Phosphorylation-dependent localization of microtubule-associated protein MAP2c to the actin cytoskeleton

Microtubule-associated protein 2 (MAP2) is a neuronal phosphoprotein that promotes net microtubule growth and actin cross-linking and bundling in vitro. Little is known about MAP2 regulation or its interaction with the cytoskeleton in vivo. Here we investigate the in vivo function of three specific sites of phosphorylation on MAP2. cAMP-dependent protein kinase activity disrupts the MAP2-microtubule interaction in living HeLa cells and promotes MAP2c localization to peripheral membrane ruffles enriched in actin. cAMP-dependent protein kinase phosphorylates serines within three KXGS motifs, one within each tubulin-binding repeat. These highly conserved motifs are also found in homologous proteins tau and MAP4. Phosphorylation at two of these sites was detected in brain tissue. Constitutive phosphorylation at these sites was mimicked by single, double, and triple mutations to glutamic acid. Biochemical and microscopy-based assays indicated that mutation of a single residue was adequate to disrupt the MAP2-microtubule interaction in HeLa cells. Double or triple point mutation promoted MAP2c localization to the actin cytoskeleton. Specific association between MAP2c and the actin cytoskeleton was demonstrated by retention of MAP2c-actin colocalization after detergent extraction. Specific phosphorylation states may enhance the interaction of MAP2 with the actin cytoskeleton, thereby providing a regulated mechanism for MAP2 function within distinct cytoskeletal domains.     

Identification of a novel sequence in PDZ-RhoGEF that mediates interaction with the actin cytoskeleton

Small GTPases of the Rho family are crucial regulators of actin cytoskeleton rearrangements. Rho is activated by members of the Rho guanine-nucleotide exchange factor (GEF) family; however, mechanisms that regulate RhoGEFs are not well understood. This report demonstrates that PDZ-RhoGEF, a member of a subfamily of RhoGEFs that contain regulator of G protein signaling domains, is partially localized at or near the plasma membranes in 293T, COS-7, and Neuro2a cells, and this localization is coincident with cortical actin. Disruption of the cortical actin cytoskeleton in cells by using latrunculin B prevents the peri-plasma membrane localization of PDZ-RhoGEF. Coimmunoprecipitation and F-actin cosedimentation assays demonstrate that PDZ-RhoGEF binds to actin. Extensive deletion mutagenesis revealed the presence of a novel 25-amino acid sequence in PDZ-RhoGEF, located at amino acids 561-585, that is necessary and sufficient for localization to the actin cytoskeleton and interaction with actin. Last, PDZ-RhoGEF mutants that fail to interact with the actin cytoskeleton display enhanced Rho-dependent signaling compared with wild-type PDZ-RhoGEF. These results identify interaction with the actin cytoskeleton as a novel function for PDZ-RhoGEF, thus implicating actin interaction in organizing PDZ-RhoGEF signaling.     

Clustering of membrane raft proteins by the actin cytoskeleton

Cell membranes are laterally organized into functionally discrete domains that include the cholesterol-dependent membrane "rafts." However, how membrane domains are established and maintained remains unresolved and controversial but often requires the actin cytoskeleton. In this study, we used fluorescence resonance energy transfer to measure the role of the actin cytoskeleton in the co-clustering of membrane raft-associated fluorescent proteins (FPs) and FPs targeted to the nonraft membrane fraction. By fitting the fluorescence resonance energy transfer data to an isothermal binding equation, we observed a specific co-clustering of raft-associated donor and acceptor probes that was sensitive to latrunculin B (Lat B), which disrupts the actin cytoskeleton. Conversely, treating with jasplakinolide to enhance actin polymerization increased co-clustering of the raft-associated FPs over that of the nonraft probes. We also observed by immunoblotting experiments that the actin-dependent co-clustering coincided with regulation of the raft-associated Src family kinase Lck. Specifically, Lat B decreased the phosphorylation of the C-terminal regulatory tyrosine of Lck (Tyr505), and combining the Lat B with filipin further decreased the Tyr505 phosphorylation. Furthermore, the Lat B-dependent changes in Lck regulation required CD45 because no significant changes occurred in treated T cells lacking CD45 expression. These data define a role for the actin cytoskeleton in promoting co-clustering of raft-associated proteins and show that this property is important toward regulating raft-associated signaling proteins such as Lck.     

