Force-dependent vinculin binding of α-catenin accelerates adherens junction formation
2026 年 51 巻 1 号 p. 23-35
詳細
2026 年 51 巻 1 号 p. 23-35
Adherens junctions (AJs) mediate cell–cell adhesion and mechanical coupling in epithelial tissues. During AJ formation, punctate AJs (punctum adherens; PA) initially appear and subsequently transition into linear AJs or zonula adherens (ZAs). The mechanosensitive interaction of α-catenin with its binding partners—actin filaments and vinculin—is thought to act as a key switch that stabilizes AJs under tension. However, the physiological role of α-catenin’s force sensitivity during the early stages of AJ formation remains unclear. Here, we analyzed α-catenin mutants with altered force sensitivity: Insensitive mutant L344P lacking vinculin binding, and Hypersensitive mutant L378P binding vinculin constitutively. Using calcium-switch assays combined with fluorescence and electron microscopy, we found that cells expressing insensitive α-catenin exhibited persistent, elongated PA-like structures corresponding to lateral associations of cellular protrusions from opposing cells, accompanied by delayed ZA formation. In contrast, cells expressing the hypersensitive mutant rapidly formed ZAs, possibly bypassing the PA stage. Similar phenotypes were observed in vinculin-knockout cells, indicating that the defects in Insensitive mutants result from the lack of vinculin recruitment to α-catenin. Based on these findings, we propose a model in which clusters of the cadherin–catenin complex (CCC) along actin filaments on opposing protrusions serve as initial adhesion sites. As protrusions shorten through actomyosin contraction, CCC clusters move toward the protrusion tips along actin filaments, where stretched α-catenin recruits vinculin to reinforce the adhesion, leading to PA formation. Thus, α-catenin’s force sensitivity is crucial for smooth and timely AJ assembly, ensuring proper epithelial morphogenesis by coupling intercellular adhesion with cytoskeletal tension.
Key words: α-catenin, vinculin, adherens junction, actin filament, force sensitivity
Graphical Abstract
Epithelial tissues define the boundary between the internal and external environments of multicellular organisms and play a central role in morphogenesis. They form sheet-like structures composed of epithelial cells connected by multiple types of intercellular junctions, including adherens junctions (AJs) and tight junctions (TJs). Among these, AJs serve as key sites for mechanical coupling between cells (Lecuit and Yap, 2015; Takeichi, 2014). AJs are located near the apical region of the lateral membranes of adjacent cells, where they are linked to the actin cytoskeleton (Yonemura, 2017). Deformation of individual cells connected through AJs leads to coordinated deformation of the entire epithelial sheet (Martin and Goldstein, 2014). During morphogenetic events involving substantial contractile forces—such as ventral furrow formation in Drosophila embryos or neural tube closure in vertebrates—AJs must maintain sufficient structural integrity (Vasquez and Martin, 2016).
Cadherins function as adhesion molecules at AJs. Their cytoplasmic tails associate with β-catenin, which in turn binds to α-catenin to form the cadherin–catenin complex (CCC; Pokutta and Weis, 2007; Takeichi, 1995). α-Catenin links this complex to the actin cytoskeleton, either directly (Desai et al., 2013; Ishiyama et al., 2018) or through actin-binding proteins such as vinculin (le Duc et al., 2010; Miyake et al., 2006; Yonemura et al., 2010). α-Catenin is indispensable for AJ assembly and for maintaining the actin connection (Hirano et al., 1992; Torres et al., 1997; Vasioukhin et al., 2001; Watabe et al., 1994; Watabe-Uchida et al., 1998). Other actin-associated proteins, including afadin (Matsuzawa et al., 2018; Sakakibara et al., 2020) and ZO-1 (Itoh et al., 1997), also localize to AJs, where actin filaments are densely organized. Among these, vinculin binds to α-catenin in a force-dependent manner (le Duc et al., 2010; Miyake et al., 2006; Yonemura et al., 2010).
