Molecular characterization of the tight junction protein ZO-1 in MDCK cells

Most of the information on the structure and function of the tight junction (TJ) has been obtained in MDCK cells. Accordingly, we have sequenced ZO-1 in this cell type, because this protein is involved in the response of the TJ to changes in Ca2+, phosphorylation, and the cytoskeleton. ZO-1 of MDCK cells comprises 6805 bp with a predicted open reading frame of 1769 amino acids. This sequence is 92 and 87% homologous to human and mouse ZO-1, respectively. Two nuclear sorting signals located at the PDZ1 and GK domains and 17 SH3 putative binding sites at the proline-rich domain were detected. We found two new splicing regions at the proline-rich region: beta had not been reported in human and mouse counterparts, and gamma, which was previously sequenced in human and mouse ZO-1, is now identified as a splicing region. The expression of different beta and gamma isoforms varies according to the tissue tested. With the information provided by the sequence, Southern blot, and PCR experiments we can predict a single genomic copy of MDCK-ZO-1 that is at least 13.16 kb long. MDCK-ZO-1 mRNA is 7.4 kb long. Its expression is regulated by calcium, while the expression of MDCK-ZO-1 protein is not.     

The tight junction protein occludin and the adherens junction protein alpha-catenin share a common interaction mechanism with ZO-1

The exact sites, structures, and molecular mechanisms of interaction between junction organizing zona occludence protein 1 (ZO-1) and the tight junction protein occludin or the adherens junction protein alpha-catenin are unknown. Binding studies by surface plasmon resonance spectroscopy and peptide mapping combined with comparative modeling utilizing crystal structures led for the first time to a molecular model revealing the binding of both occludin and alpha-catenin to the same binding site in ZO-1. Our data support a concept that ZO-1 successively associates with alpha-catenin at the adherens junction and occludin at the tight junction. Strong spatial evidence indicates that the occludin C-terminal coiled-coil domain dimerizes and interacts finally as a four-helix bundle with the identified structural motifs in ZO-1. The helix bundle of occludin406-521 and alpha-catenin509-906 interacts with the hinge region (ZO-1591-632 and ZO-1591-622, respectively) and with (ZO-1726-754 and ZO-1756-781) in the GuK domain of ZO-1 containing coiled-coil and alpha-helical structures, respectively. The selectivity of both protein-protein interactions is defined by complementary shapes and charges between the participating epitopes. In conclusion, a common molecular mechanism of forming an intermolecular helical bundle between the hinge region/GuK domain of ZO-1 and alpha-catenin and occludin is identified as a general molecular principle organizing the association of ZO-1 at adherens and tight junctions.     

Truncation mutants of the tight junction protein ZO-1 disrupt corneal epithelial cell morphology

The tight junction is the most apical intercellular junction of epithelial cells and regulates transepithelial permeability through the paracellular pathway. To examine possible functions for the tight junction-associated protein ZO-1, C-terminally truncated mutants and a deletion mutant of ZO-1 were epitope tagged and stably expressed in corneal epithelial cell lines. Only full-length ZO-1 and one N-terminal truncation mutant targeted to cell borders; other mutants showed variable cytoplasmic distributions. None of the mutants initially disrupted the localization of endogenous ZO-1. However, long-term stable expression of two of the N-terminal mutants resulted in a dramatic change in cell shape and patterns of gene expression. An elongated fibroblast-like shape replaced characteristic epithelial cobblestone morphology. In addition, vimentin and smooth muscle actin expression were up-regulated, although variable cytokeratin expression remained, suggesting a partial transformation to a mesenchymal cell type. Concomitant with the morphological change, the expression of the integral membrane tight junction protein occludin was significantly down-regulated. The localizations of endogenous ZO-1 and another family member, ZO-2, were disrupted. These findings suggest that ZO-1 may participate in regulation of cellular differentiation.     