Tyrosine phosphorylation regulates the adhesions of ras-transformed breast epithelia

Transformed epithelial cells often are characterized by a fibroblastic or mesenchymal morphology. These cells exhibit altered cell-cell and cell-substrate interactions. Here we have identified changes in the adhesions and cytoskeletal interactions of transformed epithelial cells that contribute to their altered morphology. Using MCF-10A human breast epithelial cells as a model system, we have found that transformation by an activated form of ras is characterized by less developed adherens- type junctions between cells but increased focal adhesions. Contributing to the modified adherens junctions of the transformed cells are decreased interactions among beta-catenin, E-cadherin, and the actin cytoskeleton. The ras-transformed cells reveal elevated phosphotyrosine in many proteins, including beta-catenin and p120 Cas. Whereas in the normal cells beta-catenin is found in association with E- cadherin, p120 Cas is not. In the ras-transformed cells, the situation is reversed; tyrosine-phosphorylated p120 Cas, but not tyrosine- phosphorylated beta-catenin, now is detected in E-cadherin complexes. The tyrosine-phosphorylated beta-catenin also shows increased detergent solubility, suggesting a decreased association with the actin cytoskeleton. p120 Cas, whether tyrosine phosphorylated or not, partitions into the detergent soluble fraction, suggesting that it is not tightly bound to the actin cytoskeleton in either the normal or ras- transformed cells. Inhibitors of tyrosine kinases decrease the level of tyrosine phosphorylation and restore a normal epithelial morphology to the ras-transformed cells. In particular, decreased tyrosine phosphorylation of beta-catenin is accompanied by increased interaction with both E-cadherin and the detergent insoluble cytoskeletal fraction. These results suggest that elevated tyrosine phosphorylation of proteins such as beta-catenin and p120 Cas contribute to the altered adherens junctions of ras-transformed epithelia.     

The spectrin-actin complex and erythrocyte shape

The effects of phosphorylation of spectrin on the properties of the cytoskeletal network of the human erythrocyte have been studied. A suspension of the cytoskeletal residues obtained after extraction of the ghosts with the nonionic detergent Triton X-100 forms a gel on addition of membrane kinase and ATP. Phosphorylation has no effect on the association state of purified spectrin. No species higher than a tetramer of polypeptide chains is formed in vitro; in the absence of divalent cations, this tetramer is an entity liberated from and evidently present in the membrane. It has not so far proved possible to detect any F-actin in the cytoskeleton before or after phosphorylation. It is suggested that the consequence of phosphorylation is formation of additional interactions between spectrin and monomeric actin molecules. This view is supported by the formation, after phosphorylation of the Triton-extracted cytoskeleton, of an insoluble mass of protein on treatment with a cross-linking reagent. In the absence of divalent cations, a series of oligomeric species is progressively liberated from the cytoskeleton on extraction with solutions of low ionic strength. These oligomers contain actin as well as spectrin, and are thought to result from disruption of the network by random denaturation of the mono meric actin in the absence of divalent metal ions. A schematic view of the effects of phosphorylation on the structure of the cytoskeleton is presented.     

Characterization of integrin-tetraspanin adhesion complexes: role of tetraspanins in integrin signaling.

