Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Nov 27.
Published in final edited form as: Science. 2020 Nov 27;370(6520):eaba5528. doi: 10.1126/science.aba5528

Abl and Canoe/Afadin mediate mechanotransduction at tricellular junctions

Huapeng H Yu 1, Jennifer A Zallen 1,*
PMCID: PMC8559527  NIHMSID: NIHMS1736713  PMID: 33243859

Abstract

Epithelial structure is generated by the dynamic reorganization of cells in response to mechanical forces. Adherens junctions transmit forces between cells, but how cells sense and respond to these forces in vivo is not well understood. We identify a mechanotransduction pathway involving the Abl tyrosine kinase and Canoe/Afadin that stabilizes cell adhesion under tension at tricellular junctions in the Drosophila embryo. Canoe is recruited to tricellular junctions in response to actomyosin contractility, and this mechanosensitivity requires Abl-dependent phosphorylation of a conserved tyrosine in the Canoe actin-binding domain. Preventing Canoe tyrosine phosphorylation destabilizes tricellular adhesion, and anchoring Canoe at tricellular junctions independently of mechanical inputs aberrantly stabilizes adhesion, arresting cell rearrangement. These results identify a force-responsive mechanism that stabilizes tricellular adhesion under tension during epithelial remodeling.

A universal property of cells is the ability to detect and respond to mechanical stimuli, such as the stiffness of the extracellular environment, shear and osmotic stress from fluid flow, and forces generated by neighboring cells (1,2). Mechanical forces influence cell fate, division, adhesion, and behavior (3,4), and defects in cellular mechanotransduction pathways are implicated in many diseases, including deafness, atherosclerosis, and cancer (5,6). In epithelial tissues, adherens junctions are critical sites of force transmission between cells (7–9) and undergo dynamic assembly and disassembly under tension during cell rearrangement (3,4,7). Interactions between several components of adherens junction complexes are stabilized by mechanical forces in vitro and in cultured cells, and have been proposed to reinforce cell adhesion under tension (10–15). However, how adherens junctions sense and respond to mechanical forces in vivo, and how these processes influence dynamic cell behaviors in tissues, are not well understood. Mechanical forces are converted into biochemical changes in cells through a variety of mechanisms, including changes in protein conformation, localization, activity, interactions, and post-translational modifications (1,2,9). In particular, tyrosine phosphorylation is enhanced at cell-cell and cell-matrix adhesions, two essential force-bearing structures within cells (16,17). Tyrosine phosphorylation can be modulated by mechanical force in vitro, with evidence for force sensing at the level of tyrosine kinases (18–22) and substrates (22–25). However, how tyrosine kinase signaling is regulated by mechanical forces in vivo, and the effects of this regulation on cell adhesion and epithelial remodeling, are not well understood.

Tyrosine phosphorylation at tricellular junctions is tension-sensitive

To investigate whether tyrosine kinase signaling is regulated by mechanical forces in vivo, we analyzed the distribution of tyrosine-phosphorylated proteins in the Drosophila embryo. During convergent extension, epithelial cells undergo directional cell rearrangements in response to planar polarized actomyosin contractility (26–30). We observed a striking tissue-wide pattern of phosphotyrosine during this process, with the highest levels of tyrosine phosphorylation at tricellular adherens junctions where three cells meet (Fig. 1A, S1A). Tricellular adherens junctions are specialized structures that contain dedicated adhesion molecules (31–34) and are predicted to be sites of increased tension in epithelia (35–37). Phosphotyrosine was more strongly enriched at tricellular junctions than the core adherens junction proteins E-cadherin and α-catenin or the membrane marker Spider–GFP (Fig. 1B,G, S1H,I, Table S1). The enrichment of phosphotyrosine was further increased at 4-way junctions that form as intermediates during cell rearrangement (Fig. S1B,C). These results indicate that tyrosine phosphorylation is enhanced at sites where three or more cells meet.

Figure 1. Tension-sensitive localization of phosphotyrosine at tricellular junctions.

Figure 1.

Open in a new tab

(A) Phosphotyrosine (p-Tyr) and E-cadherin in the Drosophila embryo. Close-up, tricellular junction (TCJ) and bicellular junctions (BCJs). (B) TCJ ratio (ratio of the mean TCJ intensity to the mean intensity of the three connected BCJs). (C) Alexa-Fluor 594-conjugated p-Tyr antibody was injected into syncytial embryos and retained in cells after cellularization. (D,E) p-Tyr in living embryos expressing MRLC-GFP (myosin II regulatory light chain) (D) or endogenous Canoe-Venus (E). (F) p-Tyr and myosin-Venus (myosin II heavy chain) in embryos injected with water or Y-27632. Close-ups, single TCJs. (G) TCJ ratios of p-Tyr, Canoe-Venus, E-cadherin, and Spider–GFP. (H) p-Tyr TCJ ratio in embryos injected with water or Y-27632. (I,J) p-Tyr intensity correlates with MRLC-GFP (I) and Canoe-Venus (J) at TCJs. Dots, individual TCJs in a single embryo. (K) Canoe-Venus and p-Tyr intensity at a single TCJ over time (imaged every 20 s). (L) Correlation coefficient for the rates of change in Canoe-Venus and p-Tyr intensity at TCJs (mean±SEM between embryos). Canoe-Venus data were shifted by the times on the x-axis. Boxes, second and third quartiles; whiskers, 5th–95th percentile; horizontal line, median; +, mean. Intensity is a percentage of the maximum value. ** p<0.003, Welch’s t-test. Embryos are stage 7. Living embryos are shown in D, E, F (bottom panels), and I-L. Anterior left, ventral down. Bars, 10 μm. See Table S1 for details.