The tight junction protein ZO-2 has several functional nuclear export signals

The tight junction (TJ) protein ZO-2 changes its subcellular distribution according to the state of confluency of the culture. Thus in confluent monolayers, it localizes at the TJ region whereas in sparse cultures it concentrates at the nucleus. The canine sequence of ZO-2 displays four putative nuclear export signals (NES), two at the second PDZ domain (NES-0 and NES-1) and the rest at the GK region (NES-2 and NES-3). The functionality of NES-0 and NES-3 was unknown, hence here we have explored it with a nuclear export assay, injecting into the nucleus of MDCK cells peptides corresponding to the ZO-2 NES sequences chemically coupled to ovalbumin. We show that both NES-0 and NES-3 are functional and sensitive to leptomycin B. We also demonstrate that NES-1, previously characterized as a non functional NES, is rendered capable of nuclear export upon the acquisition of a negative charge at its Ser369 residue. Experiments performed injecting at the nucleus WT and mutated ZO-2-GST fusion proteins revealed the need of both NES-0 and NES-1, and NES-2 and NES-3 for attaining an efficient nuclear exit of the respective amino and middle segments of ZO-2. Moreover, the transfection of MDCK cells with full-length ZO-2 revealed that the mutation of any of the NES present in the molecule was sufficient to induce nuclear accumulation of the protein.     

Nuclear localization of the tight junction protein ZO-2 in epithelial cells

The tight junction constitutes the major barrier to solute and water flow through the paracellular space of epithelia and endothelia. It is formed by transmembrane proteins and submembranous molecules such as the MAGUKs ZOs. We have previously found that several MAGUKs, including those of the tight (ZO-1, ZO-2, and ZO-3) and septate junction (tamou and Dlg), contain one or two nuclear sorting signals located at their first PDZ and GK domains. Now we show that these proteins also contain a nuclear export signal and focus our study on the nuclear membrane shuttling of ZO-2. In sparse cultures this molecule concentrates at the nucleus in clusters, where it partially colocalizes with splicing factor SC35. Nuclear staining diminishes as the monolayer acquires confluence through a process sensitive to the nuclear export inhibitor leptomycin B. Nuclear localization can be induced by impairing cell-cell contacts, by mechanical injury. ZO-2 that shuttles from the cell periphery into the nucleus is not newly synthesized but originates from a preexistent pool. The movement of this protein is mediated by the actin cytoskeleton.     

Redistribution of the tight junction protein ZO-1 during physiological shedding of mouse intestinal epithelial cells

A novel vasodilatory influence of endothelial cell (EC) large-conductance Ca(2+)-activated K(+) (BK(Ca)) channels is present following in vivo exposure to chronic hypoxia (CH) and may exist in other pathological states. However, the mechanism of channel activation that results in altered vasoreactivity is unknown. We tested the hypothesis that CH removes an inhibitory effect of the scaffolding domain of caveolin-1 (Cav-1) on EC BK(Ca) channels to permit activation, thereby affecting vasoreactivity. Experiments were performed on gracilis resistance arteries and ECs from control and CH-exposed (380 mmHg barometric pressure for 48 h) rats. EC membrane potential was hyperpolarized in arteries from CH-exposed rats and arteries treated with the cholesterol-depleting agent methyl-β-cyclodextrin (MBCD) compared with controls. Hyperpolarization was reversed by the BK(Ca) channel antagonist iberiotoxin (IBTX) or by a scaffolding domain peptide of Cav-1 (AP-CAV). Patch-clamp experiments documented an IBTX-sensitive current in ECs from CH-exposed rats and in MBCD-treated cells that was not present in controls. This current was enhanced by the BK(Ca) channel activator NS-1619 and blocked by AP-CAV or cholesterol supplementation. EC BK(Ca) channels displayed similar unitary conductance but greater Ca(2+) sensitivity than BK(Ca) channels from vascular smooth muscle. Immunofluorescence imaging demonstrated greater association of BK(Ca) α-subunits with Cav-1 in control arteries than in arteries from CH-exposed rats, although fluorescence intensity for each protein did not differ between groups. Finally, AP-CAV restored myogenic and phenylephrine-induced constriction in arteries from CH-exposed rats without affecting controls. AP-CAV similarly restored diminished reactivity to phenylephrine in control arteries pretreated with MBCD. We conclude that CH unmasks EC BK(Ca) channel activity by removing an inhibitory action of the Cav-1 scaffolding domain that may depend on cellular cholesterol levels.     