Tetraspanins (or proteins from the transmembrane 4 superfamily, TM4SF) form membrane complexes with integrin receptors and are implicated in integrin-mediated cell migration. Here we characterized cellular localization, structural composition, and signaling properties of alpha3beta1-TM4SF adhesion complexes. Double-immunofluorescence staining showed that various TM4SF proteins, including CD9, CD63, CD81, CD82, and CD151 are colocalized within dot-like structures that are particularly abundant at the cell periphery. Differential extraction in conjunction with chemical cross-linking indicated that the cell surface fraction of alpha3beta1-TM4SF protein complexes may not be directly linked to the cytoskeleton. However, in cells treated with cytochalasin B alpha3beta1-TM4SF protein complexes are relocated into intracellular vesicles suggesting that actin cytoskeleton plays an important role in the distribution of tetraspanins into adhesion structures. Talin and MARCKS are partially codistributed with TM4SF proteins, whereas vinculin is not detected within the tetraspanin-containing adhesion structures. Attachment of serum-starved cells to the immobilized anti-TM4SF mAbs induced dephosphorylation of focal adhesion kinase (FAK). On the other hand, clustering of tetraspanins in cells attached to collagen enhanced tyrosine phosphorylation of FAK. Furthermore, ectopic expression of CD9 in fibrosarcoma cells affected adhesion-induced tyrosine phosphorylation of FAK, that correlated with the reorganization of the cortical actin cytoskeleton. These results show that tetraspanins can modulate integrin signaling, and point to a mechanism by which TM4SF proteins regulate cell motility.     

The Bcr–Abl kinase regulates the actin cytoskeleton via a GADS/Slp-76/Nck1 adaptor protein pathway

Bcr–Abl is the transforming principle underlying chronic myelogenous leukaemia (CML). Here, we use a functional interaction proteomics approach to map pathways by which Bcr–Abl regulates defined cellular processes. The results show that Bcr–Abl regulates the actin cytoskeleton and non-apoptotic membrane blebbing via a GADS/Slp-76/Nck1 adaptor protein pathway. The binding of GADS to Bcr–Abl requires Bcr–Abl tyrosine kinase activity and is sensitive to the Bcr–Abl inhibitor imatinib, while the GADS/Slp-76 and Slp-76/Nck interactions are tyrosine phosphorylation independent. All three adaptor proteins co-localize with cortical actin in membrane blebs. Downregulation of each adaptor protein disrupts the actin cytoskeleton and membrane blebbing in a similar fashion and similar to imatinib. These findings highlight the importance of protein interaction dependent adaptor protein pathways in oncogenic kinase signaling.     

Caldesmon is a cytoskeletal target for PKC in endothelium

We have previously shown that treatment of bovine endothelial cell (EC) monolayers with phorbol myristate acetate (PMA) leads to the thinning of cortical actin ring and rearrangement of the cytoskeleton into a grid-like structure, concomitant with the loss of endothelial barrier function. In the current work, we focused on caldesmon, a cytoskeletal protein, regulating actomyosin interaction. We hypothesized that protein kinase C (PKC) activation by PMA leads to the changes in caldesmon properties such as phosphorylation and cellular localization. We demonstrate here that PMA induces both myosin and caldesmon redistribution from cortical ring into the grid-like network. However, the initial step of PMA-induced actin and myosin redistribution is not followed by caldesmon redistribution. Co-immunoprecipitation experiments revealed that short-term PMA (5 min) treatment leads to the weakening of caldesmon ability to bind actin and, to the lesser extent, myosin. Prolonged incubation (15-60 min) with PMA, however, strengthens caldesmon complexes with actin and myosin, which correlates with the grid-like actin network formation. PMA stimulation leads to an immediate increase in caldesmon Ser/Thr phosphorylation. This process occurs at sites distinct from the sites specific for ERK1/2 phosphorylation and correlates with caldesmon dissociation from the actomyosin complex. Inhibition of ERK-kinase MEK fails to abolish grid-like structure formation, although reducing PMA-induced weakening of the cortical actin ring, whereas inhibition of PKC reverses PMA-induced cytoskeletal rearrangement. Our results suggest that PKC-dependent phosphorylation of caldesmon is involved in PMA-mediated complex cytoskeletal changes leading to the EC barrier compromise.     