To investigate the dynamics of tyrosine kinase signaling at tricellular junctions, we developed a method to visualize tyrosine phosphorylation in living embryos. Fluorophore-conjugated monoclonal antibodies to phosphotyrosine were injected into syncytial embryos and retained in cells after cellularization, allowing direct visualization of tyrosine-phosphorylated proteins (Fig. 1C–E, Movie S1). Phosphotyrosine levels at tricellular junctions were positively correlated with myosin II regulatory light chain (MRLC-GFP), an indicator of cortical contractility (Fig. 1D,I, S1E). Inhibiting cortical myosin activity by injecting the Rho-kinase inhibitor Y-27632 (27,30) rapidly depleted phosphotyrosine levels at tricellular junctions within 5 min of injection, accompanied by an increase in phosphotyrosine levels at bicellular junctions (Fig. 1F,H, S1F,G). The acute release of tension by Y-27632 did not affect the overall level of tyrosine phosphorylation in the embryo (Fig. S1D), adherens junction integrity (Fig. S1H–K), or the integrity of tricellular junctions marked by the tricellular adherens junction protein Sidekick (31–34) (Fig. S1L–O), which is sensitive to actomyosin perturbations over longer timescales (32). These results demonstrate that increased tyrosine phosphorylation at tricellular junctions requires cytoskeletal tension.

Canoe/Afadin is a tyrosine kinase substrate at tricellular junctions

To investigate the molecular basis of tension-sensitive tyrosine phosphorylation at tricellular junctions, we screened >100 Venus-tagged proteins (31) to identify factors that colocalize with phosphotyrosine at tricellular junctions. This screen identified Canoe/Afadin, a conserved actin-binding protein that links cell-surface receptors to the actomyosin cytoskeleton in Drosophila and mammals (38–43). A functional Canoe-Venus fusion expressed from the endogenous locus (31) was enriched at tricellular junctions in wild-type embryos (Fig. 1G), consistent with previous findings (40,42). Canoe-Venus localization at tricellular junctions was highly dynamic, accumulating and dissipating on a timescale of seconds (Movie S2), and these changes were strongly correlated with changes in phosphotyrosine (Fig. 1J–L, S2A–C). Canoe-Venus was tyrosine phosphorylated in embryonic lysates, although Canoe protein accounted for only a minority of tyrosine-phosphorylated proteins in the embryo (Fig. S2D,I,J). Moreover, phosphotyrosine levels at tricellular junctions were significantly reduced in canoeR2 maternal mutants (referred to as canoe mutants) and these levels were restored by reintroducing wild-type N-terminally tagged Canoe protein (Fig. S2E–H). These results indicate that Canoe, or a Canoe-dependent protein, is a substrate for tyrosine kinase signaling at tricellular junctions.

Canoe/Afadin is a critical regulator of epithelial morphogenesis and animal viability in Drosophila and mammals (38–41). However, although mammalian Afadin can influence tricellular adhesion under certain conditions in culture (42), an in vivo requirement for Canoe/Afadin in regulating tricellular adhesion has not been directly demonstrated. To address this question, we analyzed cell adhesion in Canoe-deficient embryos using E-cadherin and Sidekick-Venus to visualize bicellular and tricellular junctions, respectively. In wild-type embryos, cells were in close contact at tricellular junctions and Sidekick-Venus localization in all three cells appeared as a single spot by confocal microscopy, indicating strong adhesion (Fig. 2A). By contrast, gaps in E-cadherin and Sidekick-Venus localization at tricellular junctions were detected in canoe mutants and canoe knockdown (KD) embryos that express an shRNA targeting canoe (Fig. 2A,I,K, S3A–C). Defects in E-cadherin localization at tricellular junctions were first detected in stage 7 in Canoe-deficient embryos, when E-cadherin localization at bicellular junctions was largely intact. By contrast, defects in E-cadherin localization at bicellular junctions were not prevalent until stage 8 (Fig. S3D–F). These results demonstrate an early, specific requirement for Canoe in regulating cell adhesion at tricellular junctions.

Figure 2. Tyrosine phosphorylation is required for Canoe localization and function at tricellular junctions.

Figure 2.

Open in a new tab

(A) E-cadherin and Sidekick-Venus in wild-type (WT) and canoe KD embryos. Close-ups, single edges (rotated 30°). Arrowheads, TCJs and 4-way junctions. (B) Endogenous Canoe-Venus in WT and Abl KD embryos. (C) Canoe-Venus TCJ ratios. (D) N-terminally tagged Venus-Canoe variants expressed using the Gal4/UAS system in a canoe mutant background. (E) Venus-Canoe TCJ ratios. (F) Canoe domain structure. Tyrosines, vertical lines. (G) Venus-Canoe-NC variants in which the indicated tyrosines were mutated to unphosphorylatable phenylalanine (F) residues were coexpressed with HA-Abl in S2R+ cells, immunoprecipitated (IP) with anti-GFP, and immunoblotted (IB) with anti-p-Tyr or anti-GFP. Protein size in kDa. The tyrosines mutated in 11YF and 4YF are indicated in (F). (H) Stills from movies of canoe mutant embryos expressing full-length Venus-Canoe variants using the Gal4/UAS system. (I-K) Percentage of TCJs with gaps in E-cadherin signal in canoe KD (I), Abl KD (J), or canoe mutants expressing full-length Venus-Canoe variants (K) (mean±SEM between embryos). Cno WT, canoe mutant expressing Venus-Canoe-WT; –, canoe mutant alone. * p<0.04, ** p≤0.001, Welch’s t-test. Embryos are stages 7–8. Living embryos are shown in D, E, and H. Anterior left, ventral down. Bars, 10 μm.