Molecular characterization and tissue distribution of ZO-2, a tight junction protein homologous to ZO-1 and the Drosophila discs-large tumor suppressor protein

ZO-1 is a 210-225-kD peripheral membrane protein associated with cytoplasmic surfaces of the zonula occludens or tight junction. A 160- kD polypeptide, designated ZO-2, was found to coimmunoprecipitate with ZO-1 from MDCK cell extracts prepared under conditions which preserve protein associations (Gumbiner, B., T. Lowenkopf, and D. Apatira. 1991. Proc. Natl. Acad. Sci. USA. 88: 3460-3464). We have isolated ZO-2 from MDCK cell monolayers by bulk coimmunoprecipitation with ZO-1 followed by electroelution from preparative SDS-PAGE gel slices. Amino acid sequence information obtained from a ZO-2 tryptic fragment was used to isolate a partial cDNA clone from an MDCK library. The deduced amino acid sequence revealed that canine ZO-2 contains a region that is very similar to sequences in human and mouse ZO-1. This region includes both a 90-amino acid repeat domain of unknown function and guanylate kinase- like domains which are shared among members of the family of proteins that includes ZO-1, erythrocyte p55, the product of the lethal(1)discs- large-1 (dlg) gene of Drosophila, and a synapse-associated protein from rat brain, PSD-95/SAP90. The dlg gene product has been shown to act as a tumor suppressor in the imaginal disc of the Drosophila larva, although the functions of other family members have not yet been defined. A polyclonal antiserum was raised against a unique region of ZO-2 and found to exclusively label the cytoplasmic surfaces of tight junctions in MDCK plasma membrane preparations, indicating that ZO-2 is a tight junction-associated protein. Immunohistochemical staining of frozen sections of whole tissue demonstrated that ZO-2 localized to the region of the tight junction in a number of epithelia, including liver, intestine, kidney, testis, and arterial endothelium, suggesting that this protein is a ubiquitous component of the tight junction. Double- label immunofluorescence microscopy performed on cryosections of heart, a nonepithelial tissue, revealed the presence of ZO-1 but no ZO-2 staining at the fascia adherens, a specialized junction of cardiac myocytes which has previously been shown to contain ZO-1 (Itoh, M., S. Yonemura, A. Nagafuchi, S. Tsukita, and Sh. Tsukita. 1991. J. Cell Biol. 115:1449-1462). Thus it appears that ZO-2 is not a component of the fascia adherens, and that unlike ZO-1, this protein is restricted to the epithelial tight junction.     

The tight junction protein ZO-2 and Janus kinase 1 mediate intercellular communications in vascular smooth muscle cells

Recent evidence points to a multifunctional role of ZO-2, the tight junction protein of the MAGUK (membrane-associated guanylate kinase-like) family. Though ZO-2 has been found in cell types lacking tight junction structures, such as vascular smooth muscle cells (VSMC), little is known about ZO-2 function in these cells. We provide evidence that ZO-2 mediates specific homotypic cell-to-cell contacts between VSMC. Using mass spectrometry we found that ZO-2 is associated with the non-receptor tyrosine kinase Jak1. By generating specific ZO-2 constructs we further found that the N-terminal fragment of ZO-2 molecule is responsible for this interaction. Adenovirus-based expression of Jak1 inactive mutant demonstrated that Jak1 mediates ZO-2 tyrosine phosphorylation. By means of RNA silencing, expression of Jak1 mutant form and fluorescently labeled ZO-2 fusion protein we further specified that active Jak1, but not Jak1 inactive mutant, mediates ZO-2 localization to the sites of intercellular contacts. We identified the urokinase receptor uPAR as a pre-requisite for these cellular events. Functional requirement of the revealed signaling complex for VSMC network formation was confirmed in experiments using Matrigel and in contraction assay. Our findings imply involvement of the ZO-2 tight junction independent signaling complex containing Jak1 and uPAR in VSMC intercellular communications. This mechanism may contribute to vascular remodeling in occlusive cardiovascular diseases and in arteriogenesis.     

Localization of the tight junction protein, ZO-1, is modulated by extracellular calcium and cell-cell contact in Madin-Darby canine kidney epithelial cells