The Cdk5-p35 kinase associates with the Golgi apparatus and regulates membrane traffic

We show here that an active Cdk5-p35 kinase is present in Golgi membranes, where it associates with a detergent-insoluble fraction containing actin. In addition, Cdk5-p35-dependent phosphorylation of α-PAK immunoreactive protein species was detected in Golgi membranes, as well as an interaction with the small GTPase, Cdc42. Moreover, antisense oligonucleotide suppression of Cdk5 or p35 in young cultured neurons, as well as inhibition of Cdk5 activity with olomoucine, blocks the formation of membrane vesicles from the Golgi apparatus. Taken together, these results show a novel subcellular localization of this kinase and suggest a role for Cdk5-p35 in membrane traffic during neuronal process outgrowth.     

Integrity of the cortical actin ring is required for activation of the PI3K/Akt and p38 MAPK signaling pathways in redifferentiation of chondrocytes on chitosan

Cell shape alterations and accompanying cytoskeletal changes have diverse effects on cell function. We have already shown that dedifferentiated chondrocytes have a round cell morphology and undergo redifferentiation when cultured on chitosan membrane. In the present study, we investigate the role of the cytoskeleton in chondrocyte redifferentiation. Chondrocytes obtained from a micromass culture of chick limb bud mesenchymal cells were subcultured four times. Immunofluorescence analysis of F-actin showed cortical distribution of the actin cytoskeleton upon subculture of dedifferentiated chondrocytes on chitosan membrane. Treatment with cytochalasin D disrupted the cortical actin ring formed during cultivation of chondrocytes on the chitosan membrane, and inhibited chondrocyte redifferentiation. Moreover, cytochalasin D inhibited the phosphorylation of Akt and p38 mitogen activated protein kinase (MAPK), induced during redifferentiation on chitosan membrane. LY294002, an inhibitor of phosphatidylinositol-3-OH-kinase (PI3K), suppressed chondrocyte redifferentiation. These findings suggest that integrity of the actin cytoskeleton is a crucial requirement for PI3K/Akt and p38 MAPK in chondrocyte redifferentiation.     

Evidence that the middle T antigen of polyomavirus interacts with the membrane skeleton.

The transforming protein of polyomavirus, middle T antigen, is associated with cellular membranes. We have examined the subcellular location of the middle T antigen in two different cell types by fractionation and detergent phase partitioning. Middle T antigen expressed in human cells by a recombinant adenovirus was detected primarily in the membrane skeleton. Sucrose gradient fractionation revealed that the middle T antigen was associated with complexes with molecular weights of 500,000 to 1,000,000. Several markers for cytoskeleton cofractionate with these complexes, including actin, tubulin, and vimentin. Electron micrographs of membrane skeleton prepared from cells expressing middle T antigen demonstrated that this material contained primarily fibrous structures and was clearly devoid of bilayer membranes. These structures were distinct from the filamentous structures observed in fractions enriched for cytoskeleton. Consistent with a role for membrane skeleton localization in transformation, middle T antigen was detected exclusively in fractions enriched for membrane skeleton in middle T antigen-transformed Rat-2 cells. Our results may resolve the apparent difference between middle T antigen localization as determined by immunomicroscopy and that determined by subcellular fractionation.     

