Planar polarized actomyosin contractile flows control epithelial junction remodelling

Force generation by Myosin-II motors on actin filaments drives cell and tissue morphogenesis. In epithelia, contractile forces are resisted at apical junctions by adhesive forces dependent on E-cadherin, which also transmits tension. During Drosophila embryonic germband extension, tissue elongation is driven by cell intercalation, which requires an irreversible and planar polarized remodelling of epithelial cell junctions. We investigate how cell deformations emerge from the interplay between force generation and cortical force transmission during this remodelling in Drosophila melanogaster. The shrinkage of dorsal-ventral-oriented ('vertical') junctions during this process is known to require planar polarized junctional contractility by Myosin II (refs 4, 5, 7, 12). Here we show that this shrinkage is not produced by junctional Myosin II itself, but by the polarized flow of medial actomyosin pulses towards 'vertical' junctions. This anisotropic flow is oriented by the planar polarized distribution of E-cadherin complexes, in that medial Myosin II flows towards 'vertical' junctions, which have relatively less E-cadherin than transverse junctions. Our evidence suggests that the medial flow pattern reflects equilibrium properties of force transmission and coupling to E-cadherin by α-Catenin. Thus, epithelial morphogenesis is not properly reflected by Myosin II steady state distribution but by polarized contractile actomyosin flows that emerge from interactions between E-cadherin and actomyosin networks.     

Drosophila Crumbs is a positional cue in photoreceptor adherens junctions and rhabdomeres

Drosophila Crumbs (Crb) is required for apical-basal polarity and is an apical determinant in embryonic epithelia. Here, we describe properties of Crb that control the position and integrity of the photoreceptor adherens junction and photosensitive organ, or rhabdomere. In contrast to normal photoreceptor adherens junctions and rhabdomeres, which span the depth of the retina, adherens junctions and rhabdomeres of Crb-deficient photoreceptors initially accumulate at the top of the retina and fail to maintain their integrity as they stretch to the retinal floor. We show that Crb controls localization of the adherens junction through its intracellular domain containing a putative binding site for a protein 4.1 superfamily protein (FERM). Although loss of Crb or overexpression of the FERM binding domain causes mislocalization of adherens junctions, they do not result in a significant loss of photoreceptor polarity. Mutations in CRB1, a human homologue of crb, are associated with photoreceptor degeneration in retinitis pigmentosa 12 (RP12) and Leber congenital amaurosis (LCA). The intracellular domain of CRB1 behaves similarly to its Drosophila counterpart when overexpressed in the fly eye. Our studies may provide clues for mechanisms of photoreceptor degeneration in RP12 and LCA.     

Localization of apical epithelial determinants by the basolateral PDZ protein Scribble

The generation of membrane domains with distinct protein constituents is a hallmark of cell polarization. In epithelia, segregation of membrane proteins into apical and basolateral compartments is critical for cell morphology, tissue physiology and cell signalling. Drosophila proteins that confer apical membrane identity have been found, but the mechanisms that restrict these determinants to the apical cell surface are unknown. Here we show that a laterally localized protein is required for the apical confinement of polarity determinants. Mutations in Drosophila scribble (scrib), which encodes a multi-PDZ (PSD-95, Discs-large and ZO-1) and leucine-rich-repeat protein, cause aberrant cell shapes and loss of the monolayer organization of embryonic epithelia. Scrib is localized to the epithelial septate junction, the analogue of the vertebrate tight junction, at the boundary of the apical and basolateral cell surfaces. Loss of scrib function results in the misdistribution of apical proteins and adherens junctions to the basolateral cell surface, but basolateral protein localization remains intact. These phenotypes can be accounted for by mislocalization of the apical determinant Crumbs. Our results show that the lateral domain of epithelia, particularly the septate junction, functions in restricting apical membrane identity and correctly placing adherens junctions.     