Using the monoclonal antibody R26.4, we have previously identified a approximately 225-kD peripheral membrane protein, named ZO-1, that is uniquely associated with the tight junction (zonula occludens) in a variety of epithelia including the Madin-Darby canine kidney (MDCK) epithelial cell line (Stevenson, B. R., J. D. Siliciano, M. S. Mooseker, and D. A. Goodenough. 1986. J. Cell Biol. 103:755-766). In this study we have analyzed the effects of cell-cell contact and extracellular calcium on the localization and the solubility of ZO-1. In confluent monolayers under normal calcium conditions, ZO-1 immunoreactivity is found exclusively at the plasma membrane in the region of the junctional complex. If MDCK cells are maintained in spinner culture under low calcium conditions, ZO-1 is diffusely organized within the cytoplasm. After the plating of suspension cells at high cell density in medium with normal calcium concentrations, ZO-1 becomes localized to the plasma membrane at sites of cell-cell contact within 5 h in a process that is independent of de novo protein synthesis. However, if suspension cells are plated at high density in low calcium medium or if suspension cells are plated at low cell density in normal calcium growth medium, ZO-1 remains diffusely organized. ZO-1 localization also becomes diffuse in monolayers that have been established in normal calcium medium and then subsequently switched into low calcium medium. These results suggest that both extracellular calcium and cell-cell contact are necessary for normal localization of ZO-1 to the plasma membrane. An analysis of the solubility properties of ZO-1 from suspension cells and monolayers revealed that high salt, nonionic detergent, and a buffer containing chelators were somewhat more effective at solubilizing ZO-1 from suspension cells than from monolayers.     

Characterization of ZO-1, a protein component of the tight junction from mouse liver and Madin-Darby canine kidney cells

ZO-1, originally identified by mAb techniques, is the first protein shown to be specifically associated with the tight junction. Here we describe and compare the physical characteristics of ZO-1 from mouse liver and the Madin-Darby canine kidney (MDCK) epithelial cell line. The ZO-1 polypeptide has an apparent size of 225 kD in mouse tissues and 210 kD in canine-derived MDCK cells as determined by SDS-PAGE/immunoblot analysis. ZO-1 from both sources is optimally solubilized from isolated plasma membranes by either 6 M urea or high pH conditions; partial solubilization occurs with 0.3 M KCl. The nonionic detergents, Triton X-100 and octyl-beta-D-glucopyranoside, do not solubilize ZO-1. These solubility properties indicate that ZO-1 is a peripherally associated membrane protein. ZO-1 was purified to electrophoretic homogeneity from [35S]methionine metabolically labeled MDCK cells by a combination of gel filtration and immunoaffinity chromatography. Purified ZO-1 has an s20,w of 5.3 and Stokes radius of 8.6 nm. These values suggest that purified ZO-1 is an asymmetric monomeric molecule. Corresponding values for mouse liver ZO-1, characterized in impure protein extracts, were 6 s20,w and 9 nm. ZO-1 was shown to be a phosphoprotein in MDCK cells metabolically labeled with [32P]orthophosphate; analysis of phosphoamino acids from purified ZO-1 revealed only phosphoserine. ZO-1 epitope number was determined by Scatchard analysis of competitive and saturable binding of two different 125I-mAbs to SDS-solubilized proteins from liver and MDCK cells immobilized on nitrocellulose. Saturation binding occurs at 26 ng mAb/mg liver and 63 ng/mg of MDCK cell protein. This is equivalent to 30,000 ZO-1 molecules per MDCK cell assuming a single epitope/ZO-1 molecule.     

Exogenous expression of the amino-terminal half of the tight junction protein ZO-3 perturbs junctional complex assembly

The functional characteristics of the tight junction protein ZO-3 were explored through exogenous expression of mutant protein constructs in MDCK cells. Expression of the amino-terminal, PSD95/dlg/ZO-1 domain-containing half of the molecule (NZO-3) delayed the assembly of both tight and adherens junctions induced by calcium switch treatment or brief exposure to the actin-disrupting drug cytochalasin D. Junction formation was monitored by transepithelial resistance measurements and localization of junction-specific proteins by immunofluorescence. The tight junction components ZO-1, ZO-2, endogenous ZO-3, and occludin were mislocalized during the early stages of tight junction assembly. Similarly, the adherens junction proteins E-cadherin and beta-catenin were also delayed in their recruitment to the cell membrane, and NZO-3 expression had striking effects on actin cytoskeleton dynamics. NZO-3 expression did not alter expression levels of ZO-1, ZO-2, endogenous ZO-3, occludin, or E-cadherin; however, the amount of Triton X-100-soluble, signaling-active beta-catenin was increased in NZO-3-expressing cells during junction assembly. In vitro binding experiments showed that ZO-1 and actin preferentially bind to NZO-3, whereas both NZO-3 and the carboxy-terminal half of the molecule (CZO-3) contain binding sites for occludin and cingulin. We hypothesize that NZO-3 exerts its dominant-negative effects via a mechanism involving the actin cytoskeleton, ZO-1, and/or beta-catenin.