Immobilisation of organelles and actin bundles in the cortical cytoplasm of the alga Chara corallina Klein ex. Wild

The mechanism by which sub-cortical actin bundles and membranous organelles are immobilised in the cortical cytoplasm of the alga Chara was studied by perfusing cells with a solution containing 1% Triton X-100. Light and scanning electron microscopy and the release of starch grains and chlorophyll-protein complexes indicated that the detergent extensively solubilised the chloroplasts. However, the sub-cortical actin bundles remained in situ even though they were originally separated from the plasma membrane by the chloroplasts. A fibrous layer between chloroplasts and plasma membrane became readily visible after detergent extraction of the cells and could be released by low-ionic-strength ethylenediaminetetraacetic acid, thioglycollate and trypsin. The same treatments applied to cells not subject to detergent extraction released the membrane-bound organelles and actin bundles and no fibrous meshwork was visible on subsequent extraction with Triton. It is, therefore, concluded that a detergent-insoluble cortical cytoskeleton exists and contributes to the immobility of the actin and cortical organelles in the cells.Abbreviation EDTA ethylenediaminetetraacetic acid     

Calyculin A-induced actin phosphorylation and depolymerization in renal epithelial cells

This study reports actin phosphorylation and coincident actin cytoskeleton alterations in renal epithelial cell line, LLC-PK1. Serine phosphorylation of actin was first observed in vitro after the cell lysate was incubated with phosphatase inhibitors and ATP. Both the phosphorylated actin and actin kinase activities were found in the cytoskeletal fraction. Actin phosphorylation was later detected in living LLC-PK1 cells after incubation with the phosphatase inhibitor calyculin A. Calyculin A-induced actin phosphorylation was associated with reorganization of the actin cytoskeleton, including net actin depolymerization, loss of cell-cell junction and stress fiber F-actin filaments, and redistribution of F-actin filaments in the periphery of the rounded cells. Actin phosphorylation was abolished by 3-h ATP depletion but not by the non-specific kinase inhibitor staurosporine. These results demonstrate that renal epithelial cells contain kinase/phosphatase activities and actin can be phosphorylated in LLC-PK1 cells. Actin phosphorylation may play an important role in regulating the organization of the actin cytoskeleton in renal epithelium.     

Endomitotic Megakaryocytes form a Midzone in Anaphase but Have a Deficiency in Cleavage Furrow Formation

Megakaryocyte differentiation is marked by development of progressive polyploidy and accumulation of large nuclear mass and cytoplasmic volume. During differentiation, megakaryocytes undergo repeated incomplete cell cycles in which mitosis is aborted in late anaphase with failure of cytokinesis, termed endomitosis. Recent studies have postulated that failure of Aurora-B kinase to localize to the spindle midzone is responsible for endomitosis in megakaryocytes. In diploid cells, the translocation of Aurora-B kinase is critical for positioning of the cleavage furrow, in part through its phosphorylation of the Rho family GTPase activating protein MgcRacGAP which in turn alters activity of RhoA. However, we have previously demonstrated that Aurora-B kinase localizes to centromeres and is functional in endomitotic megakaryocytes. Here, we show that endomitotic megakaryocytes form midzone structures that recruit Aurora-B kinase and its substrate MgcRacGAP. Although many cells with polyploid anaphases showed cortical localization of Aurora-B kinase, we did not observe accumulation of RhoA in furrows or formation of an actin ring. When mitotic exit was induced by inhibition of cdk1, diploid control cells formed furrows exhibiting cortical RhoA but megakaryocytes exited endomitosis without evidence of furrowing. Therefore, localization of Aurora-B kinase to the midzone is normal in endomitotic megakaryocytes but furrowing is abnormal. These data suggest that endomitotic MKs fail to complete cytokinesis due to aberrant regulation of furrowing at a step subsequent to the localization of Aurora-B kinase, possibly involving the activation or localization of RhoA. This work explores the mechanism of a normally occurring furrowing defect in a non-malignant primary cell.     