Vinculin is an actin-binding protein found in both focal adhesions and adherens junctions, and it is essential for embryonic development (Xu et al., 1998). In the absence of tension, α-catenin does not interact with vinculin because the α-helical bundle containing the vinculin-binding helix is stabilized by intramolecular contacts and remains inaccessible (Hirano et al., 2018; Rangarajan and Izard, 2013). When a force exceeding approximately 5 pN is applied between the N- and C-termini, this bundle becomes destabilized, exposing the vinculin-binding helix, which can then engage vinculin (Hirano et al., 2018; Yao et al., 2014). Single-molecule force spectroscopy using atomic force microscopy, together with real-time total internal reflection fluorescence microscopy, has directly demonstrated that vinculin is recruited upon mechanical tension applied to α-catenin (Maki et al., 2018).
The physiological relevance of α-catenin’s force sensitivity has been increasingly recognized. Structural studies have identified residues critical for vinculin binding and for force-induced conformational transitions, enabling the generation of α-catenin mutants with altered force sensitivity (Hirano et al., 2018; Li et al., 2015; Peng et al., 2012). Several studies using such mutants have shown that constitutive vinculin binding slows collective epithelial migration (Matsuzawa et al., 2018; Seddiki et al., 2018). Collective migration and coordinated cell rearrangements are fundamental to convergent extension (Tada and Heisenberg, 2012). Conversely, an α-catenin mutant lacking the vinculin-binding site impairs migration and disrupts convergent extension in zebrafish embryos (Han et al., 2016). α-Catenin has also been reported to interact with vinculin within integrin-based adhesions, thereby regulating not only cell–cell but also cell–matrix adhesions (Mukherjee et al., 2022). We previously demonstrated that the force sensitivity of α-catenin is crucial for epithelial cells to form spherical spheroids through proper multicellular rearrangements involving AJ remodeling (Nishimura et al., 2022). Vinculin has been utilized as a molecular force sensor at cell–cell junctions (Kale et al., 2018). Its recruitment to the CCC enhances the actin filament–binding capacity of α-catenin (Bax et al., 2023) and facilitates force transmission (Mezher et al., 2023).
Because AJ formation and maturation depend on the tension transmitted via associated actin filaments, elucidating how α-catenin’s mechanosensitivity—particularly in relation to vinculin binding—contributes to AJ assembly is of particular interest. α-Catenin mutants can eventually support both AJ and TJ formation (Hirano et al., 2018; Nishimura et al., 2022) and vinculin knockout (KO) EpH4 cells can form an unstable but largely normal TJ network (Konishi et al., 2019). However, the specific role of α-catenin’s force sensitivity and vinculin binding during the early stages of AJ formation remains largely unexplored. Here, we examined the functional significance of α-catenin’s force sensitivity during AJ assembly using multiple two force-sensitivity mutants. Analyses at both light- and electron-microscopic levels revealed that force-dependent vinculin recruitment to α-catenin is essential for smooth and timely AJ formation.
DLD-1 cells (ATCC), its subclone R2/7 cells (Watabe-Uchida et al., 1998; provided by F. van Roy, Ghent University, Belgium), and R2/7-derived cell lines expressing wild-type, L378P, or L344P mutant α-catenin (Yonemura et al., 2010; Peng et al., 2012; Nishimura et al., 2022) were maintained in Dulbecco’s Modified Eagle Medium (DMEM; Wako) supplemented with 10% fetal bovine serum (FBS). EpH4 cells (provided by E. Reichman, University Children’s Hospital, Switzerland) and vinculin knockout (KO) EpH4 cells (Konishi et al., 2019) were cultured under the same conditions.
Antibodies and reagentsThe following antibodies were used: rat monoclonal anti-E-cadherin (ECCD-2; a gift from M. Takeichi, RIKEN, Japan) and mouse monoclonal anti-ZO-1 (T8-754). Secondary antibodies included Alexa Fluor 555-conjugated donkey anti-rat IgG (Abcam) and Alexa Fluor 647-conjugated donkey anti-mouse IgG (Thermo Fisher Scientific). Actin filaments were stained with Alexa Fluor 488-phalloidin (Thermo Fisher Scientific).
