Dynamic contacts: rearranging adherens junctions to drive epithelial remodelling.

Nature Reviews Molecular Cell Biology | AOP, published online 14 May 2014; doi:10.1038/nrm3802

REVIEWS

Dynamic contacts: rearranging adherens junctions to drive epithelial remodelling Masatoshi Takeichi

Abstract | Epithelial cells display dynamic behaviours, such as rearrangement, movement and shape changes, particularly during embryonic development and in equivalent processes in adults. Accumulating evidence suggests that the remodelling of cell junctions, especially adherens junctions (AJs), has major roles in controlling these behaviours. AJs comprise cadherin adhesion receptors and cytoplasmic proteins that associate with them, including catenins and actin filaments, and exhibit various forms, such as linear or punctate. Remodelling of AJs induces epithelial reshaping in various ways, including by planar-polarized apical constriction that is driven by the contraction of AJ‑associated actomyosin and that occurs during neural plate bending and germband extension. RHO GTPases and their effectors regulate actin polymerization and actomyosin contraction at AJs during the epithelial reshaping processes. Apical constriction A process to induce the bending of epithelial sheets during various morphogenetic processes, such as gastrulation and neural tube formation. Epithelial cells shrink specifically at the apical ends in response to external or internal signals, and as a result their sheets bend towards the apical ends.

RIKEN Center for Developmental Biology, 2‑2‑3 Chuo‑ku, Kobe 650–0047, Japan. e‑mail: [email protected] doi:10.1038/nrm3802 Published online 14 May 2014

Epithelial cells are tightly interconnected and form a twodimensional sheet. Although intercellular adhesions are stable in mature epithelia, the junctions between cells are dynamic during morphogenetic processes, which enables their rearrangement and even movement. This is also true in adult organs during regenerative processes such as wound healing or during pathogenetic processes such as cancer invasion. If cell junctions were not dynamic in nature, cells would be ‘fixed’ within tissues and tissues would be unable to undergo remodelling as a result. A principal intercellular structure that links cells together is the adherens junction (AJ). The AJ consists of cadherin adhesion receptors and cytoplasmic proteins that associate with them, including actin filaments (FIG. 1). The fact that the physiological states of cadherin-associated cytoplasmic components are highly regulated makes AJs plastic and enables junctions to contract or extend, or to stabilize or destabilize. In response to these changes in AJs, epithelial tissues undergo various forms of reshaping, such as apical constriction, folding and migration, which in turn result in global changes in tissue morphology 1,2. Therefore, to understand how tissues can be remodelled, it is crucial to elucidate the mechanism by which AJs are regulated. AJs or AJ‑like structures are detected in both vertebrates and invertebrates, which suggests that AJs have a fundamental role in establishing tissues. Cadherin genes

have been detected in most animal species3, and these genes encode diverse proteins that contain a characteristic Ca2+-binding amino acid sequence motif and are classified as the cadherin superfamily. A specific group of cadherins within this superfamily exhibit a preserved cytoplasmic domain that can interact with a unique set of intracellular proteins that are collectively called catenins; the interaction between cadherins and catenins has a crucial role in the formation and function of AJs. This group of cadherins is often referred to as the ‘classic cadherins’ to distinguish them from other cadherin superfamily members, the biological functions of which are diverse4. Vertebrate genes encode approximately 20 subtypes of classic cadherins, such as epithelial cadherin (E‑cadherin) and neural cadherin (N‑cadherin), whereas the genes of invertebrates, such as Drosophila melanogaster or Caenorhabditis elegans, encode only a few subtypes. The extracellular domain of the vertebrate classic cadherins is subdivided into five repetitive extracellular cadherin (EC) domains, whereas that of invertebrate versions contains more EC domains, the number of which varies by cadherin subtype3. Despite this difference in the extracellular domain, vertebrate and invertebrate molecules form AJs of similar molecular architecture, owing to the conserved cytoplasmic domain. For convenience, the classic cadherins are referred to as ‘cadherin’ or ‘cadherins’ in this Review, unless otherwise specified.

NATURE REVIEWS | MOLECULAR CELL BIOLOGY

ADVANCE ONLINE PUBLICATION | 1 © 2014 Macmillan Publishers Limited. All rights reserved

REVIEWS a

E-cadherin O-linked glycans Ca2+

EC2

b EC1 EC2

E-cadherin

E-cadherin

EC3

E-cadherin EC4 EC5 Plasma membrane

Plasma membrane Cytoplasm

p120 catenin β-catenin N

N Cytoplasm

αE-catenin

M2 M3

p120 catenin isoform 4A

M1

Unfurled M1

M2

VCL M3

C β-catenin

F-actin

C

αE-catenin

Figure 1 | Molecular architecture of adherens junctions. a | A model of the cadherin–catenin core complex is shown and is based on the following crystal structures: the strand-swapped trans dimerNature of the epithelial cadherin (E‑cadherin) Reviews | Molecular Cell Biology ectodomain (light pink and blue) with bound Ca2+ ions (green spheres) and O‑linked glycans (brown spheres) (Protein Data Bank (PDB) accession identification (ID): 3Q2V)146; p120 catenin isoform 4A (orange) bound to the E‑cadherin juxtamembrane domain (blue) (PDB ID: 3L6X)147; β‑catenin (dark pink) bound to the E‑cadherin–catenin-binding domain (blue) (PDB ID: 1I7W)148; and αE‑catenin (green) (PDB ID: 1DOW, 4IGG and 4K1N)22,25,149. b | A model of the interaction between the cadherin–catenin complex, vinculin (VCL) and actin filaments. Crystal structures of mammalian αE‑catenin19,22 revealed that the modulatory 1 (M1)–M3 domains are autoinhibited, which prevents them from interacting with VCL, as shown on the left. The closed conformation of the autoinhibited state requires the M3 domain, which forms intramolecular electrostatic interactions with the M1 and M2 domains (grey lines). On the right, a putative open form of αE‑catenin is illustrated, in which the M2–M3 interface is disrupted and the M1 domain is unfurled to expose the VCL-binding site26,27. The αE‑catenin–VCL complex is thought to bind to more actin filaments than autoinhibited αE‑catenin. C, carboxyl terminus; EC, extracellular cadherin domain; N, amino terminus. Original drawings courtesy of N. Ishiyama and M. Ikura, Ontario Cancer Institute, Canada.

This article reviews our current understanding of how epithelial AJs are controlled and how AJ dynamics participate in epithelial morphogenesis, focusing on the roles of cadherin–actin interactions and related issues. Most studies in this field have used vertebrate or D. melanogaster epithelial cells, which mainly express E‑cadherin (N‑cadherin in the case of neuroepithelium) and D. melanogaster epithelial cadherin (DE‑cadherin), respectively. In C. elegans, the cadherin-based junctions seem to have limited roles in tissue construction, possibly owing to the presence of other adhesion systems5. Thus, the

following discussion mainly relies on data that are derived from vertebrates and D. melanogaster, except for a few examples that are derived from C. elegans studies. Both epithelial AJs and vascular endothelial AJs have been well studied to elucidate the mechanism by which AJs are stabilized or destabilized. As endothelial AJs contain vascular endothelial cadherin (VE‑cadherin) instead of E‑cadherin, their structural regulation may not be identical to that of epithelial AJs. However, data obtained from the study of endothelial AJs are useful and are cited occasionally throughout this Review.

2 | ADVANCE ONLINE PUBLICATION

www.nature.com/reviews/molcellbio © 2014 Macmillan Publishers Limited. All rights reserved

REVIEWS a

Linear AJs

Punctate AJs

LC

ZA

Free edge

10 μm

b

10 μm

c

b

TJ ZA ZA

Actin bundle

Cadherin

Free edge

DM

Actin

Punctate AJ 0.1 μm

Figure 2 | Structural details of adherens junctions. a | Two types of adherens junctions (AJs) were detected by Reviewsby | Molecular Cell Biology immunolabelling of epithelial cadherin (E‑cadherin; magenta) and F‑actin (green),Nature and visualized microscopy. Linear or punctate AJs in a colony of EpH4 cells are shown. In the linear AJs, which are found in the interior of the colony, the plasma membranes are lined with circumferential actin filaments at an apical portion of cell–cell contacts. E‑cadherin colocalizes with these actin filaments to organize the zonula adherens (ZA). E‑cadherin is also observed at lateral cell–cell contacts (LCs) below the ZA that overlap with amorphous actin networks, which are often visualized as tilted intercellular boundaries. Punctate AJs are found near the edge of the cell colony. Actin filaments perpendicularly terminate at the plasma membranes and pull E‑cadherin puncta. b | A schematic drawing of linear AJs (organized into the ZA) and punctate AJs, which are detected at the interior and periphery of a colony, respectively. c | Electron microscopic image of mouse intestinal epithelium. A tight junction (TJ), ZA and desmosome (DM) are arranged in this order at the apical-most region of cell–cell contacts. Images in part a courtesy of T. Nishimura, Kobe University, Japan. Image in part c courtesy of K. Misaki, RIKEN Center for Developmental Biology, Japan.

The AJ core: cadherin–actin linkages The dynamics of AJs are mostly mediated through their interaction with the actin cytoskeleton; therefore, understanding the molecular basis of this interaction is crucial for understanding epithelial remodelling. Epithelial AJs exhibit various morphologies, which can conventionally be classified into two forms, linear and punctate, depending on how AJs associate with

actin filaments (FIG. 2a,b). In mature epithelial sheets, a bundle of linear actin filaments runs parallel to the cell borders. This bundle encircles the individual cells at a near-apical portion and organizes itself into a belt-like structure that is known as the circumferential actin belt, bundle or ring. Cadherins accumulate along these actin filaments as nanometric clusters6, and this form of junction is defined as a linear AJ in this



REVIEWS Review. At the ultrastructural level, where the linear AJ is known as the zonula adherens (ZA), actin bundles lie in close proximity to the plasma membrane of cells (FIG. 2c). The ZA is located between tight junctions and desmosomes, and these three junctions are collectively called the apical junctional complex 7. Below the ZA, cadherin molecules are detected throughout lateral cell–cell contacts, which colocalize with amorphous actin networks. It is not determined whether these contacts are also called AJs. At the edges of epithelial colonies, or in other cell types such as mesenchymal cells, AJs are morphologically distinct from linear AJs (or the ZA). In these AJs, actin filaments perpendicularly terminate at the plasma membrane. These actin filaments pull cadherins from both sides of the junction, which causes a discontinuous or ‘zigzag’ appearance of cadherin distribution. This type of AJ, which is also known as a spot, radial, discontinuous or focal AJ, is defined as a punctate AJ in this Review. Various intermediate forms of AJs seem to exist between the linear and punctate AJs, which are actually interchangeable8. Transformed epithelial cells occasionally exhibit indefinable irregular AJ morphologies9. How AJ–actin interactions are established and how these interactions affect AJ shapes are now discussed.