Abl-dependent phosphorylation of Canoe is required for tricellular adhesion

The correlation between Canoe and phosphotyrosine localization at tricellular junctions, and the requirement for Canoe in maintaining tricellular adhesion, raise the possibility that Canoe is part of a tyrosine kinase signaling pathway that stabilizes cell adhesion at tricellular junctions. Consistent with this idea, injection of the tyrosine kinase inhibitor bosutinib rapidly decreased Canoe localization at tricellular junctions (Fig. S4A–G) and disrupted tricellular adhesion (Fig. S4H–J). Abl, a nonreceptor tyrosine kinase that is inhibited by bosutinib in mammals (44), has conserved functions in epithelial remodeling and phosphorylates a large number of substrates involved in actin organization, cell adhesion dynamics, and cell matrix attachment (45–49). We therefore tested whether Abl is required for Canoe localization and function at tricellular junctions using a maternally expressed shRNA that effectively inhibits Abl expression and activity (48,49). We found that Abl KD embryos displayed a significantly reduced enrichment of Canoe and phosphotyrosine at tricellular junctions (Fig. 2B,C, S5A–E), as well as defects in tricellular adhesion that were similar to, but not quite as severe as, the defects in Canoe-deficient embryos (Fig. 2I–K). These results demonstrate that Abl is required for Canoe localization and cell adhesion at tricellular junctions.

As Abl is a tyrosine kinase, we hypothesized that Abl could regulate Canoe localization and function by promoting Canoe tyrosine phosphorylation. In support of this model, Abl enhanced Canoe tyrosine phosphorylation when coexpressed with Canoe in Drosophila S2R+ cells (Fig. S6A,B). Canoe contains 47 tyrosines, making identifying the Abl target sites in Canoe a daunting task. To expedite this process, we performed a structure/function analysis to identify the minimal sequences required for Canoe function. A truncated protein that contains the N- and C-terminal domains of Canoe (Canoe-NC), but lacks the central region including the PDZ domain that interacts with several cell-surface receptors (39), localized correctly to tricellular junctions (Fig. S7A,D) and fully rescued myosin localization in canoe mutants (Fig. S7E,F), a well-established readout for Canoe function (40,41). By contrast, further deletion of the N-terminal domain of Canoe that binds to the Rap1 GTPase (43,50), or the C-terminal domain that binds to F-actin (38,40), eliminated Canoe localization to tricellular junctions and rendered Canoe nonfunctional (Fig. S7A,D–F). Moreover, tyrosine phosphorylation of Canoe-NC, but not the full-length protein, was significantly reduced in Abl KD embryos (Fig. S6C). Therefore, even though other tyrosine kinases can phosphorylate the nonessential Canoe central domain, Abl is required for the phosphorylation of the minimal Canoe-NC protein that is fully functional during convergent extension. Further analysis of Canoe-NC variants lacking specific tyrosines demonstrated that Abl-dependent tyrosine phosphorylation of the Canoe-NC protein in S2R+ cells was significantly reduced by mutating the 11 C-terminal tyrosines (Canoe-NC11YF), the last 4 tyrosines (Canoe-NC4YF), or the most C-terminal tyrosine of Canoe (Canoe-NCY1987F), which is located in a predicted Abl target site (L/I/V-Y-x-x-P) (46) (Fig. 2F,G, S6A–C). These findings indicate that Abl promotes the phosphorylation of the C-terminal tyrosine of Canoe, which is conserved from worms to mammals.

To determine whether tyrosine phosphorylation is required for Canoe function, we tested whether full-length Canoe phosphovariants rescue the defects in canoe mutant embryos. Both wild-type Canoe and a full-length Canoe variant lacking the N-terminal tyrosines (Canoe9YF) localized correctly to tricellular junctions (Fig. S7A–D) and fully rescued tricellular adhesion in canoe mutants (Fig. 2K). By contrast, full-length Canoe proteins lacking the 11 C-terminal tyrosines (Canoe11YF) or the most C-terminal tyrosine (CanoeY1987F) were less enriched at tricellular junctions and failed to rescue tricellular adhesion and myosin localization in canoe mutants (Fig. 2D,E,H,K, S7A,D–F). CanoeY1987F embryos had more severe defects in tricellular adhesion than Abl KD embryos (Fig. 2J,K), likely due to residual Abl expression or the ability of other kinases to phosphorylate the Canoe C-terminal domain. All Canoe phosphovariants localized correctly to bicellular junctions (Fig. S7A,B), indicating that a general localization to adherens junctions, in the absence of a selective enrichment at tricellular junctions, is not sufficient for Canoe function. These results demonstrate that phosphorylation of the most C-terminal tyrosine of Canoe is necessary for its localization and function in vivo.

