Limits...
Wingless signalling alters the levels, subcellular distribution and dynamics of Armadillo and E-cadherin in third instar larval wing imaginal discs.

Somorjai IM, Martinez-Arias A - PLoS ONE (2008)

Bottom Line: Surprisingly, DeltaNArm(1-155) caused displacement of both Armadillo and E-Cadherin, results supported by our novel method of quantification.Taken together, our results provide in vivo evidence for a complex non-linear relationship between Armadillo levels, subcellular distribution and Wingless signalling.Moreover, this study highlights the importance of Armadillo in regulating the subcellular distribution of E-Cadherin.

View Article: PubMed Central - PubMed

Affiliation: Department of Genetics, University of Cambridge, Cambridge, United Kingdom. ildiko.somorjai@obs-banyuls.fr

ABSTRACT

Background: Armadillo, the Drosophila orthologue of vertebrate ss-catenin, plays a dual role as the key effector of Wingless/Wnt1 signalling, and as a bridge between E-Cadherin and the actin cytoskeleton. In the absence of ligand, Armadillo is phosphorylated and targeted to the proteasome. Upon binding of Wg to its receptors, the "degradation complex" is inhibited; Armadillo is stabilised and enters the nucleus to transcribe targets.

Methodology/principal findings: Although the relationship between signalling and adhesion has been extensively studied, few in vivo data exist concerning how the "transcriptional" and "adhesive" pools of Armadillo are regulated to orchestrate development. We have therefore addressed how the subcellular distribution of Armadillo and its association with E-Cadherin change in larval wing imaginal discs, under wild type conditions and upon signalling. Using confocal microscopy, we show that Armadillo and E-Cadherin are spatio-temporally regulated during development, and that a punctate species becomes concentrated in a subapical compartment in response to Wingless. In order to further dissect this phenomenon, we overexpressed Armadillo mutants exhibiting different levels of activity and stability, but retaining E-Cadherin binding. Arm(S10) displaces endogenous Armadillo from the AJ and the basolateral membrane, while leaving E-Cadherin relatively undisturbed. Surprisingly, DeltaNArm(1-155) caused displacement of both Armadillo and E-Cadherin, results supported by our novel method of quantification. However, only membrane-targeted Myr-DeltaNArm(1-155) produced comparable nuclear accumulation of Armadillo and signalling to Arm(S10). These experiments also highlighted a row of cells at the A/P boundary depleted of E-Cadherin at the AJ, but containing actin.

Conclusions/significance: Taken together, our results provide in vivo evidence for a complex non-linear relationship between Armadillo levels, subcellular distribution and Wingless signalling. Moreover, this study highlights the importance of Armadillo in regulating the subcellular distribution of E-Cadherin.

Show MeSH

Related in: MedlinePlus

Theoretical model of the localization and interactions of wild type and mutant forms of Armadillo and E-Cadherin.In a wild type cell in its basal state (no signalling, left), Armadillo (blue dots) and E-Cadherin (green dots) are targeted to the membrane, possibly via the exocyst complex [43]. From there, Armadillo/E-Cadherin cycle to the Adherens Junction (AJ, box), or are hypothesised to remain in the cytoplasm as a complex free from degradation. There is little Armadillo in the nucleus (grey oval). During Wingless signalling (right), Armadillo accumulates in subapical puncta and can enter the nucleus (blue oval) to activate Wg targets. We hypothesise that Armadillo/E-Cadherin are released from the AJ (black arrows); alternatively the rate of cycling of Armadillo/E-Cadherin may also be increased. In our proposed model, either in its basal state or during signalling, Armadillo/E-Cadherin may be able to bind α-catenin (yellow hexagon). The activated N-terminal deletion mutants (centre, red boxes and dots) represent a range of effects in the spectrum between the basal state and Wingless signalling, depending at least in part on overexpression levels (bottom, red bar). ΔNArm1–128 (centre, top left) most closely resembles the basal state, with little Armadillo in the nucleus or signalling. At highest levels of overexpression, ΔNArm1–128 excludes Armadillo from the basolateral membrane and AJ, and is unlikely to efficiently bind α-catenin (grey hexagon; [76]). ArmS10 (centre, top right) excludes Armadillo from the AJ, as well as basolateral membranes at lowest expression levels. However, ArmS10 has no effect on E-Cadherin levels and should have the capacity to bind α-catenin [30], [76], thus ensuring appropriate adhesive function, as well as signalling (red and blue nucleus). In this case, we propose that Armadillo/E-Cadherin at the basolateral membrane may not be sufficiently stable to enter the AJ (grey arrow). Several sources indicate that Armadillo/E-Cadherin is first targeted to the basolateral membrane, and from there to the AJ [42], [78]. Myr-ΔNArm1–155, while unable to bind α-catenin [76], nevertheless ensures adhesive function through Armadillo/E-Cadherin at the AJ. These would be able to efficiently cycle between the subcellular compartments and enter the nucleus (blue oval). Additionally, filopodia extend into the environment from the basal side of cells expressing Myr-ΔNArm1–155. ΔNArm1–155, in contrast, impedes proper Armadillo/E-Cadherin function at the basolateral membrane and AJ through a reduction in their levels (fewer blue dots, pale green). It is unable to bind α-catenin, but enters the nucleus where it may interact with the transcriptional machinery. Note that all representations of cytoplasmic protein localisation are inferred from changes in the other more easily quantifiable compartments (e.g. “cellular” = nucleus+basolateral membrane+cytoplasm).
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2483348&req=5