Moesin functions antagonistically to the Rho pathway to maintain epithelial integrity

Two prominent characteristics of epithelial cells, apical-basal polarity and a highly ordered cytoskeleton, depend on the existence of precisely localized protein complexes associated with the apical plasma membrane, and on a separate machinery that regulates the spatial order of actin assembly. ERM (ezrin, radixin, moesin) proteins have been proposed to link transmembrane proteins to the actin cytoskeleton in the apical domain, suggesting a structural role in epithelial cells, and they have been implicated in signalling pathways. Here, we show that the sole Drosophila ERM protein Moesin functions to promote cortical actin assembly and apical-basal polarity. As a result, cells lacking Moesin lose epithelial characteristics and adopt invasive migratory behaviour. Our data demonstrate that Moesin facilitates epithelial morphology not by providing an essential structural function, but rather by antagonizing activity of the small GTPase Rho. Thus, Moesin functions in maintaining epithelial integrity by regulating cell-signalling events that affect actin organization and polarity. Furthermore, our results show that there is negative feedback between ERM activation and activity of the Rho pathway.     

Adherens junctions and beta-catenin-mediated cell signalling in a non-metazoan organism

Mechanical forces between cells have a principal role in the organization of animal tissues. Adherens junctions are an important component of these tissues, connecting cells through their actin cytoskeleton and allowing the assembly of tensile structures. At least one adherens junction protein, beta-catenin, also acts as a signalling molecule, directly regulating gene expression. To date, adherens junctions have only been detected in metazoa, and therefore we looked for them outside the animal kingdom to examine their evolutionary origins. The non-metazoan Dictyostelium discoideum forms a multicellular, differentiated structure. Here we describe the discovery of actin-associated intercellular junctions in Dictyostelium. We have isolated a gene encoding a beta-catenin homologue, aardvark, which is a component of the junctional complex, and, independently, is required for cell signalling. Our discovery of adherens junctions outside the animal kingdom shows that the dual role of beta-catenin in cell-cell adhesion and cell signalling evolved before the origins of metazoa.     

A two-tiered mechanism for stabilization and immobilization of E-cadherin

Epithelial tissues maintain a robust architecture which is important for their barrier function, but they are also remodelled through the reorganization of cell-cell contacts. Tissue stability requires intercellular adhesion mediated by E-cadherin, in particular its trans-association in homophilic complexes supported by actin filaments through beta- and alpha-catenin. How alpha-catenin dynamic interactions between E-cadherin/beta-catenin and cortical actin control both stability and remodelling of adhesion is unclear. Here we focus on Drosophila homophilic E-cadherin complexes rather than total E-cadherin, including diffusing 'free' E-cadherin, because these complexes are a better proxy for adhesion. We find that E-cadherin complexes partition in very stable microdomains (that is, bona fide adhesive foci which are more stable than remodelling contacts). Furthermore, we find that stability and mobility of these microdomains depend on two actin populations: small, stable actin patches concentrate at homophilic E-cadherin clusters, whereas a rapidly turning over, contractile network constrains their lateral movement by a tethering mechanism. alpha-Catenin controls epithelial architecture mainly through regulation of the mobility of homophilic clusters and it is largely dispensable for their stability. Uncoupling stability and mobility of E-cadherin complexes suggests that stable epithelia may remodel through the regulated mobility of very stable adhesive foci.     

Adherens junctions inhibit asymmetric division in the Drosophila epithelium

Asymmetric division is a fundamental mechanism for generating cellular diversity. In the central nervous system of Drosophila, neural progenitor cells called neuroblasts undergo asymmetric division along the apical-basal cellular axis. Neuroblasts originate from neuroepithelial cells, which are polarized along the apical-basal axis and divide symmetrically along the planar axis. The asymmetry of neuroblasts might arise from neuroblast-specific expression of the proteins required for asymmetric division. Alternatively, both neuroblasts and neuroepithelial cells could be capable of dividing asymmetrically, but in neuroepithelial cells other polarity cues might prevent asymmetric division. Here we show that by disrupting adherens junctions we can convert the symmetric epithelial division into asymmetric division. We further confirm that the adenomatous polyposis coli (APC) tumour suppressor protein is recruited to adherens junctions, and demonstrate that both APC and microtubule-associated EB1 homologues are required for the symmetric epithelial division along the planar axis. Our results indicate that neuroepithelial cells have all the necessary components to execute asymmetric division, but that this pathway is normally overridden by the planar polarity cue provided by adherens junctions.     

In vitro formation of a complex between cytoskeletal proteins of the human erythrocyte.

