A 130-kD protein that coimmunoprecipitates with the tight junction protein ZO-1 was bulk purified from Madin-Darby canine kidney (MDCK) cells and subjected to partial endopeptidase digestion and amino acid sequencing. A resulting 19–amino acid sequence provided the basis for screening canine cDNA libraries. Five overlapping clones contained a single open reading frame of 2,694 bp coding for a protein of 898 amino acids with a predicted molecular mass of 98,414 daltons. Sequence analysis showed that this protein contains three PSD-95/SAP90, discs-large, ZO-1 (PDZ) domains, a src homology (SH3) domain, and a region similar to guanylate kinase, making it homologous to ZO-1, ZO-2, the discs large tumor suppressor gene product of Drosophila, and other members of the MAGUK family of proteins. Like ZO-1 and ZO-2, the novel protein contains a COOH-terminal acidic domain and a basic region between the first and second PDZ domains. Unlike ZO-1 and ZO-2, this protein displays a proline-rich region between PDZ2 and PDZ3 and apparently contains no alternatively spliced domain. MDCK cells stably transfected with an epitope-tagged construct expressed the exogenous polypeptide at an apparent molecular mass of ∼130 kD. Moreover, this protein colocalized with ZO-1 at tight junctions by immunofluorescence and immunoelectron microscopy. In vitro affinity analyses demonstrated that recombinant 130-kD protein directly interacts with ZO-1 and the cytoplasmic domain of occludin, but not with ZO-2. We propose that this protein be named ZO-3.

Defining how the molecular constituents of the tight junction interact is a prerequisite to understanding tight junction physiology. We utilized in vitro binding assays with purified recombinant proteins and immunoprecipitation analyses to define interactions between ZO-1, ZO-2, ZO-3, occludin, and the actin cytoskeleton. Actin cosedimentation studies showed that ZO-2, ZO-3, and occludin all interact directly with F-actin in vitro, indicating that actin is engaged in multiple interactions at the tight junction. Low speed sedimentation analyses demonstrated that neither ZO-2, ZO-3, nor occludin act as F-actin cross-linking proteins, and further evidence indicates that these proteins do not bind to actin filament ends. The binding interactions of ZO-2, ZO-3, and occludin were corroborated in vivo by immunofluorescence colocalization experiments which showed that all three proteins colocalized with actin aggregates at cell borders in cytochalasin D-treated Madin-Darby canine kidney cells. Exploration of other tight junction protein interactions demonstrated that ZO-2 binds directly to both ZO-1 and occludin. Contrary to previous beliefs, our immunoprecipitation results indicate that ZO-1, ZO-2, and ZO-3 exist in situ primarily as independent ZO-1⅐ZO-2 and ZO-1⅐ZO-3 complexes rather than a trimeric ZO-1⅐ZO-2⅐ZO-3 grouping. These studies elucidate direct binding interactions among tight junction-associated proteins, giving insight into their organization as a multimolecular structure. The tight junction (TJ)1 is the apical-most element of the junctional complex in epithelial and endothelial cells. This complex includes the TJ, adherens junctions, desmosomes, and gap junctions (1). The TJ forms a barrier to paracellular movement of substances, keeping the apical and basolateral fluid compartments on opposite sides of the epithelial cell layer distinct. The TJ also acts in the maintenance of apical versus basolateral plasma membrane compositional asymmetry by restricting the movement of lipids and integral membrane proteins within the plane of the membrane.The molecular composition of the TJ is now beginning to be described. Since the identification and characterization of the first TJ protein ZO-1 (2), a host of other proteins which localize to the TJ have been identified. The peripheral membrane proteins ZO-2 (3, 4) and ZO-3 (5), together with ZO-1, are members of the MAGUK family of proteins (6) displaying a characteristic multidomain structure comprised of the PDZ, SH3, and guanylate kinase-like (GUK) domains. ZO-1, ZO-2, and ZO-3, unlike other members of the MAGUK family, each also have an acidic domain, a basic arginine-rich region, and a proline-rich domain. Other peripheral membrane proteins localized to the cytoplasmic surface of the TJ include cingulin (7), symplekin (8), 7H6 (9), Rab3B (10), the Ras target AF-6 (11), and ASIP, an atypical protein kinase C interacting protein (12). Some of these proteins are believed to act in signal transduction events, indicating that TJ barrier dynamics may be regulated lo...

In mammalian cells, the GW182 protein localizes to cytoplasmic bodies implicated in the regulation of messenger RNA (mRNA) stability, translation, and the RNA interference pathway. Many of these functions have also been assigned to analogous yeast cytoplasmic mRNA processing bodies. We have characterized the single Drosophila melanogaster homologue of the human GW182 protein family, which we have named Gawky (GW). Drosophila GW localizes to punctate, cytoplasmic foci in an RNA-dependent manner. Drosophila GW bodies (GWBs) appear to function analogously to human GWBs, as human GW182 colocalizes with GW when expressed in Drosophila cells. The RNA-induced silencing complex component Argonaute2 and orthologues of LSm4 and Xrn1 (Pacman) associated with 5′–3′ mRNA degradation localize to some GWBs. Reducing GW activity by mutation or antibody injection during syncytial embryo development leads to abnormal nuclear divisions, demonstrating an early requirement for GWB-mediated cytoplasmic mRNA regulation. This suggests that gw represents a previously unknown member of a small group of genes that need to be expressed zygotically during early embryo development.

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 β-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 β-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 β-catenin.

We previously demonstrated that exogenous expression of a truncated form of the tight junction protein ZO-3 affected junctional complex assembly and function. Current results indicate that this ZO-3 construct influences actin cytoskeleton dynamics more globally. We show that expression of the amino-terminal half of ZO-3 (NZO-3) in Madin-Darby canine kidney cells results in a decreased number of stress fibers and focal adhesions and causes an increased rate of cell migration in a wound healing assay. We also demonstrate that RhoA activity is reduced in NZO-3-expressing cells. We determined that ZO-3 interacts with p120 catenin and AF-6, proteins localized to the junctional complex and implicated in signaling pathways important for cytoskeleton regulation and cell motility. We also provide evidence that NZO-3 interacts directly with the C terminus of ZO-3, and we propose a model where altered interactions between ZO-3 and p120 catenin in NZO-3-expressing cells affect RhoA GTPase activity. This study reveals a potential link between ZO-3 and RhoA-related signaling events.