pone-0002893-g015: Theoretical model of the localization and interactions of wild type and mutant forms of Armadillo and E-Cadherin.In a wild type cell in its basal state (no signalling, left), Armadillo (blue dots) and E-Cadherin (green dots) are targeted to the membrane, possibly via the exocyst complex [43]. From there, Armadillo/E-Cadherin cycle to the Adherens Junction (AJ, box), or are hypothesised to remain in the cytoplasm as a complex free from degradation. There is little Armadillo in the nucleus (grey oval). During Wingless signalling (right), Armadillo accumulates in subapical puncta and can enter the nucleus (blue oval) to activate Wg targets. We hypothesise that Armadillo/E-Cadherin are released from the AJ (black arrows); alternatively the rate of cycling of Armadillo/E-Cadherin may also be increased. In our proposed model, either in its basal state or during signalling, Armadillo/E-Cadherin may be able to bind α-catenin (yellow hexagon). The activated N-terminal deletion mutants (centre, red boxes and dots) represent a range of effects in the spectrum between the basal state and Wingless signalling, depending at least in part on overexpression levels (bottom, red bar). ΔNArm1–128 (centre, top left) most closely resembles the basal state, with little Armadillo in the nucleus or signalling. At highest levels of overexpression, ΔNArm1–128 excludes Armadillo from the basolateral membrane and AJ, and is unlikely to efficiently bind α-catenin (grey hexagon; [76]). ArmS10 (centre, top right) excludes Armadillo from the AJ, as well as basolateral membranes at lowest expression levels. However, ArmS10 has no effect on E-Cadherin levels and should have the capacity to bind α-catenin [30], [76], thus ensuring appropriate adhesive function, as well as signalling (red and blue nucleus). In this case, we propose that Armadillo/E-Cadherin at the basolateral membrane may not be sufficiently stable to enter the AJ (grey arrow). Several sources indicate that Armadillo/E-Cadherin is first targeted to the basolateral membrane, and from there to the AJ [42], [78]. Myr-ΔNArm1–155, while unable to bind α-catenin [76], nevertheless ensures adhesive function through Armadillo/E-Cadherin at the AJ. These would be able to efficiently cycle between the subcellular compartments and enter the nucleus (blue oval). Additionally, filopodia extend into the environment from the basal side of cells expressing Myr-ΔNArm1–155. ΔNArm1–155, in contrast, impedes proper Armadillo/E-Cadherin function at the basolateral membrane and AJ through a reduction in their levels (fewer blue dots, pale green). It is unable to bind α-catenin, but enters the nucleus where it may interact with the transcriptional machinery. Note that all representations of cytoplasmic protein localisation are inferred from changes in the other more easily quantifiable compartments (e.g. “cellular” = nucleus+basolateral membrane+cytoplasm).