Calcium switch experimentCells were grown on coverslips until confluent, then washed briefly with PBS containing 1 mM EDTA and incubated in low-Ca2+ DMEM supplemented with 10% FBS extensively dialyzed against PBS to deplete Ca2+. DLD-1 and R2/7 cells expressing α-catenin mutants were incubated for 3 h to disrupt cell–cell junctions, whereas EpH4 and vinculin KO EpH4 cells were incubated for 13 h to ensure complete junctional disruption. The low-Ca2+ medium was then replaced with normal medium, and cells were fixed at indicated time points for immunofluorescence or electron microscopy.
ImmunostainingImmunofluorescence staining was performed as described previously (Miyake et al., 2006; Watanabe et al., 2007). Briefly, cells were fixed in 1% formaldehyde in 0.1 M HEPES buffer (pH 7.5) for 15 min, permeabilized for 10 min, and washed (once after permeabilization and three times after antibody incubations). Primary and secondary antibody incubations were carried out for 1 h each at room temperature. Images were acquired using an Olympus BX51 microscope equipped with a CCD camera (ORCA ER, Hamamatsu Photonics) controlled by IPLab Spectrum v3.5.4 (Scanalytics) and with 40× Ph2 NA 0.75, 60× Ph3 NA 1.40, or 100× NA 1.45 objectives.
Electron microscopyFor thin-section ultrastructure analysis, cells were fixed in 2% formaldehyde and 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 2 h, followed by postfixation in 1.5% potassium ferrocyanide and 1% osmium tetroxide in 0.1 M cacodylate buffer for 1 h on ice. Samples were then washed in water, treated with 1% thiocarbohydrazide at 60°C for 1 h, washed again, and fixed in 1% osmium tetroxide on ice for 1 h. After successive washes, samples were stained with 4% uranyl acetate (2 h at 4°C) and 0.66% lead nitrate in 0.03 M aspartic acid (pH 5.5) at 60°C for 1 h. Samples were dehydrated through graded ethanols and embedded in Epon 812 resin (TAAB). Coverslips were detached by alternating hot-water and liquid-nitrogen treatments. Serial ultrathin sections (70–200 nm) were cut with a diamond knife on a Leica EM UC7 ultramicrotome and mounted on silicon wafers. Sections were examined using a field-emission scanning electron microscope (FE-SEM; JSM-IT800SHL, JEOL) at 3 kV with a backscattered-electron detector.
For surface morphology, cell sheets cultured on 35 mm dishes were fixed as above, postfixed with 1% osmium tetroxide for 1 h on ice, dehydrated through ethanol, and treated briefly with propylene oxide. Detached cell sheets were washed repeatedly in propylene oxide and ethanol, then dried using a critical-point dryer (EM CPD300, Leica). Dried fragments were mounted on carbon-taped metal stubs, coated with osmium using a Tennant 20 osmium coater (Meiwafosis), and imaged with a JEOL JSM-IT800SHL FE-SEM at 3 kV conventionally (without a backscattered-electron detector).
Image analysesImages were processed and analyzed using Fiji (National Institute of Health) (Schindelin et al., 2012). E-cadherin–positive puncta or rod-like structure aligned with F-actin bundles perpendicular to cell–cell interfaces was defined as punctum adherens (PA). The maximum Feret diameter of each PA was manually measured. ChatGTP-5 (OpenAI) was used to combine 3 histograms in the same graph. For cell surface morphology analysis, EM images at low magnification (500×) were visually inspected. When an obvious bulge was observed at the central region of a cell, it was classified as a “round” cell; all others were classified as “flat” cells. Although this analysis was not strictly quantitative, approximately 300–500 cells were counted for each sample (Fig. 5B).
We used wild-type (WT) α-catenin and two force-sensitivity mutants—Insensitive and Hypersensitive—in this study. WT α-catenin binds to vinculin only when appropriate tension is applied, reflecting physiological force transmission at adherens junctions (AJs). The Insensitive mutant (L344P) cannot bind vinculin even under strong tension, whereas the Hypersensitive mutant (L378P) constitutively binds vinculin even in the absence of force. When expressed in R2/7 cells, which lack endogenous α-catenin, all three forms have been reported to ultimately form epithelial sheets with mature AJs and TJs (Hirano et al., 2018; Nishimura et al., 2022).