Zonula adherens (ZA). A cell–cell adherens junction that forms a circumferential belt around the apical pole of epithelial cells.

Tight junctions Circumferential rings at the apex of epithelial cells that seal adjacent cells to one another. Tight junctions regulate solute and ion flux between adjacent epithelial cells.

Desmosomes Junctional structures that are formed by transmembrane proteins that are homologous to cadherins and are called desmocollins and desmogleins. These are linked to plakoglobin and desmoplakin, and are anchored to intermediate filaments.

Interactions between cadherins and actin filaments. At cell–cell boundaries, cadherins undergo homophilic interactions through their amino‑terminal EC1 domain, which results in a trans dimer (also known as an adhesive dimer)10 (FIG. 1a). Inside cells, cadherin binds to p120 catenin (also known as catenin δ1) and β‑catenin, and β‑catenin in turn binds to α‑catenin, forming the cadherin–catenin core complex (CCC). Although the trans dimer, which is formed between cadherins, produces a primary force that ‘sticks’ cells together, this is not sufficient for establishing typical epithelial cell–cell contacts: α‑catenin is required to firmly ‘zip up’ plasma membranes11,12. α‑catenin can bind to F‑actin through its carboxy‑terminal domain13,14, which is thought to be crucial for the CCC to create firm cell adhesions. The role of the α‑catenin C‑terminal domain, however, has become a debated issue, since it was reported that α‑catenin in the CCC is unable to bind to F‑actin15 and that only free α‑catenin can bind to F‑actin by forming a homodimer 16. A later study that used the C. elegans α‑catenin homologue HMP‑1 also indicated that HMP‑1 is unable to bind to F‑actin in vitro owing to autoinhibition17. Nevertheless, the same study suggested that HMP‑1 must bind to F‑actin in vivo 17. Furthermore, various features of AJs can be best explained by assuming a physical linkage between the CCC and F‑actin18–21. For example, in punctate AJs, actin filaments apparently ‘pull’ E‑cadherin molecules (FIG. 2a). This morphological feature of AJs indicates that the actin–cadherin interactions must be established through their firm physical bindings. This story is further complicated by the fact that α‑catenin can interact with F‑actin through another actin-binding protein, vinculin (VCL; also known

as Vinc in D. melanogaster), which has homology to α‑catenin. α‑catenin binds to VCL through its modulatory 1 (M1) domain (also known as the VCL-binding domain or VCL-binding site) (FIG. 1b) . N‑terminal α‑catenin fragments that contain the M1 domain or those that are fused to the actin-binding domain of VCL are able to organize epithelial AJs even in the absence of the α‑catenin C‑terminal domain, whereas α‑catenin mutants that are devoid of the M1 domain are unable to organize AJs even when they contain the C‑terminal domain8,12,22. This suggests that the CCC interacts with F‑actin through VCL rather than the α‑catenin C‑terminal domain and that this inter action is important for AJ formation. Consistent with this, VCL is required for the maintenance of epithelial AJs8,23,24. Meanwhile, cell biological analysis suggested that the M1 domain is masked by the M2 and M3 domains when actomyosin cannot exert pulling forces on α‑catenin that is part of the CCC18; crystallographic analyses showed a structural basis of this masking 22,25–27. A current model to explain the interactions between α‑catenin, VCL and F‑actin is therefore as follows: a mechanism triggers α‑catenin to ‘open’ its conformation, which enables VCL to bind to the M1 domain22 (FIG. 1b) . A tensile force that pulls the C‑terminal domain, which is possibly produced through its direct or indirect binding to F‑actin, might take part in this opening 18. As soon as VCL binds to α‑catenin, these two molecules would cooperate to reinforce the interaction of the CCC with F‑actin. Vasodilator-stimulated phosphoprotein (VASP) and MENA (encoded by Enah) may also interact with VCL to promote actin polymer ization in this system28. Although VCL is detectable in both linear and punctate AJs8, it is barely detectable at zones outside the ZA (that is, in lateral contacts) in epithelial cells (FIG. 3), which suggests that AJs may provide special subcellular environments to support α‑catenin–VCL–F‑actin interactions. Curiously, the neural subtype of α‑catenin, αN‑catenin, seems to be more strongly autoinhibited than the epithelial subtype of α‑catenin, αE‑catenin, when exogenously expressed in epithelial cells22. Thus, there also seem to be cell type-specific mechanisms that regulate α‑catenins. It is of note that at lateral contacts, cadherins actively move along actin filaments in certain epithelial cell lines, and this process depends on the C‑terminal domain of α‑catenin9. For AJ formation in fibroblasts, the C‑terminal domain of α‑catenin is essential29. In endothelial cells, the adhesion of which normally depends on punctate AJs30,31, VCL or the M1 domain of α‑catenin is not required for their AJ formation, although VCL and the M1 domain are required for the stability of AJs30. These observations suggest that, in certain junction types and cell types, α‑catenin uses its own C‑terminal domain to support cadherin-mediated adhesion. Overall, cell biological data support the idea that α‑catenin mediates the linkage between the CCC and F‑actin, which seems to occur in various ways. The precise molecular mechanisms that underlie the CCC–actin interactions, however, remain to be further analysed.



REVIEWS a Linear AJ (ZA)

b Punctate AJ

Cadherin

Plasma membrane Cytoplasm p120 catenin β-catenin

Afadin

PLEKHA7

Open α-catenin VCL Actin

EPLIN

Nectin ?

Myosin II Actomyosin contraction

Actomyosin contraction

c Lateral contacts

Closed α-catenin ?

Figure 3 | Hypothetical differences between various cadherin-based cell–cellNature contacts. Side views of three possible Reviews | Molecular Cell Biology forms of cadherin-based cell–cell contacts in mammalian epithelial cells are illustrated together with their known molecular components. a | Actin filaments run parallel to cell borders to form linear adherens junctions (AJs), which are also known as the zonula adherens (ZA). The dashed arrow indicates an indirect interaction between α‑catenin and afadin. b | In punctate AJs, actin filaments perpendicularly terminate at the junctions. These actin filaments pull cadherins from both sides of the junction. c | Lateral contacts are non-specialized contacts that are located below the ZA. The linear and punctate AJs are interconvertible in epithelial cells, depending on the cellular context. How actin filaments interact with α‑catenin and vinculin (VCL) in punctate AJs and lateral contacts has not yet been determined, as indicated by the questions marks. Actomyosin cables contract dependending on the kinase activity of RHO-associated protein kinase (ROCK) and generate the tension of cell junctions. The shapes of each molecule are arbitrarily drawn. EPLIN, epithelial protein lost in neoplasm; PLEKHA7, PH domain-containing family A member 7.

Conversion of AJ types. How do AJs organize into two distinct types, linear and punctate? Although multiple mechanisms affect AJ morphology, one crucial mechanism that is required to maintain the ZA is discussed. The diameter of actin bundles that are associated with the ZA (FIG. 2c) is much larger than the putative molecular dimension of the CCC. Hence, it seems likely that, in addition

to α‑catenin and/or VCL, other actin-binding proteins may hold actin filaments together at the ZA. Epithelial protein lost in neoplasm (EPLIN; also known as LIMA1), a protein that can crosslink actin filaments32, seems to be one such protein. EPLIN binds to the C‑terminal domain of αΕ‑catenin, thereby localizing along the ZA33. Remarkably, depletion of EPLIN disrupts the ZA,



REVIEWS which transforms it into punctate AJs. This indicates that EPLIN-mediated actin assembly is essential for the formation of the ZA. The resultant AJs are much more unstable and more mobile than the original ZA (Supplementary information S1 (movie)), which confirms that the linear ZA is necessary to ensure stable epithelial architecture. It has been proposed that EPLIN ‘ties up’ actin filaments, which simultaneously links them to the CCC through αE‑catenin8 (FIG. 3a). Although the principal epithelial AJ is the linear AJ (or ZA), AJs in cells that occupy the edge of epithelial colonies are converted into the punctate type at the distal sides of the cell (FIG. 2a,b). Overexpressed EPLIN does not localize to these peripheral AJs, which indicates that EPLIN recruitment to these sites is actively prohibited. When the perpendicular actin filaments targeting the AJs were ablated with a laser, the punctate AJs were immediately converted into linear AJs; this was accompanied by the simultaneous recruitment of EPLIN to the junctions8. These observations suggest that a tensile force that is exerted on punctate AJs by perpendicular actin filaments prevents EPLIN from binding to them. Mechanisms that underlie this phenomenon remain unknown, although it has been speculated that the C‑terminal domain of αΕ‑catenin might be folded in distinct ways in linear and punctate AJs that alters its affinity for EPLIN8. These perpendicular actin filaments resemble ‘purse string’ actomyosin cables, which are formed at the edges of cell sheets that are undergoing wound closure34. As EPLINfree punctate AJs are more dynamic than the ZA, this remodelling of AJs at colony peripheries might be beneficial for cells that are forming new contacts with neighbours. A modified model for tension-dependent EPLIN binding to α‑catenin has also been proposed through the analysis of endothelial cells35.

Actomyosin cables Subcellular structures that consist of accumulated actin filaments and myosin II. Sliding of myosin II motors along the actin filaments provides a force to contract the cables. These contracting cables have various roles in cell shape changes, which are dependent on where the cables are anchored.