The formation of a high-molecular weight complex between spectrin and F-actin depends on the presence of a third cytoskeletal constituent, protein 4.1. Electron microscopy shows that in this ternary complex the actin filaments are linked by bridges, which have the appearance of spectrin. The spectrin must be in the tetrameric state for such bridges to form: the dimer is evidently univalent, for it binds but forms no cross-links. G-actin also fails to form extended complexes. It is inferred that in the native cytoskeleton the spectrin is tetrameric and associated with 4.1 and probably oligomers of actin.     

Modulation of spectrin-actin assembly by erythrocyte adducin

The spectrin-based membrane skeleton, an assembly of proteins tightly associated with the plasma membrane, determines the shape and mechanical properties of erythrocytes. Spectrin, the most abundant component of this assembly, is an elongated and flexible molecule that, with potentiation by protein 4.1, is cross-linked at its ends by short actin filaments to form a lattice beneath the membrane. These and other proteins stabilize the plasma membrane, organize integral membrane proteins and maintain specialized regions of the cell surface. A membrane-skeleton-associated calmodulin-binding protein of erythrocytes is a major substrate for Ca2+- and phospholipid-dependent protein kinase C (ref. 5), and thus is a target for Ca2+ by two regulatory pathways. Here we demonstrate that this protein, called adducin: (1) binds tightly in vitro to spectrin-actin complexes but with much less affinity either to spectrin or to actin alone; (2) promotes assembly of additional spectrin molecules onto actin filaments; and (3) is inhibited in its ability to induce the binding of additional spectrin molecules to actin by micromolar concentrations of calmodulin and Ca2+. Adducin may be involved in the action of Ca2+ on erythrocyte membrane skeleton and in the assembly of spectrin-actin complexes.     

Disruption of the actin cytoskeleton in yeast capping protein mutants

Capping protein controls the addition of actin subunits to the barbed end of actin filaments and nucleates actin polymerization in vitro. Capping protein has been identified in all eukaryotic cells examined so far; it is a heterodimer with subunits of relative molecular masses 32,000-36,000 (alpha-subunit) and 28,000-32,000 (beta-subunit). In skeletal muscle, capping protein (CapZ) probably binds the barbed ends of actin filaments at the Z line. The in vivo role of this protein in non-muscle cells is not known. We report here the characterization of CAP2, the single gene encoding the beta-subunit of capping protein in Saccharomyces cerevisiae. Yeast cells in which the CAP2 gene was disrupted by an insertion or a deletion had an abnormal actin distribution, including the loss of actin cables. The mutant cells were round and large, with a heterogeneous size distribution, and, although viable, grew more slowly than congenic wild-type cells. Chitin, a cell wall component restricted to the mother-bud junction in wild-type budding yeast, was found on the entire mother cell surface in the mutants. The phenotype of CAP2 disruption resembled that of temperature-sensitive mutations in the yeast actin gene ACT1, indicating that capping protein regulates actin-filament distribution in vivo.     

Myosin-dependent junction remodelling controls planar cell intercalation and axis elongation

Shaping a developing organ or embryo relies on the spatial regulation of cell division and shape. However, morphogenesis also occurs through changes in cell-neighbourhood relationships produced by intercalation. Intercalation poses a special problem in epithelia because of the adherens junctions, which maintain the integrity of the tissue. Here we address the mechanism by which an ordered process of cell intercalation directs polarized epithelial morphogenesis during germ-band elongation, the developmental elongation of the Drosophila embryo. Intercalation progresses because junctions are spatially reorganized in the plane of the epithelium following an ordered pattern of disassembly and reassembly. The planar remodelling of junctions is not driven by external forces at the tissue boundaries but depends on local forces at cell boundaries. Myosin II is specifically enriched in disassembling junctions, and its planar polarized localization and activity are required for planar junction remodelling and cell intercalation. This simple cellular mechanism provides a general model for polarized morphogenesis in epithelial organs.     