To clarify the significance of α-catenin force sensitivity during the early stages of AJ formation, we performed calcium-switch experiments. Confluent monolayers were transferred to calcium-free medium to disrupt intercellular junctions and then returned to normal medium to initiate AJ reassembly. In WT cells, punctate E-cadherin accumulations associated with actin bundles were evident as early as 30 min after calcium restoration (blue arrows, Fig. 1A). These puncta correspond to punctum adherens (PA), the initial form of epithelial AJs (Fig. 1B; Yonemura et al., 1995, 2011). Linear and continuous AJs, or zonula adherens (ZA), began to form at 1 h (yellow arrows, Fig. 1A) and became predominant by 2 h.
AJ formation in cells expressing wild-type or force-sensitivity mutant of α-catenin at the fluorescence microscopy level
(A) R2/7 cells expressing wild-type α-catenin (WT; vinculin-binding force-sensitive), L344P mutant (Insensitive; unable to bind vinculin even under high tension), or L378P mutant (Hypersensitive; constitutively vinculin-bound) after calcium switch (30 min, 1 h, and 2 h). Green, actin filaments; magenta, E-cadherin. Light-blue arrows indicate PA-like structures; yellow arrows indicate ZA. Scale bar, 10 μm. (B) Schematic illustration of AJ classification. Green, actin filaments; magenta, E-cadherin. (C) Histogram of PA length at 1 h after calcium switch. WT (mean ± SD, 0.80 ± 0.55 μm, n = 171), Hypersensitive (0.63 ± 0.46 μm, n = 103), and Insensitive (3.62 ± 1.61 μm, n = 122).
In Insensitive cells, numerous PA-like structures with rod-shaped E-cadherin accumulations along actin filaments were observed (Fig. 1A). These appeared as elongated PAs compared with compact WT PAs (Fig. 1B), and the transition from PA to ZA was markedly delayed, with PAs persisting for at least 2 h after calcium switch. In contrast, Hypersensitive cells exhibited fewer PAs and more rapid ZA formation (Fig. 1A). Measurement of PA length along actin bundles at 1 h confirmed these trends (WT, 0.80 μm; Insensitive, 3.62 μm; Hypersensitive, 0.63 μm), as summarized in the histogram (Fig. 1C). Individual channel images corresponding to Fig. 1A are shown in Fig. S1–S3.
Cadherin-based adhesions mediated by lateral association of cell protrusions in Insensitive cellsTo further examine these differences, we analyzed cell ultrastructure by electron microscopy. At 1 h after the calcium switch, both WT and Hypersensitive cells displayed typical ZAs. In contrast, Insensitive cells exhibited numerous fine cellular protrusions and few ZAs, even after 2 h (Fig. 2). At later time points (1–6 h), protrusions from adjacent Insensitive cells frequently contacted each other laterally (arrowheads, Fig. 3A). As shown schematically in Fig. 3B, the structures that we initially described as “elongated PAs” represent cadherin-based adhesions formed along laterally associated protrusions, rather than canonical, actin-anchored PAs. By 24 h, Insensitive cells eventually developed normal ZAs (large arrows, Fig. 3A). Notably, transitional regions showing progressive conversion from lateral contacts of protrusions to PAs and ZAs were observed at 6 h (Fig. 3A). Thin protrusions appeared to thicken and shorten before ZA development. Throughout this process, typical PAs were rarely detected in Insensitive cells.
Electron microscopy (EM) images of cells 1–2 h after calcium switch
Arrows indicate developing ZAs (WT and Hypersensitive). In Insensitive cells, abundant cellular protrusions are evident. Scale bars, 1 μm (WT and Hypersensitive); 2 μm (Insensitive).
Initial adhesion in force-insensitive mutant–expressing cells begins with lateral association of cellular protrusions
(A) EM images of Insensitive cells after calcium switch. Arrowheads indicate lateral associations of protrusions. Small arrows, PAs. Large arrows, ZAs. Scale bars, 2 μm (left panels); 1 μm (right panels). (B) Schematic representation showing the difference between lateral association of protrusions and typical PA structure.
Cell-surface morphology analysis revealed abundant laterally associated protrusions on the surface of Insensitive cells (arrowheads). In WT and Hypersensitive cells, such protrusions were absent at 30 min and were replaced by short, microvilli-like structures at cell interfaces (arrows) by 2 h after the calcium switch (Fig. 4). At low magnification, Insensitive cells appeared rounded even at 6 h, whereas WT and Hypersensitive cells flattened within 2 h (Fig. 5).