Cadherin turnover maintains AJ plasticity AJs are also modulated by the turnover of cadherin molecules, which is elicited by endocytosis. Cadherins are recycled or degraded through various pathways, including endocytosis and proteolytic cleavage36,37. How are cadherins stabilized in the plasma membranes or subjected to endocytosis? The role of p120 catenin, which binds to the juxtamembrane domain of cadherins (FIG. 1a), in regulating cadherin stability has been studied most extensively. The removal of p120 catenin greatly increases cadherin internalization in mammalian cells38,39, although the importance of this catenin in cell adhesion is less clear in D. melanogaster and C. elegans40,41. Recent studies have started to uncover the mechanisms by which p120 catenin suppresses cadherin internalization. The p120 catenin juxtam embrane domain contains the amino acid sequence DEEGGGE, which is conserved among classic cadherins42. Mutations in the gene that encodes the juxta membrane domain block the binding of p120 catenin to cadherins, which induces their clathrin-dependent endocytosis. Analysis using VE‑cadherin revealed that the DEE portion of DEEGGGE is an endocytic signal that is necessary for cadherin internalization, which suggests that p120 catenin blocks VE‑cadherin internalization by masking this endocytic signal42. In the case of

E‑cadherin, the dileucine motif that is located distal to the core p120 catenin-binding region has a similar role43. The molecules that recognize these endocytosis‑inducing motifs, however, remain to be identified. Although cadherins need to stay within the plasma membrane to maintain cell junctions, their endocytosis is crucial for AJ plasticity. The inhibition of endocytosis blocks the redistribution of E‑cadherin into membranes44 and suppresses the formation of new E‑cadherin linkages across plasma membranes45. At the cellular level, endothelial cells that express endocytosis-resistant VE‑cadherin mutants exhibit decreased migration in wound-healing assays42. It is likely that, within migrating cell layers, cells require the progressive renewal of their contacts with neighbours. Endocytosis-dependent AJ plasticity must assist such processes. Studies using zebrafish embryos are providing results that are consistent with this hypothesis: E‑cadherin is required for epiboly, a developmental process of collective cell movement 46; blocking the endocytosis of E‑cadherin in these embryos results in cell migration defects47,48. Another study demonstrated, in E‑cadherin-deficient embryos, that cell motility is not impaired but directed and coordinated migration is abolished49. This observation suggests that an important role of cadherin-mediated contacts is to prevent cells from random movement. Thus, although cadherins are required for the coordinated migration of cells, they must be dynamically turned over to promote cell migration. Whether p120 catenin is actively involved in regulating collective cell migration remains to be investigated.

Small GTPases regulate AJ integrity The polymerization and stability of actin filaments at AJs is regulated by multiple mechanisms in which small GTPases, including the RHO GTPases RHO, RAC and CDC42, and their effectors have important roles (FIG. 4). Junctional actin nucleation seems to depend on the WAVE2 (also known as WASF2)–actin-related protein 2/3 (ARP2/3) complex, a RAC effector 50,51, although ARP2/3 also promotes the endocytosis of junction components52 (FIG. 4a). Neural Wiskott–Aldrich syndrome protein (NWASP), a CDC42 effector, is important for the stabilization of actin networks at AJs53,54. Formins, another group of actin regulators that function downstream of small GTPases, also promote junctional actin polymerization, as mentioned below. Actin filaments that are organized through these processes become coupled with non-muscle myosin IIA, myosin IIB and/or myosin IIC at the ZA55 as well as at other AJs (FIG. 3). These actomyosin filaments crucially regulate the structure and function of the ZA by having a dual role: they maintain cell junctions in stationary epithelial layers and also dynamically contract the ZA, especially during embryonic morphogenesis. In this section, the homeostatic role of actomyosin in ZA maintenance is discussed. Although actomyosin is regulated by numerous factors, the factors that are particularly important in controlling the circumferential actomyosin belt are the small GTPases RHO and repressor/activator protein 1 (RAP1; also known as TERF2IP) homologue, which belong to the RHO and RAS GTPase families, respectively. This Review, therefore, focuses on these two GTPases.



REVIEWS a

b

Plasma membrane

Lulu 2

Cytoplasm PAR3 Shroom 3

TJ

PATJ

ROCK

p114RHOGEF

GEFH1

ZO1/ZO2 AJC

p120 catenin β-catenin

ZA

Cadherin

Centralspindlin

p190B ECT2

Cross-section of actomyosin bundles

VCL

Open α-catenin

Myosin II

Cingulin

Myosin IIA

Cadherin or α-catenin DLC1 ARHGAP10

Myosin II Actin • Formins • ARP2/3 • NWASP

TEM4

c

PLEKHA7

p120 catenin

Afadin

Actin

Cadherin endocytosis

p190A

Myosin II LCs

RAP1 GEFs

d

RAP1

Myosin II

RHO GEFs • p114RHOGEF • TEM4 • ECT2 • GEFH1 • PDZRHOGEF • RhoGEF2 (Drosophila melanogaster)

RHO GAPs • p190A and p190B • DLC1 and DLC3 • ARHGAP10

Figure 4 | Pathways to regulate actomyosin attachment to the cadherin–catenin complex. a | A side view of an epithelial junction in which the vertebrate intestinal epithelium is adopted as a model. Actin filaments are shown as Nature Reviews | Molecular Cell Biology cross-sections at the apical junctional complex (AJC) zone. The desmosome is omitted from the illustration of the AJC. The positions of the molecules are arbitrarily illustrated and do not necessarily represent their actual subcellular localization. The polymerization of actin filaments is regulated by various molecules, including formins, neural Wiskott– Aldrich syndrome protein (NWASP) and actin-related protein 2/3 (ARP2/3). The contraction of actomyosin cables at the zonula adherens (ZA) is induced by RHO-associated protein kinase (ROCK), which is anchored at the AJC through Shroom 3; this process can be suppressed by zonula occludens 1 (ZO1) and ZO2. b | Guanine nucleotide exchange factors (GEFs), several of which localize at the AJC zone, regulate RHO, which in turn activates ROCK. Two pathways by which the RHO GEFs p114RHOGEF and ECT2 are recruited to AJCs are shown; additional subpathways that have been published are also included in the diagram. c | Pathways that are known to activate repressor/activator protein 1 (RAP1) and to recruit it to AJCs, where RAP1 regulates the ZA, are summarized. The afadin–PLEKHA7 (PH domain-containing family A member 7)– p120 catenin complex is thought to merge with the RAP1‑dependent myosin II-regulating pathway, as afadin can bind to RAP1. Lines indicate binding between molecules, and arrows represent activation or inhibition. The dashed line represents transient binding. d | RHO GEFs and RHO GTPase-activating proteins (GAPs) that have been detected at AJCs or shown to bind to a component of the cadherin–catenin core complex are listed. DLC1, deleted in liver cancer 1; LCs, lateral cell–cell contacts; Lulu 2, erythrocyte membrane protein band 4.1-like 4B variant; PAR3, partitioning defective 3 homologue; PATJ, PALS1‑associated tight junction; TEM4, tumour endothelial marker 4; TJ, tight junction; VCL, vinculin.

The roles of RHO and its effectors in ZA integrity. Early studies demonstrated that RHO inhibition disrupts the ZA56,57. At the ZA, RHO positively regulates at least two effectors, diaphanous homologue 1 (DIA1; also known as DIAPH1) and RHO-associated protein kinase (ROCK; also known as RHO kinase or Rok in D. melanogaster). DIA1 is one of the formin family proteins that promotes actin polymerization under the control of small RHO GTPases58. Depletion of DIA1 disrupts the ZA, whereas its overexpression increases actin accumulation at junctions59–61. Another formin, formin 1, is also important

for the maintenance of AJs62. Hence, RHO seems to sustain actin polymerization through formins at the ZA to promote ZA stability. Another RHO target, ROCK, is a Ser/Thr kinase that phosphorylates myosin light chain (MLC) and MLC phosphatase, which leads to actomyosin contraction63. In epithelial cell lines, ROCK is detectable along the ZA only at moderate levels and an atypical protein kinase C (aPKC)-dependent mechanism to suppress junctional ROCK levels has been found64. Even the tight junction proteins zonula occludens 1 (ZO1) and ZO2 suppress



REVIEWS ROCK activity at the ZA54,65. Treatment of cells with the ROCK inhibitor Y-27632 does not generally disrupt the ZA, but it does tend to induce a wavy appearance in the junction outlines66,67, which suggests that ROCK functions to produce tension in the ZA. However, by contrast, constitutive activation of ROCK disrupts linear junctions60. This phenomenon was explained by assuming that increased ROCK activity induces the contraction of actomyosin filaments that are perpendicularly attached to AJs, which enables them to pull the junctions in radial directions, as seen in peripheral AJs (FIG. 3b). Thus, ROCK could function as an inhibitor of ZA formation when it is over-activated. In normal epithelial cells, the activities of the two RHO effectors DIA1 and ROCK must be precisely balanced for ZA maintenance. RHO also seems to suppress cadherin endocytosis68. The activity of RHO GTPases is regulated by guanine nucleotide exchange factors (GEFs), and several RHO GEFs have been found to localize at the apical junctional complex zone. These include p114RHOGEF (also known as ARHGEF18)69–71, tumour endothelial marker 4 (TEM4; also known as ARHGEF17)72, ECT2 (REF. 73), GEFH1 (also known as ARHGEF2)74, and PDZRHOGEF (also known as ARHGEF11) and its D. melanogaster homologue RhoGEF2 (REFS 75,76). Removal of these GEFs perturbs ZA integrity or suppresses its contraction (see below), which suggests that multiple RHO GEFs regulate different functions of cell junctions. In the case of GEFH1, its activity is inhibited by the tight junction component cingulin at the apical junctional complex, and the inhibition but not the activation of GEFH1 seems to be important for junction formation74. RHO GTPase-activating proteins (GAPs), which can counteract RHO GEFs, were also detected at AJs or shown to bind to components of the CCC. These include the p190 RHO GAPs73,77 deleted in liver cancer 1 (DLC1), DLC3 (encoded by STARD8) (REFS 78,79) and ARHGAP10 (REF. 80). DLC proteins are important in maintaining cadherins at AJs, which suggests that the balancing of RHO activity by GEFs and GAPs is crucial for AJ integrity. Proposed pathways by which these molecules are recruited to AJs and/or activated are summarized in FIG. 4b. In summary, many RHO GEFs and RHO GAPs seem to contribute to AJ regulation, which means that the entire system is complicated. The localization and activation of each molecule must be tightly controlled in a spatiotemporal manner by upstream signals.

Guanine nucleotide exchange factors (GEFs). Proteins that facilitate the exchange of GDP for GTP in the nucleotide-binding pocket of a GTP-binding protein.

GTPase-activating proteins (GAPs). Proteins that inactivate small GTP-binding proteins, such as RAS family members, by increasing their rate of GTP hydrolysis.