Roles of ATP and cytoskeleton in the regulation of Na+/H+ exchanger along the nephron luminal membrane

Although in LLC-PK cells ATP depletion has been shown to result in alterations of cytoskeleton actin and an inhibition of Na+/H+ exchanger activity, there is little information concerning the regulation of this exchanger in the distal luminal membrane by ATP and actin filaments. The present study examined the direct effect of ATP and cytochalasin B on the Na+/H+ exchanger activity in the proximal and distal tubule luminal membranes. The presence of 100 microM ATP in the luminal membrane vesicles from rabbit proximal tubules did not influence the Ethyl Isopropyl Amiloride sensitive Na+ uptake by these membranes. In contrast, the same treatment of luminal membranes from distal tubules significantly enhanced the exchanger activity from 0.22 +/- 0.04 to 0.39 +/- 0.08 pM/microg/10 sec (P < 0.02). When ATP was replaced by its nonhydrolysable form, ATPgammas, the effect on the distal luminal membrane was strongly diminished suggesting that the action of the nucleotide implicates a phosphorylation step. Confirming this hypothesis, addition of 300-microM-Rp cAMP, a protein kinase A inhibitor, completely abolished the effect of ATP. In view of the fact that a tight relationship has been described between ATP, the cytoskeleton complex and the exchanger activity, we studied the effect of cytochalasin B on this activity. The presence of 20 microM cytochalasin B in the distal luminal membrane vesicles induced, as observed with ATP, a significant increase in the Na+ uptake. However, the actions of ATP and cytochalasin B were not additive. These results suggest that firstly, ATP and short actin filaments of the cytoskeleton regulate the distal luminal isoform through an intramembranous mechanism and secondly, a phosphorylation mechanism is, at least partially, implicated in the action of ATP. In contrast, the proximal tubule exchanger is regulated through different mechanisms.     

Co-localization of polysomes,cytoskeleton, and membranes with protein bodies from corn endosperm

Summary Frozen corn endosperm was homogenized in a cytoskeleton-stabilizing buffer and stained directly (without pelleting) with rhodamine-phalloidin for actin and either thiazole orange to stain RNA or DiOC₆ to stain membranes prior to examination under the fluorescence microscope. Other samples were treated with a non-ionic detergent alone or in conjunction with a ionic detergent prior to staining and fluorescence microscopy. Very gentle homogenization in unsupplemented buffer yielded a massive aggregate containing protein bodies that fluoresced after treatment with the ER stain DiOC₆. This aggregate was capped by an aggregate of unstained starch grains. More vigorous homogenization yielded more disperse patterns showing almost identical co-localization of ER, actin and RNA (polysomes). Homogenization in buffer plus non-ionic detergent removed most of the membrane yet maintained co-localization of actin and polysomes, while homogenization in double detergent removed the last traces of membrane and actin, and released over 70% of the polysomes. We interpret these results to suggest that protein bodies are surrounded by membranes, cytoskeleton and RNA (polysomes) and that the majority of the polysomes are attached more firmly to the cytoskeleton than to the membrane. This provides evidence from fluorescence microscopy to supplement that from biochemical analyses for the existence of cytomatrix-bound polysomes in plants.Abbreviations CBP cytoskeleton-bound polysomes - CMBP cyto-matrix-bound polysomes - CSB cytoskeleton-stabilizing buffer - DOC sodium deoxycholate - DiOC₆ 3,3-dihexyloxacarbocyanine iodide - DTE dithioerythritol - MBP membrane-bound polysomes - FP free polysomes - PMSF phenylmethyl-sulfonyl fluoride - PTE polyoxy-ethylene-10-tridecyl ether - Rh-Ph rhodamine-phalloidin - TO thiazole orange - Tris tris-(hydroxymethyl) aminomethane     

A rapid,nongenomic, signaling pathway regulates the actin reorganization induced by activation of membrane testosterone receptors