Changes in cell surface morphology after calcium switch
WT, Insensitive, and Hypersensitive cells were analyzed at indicated time points after calcium switch. Arrows, microvilli-like structures at cell interfaces; arrowheads, protrusions connecting adjacent cells. Scale bars, 1 μm; 2 μm (insets).
Force-insensitive mutant–expressing cells retain a rounded morphology for an extended period after calcium switch
(A) Low-magnification surface images of WT, Insensitive, and Hypersensitive cell sheets at indicated time points. Scale bar, 10 μm. (B) Quantification of flat cells in each cell sheet. A considerable fraction of Insensitive cells remained round even 6 h after calcium switch. Error bars indicate SEM (n = 300–500 cells).
A well-known feature of the Insensitive mutant is its inability to bind vinculin. If this property underlies the phenotype of Insensitive cells, vinculin-knockout (KO) cells should exhibit similar behavior. To test this, we used vinculin-KO EpH4 cells that we previously generated (Konishi et al., 2019), along with control EpH4 cells. In control EpH4 cells, PAs were present together with ZAs at 4 h after calcium switch, and ZAs were almost complete by 8 h (Fig. 6). In contrast, vinculin-KO EpH4 cells exhibited elongated PA-like structures similar to those observed in Insensitive cells. These structures remained abundant even 8 h after the calcium switch, indicating that the transition to ZAs was markedly delayed. Electron microscopy of control EpH4 cells revealed ZAs as early as 5 h after calcium switch (Fig. 7), whereas vinculin-KO EpH4 cells displayed numerous cellular protrusions and lateral contacts between 2.5 h and 8 h (Fig. 8A–C). Transitional forms from lateral associations of protrusions to ZAs were also observed at 8–24 h (Fig. 8C, D). These findings support the notion that the abundant lateral associations of protrusions and delayed AJ maturation seen in Insensitive cells result from the lack of vinculin recruitment to α-catenin at AJs. PAs are not obvious possibly due to the difficulty in visualizing membranes and actin filament bundles in these cells.
Vinculin-KO cells show AJ developmental processes similar to those of force-insensitive mutant–expressing cells
EpH4 and vinculin-KO EpH4 cells were compared after calcium switch. From 4–8 h after calcium restoration, EpH4 cells formed nearly complete ZAs, whereas vinculin-KO cells exhibited elongated PA-like structures. Green, actin filaments; magenta, E-cadherin. Scale bar, 10 μm.
EM images of EpH4 cells after calcium switch
(A) EpH4 cells 2.5 h after calcium switch (low magnification). Boxed regions are magnified below. (B, C) EpH4 cells 5 h (B) and 24 h (C) after calcium switch. Arrows indicate ZAs. Scale bars, 5 μm (A); 1 μm (A, higher magnification; B, C).
EM images of Vinculin-KO cells after calcium switch
Vinculin-KO EpH4 cells after 2.5 h (A), 5 h (B), 8 h (C), and 24 h (D). Fine protrusions remained abundant even 8 h after calcium switch. The boxed area in (B) is magnified in the inset, showing lateral associations of protrusions. Arrowheads indicate ZAs; arrows, protrusion contact sites. Scale bars, 5 μm (A), 2 μm (B), 1 μm (inset in B), 2 μm (C), 1 μm (D).
We next investigated whether lateral associations of protrusions also occur transiently in WT cells. At 15 min after the calcium switch, numerous protrusions and their lateral contacts were observed, together with occasional PAs (Fig. 9). By 20 min, protrusions appeared thickened and shortened, and laterally aligned PAs became evident. These observations suggest that such protrusive associations represent a normal, transient intermediate during AJ formation in WT cells. When α-catenin possesses normal force sensitivity, protrusion contacts rapidly transition into PAs and subsequently into ZAs. In Hypersensitive cells, protrusions and their lateral associations were also detected at 15 min, but ZAs were already evident (Fig. 10). Continuous vinculin binding to α-catenin likely accelerates this transition or enables direct ZA formation, bypassing the intermediate stages.