Canoe localization to tricellular junctions is mechanosensitive

As Canoe is dynamically recruited to and stabilizes adhesion at tricellular junctions, we hypothesized that Canoe could be part of a force-regulated mechanism that strengthens adhesion at tricellular junctions under tension. To test this model, we used three approaches to determine whether Canoe localization is regulated by force. First, we analyzed the relationship between Canoe and myosin localization at tricellular junctions in living embryos. Canoe-Venus localization to tricellular junctions was pulsatile (Fig. 3A,B), a characteristic feature of contractile actomyosin networks (51,52). Moreover, Canoe recruitment to tricellular junctions was strongly correlated with changes in myosin localization, with no delay between Canoe-Venus and MRLC-mCherry recruitment by cross-correlation analysis (Fig. 3C, Movie S3). Consistent with this relationship, the enrichment of Canoe at tricellular junctions was significantly reduced in stage 15 embryos that have less cortical myosin (Fig. S8A–C). By contrast, less Canoe protein was recruited with each myosin pulse in Abl KD embryos, even though myosin localization and dynamics at tricellular junctions occurred normally (Fig. S5H–N, Movie S4). Thus, Canoe recruitment correlates with bursts of myosin accumulation at tricellular junctions in wild-type embryos, and Canoe localization appears to be uncoupled from myosin in the absence of Abl activity.

Figure 3. Canoe is recruited to tricellular junctions by cytoskeletal tension.

Figure 3.

Open in a new tab

(A,B) Kymograph (A) and plot (B) of Canoe-Venus and MRLC-mCherry intensity at a single TCJ over time (imaged every 20 s). Intensity is a percentage of the maximum value. Bracket, time window in A. (C) Correlation coefficient for the rates of change in Canoe-Venus and MRLC-mCherry intensity at TCJs. Canoe-Venus data were shifted by the times on the x-axis (mean±SEM between embryos). (D) Canoe-Venus in embryos injected with water or Y-27632. (E) Canoe-Venus TCJ ratios. (F) Laser ablation schematic. (G) Stills from movies of Canoe-Venus before and after ablation of an edge connecting two TCJs (arrowheads) (rotated 30°). (H,I) Canoe-Venus intensity at TCJs (H) and BCJs (I) before and after ablation, normalized to the value at t=0 (mean±SD between ablations). * p=0.01, ** p≤0.0001, Welch’s t-test (E), one-way ANOVA (H,I). Embryos are stage 7. Living embryos expressing endogenous Canoe-Venus are shown in all panels. Anterior left, ventral down. Bars, 10 μm.

To further investigate this model, we tested whether Canoe localization to tricellular junctions requires actomyosin activity. Reducing actomyosin contractility by injecting a Rho-kinase inhibitor resulted in a rapid loss of Canoe-Venus from tricellular junctions within 5 min of injection and suppressed fluctuations in Canoe-Venus intensity (Fig. 3D,E, S8D–F, Movie S5). By contrast, Rho-kinase inhibitor did not decrease, and in fact slightly increased, Canoe-Venus levels at bicellular junctions (Fig. 3D, S8G,I,J). These results indicate that actomyosin contractility is not necessary for Canoe cortical localization, but is required for the selective accumulation of Canoe at tricellular junctions.

To directly test whether Canoe localization is regulated by mechanical force, we used laser ablation methods (30,35) to acutely release tension at individual tricellular junctions by severing one of the three connected bicellular junctions (Fig. 3F,G, Movie S6). To quantify the effects of releasing tension, we measured Canoe-Venus intensity at the tricellular junctions and bicellular junctions attached to the cut edge before and after ablation. To control for photobleaching, we analyzed Canoe-Venus intensity in cells located two or more cells away from the ablation site for comparison (Fig. S8H). Laser ablation led to a 23±9% (mean±SD) decrease in Canoe-Venus intensity at tricellular junctions within 70 s after ablation, whereas Canoe-Venus localization at bicellular junctions was unaffected (Fig. 3H,I). Taken together, these results demonstrate that Canoe-Venus enrichment at tricellular junctions requires mechanical forces generated by actomyosin contractility.

Dynamic Canoe localization to tricellular junctions is required for cell intercalation

The findings that Canoe is recruited to tricellular junctions by tyrosine phosphorylation and actomyosin contractility raise two questions. First, what is the physiological relevance of coupling Canoe recruitment with changes in myosin activity? Second, does tyrosine phosphorylation only regulate Canoe localization to tricellular junctions, or is it also required for Canoe function at these sites? To address these questions, we developed a method to target Canoe to tricellular junctions independently of endogenous mechanical inputs. This “vertex trap” method involved expressing VhhGFP4 (VHH), a genetically encoded nanobody with high affinity for GFP (53), fused to the C-terminus of the tricellular junction protein Sidekick (Sdk) (32–34). Sdk-VHH expression is predicted to recruit GFP- and Venus-tagged proteins to tricellular junctions through interactions with the localized nanobody (Fig. 4B). The localization of Canoe-Venus was not significantly altered by Sdk-VHH expression in wild-type embryos (Fig. 4A,E), and cell adhesion and myosin localization in these embryos occurred normally (Fig. S9A–E). By contrast, Sdk-VHH allowed Canoe-Venus to remain associated with tricellular junctions when myosin activity was inhibited by Y-27632 injection, a situation in which this enrichment is normally lost in wild-type embryos (Fig. 4A,F). Thus, the vertex trap system uncouples Canoe localization from cytoskeletal tension, providing an opportunity to investigate the effects of eliminating Canoe mechanosensitivity on cell rearrangement.

Figure 4. Anchoring Canoe at tricellular junctions impedes cell rearrangement.

Figure 4.