The role of RAP1 and its effectors in ZA integrity. RAP1, another small GTPase, also regulates the ZA81–84. The RAP1 GEFs C3G (also known as RAPGEF1), PDZGEF2 (also known as RAPGEF6) or exchange protein directly activated by cAMP 1 (EPAC1; also known as RAPGEF3) all activate junctional RAP1. Depletion of RAP1 or PDZGEF2 increases the formation of punctate AJs85. Conversely, constitutively active RAP1 can restore the typical ZA in RAS-transformed Madin–Darby canine kidney (MDCK) cells that exhibit disorganized AJs83. Endothelial cells, although VE‑cadherin normally organizes punctate AJs in these cells30,31, can assume linear junctions when RAP1 is activated: RAP1 induces

FGD5-dependent CDC42 activation, which in turn activates myosin II through myotonic dystrophy kinaserelated CDC42‑binding kinase (MRCK)86. RAP1 was also shown to bind to and activate afadin, an adaptor protein that binds to nectins as well as the CCC87. The activated afadin interacts with p120 catenin and strengthens its binding to E‑cadherin, which results in reduced E‑cadherin endocytosis88 (FIG. 4c). The signalling relations between myosin activation and E‑cadherin stabilization by RAP1, however, remain unknown. In summary, two distinct small GTPases, RHO and RAP1, have important roles in linear AJ formation. The roles of the RHO effector ROCK are, however, complicated: ROCK functions as a tension inducer of the ZA, and it also counteracts ZA formation by activating radial actomyosin filaments. It is of note that myosin IIA and myosin IIB have differential sensitivities to RAP1 and ROCK: RAP1 seems to only regulate myosin IIB, whereas ROCK seems to selectively regulate myosin IIA67. Moreover, myosin IIA depletion severely disrupts the ZA, whereas myosin IIB loss only decreases the dynamics of ZA‑associated actin filaments, which hardly affects cadherin accumulation8,55,67. These findings suggest the possibility that the role of the RAP1–myosin IIB system is more restricted to the regulation of junctional actomyosin compared with the broader functions of the ROCK–myosin IIA system.

AJs in epithelial remodelling Epithelial folding is a major morphogenetic event during animal development. The bending of cell layers is induced by apical constriction of individual cells, and this constriction is typically induced by the contraction of actomyosin belts that line the ZA. Importantly, such modulation of the ZA occurs in polarized patterns. How these processes are controlled is reviewed below. The mechanisms that underlie apical constriction do not seem to be fully identical in vertebrate and invertebrate cells. Therefore, the vertebrate and invertebrate versions are discussed separately. It is of note that there are other categories of AJ‑dependent morphogenesis, including the apical constriction processes that do not depend on ZA contraction. Furthermore, cooperation between cell biological processes and transcription factors leads to complex epithelial remodelling, such as the rotating or tilting of tissue. These topics are also discussed. ROCK-mediated ZA contraction. In stationary epithelial cells, the impact of ROCK on ZA integrity seems to be minimal (see above). However, as soon as marked levels of ROCK are recruited to the ZA, it induces contraction of the circumferential actomyosin belt. The postsynaptic density 95, discs large and ZO1 (PDZ) domain-containing protein Shroom 3 can mediate the recruitment of ROCK to the ZA. Shroom 3 was originally identified as a key protein in neural tube formation89. Later, Shroom 3 or its isoforms were found to be important for various morphogenetic events that require apical constriction90–94. Shroom 3 is concentrated along the apical junctional complex in epithelial cells when ectopically expressed95, and it recruits ROCK to the junctions through direct



REVIEWS binding 96,97 (FIGS 4a). As a consequence, the circumferential actin belts contract, which induces apical constriction in the cell layers93,95 (FIG. 5a). A recent study showed that Shroom is also required for apical constriction in D. melanogaster embryos98. A RHO GEF, Trio, functions together with Shroom 3 to induce actomyosin contraction99. Band 4.1-like protein 5 (also known as Lulu), a four-point-one, ezrin, radixin, moesin (FERM) domain protein that has been implicated in apical constriction and actomyosin contractility 70, seems to be required for the Shroom 3–ROCK system to function properly 100. Polarized ZA contraction and epithelial bending. Epithelial sheets bend in various patterns in developing embryos. In the case of neural tube formation in mammalian or avian embryos, a layer of neuroepithelial cells bends inwards, but only along the mediolateral axis of the embryo, to form the future neural tube. How does this planar-polarized bending take place? In the bending neural plates, Shroom 3–ROCK complexes are concentrated at the ZA, and this ROCK is required for the bending of the plates97,101–103. However, this process in vivo is not as simple as that observed in cell lines. During neural plate bending, actomyosin that lines the ZA organizes into cables that extend across multiple cells, and, notably, their contraction mainly occurs towards the mediolateral axis76,97 (FIG. 5b). Further studies identified upstream signals: the atypical cadherin CELSR1 (cadherin EGF LAG seven-pass G-type receptor 1), a vertebrate homologue of D. melanogaster Flamingo that regulates planar cell polarity104, is concentrated along the AJs that are oriented towards the mediolateral axis. CELSR1 recruits Frizzled and Dishevelled, core members of planar cell polarity signalling 105, and they in turn activate PDZRHOGEF, a RHO GEF that cooperates with Dishevelled-associated activator of morphogenesis 1 (DAAM1) of the formin family. This series of signalling events eventually activates ROCK at the corresponding portions of the ZA. It has been proposed that ROCK-dependent actomyosin contraction at these sites produces a force to bend the plate in the mediolateral direction. How CELSR1 is recruited to particular zones of the ZA remains an important subject for future research.

Planar cell polarity A mechanism of cellular organization by which cells acquire information about their orientation within the tissue in the plane of the epithelium. It is distinct from apical–basal polarity.

Amnioserosa cells Cells that form the amnioserosa, an extra-embryonic epithelial sheet that covers the dorsal side of fly embryos at the blastoderm stage.

Cortical actomyosin-mediated AJ modulation in remodelling. Curiously, AJ contraction is induced by a different mechanism in invertebrate embryos. Three developmental systems — gastrulation, dorsal closure and germband extension — have been analysed in studies of epithelial apical constriction in D. melanogaster embryos. In these developmental systems, AJ shrinkage is mainly controlled by actomyosin networks that are located at the apical cortex rather than by self-contraction of AJs, although RHO GEF–ROCK signals are commonly involved in AJ shrinkage in D. melanogaster and vertebrate cells75,106. During D. melanogaster gastrulation, the extra cellular ligand Folded gastrulation (Fog) initiates apical constriction in future mesodermal cells in conjunction with RhoGEF2 and other cooperators75,107. It has been shown that this apical constriction exhibits pulsation108.

These pulses are generated by the ‘myosin coalescence’ that is associated with actin networks that are located at the medial portion of the apical cortex of the cells. These cortical actomyosin networks attach to the edges of the ZA and pull them inwards, which leads to the eventual constriction of the entire ZA (FIG. 5c). The PDZcontaining GEF protein Dizzy and its downstream GTPase Rap1 are required to complete these apical constriction processes109. Similarly, in amnioserosa cells, during dorsal closure, cortical actomyosin foci induce local displacement of AJs towards themselves110–112. In C. ele gans endodermal precursor cells that are undergoing apical constriction, the cortical actomyosin contraction begins independently of junctional contraction, and later the actomyosin contraction coordinates with the inwards movement of the junctions113. These observations suggest the presence of a molecular ‘clutch’ to connect cortical actomyosin filaments with AJs, which functions only at later stages of apical constriction. The CCC is thought to be a candidate for the clutch, as removal of HMR‑1, a C. elegans classic cadherin, impairs the coupled junctional constriction without affecting actomyosin contraction. Canoe, a D. melanogaster homologue of afadin87, also functions to maintain the interaction between actomyosin coalescences and AJs114,115. Canoe and afadin are able to bind to F‑actin and the CCC114,116. Thus, Canoe may also be involved in the putative clutch machinery. Overall, in these apical constriction systems, shrinking of AJs is led by cortical actomyosin. Germband elongation is governed by the intercalation of cells, the junctions of which have planar polarity and are enriched in myosin II and Bazooka (known as PAR3 in humans) at the cell edges, which are vertically (dorsal–ventral) and transversely (anterior–posterior) oriented, respectively 117,118. The differential distribution of these molecules is regulated by Rok119. The vertical myosin II-bearing junctions shrink and then extend horizontally through a process that involves the transient rosette-like rearrangement of cells120, and each component of the junctions is required for normal convergent extension119,121,122. In this cellular system, cortical actomyosin coalescences were also detected123. Intriguingly, these coalescences flow towards a vertical junction and then induce shrinkage of the junction (FIG. 5d). Further studies revealed that levels of DE‑cadherin fluctuate among the vertical junctions in a cell, owing to asymmetrical endocytosis of this molecule, and the acto myosin flow is oriented towards the junctions with higher levels of DE‑cadherin124. The authors of this study proposed that the actomyosin flow emerges from a mechanical imbalance that is caused by transient DE‑cadherin asymmetries at cell junctions and that this process oscillates along the anterior–posterior axis to ensure that all vertical junctions shrink in a stepwise manner. The asymmetrical endocytosis of DE‑cadherin is regulated by myosin II and the formin Dia125. In vertical junctions, β-catenin turnover is also increased in an Abl Tyr kinase-dependent manner 126, which suggests that both cadherin and β-catenin must be turned over to ensure AJ shrinkage. A recent review discusses the roles of such AJ contraction patterns in cell intercalation127.



REVIEWS a Apical constriction

b Bending of the neural plate Actin belt

Contracting AJs Actin ﬁlament Myosin II ZA

AJ contraction

AJ sliding

c Invaginating mesoderm

CELSR1

Bending direction (mediolateral axis)

ZA

d Extending germband epithelium

Contraction Inward pulling of membrane

ZA

DE-cadherin endocytosis

Potential linkage between cadherin and F-actin Extension direction (anterior–posterior axis) Nature Reviews | Molecular Cell Biology Dorsal folds Epithelial structures that form on the dorsal side of the gastrulating Drosophila melanogaster embryo. Dorsal epithelial sheets are folded at an anterior and posterior portion of the embryo during development. The anterior fold is shallower than the posterior fold.