EM images of WT cells at early stages of intercellular adhesion
R2/7 cells expressing WT α-catenin 15 min and 20 min after calcium switch. Arrowheads indicate contacts between protrusions; arrows, PAs. Scale bar, 0.5 μm.
EM images of Hypersensitive cells at early stages of intercellular adhesion
R2/7 cells expressing Hypersensitive α-catenin 15 min after calcium switch. Both protrusion contacts (arrowheads) and ZAs (arrows) are evident. Scale bar, 1 μm.
The force sensitivity of α-catenin was first characterized through vinculin binding as a measurable molecular output (Yonemura et al., 2010). Although the force-insensitive L344P mutant cannot bind vinculin, it might theoretically possess other structural or binding defects. In this study, we found that the phenotypes of Insensitive cells—abundant elongated PA-like structures and delayed AJ maturation—were faithfully reproduced in vinculin-knockout (KO) EpH4 cells. This strongly supports the notion that these phenotypes primarily result from the loss of vinculin recruitment to AJs, rather than from other structural defects in α-catenin. It should be noted, however, that these results do not fully exclude the possible involvement of other α-catenin–binding proteins in these phenotypes; for example, afadin has been reported to participate in AJ formation from the early stages (Sato et al., 2006; Sawyer et al., 2009).
Vinculin-KO mice exhibit embryonic lethality accompanied by multiple developmental abnormalities, including impaired heart and brain morphogenesis, reduced body size, and fragile ectoderm (Xu et al., 1998). However, the specific consequences of vinculin recruitment to α-catenin in a force-dependent manner have remained unclear. R2/7 cells expressing an α-catenin mutant in which the vinculin-binding domain is replaced with the homologous region of vinculin, thereby abolishing vinculin binding, displayed no obvious differences in AJ organization (Huveneers et al., 2012). Yet, when this mutant was introduced into HUVECs in which endogenous α-catenin was knocked down, the cells exhibited transiently elongated cadherin-based adhesions along actin filaments after thrombin stimulation—consistent with our observations in the present study.
Vinculin-KO EpH4 cells can form nearly normal AJs and TJs, although their TJs are unstable (Konishi et al., 2019). Thus, vinculin is not absolutely required for the establishment of epithelial sheets with junctional structures. Nonetheless, vinculin appears indispensable for the proper formation and stabilization of AJs, explaining both the defects observed in KO cells and the embryonic lethality seen in KO mice.
In this study, Insensitive cells exhibited persistent lateral associations of cadherin-positive protrusions, which delayed AJ maturation beyond the normal time frame. Given the importance of rapid AJ assembly for epithelial barrier integrity and tissue homeostasis, vinculin-mediated reinforcement likely plays a critical role in ensuring timely junctional maturation. In contrast, Hypersensitive cells, in which vinculin is constitutively bound, displayed accelerated AJ formation, potentially bypassing intermediate stages (Fig. 10).
A previous study showed that R2/7 cells expressing a chimeric molecule composed of α-catenin residues 1–325 (including only the β-catenin–binding domain) fused to the actin-binding region of vinculin (residues 822–1067)—but not to the α-catenin C-terminus lacking the vinculin-binding region (residues 510–890)—were able to form TJs (Watabe-Uchida et al., 1998). This finding implies that vinculin’s actin-binding domain can promote TJ—and likely preceding ZA—formation even in the absence of α-catenin’s own actin-binding domain. Considering that the Insensitive α-catenin mutant retains a functional actin-binding domain yet exhibits markedly delayed AJ maturation, the vinculin actin-binding domain likely has an additional regulatory role beyond simple actin attachment. Moreover, although both α-catenin and vinculin exhibit directional catch-bond behavior between their actin-binding domains and actin filaments (Buckley et al., 2014; Chirasani et al., 2023), differences in their molecular characteristics may underlie their distinct contributions to ZA and TJ formation.
Insensitive cells also maintained a rounded morphology for an extended period after the calcium switch, whereas WT and Hypersensitive cells flattened rapidly (Fig. 5). In epithelial sheets, cells typically become rounded when cadherin-mediated adhesion is disrupted under low-calcium conditions and flatten again as junctions reform. The persistent lateral associations of protrusions observed in Insensitive cells likely transmit insufficient intercellular tension for effective cell flattening.