Open in a new tab

(A) Embryos expressing endogenous Canoe-Venus (wild-type Canoe, left) or Canoe-Venus trapped at tricellular junctions by Sdk-VHH (vertex-trapped Canoe, right). Top panels, uninjected embryos. Bottom panels, embryos injected with Y-27632. (B) Schematic of the vertex trap technique. The GFP nanobody (VHH) fused to the TCJ protein Sidekick (Sdk) recruits Venus-tagged proteins to tricellular junctions. (C,D) Endogenous Canoe-Venus intensity in the absence (C) or presence (D) of Sdk-VHH. Two TCJs (gray) converge to form a 4-way vertex (magenta), which resolves to form two new TCJs (imaged every 30 s). (E-G) Canoe-Venus TCJ ratio (E,F) and 4-way vertex duration (G) in the absence (WT Canoe) or presence (vertex trap) of Sdk-VHH. (H) 4-way vertex duration in embryos that overexpress vertex-trapped Venus-Canoe variants in a wild-type background. (I) Stills from movies of embryos expressing endogenous Canoe-Venus in the absence (wild-type Canoe) or presence (vertex-trapped Canoe) of Sdk-VHH. (J) Model. Actomyosin contractility and Abl-mediated tyrosine phosphorylation coordinate Canoe levels with mechanical tension at tricellular junctions, promoting the remodeling of cell adhesion during cell rearrangement. ** p<0.0001, Welch’s t-test. Embryos are stages 7–8. Living embryos are shown in all panels. Anterior left, ventral down. Bars, 10 μm.

To investigate the effects of uncoupling Canoe localization from mechanical inputs, we performed time-lapse imaging of embryos expressing vertex-trapped Canoe. During convergent extension, epithelial cells undergo cell rearrangements that are driven by planar polarized actomyosin contractility (26–30). In a stereotypical cell rearrangement during this process (27), tricellular junctions merge to form 4-way vertices that subsequently resolve to form two new tricellular junctions, a process that requires active remodeling of cell adhesion (Fig. 4C,I, S10A). Protein localization is highly dynamic during this process, with MRLC-mCherry and Canoe-Venus reproducibly dissociating from 4-way vertices just before resolution (Fig. 4C, S10B,C). In embryos that expressed vertex-trapped Canoe, cells were frequently arrested at the 4-way vertex stage, resulting in an increase in the average vertex duration time to 9.1±0.7 min compared with 5.1±0.4 min in wild-type embryos (mean±SEM) (Fig. 4D,G,I, Movie S7). Embryos that expressed vertex-trapped Canoe also often failed to complete resolution and displayed increased cell stretching, consistent with a reduction in cell rearrangement (Fig. S9F,G). Further increasing the levels of Canoe at tricellular junctions by coexpressing Sdk-VHH with overexpressed Venus-Canoe extended the vertex duration time to 17.8±0.9 min, more than three times longer than in wild-type embryos (Fig. 4H, Movie S8). By contrast, Abl KD embryos that recruited less Canoe to tricellular junctions accelerated vertex resolution, reducing the average vertex duration time to 2.6±0.2 min (Fig. S10D,E) (48). It was not possible to analyze vertex duration in Canoe-deficient embryos, due to widespread defects in tricellular adhesion. These results indicate that varying Canoe levels at tricellular junctions modulates the timing of cell rearrangement.

We next tested whether tyrosine phosphorylation is solely required for Canoe localization to tricellular junctions, or if it is also necessary for Canoe function at these sites. If tyrosine phosphorylation is only required to recruit Canoe to tricellular junctions, then trapping Canoe phosphovariants at these sites would be predicted to bypass the requirement for phosphorylation. Using the vertex trap assay, we found that Canoe11YF-Venus, which lacks the 11 C-terminal tyrosines of Canoe including Y1987, stabilized cell adhesion at 4-cell vertices and delayed cell rearrangement to the same extent as wild-type Canoe (Fig. 4H, Movie S9). By contrast, deleting the entire C-terminal region of Canoe (Canoe-ΔC) including the F-actin-binding domain eliminated the effects of vertex-trapped Canoe on cell adhesion (Fig. 4H, S9H,I). Thus, tyrosine phosphorylation is essential for Canoe localization to tricellular junctions, but is not required for Canoe to stabilize adhesion at these sites when Canoe is targeted to tricellular junctions through a heterologous mechanism. Together, these results demonstrate that tyrosine phosphorylation modulates Canoe levels at tricellular junctions, which influences the rate of cell rearrangement during epithelial remodeling.

Discussion

Here we show that Canoe and Abl function in a mechanotransduction pathway that regulates dynamic changes in cell adhesion at tricellular junctions during epithelial remodeling. Canoe localization to tricellular junctions is acutely disrupted by laser ablation, demonstrating that Canoe localization is rapidly modulated by mechanical perturbation. Canoe mechanosensitivity requires Abl-dependent phosphorylation of a conserved tyrosine (Y1987) in the Canoe actin-binding domain, and Canoe localization to tricellular junctions is required to stabilize tricellular adhesion. Conversely, constitutively anchoring Canoe at tricellular junctions arrests cell rearrangement, indicating that Canoe levels tune the rate of junctional remodeling (Fig. 4J). These results demonstrate that the mechanosensitivity of this critical junctional regulator is modulated by phosphorylation of a single tyrosine residue, and reveal an essential role of Canoe in coupling tricellular adhesion with mechanical forces during epithelial remodeling.