Epithelial folding due to AJ sliding. During D. melano gaster gastrulation, the anterior and posterior dorsal folds form on the dorsal side of the embryo. Unlike during the ordinary process of epithelial folding, the levels of apical myosin remain low throughout the generation of dorsal folds, which indicates that myosin contractility is not a major regulator of this process. Before the initiation of dorsal folds, AJs shift basally in the cells that will form the folds but maintain their original positioning in the neighbouring cells128 (FIG. 5a). This basal shift of AJs

is induced by a decrease in the ratio of Par1, a micro tubule affinity regulating kinase (MARK) family kinase, to Bazooka in the lateral cell membranes. The authors of this study proposed that the basal shift of junctions not only alters the apical shape of the initiating cells but also forces the lateral membranes of the adjacent cells to bend towards the initiating cells, thereby facilitating tissue deformation. Further analysis of dorsal fold formation explained why the posterior fold invaginates more extensively than the anterior fold129. Specifically, α‑Catenin



REVIEWS ◀ Figure 5 | Various forms of adherens junction modulation induce epithelial

reshaping. a | Apical constriction can be induced by the contraction of linear adherens junctions (AJs) (bottom left) that are known as the zonula adherens (ZA), which contains a circumferential actin belt, or by its downwards sliding (bottom right). b | Polarized contraction of the ZA occurs during the bending of the neural plate in chicken embryos. An apical junction pattern is shown, viewed from the apical side of the neural plate. Actomyosin organizes into cables across multiple cells and contracts towards the mediolateral axis of the embryo. Although actin filaments are detectable throughout the junctions, non-contracting filaments are omitted from the illustration. CELSR1 (cadherin EGF LAG seven-pass G-type receptor 1), a member of the cadherin superfamily, localizes at portions of the ZA that are oriented towards the mediolateral axis of the bending neural plate. It upregulates RHO-associated protein kinase (ROCK) at these sites through a signalling pathway that comprises Dishevelled, Dishevelled-associated activator of morphogenesis 1 (DAAM1) and PDZRHOGEF, which leads to the contraction of actomyosin76. c | Cortical actomyosin-dependent contraction of junctions occurs in mesodermal cells during gastrulation (that is, the invagination of mesoderm) of Drosophila melanogaster embryos. Pulsing actomyosin coalescences move towards cell edges and pull them. A schematic of a sectioned embryo undergoing gastrulation is depicted, in which the direction for viewing the apical profile of cells is shown. This is expanded to show an apical view of a cell in which actomyosin networks are drawn. d | Planar-polarized contraction of cell edges occurs during germband extension. Actomyosin coalescences move towards a junction to induce contraction. The movement is controlled by the endocytosis of D. melanogaster epithelial cadherin (DE‑cadherin). A cell pattern in the germband that undergoes convergent extension is shown. An area is expanded to give an apical view of cells in which actomyosin networks are drawn. The diagrams in part c and part d were drawn on the basis of findings that are published in REFS 108,123,124.

knockdown leads to a deeper invagination of the anterior fold. By contrast, expression of a constitutively active Rap1 suppresses invagination of both dorsal folds in an α‑Catenin-dependent manner, which suggests that Rap1 stabilizes the junctions through a mechanism that involves α‑Catenin. RapGAP1, a modulator of Rap1 activity, is expressed at higher levels in cells that flank the posterior fold than in those of the anterior fold. The resulting model thus proposes that distinct activities of Rap1 modulate α‑Catenin-dependent coupling between junctions and actin to control the extent of epithelial invagination. Whether other epithelial-folding phenomena also adopt these mechanisms remains to be investigated. Epithelial tubule rotation by asymmetrical AJ modulation. AJ modulation regulates other types of epithelial patterning. Developing guts rotate in a left–right asymmetrical manner to fit into a limited body space. Studies that use D. melanogaster hindgut epithelium suggested that the fly homologue of myosin ID controls the endo cytosis of DE‑cadherin in a planar-polarized manner, which creates asymmetrical tension on individual junctions and chirality of planar cells130. Computer simulation supports the idea that this chirality induces the rotation of the gut. This type of morphogenesis can be considered as a modified version of planar-polarized AJ contraction‑dependent rearrangement of cells. In vertebrates, gut rotation is driven by the asymmetrical thickening of epithelial layers and the condensation of mesenchymal cells at the left side of the dorsal mesentery. This process requires N‑cadherin, the expression of which is induced by pituitary homeobox 2 (PITX2), a central transcription factor that determines the left– right asymmetry of the body 131,132, and it also requires

Shroom 3 (REF. 133). A recent study showed that DAAM2, a member of the formin family, is only expressed at the left side of the dorsal mesentery under the control of PITX2. DAAM2 strengthens cell–cell adhesion through its interaction with the N‑cadherin–α‑catenin complex 134, which induces asymmetrical mesenchymal condensation. This is another example of how the local modulation of AJs and their associated actomyosin regulators can contribute to global tissue patterning.

Conclusions and perspectives AJs are important not only for the formation of mechanical connections between cells but also for the dynamic regulation of these connections. A key feature of AJs and classic cadherins is their structural and functional linkage with actin or actomyosin bundles. Among the numerous cadherin superfamily members, only the classic cadherins exhibit this property, owing to their ability to bind to the β‑catenin–α‑catenin complex 3,4. This unique feature of AJs is therefore based on their linkage with actomyosin bundles, which can be controlled by various mechanisms. Small RHO GTPases and their effectors regulate not only actin polymerization but also actomyosin contractility at AJs. ROCK is particularly important for inducing the robust contraction of circumferential actomyosin belts, which leads to the apical constriction of epithelial cells. RHO GEFs and RHO GAPs regulate the activity of ROCK, although how these regulators are recruited to AJs is only partly understood. A recent study demonstrated that the lateral cell contacts below the ZA also show a limited level of cadherin- and actomyosin-dependent contractility, and the difference between the ZA and lateral contacts is generated by NWASP, which is apically located135. This novel feature of junctional contractility needs to be kept in mind during the future analysis of AJ‑dependent epithelial reshaping. The definition of AJ also needs further clarification, considering the structural and molecular diversity of cadherin-mediated cell contacts. Molecular mechanisms by which the CCC is anchored to actin filaments are still a topic of debate. Many cell biological phenomena can be explained most easily by assuming physical linkages between the CCC and F‑actin, despite the failure to detect such linkages in vitro15. It is very likely that direct linkages between F‑actin and α‑catenin (in the form of the CCC) through its C‑terminal domain or associated VCL occur in vivo, at least in specific subcellular environments; it is therefore worth investigating the in vitro conditions that can induce these linkages. As α‑catenin seems to be autoinhibited, it is especially important to identify the conditions that can unfold its conformation to mimic the processes that occur in vivo. Such experiments could lead to the reconstitution of AJs in vitro and enable us to learn more about the detailed mechanisms of molecular interactions that occur within the AJ. VCL has been highlighted as a mediator of CCC– actin linkages, and it is very likely to be involved in the formation of the ZA. However, Vinc is not essential for D. melanogaster development 136. Even in vertebrates, Vlc-knockout mice survive longer than α‑catenin-null mice137,138, although Vlc-knockout mice show various histological defects. Hence, it is important to further



REVIEWS Box 1 | Interaction of adherens junctions with microtubules Microtubules interact with adherens junctions (AJs) through their minus or plus ends. The cadherin–catenin core complex binds to PLEKHA7 (PH domain-containing family A member 7) through p120 catenin, and PLEKHA7 in turn binds to CAMSAP3 (calmodulin-regulated spectrin-associated protein 3; also known as Nezha) (see the figure). CAMSAP3 interacts with the minus end of non-centrosmal microtubules — a population of microtubules that is not anchored to the centrosome141. It is also thought that microtubules interact with AJs through their plus ends; this interaction is dependent on the plus end-tracking protein end-binding 1 (EB1; also known as MAPRE1) and dynein motors143–145. Microubules that interact with AJs through their plus ends are supposed to stabilize AJs140,141, possibly by promoting the kinesin-mediated delivery of AJ components or their regulators to the AJs. The actual roles of AJ‑associated microtubules, however, remain to be elucidated.

elucidate the VCL-independent parts of the mechanisms that regulate the CCC–actin interactions. The idea that actomyosin-dependent AJ contraction is a mechanism of epithelial remodelling has been established. However, an intriguing difference in actomyosinmediated mechanisms has been observed between vertebrates and invertebrates. In invertebrates, the actomyosin networks that are located at the apical cortex control AJ contraction, whereas such mechanisms have not been reported for vertebrate cells. I consider two possible reasons for this difference. First, vertebrate cells might not use cortical actomyosin, as different mechanisms may have evolved between vertebrates and invertebrates. Second, vertebrate cells may use similar mechanisms to invertebrates that have been merely missed in past observations. A problem with vertebrate studies is that most use cell lines that may not retain all of the original in vivo properties. High-resolution live images of AJ components need to be collected using tissues to obtain more information about junction dynamics in vertebrate cells. In addition, it should be stressed that the contraction of junctions themselves is also involved in the apical constriction of D. melanogaster cells. It is thus possible that the junctional and cortical actomyosin contractions in principle need to cooperate, but different animals or cell types use the two processes in different proportions, depending on the morphogenetic context. Lecuit, T. Adhesion remodeling underlying tissue morphogenesis. Trends Cell Biol. 15, 34–42 (2005). Baum, B. & Georgiou, M. Dynamics of adherens junctions in epithelial establishment, maintenance, and remodeling. J. Cell Biol. 192, 907–917 (2011). 3. Oda, H. & Takeichi, M. Evolution: structural and functional diversity of cadherin at the adherens junction. J. Cell Biol. 193, 1137–1146 (2011). 4. Hirano, S. & Takeichi, M. Cadherins in brain morphogenesis and wiring. Physiol. Rev. 92, 597–634 (2012). 5. Lynch, A. M. et al. A genome-wide functional screen shows MAGI 1 is an L1CAM dependent stabilizer of apical junctions in C. elegans. Curr. Biol. 22, 1891–1899 (2012). 6. Truong Quang, B. A., Mani, M., Markova, O., Lecuit, T. & Lenne, P. F. Principles of E-adherin supramolecular organization in vivo. Curr. Biol. 23, 2197–2207 (2013).

Cadherin

Plasma membrane Cytoplasm p120 catenin β-catenin EB1 +

7.

2.

8.

CAMSAP3 –

Microtubule –

+

We have learned that the endocytosis of cadherins Nature Reviews | Molecular Cell Biology cooperates with actomyosin contractility to induce AJ shrinkage during development. The cooperation of endocytosis and actomyosin contractility could operate in many other systems that involve epithelial remodelling. It will be particularly interesting to test whether p120 catenin is involved in these processes, as it has thus far only been studied with regard to its role in maintaining AJ integrity. Not surprisingly, AJs are regulated by transcription factors: this is particularly evident during the processes that involve cell fate changes. Notably, it seems that cell biological processes themselves are controlled by transcription factors, such as Twist and PITX2, as observed during gastrulation of D. melanogaster embryos139 and gut coiling in vertebrates134, respectively. Collecting further examples of such cases will be useful for determining the combinatory actions of genetic and cell biological elements. In this Review, I have not discussed the potential roles of microtubules in AJ regulation. Microtubules interact with AJs in various ways140–145, and this cyto skeletal system is apparently important for AJ dynamics (BOX 1). Elucidating the mechanisms of AJ dynamics by which the two major cytoskeletal elements, actin and microtubules, interact would probably paint a more detailed picture of AJ regulation.