Abolition of actin-bundling by phosphorylation of human erythrocyte protein 4.9

Protein 4.9, first identified as a component of the human erythrocyte membrane skeleton, binds to and bundles actin filaments. Protein 4.9 is a substrate for various kinases, including a cyclic AMP(cAMP)-dependent one, in vivo and in vitro. We show here that phosphorylation of protein 4.9 by the catalytic subunit of cAMP-dependent protein kinase reversibly abolishes its actin-bundling activity, but phosphorylation by protein kinase C has no such effect. A quantitative immunoassay showed that human erythrocytes contain 43,000 trimers of protein 4.9 per cell, which is equivalent to one trimer for each actin oligomer in these red blood cells. As analogues of protein 4.9 have been identified together with analogues of other erythroid skeletal proteins in non-erythroid tissues of numerous vertebrates, phosphorylation and dephosphorylation of protein 4.9 may be the basis for a mechanism that regulates actin bundling in many cells.     

Drosophila Spire is an actin nucleation factor

The actin cytoskeleton is essential for many cellular functions including shape determination, intracellular transport and locomotion. Previous work has identified two factors--the Arp2/3 complex and the formin family of proteins--that nucleate new actin filaments via different mechanisms. Here we show that the Drosophila protein Spire represents a third class of actin nucleation factor. In vitro, Spire nucleates new filaments at a rate that is similar to that of the formin family of proteins but slower than in the activated Arp2/3 complex, and it remains associated with the slow-growing pointed end of the new filament. Spire contains a cluster of four WASP homology 2 (WH2) domains, each of which binds an actin monomer. Maximal nucleation activity requires all four WH2 domains along with an additional actin-binding motif, conserved among Spire proteins. Spire itself is conserved among metazoans and, together with the formin Cappuccino, is required for axis specification in oocytes and embryos, suggesting that multiple actin nucleation factors collaborate to construct essential cytoskeletal structures.     

Drosophila Stardust interacts with Crumbs to control polarity of epithelia but not neuroblasts.

Establishing cellular polarity is critical for tissue organization and function. Initially discovered in the landmark genetic screen for Drosophila developmental mutants, bazooka, crumbs, shotgun and stardust mutants exhibit severe disruption in apicobasal polarity in embryonic epithelia, resulting in multilayered epithelia, tissue disintegration, and defects in cuticle formation. Here we report that stardust encodes single PDZ domain MAGUK (membrane-associated guanylate kinase) proteins that are expressed in all primary embryonic epithelia from the onset of gastrulation. Stardust colocalizes with Crumbs at the apicolateral boundary, although their expression patterns in sensory organs differ. Stardust binds to the carboxy terminus of Crumbs in vitro, and Stardust and Crumbs are mutually dependent in their stability, localization and function in controlling the apicobasal polarity of epithelial cells. However, for the subset of ectodermal cells that delaminate and form neuroblasts, their polarity requires the function of Bazooka, but not of Stardust or Crumbs.     

Modulation of gelsolin function by phosphatidylinositol 4,5-bisphosphate

The actin-binding protein gelsolin requires micromolar concentrations of calcium ions to sever actin filaments, to potentiate its binding to the end of the filament and to promote the polymerization of monomeric actin into filaments. Because transient increases in both intracellular [Ca2+] and actin polymerization accompany the cellular response to certain stimuli, it has been suggested that gelsolin regulates the reversible assembly of actin filaments that accompanies such cellular activations. But other evidence suggests that these activities do not need increased cytoplasmic [Ca2+] and that once actin-gelsolin complexes form in the presence of Ca2+ in vitro, removal of free Ca2+ causes dissociation of only one of two bound actin monomers from gelsolin and the resultant binary complexes cannot sever actin filaments. The finding that cellular gelsolin-actin complexes can be dissociated suggests that a Ca2+-independent regulation of gelsolin also occurs. Here we show that, like the dissociation of profilin-actin complexes, phosphatidylinositol 4,5-bisphosphate, which undergoes rapid turnover during cell stimulation, strongly inhibits the actin filament-severing properties of gelsolin, inhibits less strongly the nucleating ability of this protein and restores the potential for filament-severing activity to gelsolin-actin complexes.     