Mentions: One of the possibilities raised by our study that might help to explain discrepancies in the signalling ability of our N-terminal deletion mutants is that some cell-surface protein is required for the proper shuttling of Armadillo/E-Cadherin among the subcellular compartments. A putative candidate might be α-catenin, whose passive role in linking the E-Cadherin/ß-catenin complex to the actin cytoskeleton has recently been put into question [49], [50]. Indirect evidence for a role of α-catenin in mediating the signalling or stability of Armadillo arises from several observations. First, ArmS10, which is functional in both signalling and adhesion, is likely the only activated form able to efficiently bind α-catenin, with perhaps the exception of ArmΔN1–128 [30], [75], [76]. Second, although Myr-ΔNArm1–155 lacks α-catenin binding sites, its targeting to the membrane may overcome this limitation by bringing it into proximity with cell surface molecules. Additionally, junctional function is not compromised as endogenous Armadillo and E-Cadherin occupy the AJ. ΔNArm1–155, on the other hand, can neither bind α-catenin nor is able to effectively recruit proteins near the cell surface, and furthermore impedes normal Armadillo/E-Cadherin function at the AJ. Finally, our data suggest that actin accumulates at the A/P boundary in cells depleted of E-Cadherin, most notably upon overexpression of ΔNArm1–155. We therefore see a correlation between signalling of mutants, either alone or through endogenous Armadillo, and their ability to interact with α-catenin. It is interesting in this context to note that α-catenin may inhibit CK1 phosphorylation-dependent degradation of β-catenin, and that the region encompassing the junction of the N-terminus and first Armadillo repeat are critical for this regulation [77].Thus a possible model of Armadillo/E-Cadherin movement upon signalling can be derived from our mutant data, from which we infer a regulatory input from α-catenin (Figure 15). In support of our model, plasma membrane recruitment of a punctate, signalling-competent species of ß-catenin appears to be an important step in the transcriptional activation of Wnt signalling both in vitro and in vivo, and occurs independently of E-Cadherin [81]. Additional studies will be required to confirm the validity of this model, as well as any putative role of α-catenin in mediating these interactions.


Wingless signalling alters the levels, subcellular distribution and dynamics of Armadillo and E-cadherin in third instar larval wing imaginal discs.

Somorjai IM, Martinez-Arias A - PLoS ONE (2008)

Theoretical model of the localization and interactions of wild type and mutant forms of Armadillo and E-Cadherin.In a wild type cell in its basal state (no signalling, left), Armadillo (blue dots) and E-Cadherin (green dots) are targeted to the membrane, possibly via the exocyst complex [43]. From there, Armadillo/E-Cadherin cycle to the Adherens Junction (AJ, box), or are hypothesised to remain in the cytoplasm as a complex free from degradation. There is little Armadillo in the nucleus (grey oval). During Wingless signalling (right), Armadillo accumulates in subapical puncta and can enter the nucleus (blue oval) to activate Wg targets. We hypothesise that Armadillo/E-Cadherin are released from the AJ (black arrows); alternatively the rate of cycling of Armadillo/E-Cadherin may also be increased. In our proposed model, either in its basal state or during signalling, Armadillo/E-Cadherin may be able to bind α-catenin (yellow hexagon). The activated N-terminal deletion mutants (centre, red boxes and dots) represent a range of effects in the spectrum between the basal state and Wingless signalling, depending at least in part on overexpression levels (bottom, red bar). ΔNArm1–128 (centre, top left) most closely resembles the basal state, with little Armadillo in the nucleus or signalling. At highest levels of overexpression, ΔNArm1–128 excludes Armadillo from the basolateral membrane and AJ, and is unlikely to efficiently bind α-catenin (grey hexagon; [76]). ArmS10 (centre, top right) excludes Armadillo from the AJ, as well as basolateral membranes at lowest expression levels. However, ArmS10 has no effect on E-Cadherin levels and should have the capacity to bind α-catenin [30], [76], thus ensuring appropriate adhesive function, as well as signalling (red and blue nucleus). In this case, we propose that Armadillo/E-Cadherin at the basolateral membrane may not be sufficiently stable to enter the AJ (grey arrow). Several sources indicate that Armadillo/E-Cadherin is first targeted to the basolateral membrane, and from there to the AJ [42], [78]. Myr-ΔNArm1–155, while unable to bind α-catenin [76], nevertheless ensures adhesive function through Armadillo/E-Cadherin at the AJ. These would be able to efficiently cycle between the subcellular compartments and enter the nucleus (blue oval). Additionally, filopodia extend into the environment from the basal side of cells expressing Myr-ΔNArm1–155. ΔNArm1–155, in contrast, impedes proper Armadillo/E-Cadherin function at the basolateral membrane and AJ through a reduction in their levels (fewer blue dots, pale green). It is unable to bind α-catenin, but enters the nucleus where it may interact with the transcriptional machinery. Note that all representations of cytoplasmic protein localisation are inferred from changes in the other more easily quantifiable compartments (e.g. “cellular” = nucleus+basolateral membrane+cytoplasm).
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2483348&req=5