Farquhar, M. G. & Palade, G. E. Junctional complexes in various epithelia. J. Cell Biol. 17, 375–412 (1963). Taguchi, K., Ishiuchi, T. & Takeichi, M. Mechanosensitive EPLIN-dependent remodeling of adherens junctions regulates epithelial reshaping. J. Cell Biol. 194, 643–656 (2011). Shows that epithelial linear junctions are converted into a punctate type at colony peripheries, where radial actin filaments prevent EPLIN from binding to AJs. 9. Kametani, Y. & Takeichi, M. Basal to apical cadherin flow at cell junctions. Nature Cell Biol. 9, 92–98 (2007). 10. Brasch, J., Harrison, O. J., Honig, B. & Shapiro, L. Thinking outside the cell: how cadherins drive adhesion. Trends Cell Biol. 22, 299–310 (2012). 11. Hirano, S., Kimoto, N., Shimoyama, Y., Hirohashi, S. & Takeichi, M. Identification of a neural α-catenin as a key

1.

PLEKHA7


regulator of cadherin function and multicellular organization. Cell 70, 293–301 (1992). 12. Watabe-Uchida, M. et al. α-Catenin-vinculin interaction functions to organize the apical junctional complex in epithelial cells. J. Cell Biol. 142, 847–857 (1998). 13. Rimm, D. L., Koslov, E. R., Kebriaei, P., Cianci, C. D. & Morrow, J. S. α1(E)-catenin is an actin-binding and -bundling protein mediating the attachment of F actin to the membrane adhesion complex. Proc. Natl Acad. Sci. USA 92, 8813–8817 (1995). 14. Hansen, S. D. et al. αE-catenin actin-binding domain alters actin filament conformation and regulates binding of nucleation and disassembly factors. Mol. Biol. Cell 24, 3710–3720 (2013). 15. Yamada, S., Pokutta, S., Drees, F., Weis, W. I. & Nelson, W. J. Deconstructing the cadherin–catenin– actin complex. Cell 123, 889–901 (2005).


REVIEWS Reports the unexpected finding that, despite the well-known ability of α-catenin to bind to F-actin, the cadherin–α-catenin complex does not do so. 16. Drees, F., Pokutta, S., Yamada, S., Nelson, W. J. & Weis, W. I. α-catenin is a molecular switch that binds E cadherin-β-catenin and regulates actin-filament assembly. Cell 123, 903–915 (2005). 17. Kwiatkowski, A. V. et al. In vitro and in vivo reconstitution of the cadherin–catenin–actin complex from Caenorhabditis elegans. Proc. Natl Acad. Sci. USA 107, 14591–14596 (2010). 18. Yonemura, S. Wada, Y., Watanabe, T., Nagafuchi, A. & Shibata, M. α-Catenin as a tension transducer that induces adherens junction development. Nature Cell Biol. 12, 533–542 (2010). First to propose that the binding of VCL to α-catenin is inhibited by self-folding of α-catenin, and this autoinhibited state is reversed by a force that pulls the α-catenin C-terminal domain. 19. Desai, R. et al. Monomeric α-catenin links cadherin to the actin cytoskeleton. Nature Cell Biol. 15, 261–273 (2013). 20. 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). 21. Twiss, F. et al. Vinculin-dependent cadherin mechanosensing regulates efficient epithelial barrier formation. Biol. Open 1, 1128–1140 (2012). 22. Ishiyama, N. et al. An autoinhibited structure of α-catenin and its implications for vinculin recruitment to adherens junctions. J. Biol. Chem. 288, 15913–15925 (2013). 23. Maddugoda, M. P., Crampton, M. S., Shewan, A. M. & Yap, A. S. Myosin VI and vinculin cooperate during the morphogenesis of cadherin cell cell contacts in mammalian epithelial cells. J. Cell Biol. 178, 529–540 (2007). 24. Peng, X., Cuff, L. E., Lawton, C. D. & DeMali, K. A. Vinculin regulates cell-surface E cadherin expression by binding to β-catenin. J. Cell Sci. 123, 567–577 (2010). 25. Rangarajan, E. S. & Izard, T. Dimer asymmetry defines α-catenin interactions. Nature Struct. Mol. Biol. 20, 188–193 (2013). 26. Choi, H. J. et al. αE catenin is an autoinhibited molecule that coactivates vinculin. Proc. Natl Acad. Sci. USA 109, 8576–8581 (2012). 27. Rangarajan, E. S. & Izard, T. The cytoskeletal protein α-catenin unfurls upon binding to vinculin. J. Biol. Chem. 287, 18492–18499 (2012). 28. Vasioukhin, V., Bauer, C., Yin, M. & Fuchs, E. Directed actin polymerization is the driving force for epithelial cell–cell adhesion. Cell 100, 209–219 (2000). 29. Imamura, Y., Itoh, M., Maeno, Y., Tsukita, S. & Nagafuchi, A. Functional domains of α-catenin required for the strong state of cadherin-based cell adhesion. J. Cell Biol. 144, 1311–1322 (1999). 30. Huveneers, S. et al. Vinculin associates with endothelial VE cadherin junctions to control force-dependent remodeling. J. Cell Biol. 196, 641–652 (2012). 31. Millan, J. et al. Adherens junctions connect stress fibres between adjacent endothelial cells. BMC Biol. 8, 11 (2010). 32. Maul, R. S. et al. EPLIN regulates actin dynamics by cross-linking and stabilizing filaments. J. Cell Biol. 160, 399–407 (2003). 33. Abe, K. & Takeichi, M. EPLIN mediates linkage of the cadherin–catenin complex to F actin and stabilizes the circumferential actin belt. Proc. Natl Acad. Sci. USA 105, 13–19 (2008). 34. Tamada, M., Perez, T. D., Nelson, W. J. & Sheetz, M. P. Two distinct modes of myosin assembly and dynamics during epithelial wound closure. J. Cell Biol. 176, 27–33 (2007). 35. Chervin-Petinot, A. et al. Epithelial protein lost in neoplasm (EPLIN) interacts with α-catenin and actin filaments in endothelial cells and stabilizes vascular capillary network in vitro. J. Biol. Chem. 287, 7556–7572 (2012). 36. Palacios, F., Schweitzer, J. K., Boshans, R. L. & D’SouzaSchorey, C. ARF6 GTP recruits Nm23 H1 to facilitate dynamin-mediated endocytosis during adherens junctions disassembly. Nature Cell Biol. 4, 929–936 (2002). 37. Delva, E. & Kowalczyk, A. P. Regulation of cadherin trafficking. Traffic 10, 259–267 (2009). 38. Davis, M. A., Ireton, R. C. & Reynolds, A. B. A core function for p120 catenin in cadherin turnover. J. Cell Biol. 163, 525–534 (2003). 39. Xiao, K. et al. Cellular levels of p120 catenin function as a set point for cadherin expression levels in

microvascular endothelial cells. J. Cell Biol. 163, 535–545 (2003). 40. Myster, S. H., Cavallo, R., Anderson, C. T., Fox, D. T. & Peifer, M. Drosophila p120catenin plays a supporting role in cell adhesion but is not an essential adherens junction component. J. Cell Biol. 160, 433–449 (2003). 41. Pettitt, J., Cox, E. A., Broadbent, I. D., Flett, A. & Hardin, J. The Caenorhabditis elegans p120 catenin homologue, JAC 1, modulates cadherin-catenin function during epidermal morphogenesis. J. Cell Biol. 162, 15–22 (2003). 42. Nanes, B. A. et al. p120 catenin binding masks an endocytic signal conserved in classical cadherins. J. Cell Biol. 199, 365–380 (2012). Dissects the core p120 catenin-binding region of VE-cadherin and identifies a specific sequence that is masked by this catenin to prevent cadherin internalization. 43. Miyashita, Y. & Ozawa, M. Increased internalization of p120 uncoupled E cadherin and a requirement for a dileucine motif in the cytoplasmic domain for endocytosis of the protein. J. Biol. Chem. 282, 11540–11548 (2007). 44. de Beco, S., Gueudry, C., Amblard, F. & Coscoy, S. Endocytosis is required for E cadherin redistribution at mature adherens junctions. Proc. Natl Acad. Sci. USA 106, 7010–7015 (2009). 45. Troyanovsky, R. B., Sokolov, E. P. & Troyanovsky, S. M. Endocytosis of cadherin from intracellular junctions is the driving force for cadherin adhesive dimer disassembly. Mol. Biol. Cell 17, 3484–3493 (2006). 46. Kane, D. A., McFarland, K. N. & Warga, R. M. Mutations in half baked/E cadherin block cell behaviors that are necessary for teleost epiboly. Development 132, 1105–1116 (2005). 47. Song, S. et al. Pou5f1‑dependent EGF expression controls E cadherin endocytosis, cell adhesion, and zebrafish epiboly movements. Dev. Cell 24, 486–501 (2013). 48. Ulrich, F. et al. Wnt11 functions in gastrulation by controlling cell cohesion through Rab5c and E cadherin. Dev. Cell 9, 555–564 (2005). 49. Arboleda-Estudillo, Y. et al. Movement directionality in collective migration of germ layer progenitors. Curr. Biol. 20, 161–169 (2010). 50. Tang, V. W. & Brieher, W. M. α-Actinin 4/FSGS1 is required for Arp2/3 dependent actin assembly at the adherens junction. J. Cell Biol. 196, 115–130 (2012). 51. Verma, S. et al. A WAVE2 Arp2/3 actin nucleator apparatus supports junctional tension at the epithelial zonula adherens. Mol. Biol. Cell 23, 4601–4610 (2012). 52. Georgiou, M., Marinari, E., Burden, J. & Baum, B. Cdc42, Par6, and aPKC regulate Arp2/3 mediated endocytosis to control local adherens junction stability. Curr. Biol. 18, 1631–1638 (2008). 53. Kovacs, E. M. et al. N-WASP regulates the epithelial junctional actin cytoskeleton through a non-canonical post-nucleation pathway. Nature Cell Biol. 13, 934–943 (2011). 54. Otani, T., Ichii, T., Aono, S. & Takeichi, M. Cdc42 GEF Tuba regulates the junctional configuration of simple epithelial cells. J. Cell Biol. 175, 135–146 (2006). 55. Ivanov, A. I. et al. A unique role for nonmuscle myosin heavy chain IIa in regulation of epithelial apical junctions. PLoS ONE 2, e658 (2007). 56. McCormack, J., Welsh, N. J. & Braga, V. M. Cycling around cell–cell adhesion with Rho GTPase regulators. J. Cell Sci. 126, 379–391 (2013). 57. Citi, S., Spadaro, D., Schneider, Y., Stutz, J. & Pulimeno, P. Regulation of small GTPases at epithelial cell–cell junctions. Mol. Membr. Biol. 28, 427–444 (2011). 58. Goode, B. L. & Eck, M. J. Mechanism and function of formins in the control of actin assembly. Annu. Rev. Biochem. 76, 593–627 (2007). 59. Carramusa, L., Ballestrem, C., Zilberman, Y. & Bershadsky, A. D. Mammalian diaphanous-related formin Dia1 controls the organization of E cadherinmediated cell–cell junctions. J. Cell Sci. 120, 3870–3882 (2007). 60. Sahai, E. & Marshall, C. J. ROCK and Dia have opposing effects on adherens junctions downstream of Rho. Nature Cell Biol. 4, 408–415 (2002). 61. Homem, C. C. & Peifer, M. Diaphanous regulates myosin and adherens junctions to control cell contractility and protrusive behavior during morphogenesis. Development 135, 1005–1018 (2008).