Mechanical forces trigger a cascade of molecular events in cells that translate biophysical signals into altered cellular behaviors. However, the mechanosensors that directly change conformation under tension in vivo are not well defined. Cell-surface receptors are well-positioned to detect forces generated by neighboring cells, but Canoe localization and function at tricellular junctions do not require the PDZ domain that mediates its interaction with known receptors (39). One possibility is that the Canoe protein itself could act as a mechanosensor. Canoe is predicted to be anchored to the membrane through interactions with Rap1 and the actin cytoskeleton (38–40,43,50). Thus, cytoskeletal tension could stretch the Canoe protein and expose tyrosine 1987 to phosphorylation by Abl. Alternatively, Abl or its upstream activators could be regulated by tension during this process, as the activity of Abl and other tyrosine kinases such as Src and FAK has been shown to be regulated by tension in vitro (18–22). In a third possibility, the regulation of Canoe by Abl could allow Canoe to detect force-induced changes in other molecules at tricellular junctions. As tyrosine 1987 is in the Canoe actin-binding domain, phosphorylation at this site could allow Canoe to recognize distinct actin structures at tricellular junctions, force-induced conformational changes in the cadherin-catenin complex (54,55), or other specialized features of tricellular junction composition or geometry (37). Once recruited to tricellular junctions, Canoe could stabilize adhesion at these sites by reinforcing the connection between adherens junctions and the actomyosin cytoskeleton (40–42).

Abl tyrosine kinases influence many structural changes that are driven by mechanical forces in tissues, including epithelial remodeling, tissue invagination, axon guidance, and cell migration (45,46). Abl has been shown to regulate a number of proteins that act at tension-bearing structures in cells in addition to Canoe/Afadin recruitment to tricellular junctions, including vinculin localization (20,22,25) and β-catenin recycling (48) at bicellular junctions and regulation of membrane curvature by BAR-domain proteins (21). Together, these results raise the possibility that Abl could transduce mechanical forces into a wide range of structural changes within cells. For example, during cell rearrangement, Abl could simultaneously stabilize tricellular adhesion by recruiting Canoe and destabilize bicellular adhesion by enhancing β-catenin turnover (48), which could allow bicellular junctions to complete contraction before tricellular junctions are remodeled. Tricellular junctions serve many important roles in epithelial development and homeostasis, including modulating cell rearrangement (32–34,52), orienting mitotic spindles (56), balancing stem cell proliferation and differentiation (57), and maintaining epithelial barrier function (58). An understanding of how mechanical inputs affect the conformation, localization, activity, and interactions of proteins at tricellular junctions will provide insight into how these structures sense and integrate mechanical forces in epithelial tissues.

Methods summary

The methods used in this study are described in the Supplementary Materials, including Drosophila stocks and genetics, cloning and transgenic lines, immunofluorescence, time-lapse imaging, laser ablation, drug and antibody injection, immunoprecipitation, immunoblotting, and mass spectrometry, and quantification and statistical analysis.

Supplementary Material

Supplementary_material
Movie S1
Download video file (449.2KB, mp4)
Movie S2
Download video file (803.3KB, mp4)
Movie S3
Download video file (622.3KB, mp4)
Movie S4
Download video file (413.7KB, mp4)
Movie S5
Download video file (665.2KB, mp4)
Movie S6
Download video file (178.7KB, mp4)
Movie S7
Download video file (1.5MB, mp4)
Movie S8
Download video file (1.4MB, mp4)
Movie S9
Download video file (1.4MB, mp4)

Acknowledgments:

We thank Masako Tamada and Chris Fincher for initially observing the effects of increased myosin activity on tyrosine phosphorylation, Ben Glick for msVenus, Matt Laurie for the UASp vector, Marissa Gredler for help with graphic illustration, Henrik Molina for help with mass spectrometry, and Eric Brooks, Marissa Gredler, Adam Paré, Will Razzell, Germán Sabio, Masako Tamada, and Richard Zallen for comments on the manuscript. Some stocks were obtained from the Bloomington Drosophila Stock Center (NIH P4OODO18537) and the Kyoto Stock Center.

Funding:

This work was funded by NIH/NIGMS R01 grant GM079340 to JAZ. JAZ is an investigator of the Howard Hughes Medical Institute.

Footnotes

Competing interests: The authors declare no competing interests.

Supplementary Materials:

Materials and Methods

Figures S1 – S10

Movies S1 – S9

References (59–69)

Data and materials availability:

All data are available in the manuscript or supplementary material.

References and Notes:

  • 1.Schwartz MA, Integrins and extracellular matrix in mechanotransduction. Cold Spring Harb. Perspect. Biol 2, a005066 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Heisenberg C-P, Bellaïche Y, Forces in tissue morphogenesis and patterning. Cell 153, 948–962 (2013). [DOI] [PubMed] [Google Scholar]
  • 3.Petridou NI, Spiró Z, Heisenberg C-P, Multiscale force sensing in development. Nat. Cell Biol 19, 581–588 (2017). [DOI] [PubMed] [Google Scholar]
  • 4.Pinheiro D, Bellaïche Y, Mechanical force-driven adherens junction remodeling and epithelial dynamics. Dev. Cell 47, 3–19 (2018). [DOI] [PubMed] [Google Scholar]
  • 5.Hahn C, Schwartz MA, Mechanotransduction in vascular physiology and atherogenesis. Nat. Rev. Mol. Cell Biol 10, 53–62 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Schramek D et al. , Direct in vivo RNAi screen unveils myosin IIa as a tumor suppressor of squamous cell carcinomas. Science 343, 309–313 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Takeichi M, Dynamic contacts: Rearranging adherens junctions to drive epithelial remodelling. Nat. Rev. Mol. Cell Biol 15, 397–410 (2014). [DOI] [PubMed] [Google Scholar]
  • 8.Nelson WJ, Weis WI, 25 years of tension over actin binding to the cadherin cell adhesion complex: The devil is in the details. Trends Cell Biol. 26, 471–473 (2016). [DOI] [PubMed] [Google Scholar]
  • 9.Yap AS, Duszyc K, Viasnoff V, Mechanosensing and mechanotransduction at cell-cell junctions. Cold Spring Harb. Perspect. Biol 10, a028761 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.le Duc Q et al. , Vinculin potentiates E-cadherin mechanosensing and is recruited to actin-anchored sites within adherens junctions in a myosin II-dependent manner. J. Cell Biol 189, 1107–1115 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Yonemura S, Wada Y, Watanabe T, Nagafuchi A, Shibata M, α-Catenin as a tension transducer that induces adherens junction development. Nat. Cell Biol 12, 533–542 (2010). [DOI] [PubMed] [Google Scholar]
  • 12.Buckley CD et al. , Cell adhesion. The minimal cadherin-catenin complex binds to actin filaments under force. Science 346, 1254211–1254211 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Manibog K, Li H, Rakshit S, Sivasankar S, Resolving the molecular mechanism of cadherin catch bond formation. Nat. Commun 5, 3941 (2014). [DOI] [PubMed] [Google Scholar]
  • 14.Yao M et al. , Force-dependent conformational switch of α-catenin controls vinculin binding. Nat. Commun 5, 4525 (2014). [DOI] [PubMed] [Google Scholar]
  • 15.Ishiyama N et al. , Force-dependent allostery of the α-catenin actin-binding domain controls adherens junction dynamics and functions. Nat. Commun 9, 5121 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Giannone G, Sheetz MP, Substrate rigidity and force define form through tyrosine phosphatase and kinase pathways. Trends Cell Biol. 16, 213–223 (2006). [DOI] [PubMed] [Google Scholar]
  • 17.McLachlan RW, Yap AS, Not so simple: The complexity of phosphotyrosine signaling at cadherin adhesive contacts. J. Mol. Med 85, 545–554 (2007). [DOI] [PubMed] [Google Scholar]
  • 18.Wang Y et al. , Visualizing the mechanical activation of Src. Nature 434, 1040–1045 (2005). [DOI] [PubMed] [Google Scholar]
  • 19.Seong J et al. , Distinct biophysical mechanisms of focal adhesion kinase mechanoactivation by different extracellular matrix proteins. Proc. Natl. Acad. Sci. U.S.A 110, 19372–19377 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Sehgal P et al. , Epidermal growth factor receptor and integrins control force-dependent vinculin recruitment to E-cadherin junctions. J. Cell Sci 131, jcs206656 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Echarri A et al. , An Abl-FBP17 mechanosensing system couples local plasma membrane curvature and stress fiber remodeling during mechanoadaptation. Nat Commun. 10, 5828 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Bays JL et al. , Vinculin phosphorylation differentially regulates mechanotransduction at cell-cell and cell-matrix adhesions. J Cell Biol. 205, 251–263 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Tamada M, Sheetz MP, Sawada Y, Activation of a signaling cascade by cytoskeleton stretch. Dev Cell 7, 709–718 (2004). [DOI] [PubMed] [Google Scholar]
  • 24.Sawada Y et al. , Force sensing by mechanical extension of the Src family kinase substrate p130Cas. Cell 127, 1015–1026 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Bertocchi C et al. , Nanoscale architecture of cadherin-based cell adhesions. Nat Cell Biol. 19, 28–37 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Zallen JA, Wieschaus E, Patterned gene expression directs bipolar planar polarity in Drosophila. Dev. Cell 6, 343–355 (2004). [DOI] [PubMed] [Google Scholar]
  • 27.Bertet C, Sulak L, Lecuit T, Myosin-dependent junction remodelling controls planar cell intercalation and axis elongation. Nature 429, 667–671 (2004). [DOI] [PubMed] [Google Scholar]
  • 28.Blankenship JT, Backovic ST, Sanny JSP, Weitz O, Zallen JA, Multicellular rosette formation links planar cell polarity to tissue morphogenesis. Dev Cell 11, 459–470 (2006). [DOI] [PubMed] [Google Scholar]
  • 29.Rauzi M, Verant P, Lecuit T, Lenne P-F, Nature and anisotropy of cortical forces orienting Drosophila tissue morphogenesis. Nat Cell Biol. 10, 1401–1410 (2008). [DOI] [PubMed] [Google Scholar]
  • 30.Fernandez-Gonzalez R, Simoes S. de M., Röper J-C, Eaton S, Zallen JA, Myosin II dynamics are regulated by tension in intercalating cells. Dev Cell 17, 736–743 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Lye CM, Naylor HW, Sanson B, Subcellular localisations of the CPTI collection of YFP-tagged proteins in Drosophila embryos. Development 141, 4006–4017 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Letizia A et al. , Sidekick is a key component of tricellular adherens junctions that acts to resolve cell rearrangements. Dev. Cell 50, 313–326.e5 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Uechi H, Kuranaga E, The tricellular junction protein sidekick regulates vertex dynamics to promote bicellular junction extension. Dev. Cell 50, 327–338.e5 (2019). [DOI] [PubMed] [Google Scholar]
