Clues to occludin. Focus on "Knockdown of occludin expression leads to diverse phenotypic alterations in epithelial cells".

FOR MANY YEARS AFTER ITS SYSTEMATIC morphological description by Farquhar and Palade in 1963 (7), the tight junction remained a puzzle. The freeze-fracture knife exposed anastomosing ridges in platinum/carbon replicas resembling the bones of ancient fossils, and, like fossils, provided few clues to living, breathing function. Speculation abounded about the true nature of the tight junction, with those favoring protein or lipid structures squared off in competing camps (14) while others counted strands and developed models, waiting for new breakthroughs (4). All of this began to change in 1986 with the identification of ZO-1 (15), proudly named as the first bona fide component of the zonula occludens, as tight junctions were formally bestowed by Farquhar and Palade (7). Enthusiasm was tempered, however, with the recognition that ZO-1 was not an integral membrane protein, but one that associated with the cytoplasmic face of the tight junction and thus could not account for the intercellular seal or the intramembrane ridges. In 1993, Furuse et al. (8) identified occludin, the first integral membrane protein of the tight junction. Here, too, the party was short lived; deletion of the occludin gene from embryonic stem cells did not prevent differentiation of these cells into a polarized epithelium with clear tight junctions (10, 14, 16). This observation soon spurred further research that led to the discovery of claudins, a large family of membrane proteins now recognized to make up the bulk of the tight junctional strands and regulate paracellular ion flow (5, 6, 10, 14, 16). In the middle of this tremendous progress in our understanding of tight junctional structure and function, the role of occludin remained largely a mystery. In this issue of the American Journal of Physiology, Cell Physiology, Yu, Schneeberger, and colleagues now shed light on the function of occludin (17). In an elegant study, they report that occludin’s most important role may be in transducing signals from apoptotic cells through the tight junction to the actin cytoskeleton. By deftly applying small-interfering RNA technology to almost completely knock down occludin synthesis in Madin-Darby canine kidney (MDCK) cell lines, they are able to analyze tight junction function in the absence of occludin in detail. In keeping with what has now become a tradition in occludin studies, the phenotype of these cells is neither simple nor monotonic. Nevertheless, this work broadens our view of occludin and moves our concept of the tight junction even further away from that of a simple barrier toward one of a dynamic signaling machine. Occludin is tetraspanning membrane proteins with long cytoplasmic domains and a mass of 60 kDa (14, 16). Unlike the large claudin family, whose extracellular domains vary in charge, occludin’s two extracellular loops are neutral at physiological pH (14). Its long cytoplasmic tail is rich in serine, threonine, and tyrosine, many of which are frequently phosphorylated (14). In both immunoelectron microscopy and in immune replicas, occludin localizes to the tight junction and appears interspersed with claudin in strands. Unlike claudin, however, expression of occludin in cells that normally lack tight junctions does not generate the typical anastomosing network (14, 16), indicating that claudin but not occludin drives the overall structure. In tight junctions, one of the proteins that occludin associates with is ZO-1, thereby providing it with a direct linkage to the actin cytoskeleton (11, 14, 16). A variety of studies have consistently implicated occludin in the regulation of or physical participation in the paracellular barrier. Overexpression of occludin increases transepithelial resistance (TER) but, paradoxically, also increases paracellular flux, an effect that may be due to saturation of the ZO-1 pool available for complex formation (11). Expression of occludin with a truncated cytoplasmic tail also abolishes the ability of the tight junction to prevent diffusion of lipids from apical to basolateral parts of the plasma membrane (the so-called “fence” function) (1). Phosphorylation of occludin on its tail, possibly mediated by PKC isoforms or tyrosine kinases, may regulate its assembly into tight junctions as well as barrier function (3, 9, 13, 14). Of particular relevance to the paper of Yu et al. (17) are a set of observations linking occludin, paracellular permeability, and the Rho family of small GTPases. RhoA has long been known to affect tight junctions (10, 14, 16). Furthermore, in 2001, Hirase et al. (9) reported that phosphorylation of occludin and increase in paracellular permeability could be stimulated by both lysophosphatidic acid and histamine. Inhibition of either RhoA or Rho-activated kinases in endothelial cells blocked occludin phosphorylation, permeability changes, and actin cytoskeletal changes when lysophosphatidic acid was the agonist. With histamine stimulation, however, there was no effect on occludin phosphorylation or permeability, suggesting alternative pathways regulating occludin and paracellular flux (9). Recently, a Rho GEF (guanine nucleotide exchange factor), which localizes to the tight junction has been identified in MDCK cells (2). Overexpression of this GEF, called cGEFH1, a member of the Dbl family, causes increased paracellular permeability to small hydrophilic molecules, while having no effect on TER or apparent tight junctional structure. Unfortunately, the effect of cGEF-H1 on occludin phoshorylation or localization was not examined (2). The study by Yu et al. (17) now links occludin even more firmly into regulatory pathways involving RhoA and the actin cytoskeleton. Initial characterization of cell lines in which Address for reprint requests and other correspondence: K. S. Matlin, Epithelial Pathobiology, Vontz Center for Molecular Studies, ML0581, 3125 Eden Ave., Cincinnati, OH 45267-0581 (e-mail: [email protected]). Am J Physiol Cell Physiol 288: C1191–C1192, 2005; doi:10.1152/ajpcell.00067.2005.