Integrating our findings with recent studies, we propose a model for AJ formation (Fig. 11). The CCC forms oligomers along actin filaments when α-catenin binds actin through its C-terminal domain (Troyanovsky et al., 2025; Fig. 11A). Because cellular protrusions often contain bundled actin filaments with minus ends towards the protrusion tip, CCC oligomers are likely positioned preferentially on these protrusions (Fig. 11B). During the initial stages of cell–cell contact, protrusions from adjacent cells first meet (Fig. 11C). Lateral association of these protrusions may occur because CCC oligomers exhibit greater homophilic avidity than individual complexes. Actin filaments within the protrusions are connected to the cytoplasmic actomyosin network, enabling myosin II–driven contraction that shortens the protrusions. During actomyosin-driven contraction, CCC pairs associated through cadherin extracellular domains on opposing protrusions move toward the protrusion tips along actin bundles as the overlap between the two protrusions diminishes. This process stretches α-catenin within the complex, triggering vinculin recruitment and stabilization of the adhesion site. Vinculin enhances the actin filament–binding capacity of α-catenin, and the lifetimes of this quaternary complex (CCC plus vinculin) increase as additional complexes bind actin filaments when load is oriented toward the minus end of actin filaments (Bax et al., 2023). The accumulation of quaternary complexes at protrusion tips marks the onset of PA formation.
Schematic illustration of the proposed model for AJ formation
(A) Cadherin–catenin complex (CCC) clustering along actin filaments (Troyanovsky et al., 2025). (B) Expected CCC clustering on cell protrusions containing actin filaments. (C) Proposed model of initial intercellular adhesion leading to PA formation. CCC clusters on opposing protrusions associate laterally. As protrusions shorten, clusters accumulate at their tips. Actomyosin-generated forces stretch α-catenin, enabling vinculin recruitment and stabilization of tip adhesions, leading to PA formation. In Insensitive mutant cells, lack of vinculin recruitment prevents efficient CCC stabilization at protrusion tips.
These processes appear to be inherently reversible. When α-catenin possesses normal force sensitivity, the lateral protrusive association stage is brief and transitions rapidly to PA formation (Fig. 9). In Insensitive cells lacking vinculin binding, tip-to-tip adhesions cannot be efficiently stabilized, prolonging the protrusion-contact stage. Conversely, in Hypersensitive cells, continuous vinculin binding may shorten or bypass these intermediate stages, leading directly to ZA formation. Testing this model will require direct visualization of CCC oligomers along actin-containing protrusions, for example through live-cell super-resolution microscopy.
In summary, our findings demonstrate that α-catenin’s force sensitivity—manifested as force-dependent vinculin recruitment—is crucial for timely and orderly AJ formation. This mechanosensitive regulation ensures proper epithelial morphogenesis by coupling intercellular adhesion with cytoskeletal tension.
This work was supported by JSPS KAKENHI (Grant numbers 18H02617 and 24K02228 to SY), JST Moonshot R&D (Grant number JPMJMS2022 to SY).
Conflict of Interest StatementThe authors declare that they have no conflicts of interest with the contents of this article.
Data Availability StatementThe data in this study are available from the corresponding author upon reasonable request.
Author Contribution StatementQG, SY, and ST conceptualization and data curation; SY funding acquisition; QG and SY investigation; SY methodology; SY project administration; SY and ST resources; SY supervision; SY and ST validation; QG and SY visualization; QG writing-original draft; SY writing-review and editing.
Ethics Approval and Consent to ParticipateThis study did not involve human patients or identifiable personal information, and therefore neither ethical approval nor informed consent was required.
Patient Consent for PublicationNot applicable.
We thank Dr. F. Van Roy (Ghent University), and Dr. E. Reichman (University Children’s Hospital) for R2/7 cells, and EpH4 cells, respectively. We thank A. Suzuki for technical assistance (Tokushima University). We thank all laboratory members for insightful discussion. This work was also supported by the Support Center for Advanced Medical Sciences, Tokushima University Graduate School of Biomedical Sciences.
adherens junction
DMEMDulbecco’s modified Eagle medium
PApunctum adherens
TJtight junction
ZAzonula adherens