  • 34.Finegan TM et al. , The tricellular vertex-specific adhesion molecule Sidekick facilitates polarised cell intercalation during Drosophila axis extension. PLOS Biol. 17, e3000522 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Farhadifar R, Röper J-C, Aigouy B, Eaton S, Jülicher F, The influence of cell mechanics, cell-cell interactions, and proliferation on epithelial packing. Curr. Biol 17, 2095–2104 (2007). [DOI] [PubMed] [Google Scholar]
  • 36.Trichas G et al. , Multi-cellular rosettes in the mouse visceral endoderm facilitate the ordered migration of anterior visceral endoderm cells. PLoS Biol. 10, e1001256 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Higashi T, Miller AL, Tricellular junctions: How to build junctions at the TRICkiest points of epithelial cells. Mol. Biol. Cell 28, 2023–2034 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Mandai K et al. , Afadin: A novel actin filament-binding protein with one PDZ domain localized at cadherin-based cell-to-cell adherens junction. J Cell Biol. 139, 517–528 (1997). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Mandai K, Rikitake Y, Shimono Y, Takai Y, Afadin/AF-6 and canoe: Roles in cell adhesion and beyond. Prog Mol Biol Transl Sci. 116, 433–454 (2013). [DOI] [PubMed] [Google Scholar]
  • 40.Sawyer JK, Harris NJ, Slep KC, Gaul U, Peifer M, The Drosophila afadin homologue Canoe regulates linkage of the actin cytoskeleton to adherens junctions during apical constriction. J Cell Biol. 186, 57–73 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Sawyer JK et al. , A contractile actomyosin network linked to adherens junctions by Canoe/afadin helps drive convergent extension. Mol. Biol. Cell 22, 2491–2508 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Choi W et al. , Remodeling the zonula adherens in response to tension and the role of afadin in this response. J Cell Biol. 213, 243–260 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Bonello TT, Perez-Vale KZ, Sumigray KD, Peifer M, Rap1 acts via multiple mechanisms to position Canoe and adherens junctions and mediate apical-basal polarity establishment. Development 145, dev157941 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Golas JM et al. , SKI-606, a 4-anilino-3-quinolinecarbonitrile dual inhibitor of Src and Abl kinases, is a potent antiproliferative agent against chronic myelogenous leukemia cells in culture and causes regression of K562 xenografts in nude mice. Cancer Res. 63, 375–381 (2003). [PubMed] [Google Scholar]
  • 45.Bradley WD, Koleske AJ, Regulation of cell migration and morphogenesis by Abl-family kinases: Emerging mechanisms and physiological contexts. J Cell Sci. 122, 3441–3454 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Colicelli J, ABL tyrosine kinases: Evolution of function, regulation, and specificity. Sci Signal. 3, re6 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Fox DT, Peifer M, Abelson kinase (Abl) and RhoGEF2 regulate actin organization during cell constriction in Drosophila. Development 134, 567–578 (2007). [DOI] [PubMed] [Google Scholar]
  • 48.Tamada M, Farrell DL, Zallen JA, Abl regulates planar polarized junctional dynamics through β-catenin tyrosine phosphorylation. Dev Cell 22, 309–319 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Jodoin JN, Martin AC, Abl suppresses cell extrusion and intercalation during epithelium folding. Mol. Biol. Cell 27, 2822–2832 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Boettner B et al. , The AF-6 homolog canoe acts as a Rap1effector during dorsal closure of the Drosophila embryo. Genetics 165, 159–169 (2003). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Martin AC, Kaschube M, Wieschaus EF, Pulsed contractions of an actin-myosin network drive apical constriction. Nature 457, 495–499 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Vanderleest TE et al. , Vertex sliding drives intercalation by radial coupling of adhesion and actomyosin networks during Drosophila germband extension. eLife 7, e34586 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Saerens D et al. , Identification of a universal VHH framework to graft non-canonical antigen-binding loops of camel single-domain antibodies. J. Mol. Biol 352, 597–607 (2005). [DOI] [PubMed] [Google Scholar]
  • 54.Pokutta S, Drees F, Takai Y, Nelson WJ, Weis WI, Biochemical and structural definition of the l-afadin- and actin-binding sites of α-catenin. J Biol Chem. 277, 18868–18874 (2002). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Sakakibara S et al. , Afadin regulates actomyosin organization through αE-catenin at adherens junctions. J Cell Biol. 219, e201907079 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Bosveld F et al. , Epithelial tricellular junctions act as interphase cell shape sensors to orient mitosis. Nature 530, 495–498 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Resnik-Docampo M et al. , Tricellular junctions regulate intestinal stem cell behaviour to maintain homeostasis. Nat Cell Biol. 19, 52–59 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Stephenson RE et al. , Rho flares repair local tight junction leaks. Dev Cell 48, 445–459.e5 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary_material
NIHMS1736713-supplement-Supplementary_material.pdf (26.6MB, pdf)
Movie S1
Download video file (449.2KB, mp4)
Movie S2
Download video file (803.3KB, mp4)
Movie S3
Download video file (622.3KB, mp4)
Movie S4
Download video file (413.7KB, mp4)
Movie S5
Download video file (665.2KB, mp4)
Movie S6
Download video file (178.7KB, mp4)
Movie S7
Download video file (1.5MB, mp4)
Movie S8
Download video file (1.4MB, mp4)
Movie S9
Download video file (1.4MB, mp4)

Data Availability Statement

All data are available in the manuscript or supplementary material.

RESOURCES