62. Kobielak, A., Pasolli, H. A. & Fuchs, E. Mammalian formin 1 participates in adherens junctions and polymerization of linear actin cables. Nature Cell Biol. 6, 21–30 (2004). 63. Riento, K. & Ridley, A. J. Rocks: multifunctional kinases in cell behaviour. Nature Rev. Mol. Cell Biol. 4, 446–456 (2003). 64. Ishiuchi, T. & Takeichi, M. Willin and Par3 cooperatively regulate epithelial apical constriction through aPKCmediated ROCK phosphorylation. Nature Cell Biol. 13, 860–866 (2011). 65. Fanning, A. S., Van Itallie, C. M. & Anderson, J. M. Zonula occludens 1 and -2 regulate apical cell structure and the zonula adherens cytoskeleton in polarized epithelia. Mol. Biol. Cell 23, 577–590 (2012). 66. Ayollo, D. V., Zhitnyak, I. Y., Vasiliev, J. M. & Gloushankova, N. A. Rearrangements of the actin cytoskeleton and E cadherin-based adherens junctions caused by neoplasic transformation change cell–cell interactions. PLoS ONE 4, e8027 (2009). 67. Smutny, M. et al. Myosin II isoforms identify distinct functional modules that support integrity of the epithelial zonula adherens. Nature Cell Biol. 12, 696–702 (2010). 68. Warner, S. J. & Longmore, G. D. Distinct functions for Rho1 in maintaining adherens junctions and apical tension in remodeling epithelia. J. Cell Biol. 185, 1111–1125 (2009). 69. Herder, C. et al. ArhGEF18 regulates RhoA Rock2 signaling to maintain neuro-epithelial apico–basal polarity and proliferation. Development 140, 2787–2797 (2013). 70. Nakajima, H. & Tanoue, T. Lulu2 regulates the circumferential actomyosin tensile system in epithelial cells through p114RhoGEF. J. Cell Biol. 195, 245–261 (2011). 71. Terry, S. J. et al. Spatially restricted activation of RhoA signalling at epithelial junctions by p114RhoGEF drives junction formation and morphogenesis. Nature Cell Biol. 13, 159–166 (2011). 72. Ngok, S. P. et al. TEM4 is a junctional Rho GEF required for cell–cell adhesion, monolayer integrity and barrier function. J. Cell Sci. 126, 3271–3277 (2013). 73. Ratheesh, A. et al. Centralspindlin and α-catenin regulate Rho signalling at the epithelial zonula adherens. Nature Cell Biol. 14, 818–828 (2012). Describes an unforeseen pathway that recruits a RHO GEF to cell junctions; this pathway includes proteins that are involved in cytokinesis. 74. Aijaz, S., D’Atri, F., Citi, S., Balda, M. S. & Matter, K. Binding of GEF H1 to the tight junction-associated adaptor cingulin results in inhibition of Rho signaling and G1/S phase transition. Dev. Cell 8, 777–786 (2005). 75. Kolsch, V., Seher, T., Fernandez-Ballester, G. J., Serrano, L. & Leptin, M. Control of Drosophila gastrulation by apical localization of adherens junctions and RhoGEF2. Science 315, 384–386 (2007). 76. Nishimura, T., Honda, H. & Takeichi, M. Planar cell polarity links axes of spatial dynamics in neural-tube closure. Cell 149, 1084–1097 (2012). Demonstrates that, in the bending neural plate, the cadherin CELSR1 induces contraction of AJ-associated actomyosin along the mediolateral axis, which leads to the polarized bending of the plate. 77. Zebda, N. et al. Interaction of p190RhoGAP with C terminal domain of p120 catenin modulates endothelial cytoskeleton and permeability. J. Biol. Chem. 288, 18290–18299 (2013). 78. Holeiter, G. et al. The RhoGAP protein Deleted in Liver Cancer 3 (DLC3) is essential for adherens junctions integrity. Oncogenesis 1, e13 (2012). 79. Tripathi, V., Popescu, N. C. & Zimonjic, D. B. DLC1 interaction with α-catenin stabilizes adherens junctions and enhances DLC1 antioncogenic activity. Mol. Cell. Biol. 32, 2145–2159 (2012). 80. Sousa, S. et al. ARHGAP10 is necessary for α-catenin recruitment at adherens junctions and for Listeria invasion. Nature Cell Biol. 7, 954–960 (2005). 81. Kooistra, M. R., Dube, N. & Bos, J. L. Rap1: a key regulator in cell–cell junction formation. J. Cell Sci. 120, 17–22 (2007). 82. Hogan, C. et al. Rap1 regulates the formation of E cadherin-based cell–cell contacts. Mol. Cell. Biol. 24, 6690–6700 (2004). 83. Price, L. S. et al. Rap1 regulates E cadherin-mediated cell–cell adhesion. J. Biol. Chem. 279, 35127–35132 (2004).


REVIEWS 84. Knox, A. L. & Brown, N. H. Rap1 GTPase regulation of adherens junction positioning and cell adhesion. Science 295, 1285–1288 (2002). 85. Dube, N. et al. The RapGEF PDZ GEF2 is required for maturation of cell–cell junctions. Cell. Signal. 20, 1608–1615 (2008). 86. Ando, K. et al. Rap1 potentiates endothelial cell junctions by spatially controlling myosin II activity and actin organization. J. Cell Biol. 202, 901–916 (2013). 87. 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). 88. Hoshino, T. et al. Regulation of E cadherin endocytosis by nectin through afadin, Rap1, and p120ctn. J. Biol. Chem. 280, 24095–24103 (2005). 89. Hildebrand, J. D. & Soriano, P. Shroom, a PDZ domaincontaining actin-binding protein, is required for neural tube morphogenesis in mice. Cell 99, 485–497 (1999). 90. Ernst, S. et al. Shroom3 is required downstream of FGF signalling to mediate proneuromast assembly in zebrafish. Development 139, 4571–4581 (2012). 91. Chung, M. I., Nascone-Yoder, N. M., Grover, S. A., Drysdale, T. A. & Wallingford, J. B. Direct activation of Shroom3 transcription by Pitx proteins drives epithelial morphogenesis in the developing gut. Development 137, 1339–1349 (2010). 92. Haigo, S. L., Hildebrand, J. D., Harland, R. M. & Wallingford, J. B. Shroom induces apical constriction and is required for hingepoint formation during neural tube closure. Curr. Biol. 13, 2125–2137 (2003). 93. Bolinger, C., Zasadil, L., Rizaldy, R. & Hildebrand, J. D. Specific isoforms of drosophila shroom define spatial requirements for the induction of apical constriction. Dev. Dyn. 239, 2078–2093 (2010). 94. Plageman, T. F. et al. Pax6 dependent Shroom3 expression regulates apical constriction during lens placode invagination. Development 137, 405–415 (2010). 95. Hildebrand, J. D. Shroom regulates epithelial cell shape via the apical positioning of an actomyosin network. J. Cell Sci. 118, 5191–5203 (2005). 96. Mohan, S. et al. Structure of Shroom domain 2 reveals a three-segmented coiled-coil required for dimerization, Rock binding, and apical constriction. Mol. Biol. Cell 23, 2131–2142 (2012). 97. Nishimura, T. & Takeichi, M. Shroom3‑mediated recruitment of Rho kinases to the apical cell junctions regulates epithelial and neuroepithelial planar remodeling. Development 135, 1493–1502 (2008). 98. Simoes Sde, M., Mainieri, A. & Zallen, J. A. Rho GTPase and Shroom direct planar polarized actomyosin contractility during convergent extension. J. Cell Biol. 204, 575–589 (2014). 99. Plageman, T. F. Jr et al. A Trio-RhoA Shroom3 pathway is required for apical constriction and epithelial invagination. Development 138, 5177–5188 (2011). 100. Chu, C. W., Gerstenzang, E., Ossipova, O. & Sokol, S. Y. Lulu regulates Shroom-induced apical constriction during neural tube closure. PLoS ONE 8, e81854 (2013). 101. Wei, L. et al. Rho kinases play an obligatory role in vertebrate embryonic organogenesis. Development 128, 2953–2962 (2001). 102. Ybot-Gonzalez, P. et al. Convergent extension, planarcell-polarity signalling and initiation of mouse neural tube closure. Development 134, 789–799 (2007). 103. Kinoshita, N., Sasai, N., Misaki, K. & Yonemura, S. Apical accumulation of Rho in the neural plate is important for neural plate cell shape change and neural tube formation. Mol. Biol. Cell 19, 2289–2299 (2008). 104. Usui, T. et al. Flamingo, a seven-pass transmembrane cadherin, regulates planar cell polarity under the control of Frizzled. Cell 98, 585–595 (1999). 105. Gray, R. S., Roszko, I. & Solnica-Krezel, L. Planar cell polarity: coordinating morphogenetic cell behaviors with embryonic polarity. Dev. Cell 21, 120–133 (2011). 106. Dawes-Hoang, R. E. et al. folded gastrulation, cell shape change and the control of myosin localization. Development 132, 4165–4178 (2005). 107. Barrett, K., Leptin, M. & Settleman, J. The Rho GTPase and a putative RhoGEF mediate a signaling pathway for the cell shape changes in Drosophila gastrulation. Cell 91, 905–915 (1997). 108. Martin, A. C., Kaschube, M. & Wieschaus, E. F. Pulsed contractions of an actin-myosin network drive apical constriction. Nature 457, 495–499 (2009). Reports that the actomyosin networks at the apical cortex of cells display pulsed contractions, which eventually induce apical constriction of the cells owing to a linkage between cortical actomyosins and cell edges.