pone-0002893-g015: Theoretical model of the localization and interactions of wild type and mutant forms of Armadillo and E-Cadherin.In a wild type cell in its basal state (no signalling, left), Armadillo (blue dots) and E-Cadherin (green dots) are targeted to the membrane, possibly via the exocyst complex [43]. From there, Armadillo/E-Cadherin cycle to the Adherens Junction (AJ, box), or are hypothesised to remain in the cytoplasm as a complex free from degradation. There is little Armadillo in the nucleus (grey oval). During Wingless signalling (right), Armadillo accumulates in subapical puncta and can enter the nucleus (blue oval) to activate Wg targets. We hypothesise that Armadillo/E-Cadherin are released from the AJ (black arrows); alternatively the rate of cycling of Armadillo/E-Cadherin may also be increased. In our proposed model, either in its basal state or during signalling, Armadillo/E-Cadherin may be able to bind α-catenin (yellow hexagon). The activated N-terminal deletion mutants (centre, red boxes and dots) represent a range of effects in the spectrum between the basal state and Wingless signalling, depending at least in part on overexpression levels (bottom, red bar). ΔNArm1–128 (centre, top left) most closely resembles the basal state, with little Armadillo in the nucleus or signalling. At highest levels of overexpression, ΔNArm1–128 excludes Armadillo from the basolateral membrane and AJ, and is unlikely to efficiently bind α-catenin (grey hexagon; [76]). ArmS10 (centre, top right) excludes Armadillo from the AJ, as well as basolateral membranes at lowest expression levels. However, ArmS10 has no effect on E-Cadherin levels and should have the capacity to bind α-catenin [30], [76], thus ensuring appropriate adhesive function, as well as signalling (red and blue nucleus). In this case, we propose that Armadillo/E-Cadherin at the basolateral membrane may not be sufficiently stable to enter the AJ (grey arrow). Several sources indicate that Armadillo/E-Cadherin is first targeted to the basolateral membrane, and from there to the AJ [42], [78]. Myr-ΔNArm1–155, while unable to bind α-catenin [76], nevertheless ensures adhesive function through Armadillo/E-Cadherin at the AJ. These would be able to efficiently cycle between the subcellular compartments and enter the nucleus (blue oval). Additionally, filopodia extend into the environment from the basal side of cells expressing Myr-ΔNArm1–155. ΔNArm1–155, in contrast, impedes proper Armadillo/E-Cadherin function at the basolateral membrane and AJ through a reduction in their levels (fewer blue dots, pale green). It is unable to bind α-catenin, but enters the nucleus where it may interact with the transcriptional machinery. Note that all representations of cytoplasmic protein localisation are inferred from changes in the other more easily quantifiable compartments (e.g. “cellular” = nucleus+basolateral membrane+cytoplasm).
Mentions: One of the possibilities raised by our study that might help to explain discrepancies in the signalling ability of our N-terminal deletion mutants is that some cell-surface protein is required for the proper shuttling of Armadillo/E-Cadherin among the subcellular compartments. A putative candidate might be α-catenin, whose passive role in linking the E-Cadherin/ß-catenin complex to the actin cytoskeleton has recently been put into question [49], [50]. Indirect evidence for a role of α-catenin in mediating the signalling or stability of Armadillo arises from several observations. First, ArmS10, which is functional in both signalling and adhesion, is likely the only activated form able to efficiently bind α-catenin, with perhaps the exception of ArmΔN1–128 [30], [75], [76]. Second, although Myr-ΔNArm1–155 lacks α-catenin binding sites, its targeting to the membrane may overcome this limitation by bringing it into proximity with cell surface molecules. Additionally, junctional function is not compromised as endogenous Armadillo and E-Cadherin occupy the AJ. ΔNArm1–155, on the other hand, can neither bind α-catenin nor is able to effectively recruit proteins near the cell surface, and furthermore impedes normal Armadillo/E-Cadherin function at the AJ. Finally, our data suggest that actin accumulates at the A/P boundary in cells depleted of E-Cadherin, most notably upon overexpression of ΔNArm1–155. We therefore see a correlation between signalling of mutants, either alone or through endogenous Armadillo, and their ability to interact with α-catenin. It is interesting in this context to note that α-catenin may inhibit CK1 phosphorylation-dependent degradation of β-catenin, and that the region encompassing the junction of the N-terminus and first Armadillo repeat are critical for this regulation [77].Thus a possible model of Armadillo/E-Cadherin movement upon signalling can be derived from our mutant data, from which we infer a regulatory input from α-catenin (Figure 15). In support of our model, plasma membrane recruitment of a punctate, signalling-competent species of ß-catenin appears to be an important step in the transcriptional activation of Wnt signalling both in vitro and in vivo, and occurs independently of E-Cadherin [81]. Additional studies will be required to confirm the validity of this model, as well as any putative role of α-catenin in mediating these interactions.