Identification of an actin-binding protein from Dictyostelium as elongation factor 1a

Indirect evidence has implicated an interaction between the cytoskeleton and the protein synthetic machinery. Two recent reports have linked the elongation factor 1a (EF-1a) which is involved in protein synthesis, with the microtubular cytoskeleton. In situ hybridization has, however, revealed that the messages for certain cytoskeletal proteins are preferentially associated with actin filaments. ABP-50 is an abundant actin filament bundling protein of native relative molecular mass 50,000 (50K) isolated from Dictyostelium discoideum. Immunofluorescence studies show that ABP-50 is present in filopodia and other cortical regions that contain actin filament bundles. In addition, ABP-50 binds to monomeric actin in the cytosol of unstimulated cells and the association of ABP-50 with the actin cytoskeleton is regulated during chemotaxis. Through complementary DNA sequencing and subsequent functional analysis, we have identified ABP-50 as D. discoideum EF-1a. The ability of EF-1a to bind reversibly to the actin cytoskeleton upon stimulation could provide a mechanism for spatially and temporally regulated protein synthesis in eukaryotic cells.     

Resemblance of actin-binding protein/actin gels to covalently crosslinked networks

The maintainance of the shape of cells is often due to their surface elasticity, which arises mainly from an actin-rich cytoplasmic cortex. On locomotion, phagocytosis or fission, however, these cells become partially fluid-like. The finding of proteins that can bind to actin and control the assembly of, or crosslink, actin filaments, and of intracellular messages that regulate the activities of some of these actin-binding proteins, indicates that such 'gel-sol' transformations result from the rearrangement of cortical actin-rich networks. Alternatively, on the basis of a study of the mechanical properties of mixtures of actin filaments and an Acanthamoeba actin-binding protein, alpha-actinin, it has been proposed that these transformations can be accounted for by rapid exchange of crosslinks between actin filaments: the cortical network would be solid when the deformation rate is greater than the rate of crosslink exchange, but would deform or 'creep' when deformation is slow enough to permit crosslinker molecules to rearrange. Here we report, however, that mixtures of actin filaments and actin-binding protein (ABP), an actin crosslinking protein of many higher eukaryotes, form gels rheologically equivalent to covalently crosslinked networks. These gels do not creep in response to applied stress on a time scale compatible with most cell-surface movements. These findings support a more complex and controlled mechanism underlying the dynamic mechanical properties of cortical cytoplasm, and can explain why cells do not collapse under the constant shear forces that often exist in tissues.     

The cargo-binding domain regulates structure and activity of myosin 5

Myosin 5 is a two-headed motor protein that moves cargoes along actin filaments. Its tail ends in paired globular tail domains (GTDs) thought to bind cargo. At nanomolar calcium levels, actin-activated ATPase is low and the molecule is folded. Micromolar calcium concentrations activate ATPase and the molecule unfolds. Here we describe the structure of folded myosin and the GTD's role in regulating activity. Electron microscopy shows that the two heads lie either side of the tail, contacting the GTDs at a lobe of the motor domain (approximately Pro 117-Pro 137) that contains conserved acidic side chains, suggesting ionic interactions between motor domain and GTD. Myosin 5 heavy meromyosin, a constitutively active fragment lacking the GTDs, is inhibited and folded by a dimeric GST-GTD fusion protein. Motility assays reveal that at nanomolar calcium levels heavy meromyosin moves robustly on actin filaments whereas few myosins bind or move. These results combine to show that with no cargo, the GTDs bind in an intramolecular manner to the motor domains, producing an inhibited and compact structure that binds weakly to actin and allows the molecule to recycle towards new cargoes.     

Vinculin activation by talin through helical bundle conversion

Vinculin is a conserved component and an essential regulator of both cell-cell (cadherin-mediated) and cell-matrix (integrin-talin-mediated focal adhesions) junctions, and it anchors these adhesion complexes to the actin cytoskeleton by binding to talin in integrin complexes or to alpha-actinin in cadherin junctions. In its resting state, vinculin is held in a closed conformation through interactions between its head (Vh) and tail (Vt) domains. The binding of vinculin to focal adhesions requires its association with talin. Here we report the crystal structures of human vinculin in its inactive and talin-activated states. Talin binding induces marked conformational changes in Vh, creating a novel helical bundle structure, and this alteration actively displaces Vt from Vh. These results, as well as the ability of alpha-actinin to also bind to Vh and displace Vt from pre-existing Vh-Vt complexes, support a model whereby Vh functions as a domain that undergoes marked structural changes that allow vinculin to direct cytoskeletal assembly in focal adhesions and adherens junctions. Notably, talin's effects on Vh structure establish helical bundle conversion as a signalling mechanism by which proteins direct cellular responses.