms also exist to regulate AJ formation and function. For example, the membrane trafficking system plays a significant role by facilitating the transport of E-cadherin-β-catenin complexes from the endoplasmic reticulum (ER) where they form. This system also enables complex sorting in the Trans-Golgi Network, delivers the cadherin-β-catenin complex to the plasma membrane and controls recycling of adhesion components upon adhesion disassembly and cadherin internalization. The influence of these processes on cell-cell adhesion regulation has recently been reviewed [23] and is illustrated in several recent studies.
One such example used an epithelial cell line expressing mutant reggies and prion protein – which contribute to the recruitment of E-cadherin, one study proposed that AJ function is influenced by the reggie dependent regulation of EGFR internalization, which in turn controls the macropinocytosis and recycling of E-cadherin [24]. In another example the small GTPase Rab35 was shown to regulate the recycling of internalized cadherin back to the plasma membrane. Inhibition of this protein also prevented PIP5KIy from accumulating at the adhesion site which impeded PI(4,5)P2 production [25]. PI(4,5)P2 is required for complex formation and stability [26].
In summary, regulation of cell-cell junctions is controlled in a highly complex manner, being influenced by numerous cellular processes and components. The primary modes of regulation appear to be those discussed; namely posttranslational modification of core components and scaffolding proteins, as well as regulation via actin cytoskeleton dynamics. Importantly other major cellular processes such as membrane trafficking also contribute to the regulation of cell-cell adhesions and may influence any one of the steps in their formation and function.
One such example used an epithelial cell line expressing mutant reggies and prion protein – which contribute to the recruitment of E-cadherin, one study proposed that AJ function is influenced by the reggie dependent regulation of EGFR internalization, which in turn controls the macropinocytosis and recycling of E-cadherin [24]. In another example the small GTPase Rab35 was shown to regulate the recycling of internalized cadherin back to the plasma membrane. Inhibition of this protein also prevented PIP5KIy from accumulating at the adhesion site which impeded PI(4,5)P2 production [25]. PI(4,5)P2 is required for complex formation and stability [26].
In summary, regulation of cell-cell junctions is controlled in a highly complex manner, being influenced by numerous cellular processes and components. The primary modes of regulation appear to be those discussed; namely posttranslational modification of core components and scaffolding proteins, as well as regulation via actin cytoskeleton dynamics. Importantly other major cellular processes such as membrane trafficking also contribute to the regulation of cell-cell adhesions and may influence any one of the steps in their formation and function.