Bottom Line: Surprisingly, DeltaNArm(1-155) caused displacement of both Armadillo and E-Cadherin, results supported by our novel method of quantification.Taken together, our results provide in vivo evidence for a complex non-linear relationship between Armadillo levels, subcellular distribution and Wingless signalling.Moreover, this study highlights the importance of Armadillo in regulating the subcellular distribution of E-Cadherin.

View Article: PubMed Central - PubMed

Affiliation: Department of Genetics, University of Cambridge, Cambridge, United Kingdom. ildiko.somorjai@obs-banyuls.fr

ABSTRACT

Background: Armadillo, the Drosophila orthologue of vertebrate ss-catenin, plays a dual role as the key effector of Wingless/Wnt1 signalling, and as a bridge between E-Cadherin and the actin cytoskeleton. In the absence of ligand, Armadillo is phosphorylated and targeted to the proteasome. Upon binding of Wg to its receptors, the "degradation complex" is inhibited; Armadillo is stabilised and enters the nucleus to transcribe targets.

Methodology/principal findings: Although the relationship between signalling and adhesion has been extensively studied, few in vivo data exist concerning how the "transcriptional" and "adhesive" pools of Armadillo are regulated to orchestrate development. We have therefore addressed how the subcellular distribution of Armadillo and its association with E-Cadherin change in larval wing imaginal discs, under wild type conditions and upon signalling. Using confocal microscopy, we show that Armadillo and E-Cadherin are spatio-temporally regulated during development, and that a punctate species becomes concentrated in a subapical compartment in response to Wingless. In order to further dissect this phenomenon, we overexpressed Armadillo mutants exhibiting different levels of activity and stability, but retaining E-Cadherin binding. Arm(S10) displaces endogenous Armadillo from the AJ and the basolateral membrane, while leaving E-Cadherin relatively undisturbed. Surprisingly, DeltaNArm(1-155) caused displacement of both Armadillo and E-Cadherin, results supported by our novel method of quantification. However, only membrane-targeted Myr-DeltaNArm(1-155) produced comparable nuclear accumulation of Armadillo and signalling to Arm(S10). These experiments also highlighted a row of cells at the A/P boundary depleted of E-Cadherin at the AJ, but containing actin.

Conclusions/significance: Taken together, our results provide in vivo evidence for a complex non-linear relationship between Armadillo levels, subcellular distribution and Wingless signalling. Moreover, this study highlights the importance of Armadillo in regulating the subcellular distribution of E-Cadherin.

Show MeSH
Related in: MedlinePlus