The human prostate cancer cell line LNCaP bears functional membrane testosterone receptors, which modify the actin cytoskeleton and increase the secretion of prostate-specific antigen (PSA) within minutes. Membrane steroid receptors are, indeed, a newly identified element of steroid action that is different from the classical intracellular sites. In the present work, using a nonpermeable analog of testosterone (testosterone-BSA), we investigated the signaling pathway that is triggered by the membrane testosterone receptors' activation and leads to actin cytoskeleton reorganization. We report that exposure of cells to testosterone-BSA resulted in phosphorylation of focal adhesion kinase (FAK), the association of FAK with the phosphatidylinositol-3 (PI-3) kinase, and the subsequent activation of the latter as well as the activation of the small guanosine triphosphatases Cdc42/Rac1. Pretreatment of cells with the specific PI-3 kinase inhibitor wortmannin abolished both the activation of the small guanosine triphosphatases and the alterations of actin cytoskeleton, whereas it did not affect the phosphorylation of FAK. These findings indicate that PI-3 kinase is activated downstream of FAK and upstream of Cdc42/Rac1, which subsequently regulate the actin organization. Moreover, wortmannin diminished the secretion of PSA, implying that the signaling events described above are responsible for the testosterone-BSA-induced PSA secretion. Our results are discussed under the prism of a possible implication of these membrane receptors in prostate cancer chemotherapy.     

A mutation in dVps28 reveals a link between a subunit of the endosomal sorting complex required for transport-I complex and the actin cytoskeleton in Drosophila

Proteins that constitute the endosomal sorting complex required for transport (ESCRT) are necessary for the sorting of proteins into multivesicular bodies (MVBs) and the budding of several enveloped viruses, including HIV-1. The first of these complexes, ESCRT-I, consists of three proteins: Vps28p, Vps37p, and Vps23p or Tsg101 in mammals. Here, we characterize a mutation in the Drosophila homolog of vps28. The dVps28 gene is essential: homozygous mutants die at the transition from the first to second instar. Removal of maternally contributed dVps28 causes early embryonic lethality. In such embryos lacking dVps28, several processes that require the actin cytoskeleton are perturbed, including axial migration of nuclei, formation of transient furrows during cortical divisions in syncytial embryos, and the subsequent cellularization. Defects in actin cytoskeleton organization also become apparent during sperm individualization in dVps28 mutant testis. Because dVps28 mutant cells contained MVBs, these defects are unlikely to be a secondary consequence of disrupted MVB formation and suggest an interaction between the actin cytoskeleton and endosomal membranes in Drosophila embryos earlier than previously appreciated.     

Ezrin contains cytoskeleton and membrane binding domains accounting for its proposed role as a membrane-cytoskeletal linker

Ezrin, a widespread protein present in actin-containing cell-surface structures, is a substrate of some protein tyrosine kinases. Based on its primary and secondary structure similarities with talin and band 4.1 it has been suggested that this protein could play a role in linking the cytoskeleton to the plasma membrane (Gould, K.L., A. Bretscher, F.S. Esch, and T. Hunter. 1989. EMBO (Eur. Mol. Biol. Organ.), J. 8:4133-4142; Turunen, O., R. Winqvist, R. Pakkanen, K.-H. Grzeschik, T. Wahlstrom, and A. Vaheri. 1989. J. Biol. Chem. 264:16727- 16732). To test this hypothesis, we transiently expressed the complete human ezrin cDNA, or truncated cDNAs encoding the amino- and carboxy- terminal domains of the protein, in CV-1 cells. Protein epitope tagging was used to unambiguously determine the subcellular distribution of the protein encoded by the transfected cDNA. We show that this protein is concentrated underneath the dorsal plasma membrane in all actin- containing structures and is partially detergent insoluble. The amino- terminal domain displays the same localization but is readily extractable by nonionic detergent. The carboxy-terminal domain colocalizes with microvillar actin filaments as well as with stress fibers and remains associated with actin filaments after detergent extraction, and with disorganized actin structures after cytochalasin D treatment. Our results clearly demonstrate that ezrin interacts with membrane-associated components via its amino-terminal domain, and with the cytoskeleton via its carboxy-terminal domain. The amino-terminal domain could include the main determinant that restricts the entire protein to the cortical cytoskeleton in contact with the dorsal plasma membrane and its specialized microdomains such as microvilli, microspikes and lamellipodia.