109. Spahn, P., Ott, A. & Reuter, R. The PDZ-GEF protein Dizzy regulates the establishment of adherens junctions required for ventral furrow formation in Drosophila. J. Cell Sci. 125, 3801–3812 (2012). 110. Solon, J., Kaya-Copur, A., Colombelli, J. & Brunner, D. Pulsed forces timed by a ratchet-like mechanism drive directed tissue movement during dorsal closure. Cell 137, 1331–1342 (2009). 111. Blanchard, G. B., Murugesu, S., Adams, R. J., MartinezArias, A. & Gorfinkiel, N. Cytoskeletal dynamics and supracellular organisation of cell shape fluctuations during dorsal closure. Development 137, 2743–2752 (2010). 112. David, D. J., Tishkina, A. & Harris, T. J. The PAR complex regulates pulsed actomyosin contractions during amnioserosa apical constriction in Drosophila. Development 137, 1645–1655 (2010). 113. Roh-Johnson, M. et al. Triggering a cell shape change by exploiting preexisting actomyosin contractions. Science 335, 1232–1235 (2012). 114. Sawyer, J. K., Harris, N. J., Slep, K. C., 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). 115. Sawyer, J. K. et al. A contractile actomyosin network linked to adherens junctions by Canoe/afadin helps drive convergent extension. Mol. Biol. Cell 22, 2491–2508 (2011). 116. Kurita, S., Yamada, T., Rikitsu, E., Ikeda, W. & Takai, Y. Binding between the junctional proteins afadin and PLEKHA7 and implication in the formation of adherens junction in epithelial cells. J. Biol. Chem. 288, 29356–29368 (2013). 117. Bertet, C., Sulak, L. & Lecuit, T. Myosin-dependent junction remodelling controls planar cell intercalation and axis elongation. Nature 429, 667–671 (2004). 118. Zallen, J. A. & Wieschaus, E. Patterned gene expression directs bipolar planar polarity in Drosophila. Dev. Cell 6, 343–355 (2004). 119. Simoes Sde, M. et al. Rho-kinase directs Bazooka/Par 3 planar polarity during Drosophila axis elongation. Dev. Cell 19, 377–388 (2010). 120. Blankenship, J. T., Backovic, S. T., Sanny, J. S. P., Weitz, O. & Zallen, J. A. Multicellular rosette formation links planar cell polarity to tissue morphogenesis. Dev. Cell 11, 459–470 (2006). 121. Fernandez-Gonzalez, R., Simoes, S. D., Roper, J. C., Eaton, S. & Zallen, J. A. Myosin, I. I. Dynamics are regulated by tension in intercalating cells. Dev. Cell 17, 736–743 (2009). 122. Simoes, S. et al. The role of Bazooka/Par3 in epithelial intercalation - a live imaging approach. Mech. Dev. 126, S81 (2009). 123. Rauzi, M., Lenne, P. F. & Lecuit, T. Planar polarized actomyosin contractile flows control epithelial junction remodelling. Nature 468, 1110–1114 (2010). 124. Levayer, R. & Lecuit, T. Oscillation and polarity of E cadherin asymmetries control actomyosin flow patterns during morphogenesis. Dev. Cell 26, 162–175 (2013). Links two subcellular events, movement of cortical actomyosin networks and E-cadherin endocytosis, to explain how the shrinkage of polarized junctions occurs. 125. Levayer, R., Pelissier-Monier, A. & Lecuit, T. Spatial regulation of Dia and Myosin II by RhoGEF2 controls initiation of E cadherin endocytosis during epithelial morphogenesis. Nature Cell Biol. 13, 529–540 (2011). 126. Tamada, M., Farrell, D. L. & Zallen, J. A. Abl regulates planar polarized junctional dynamics through β-catenin tyrosine phosphorylation. Dev. Cell 22, 309–319 (2012). 127. Walck-Shannon, E. & Hardin, J. Cell intercalation from top to bottom. Nature Rev. Mol. Cell Biol. 15, 34–48 (2013). 128. Wang, Y. C., Khan, Z., Kaschube, M. & Wieschaus, E. F. Differential positioning of adherens junctions is associated with initiation of epithelial folding. Nature 484, 390–393 (2012). Finds a novel mechanism for epithelial reshaping in which the sliding, rather than contraction, of AJs is crucial for apical constriction in certain tissues. 129. Wang, Y. C., Khan, Z. & Wieschaus, E. F. Distinct Rap1 activity states control the extent of epithelial invagination via α-catenin. Dev. Cell 25, 299–309 (2013). 130. Taniguchi, K. et al. Chirality in planar cell shape contributes to left-right asymmetric epithelial morphogenesis. Science 333, 339–341 (2011).


131. Davis, N. M. et al. The chirality of gut rotation derives from left-right asymmetric changes in the architecture of the dorsal mesentery. Dev. Cell 15, 134–145 (2008). 132. Kurpios, N. A. et al. The direction of gut looping is established by changes in the extracellular matrix and in cell:cell adhesion. Proc. Natl Acad. Sci. USA 105, 8499–8506 (2008). 133. Plageman, T. F., Zacharias, A. L., Gage, P. J. & Lang, R. A. Shroom3 and a Pitx2 N cadherin pathway function cooperatively to generate asymmetric cell shape changes during gut morphogenesis. Dev. Biol. 357, 227–234 (2011). 134. Welsh, I. C. et al. Integration of left-right Pitx2 transcription and Wnt signaling drives asymmetric gut morphogenesis via Daam2. Dev. Cell 26, 629–644 (2013). 135. Wu, S. K. et al. Cortical F actin stabilization generates apical-lateral patterns of junctional contractility that integrate cells into epithelia. Nature Cell Biol. 16, 167–178 (2014). 136. Alatortsev, V. E., Kramerova, I. A., Frolov, M. V., Lavrov, S. A. & Westphal, E. D. Vinculin gene is nonessential in Drosophila melanogaster. FEBS Lett. 413, 197–201 (1997). 137. Xu, W., Baribault, H. & Adamson, E. D. Vinculin knockout results in heart and brain defects during embryonic development. Development 125, 327–337 (1998). 138. Torres, M. et al. An α E catenin gene trap mutation defines its function in preimplantation development. Proc. Natl Acad. Sci. USA 94, 901–906 (1997). 139. Mason, F. M., Tworoger, M. & Martin, A. C. Apical domain polarization localizes actin-myosin activity to drive ratchet-like apical constriction. Nature Cell Biol. 15, 926–936 (2013). 140. Stehbens, S. J. et al. Dynamic microtubules regulate the local concentration of E cadherin at cell–cell contacts. J. Cell Sci. 119, 1801–1811 (2006). 141. Meng, W., Mushika, Y., Ichii, T. & Takeichi, M. Anchorage of microtubule minus ends to adherens junctions regulates epithelial cell–cell contacts. Cell 135, 948–959 (2008). 142. Harris, T. J. & Peifer, M. aPKC controls microtubule organization to balance adherens junction symmetry and planar polarity during development. Dev. Cell 12, 727–738 (2007). 143. Shaw, R. M. et al. Microtubule plus-end-tracking proteins target gap junctions directly from the cell interior to adherens junctions. Cell 128, 547–560 (2007). 144. Ligon, L. A., Karki, S., Tokito, M. & Holzbaur, E. L. Dynein binds to β-catenin and may tether microtubules at adherens junctions. Nature Cell Biol. 3, 913–917 (2001). 145. Bellett, G. et al. Microtubule plus-end and minus-end capture at adherens junctions is involved in the assembly of apico–basal arrays in polarised epithelial cells. Cell Motil. Cytoskeleton 66, 893–908 (2009). 146. Harrison, O. J. et al. The extracellular architecture of adherens junctions revealed by crystal structures of type I cadherins. Structure 19, 244–256 (2011). 147. Ishiyama, N. et al. Dynamic and static interactions between p120 catenin and E-cadherin regulate the stability of cell–cell adhesion. Cell 141, 117–128 (2010). 148. Huber, A. H. & Weis, W. I. The structure of the β-catenin/E cadherin complex and the molecular basis of diverse ligand recognition by β-catenin. Cell 105, 391–402 (2001). 149. Pokutta, S. & Weis, W. I. Structure of the dimerization and β-catenin-binding region of α-catenin. Mol. Cell 5, 533–543 (2000).

Acknowledgements

The author thanks T. Nishimura for critical comments on the manuscript, K. Taguchi for editing the movie and S. Ito for suggestions on references. The author’s laboratory is supported by the programme Grants-in‑Aid for Specially Promoted Research of the Ministry of Education, Science, Sports and Culture of Japan.

Competing interests statement

The author declares no competing interests.

SUPPLEMENTARY INFORMATION See online article: S1 (movie) ALL LINKS ARE ACTIVE IN THE ONLINE PDF


Formation of adherens junctions leads to the emergence of a tissue-level tension in epithelial monolayers.

PLEKHA7 Recruits PDZD11 to Adherens Junctions to Stabilize Nectins.

The adherens junctions control susceptibility to Staphylococcus aureus α-toxin.

Microtubules CLASP to Adherens Junctions in epidermal progenitor cells.

The cytoplasmic domain of adherens-type junctions.

Alpha-catenin-dependent recruitment of the centrosomal protein CAP350 to adherens junctions allows epithelial cells to acquire a columnar shape.

Segmentation and tracking of adherens junctions in 3D for the analysis of epithelial tissue morphogenesis.

Janus kinase 3 regulates adherens junctions and epithelial mesenchymal transition through β-catenin.

Tension across adherens junctions: when less is more.

Small Rho GTPase-mediated actin dynamics at endothelial adherens junctions.

Molecular junctions: Single-molecule contacts exposed.

Correction: Dynamic behavior of rearranging carbocations - implications for terpene biosynthesis.

Mechanosensitive Adaptation of E-Cadherin Turnover across adherens Junctions.

The role of adherens junctions in the developing neocortex.

Regulation of Endothelial Adherens Junctions by Tyrosine Phosphorylation.

The mechanotransduction machinery at work at adherens junctions.

Dynamic behavior of rearranging carbocations - implications for terpene biosynthesis.

HIV-associated disruption of tight and adherens junctions of oral epithelial cells facilitates HSV-1 infection and spread.

Minus end-directed motor KIFC3 suppresses E-cadherin degradation by recruiting USP47 to adherens junctions.

SnapShot: Epithelial tight junctions.

ERK signalling modulates epigenome to drive epithelial to mesenchymal transition.

Analysis of Myosin II Minifilament Orientation at Epithelial Zonula Adherens.

Tricellular junctions: how to build junctions at the TRICkiest points of epithelial cells.

ZO-1 controls endothelial adherens junctions, cell-cell tension, angiogenesis, and barrier formation.