Maturation of tight junctions in guinea-pig cecal epithelium

Summary By means of the freeze-fracture technique and in tracer studies it is demonstrated that the structure of tight junctions and the permeability to lanthanum of the guinea-pig cecal epithelium change during maturation of cells. Height and strand number of tight junctions in the apical-basal direction increase as crypt cells migrate to the surface of the epithelium. Likewise, the interlacing of continuous strands was greater in surface than in crypt junctions. The numerous free-ends, isolated individual freestrands and maculae occludentes found in crypt cells were absent in surface epithelial cells. Goblet cells, located at the bottom of crypts, displayed tight junctions similar in characteristics to those of cells located in the middle region of crypts. Cells at the surface and in middle regions of crypts possess tight junctions impermeable to lanthanum, whereas junctions between cells located at the bottom of crypts often were permeable to the tracer, indicating that permeability decreases as the epithelial cells mature. Genesis and maturation mechanisms related to structural configuration of tight junctions are discussed.     

Relationship between plasma and uterine oestradiol in pseudopregnant rabbits.

Plasma oestradiol concentrations did not change during hCG-induced pseudopregnancy except for a slight increase on Day 18. However, a marked decline was observed in tissue oestradiol on Day 3 whether expressed on the basis of tissue wet weight or protein. The lowest concentration was found at Day 6 but the Day 18 values were comparable to those of untreated controls.     

Morphology of tight junctions in the ciliary epithelium of rabbits during arachidonic acid-induced breakdown of the blood-aqueous barrier

Summary A reversible breakdown of the blood-aqueous barrier in the iridial processes of rabbits has been induced by arachidonic acid as demonstrated by the passage of horseradish peroxidase at places through the tight junctions. Freeze-fracture images reveal very discontinuous Pface ridges. However, the analysis of complementary replicas demonstrates that discontinuities of P-face ridges are always complemented by particles or short bars found in the E-face furrows. Though the problem exists of correlating freeze-fracture images of the junctional structure to the focal passage of horseradish peroxidase, the data suggest that the discontinuities of P-face ridges cannot be the structural counterpart of the passage of horseradish peroxidase. Alternative pathways of horseradish peroxidase are discussed in context with the offset bifibrillary model of the junction.Supported by a research fellowship from the Deutsche Forschungsgemeinschaft Present address: Universitäts Augen- und Poliklinik der Freien Universität, Klinikum Steglitz, Hindenburgdamm 30, 1000 Berlin 45This paper was presented in part at the International Symposium on Membrane Transport Mechanisms in the Eye, September 1984, Berlin     

Membrane events involved in fusion of uterine epithelial cells in pseudopregnant rabbits

Summary During pseudopregnancy in the rabbit some uterine epithelial cells undergo conversion into symplasmata. This event serves as a model for studies of membrane apposition, fusion and fission of the lateral membranes with the use of different ultrastructural techniques. Apposition of lateral membranes occurs by means of proliferation of the tight-junctional belt and macular tight junctions. Membrane fusion is characterized in freeze-fracture replicas by continuously running fracture planes between neighboring membrane leaflets of epithelial cells, in general without reorganization of the particles. It is suggested that the reorganization of particles as well as the blebs or vesicles of smooth membranes, which are occasionally observed, may be artefacts. Membrane fission occurs simultaneously with fusion resulting in irregularly shaped membrane holes on freeze-fracture replicas. These events are rarely seen in thin sections. Staining with tannic acid reveals that only the layers of the plasma membrane are accessible to this agent. The fusion-fission process starts in the lower region of the lateral membranes, whereas the luminal portion with the broad tight-junctional belt remains intact.Dedicated to Professor Dr. med. Dr. phil. Karl-Heinrich Knese, Stuttgart-Hohenheim, in honour of his 70th birthday     

Freeze-fracture replica studies of tight junctions in normal human bronchial epithelium

Using freeze-fracture techniques, tight junctional networks were observed in the human normal bronchial epithelium. They were morphologically classified into three types: type I was a loosely interconnected, most complicated network consisting of 7-11 roughly parallel wavy strands and situated between ciliated cells; type II was a randomly anastomosing, simple network made up of 2-4 strands and present between goblet cells; type III was an irregularly anastomosing network composed of 4-7 strands and located between a ciliated cell and a goblet cell. Type III junctions, when a goblet cell was strongly bulged, were located on the swollen ridge, the upper surface of which was separated by a deep groove from the bulged apical surface, around the lateral surface of the cell at the level of the luminal surface. The possible relation between the orientation of strands of these networks and extra- or intracellular stress was discussed.     

Changes in tight junctions of thyroid epithelium with changes in thyroid activity

The morphology of the tight junction of rat thyroid epithelium was examined in freeze-fractured material fixed in glutaraldehyde and briefly glycerinated. In normal thyroids the overall appearance of this junctional specialization resembled that of other cell types in many respects. Short-term changes in thyroid activity and hypophysectomy for 3 wk did not obviously affect the appearance of tight junctions. Feeding of the goitrogen, thiouracil, which stimulates secretion of thyroid-stimulating hormone, resulted in the appearance of some very narrow and some very wide, tight junctions or sometimes junctions with both wide and narrow regions within the same cell.     

Development of tight junctions in the caput epididymal epithelium of the mouse

The course of development of the epithelial tight junctions of the Wolffian duct and the caput epididymal principal cells in the mouse were examined by freeze-fracture. The histogenesis of the epididymis is briefly described. In the 12-day embryo, tight junction meshworks surround the entire circumference of the columnar cells in the juxtaluminal position. During fetal life, the strands are more discontinuous than those of postnatal mice, and two or more strands frequently run together. Up to 10 days of age, the basal compartments of the tight junctions are much larger than the luminal ones. Marked increases in both the number of strands and the depth of the tight junctions appear by 20 days. Strands with a terminal loop are often observed up to 16 days, except for the newborn stage, suggesting that the formation of the terminal loop is related to the active elongation of the strands. The tight junctions increase greatly in number and depth near three-cell junctions. Up to 20 days, the strands anastomose frequently, with no particular orientation to the cell axis. After 20 to 37 days, the direction of the strands becomes parallel to the luminal surface, with a decreased number of anastomoses as the lumen widens. In the adult, the number of sealing strands is about 10 within the depth of the tight junctions. Free-ended strands are seen in all stages examined. The formation of the tight junction meshworks is discussed in the light of the findings during the development.     

Formation and barrier function of tight junctions in human ovarian surface epithelium

The normal ovarian surface epithelium (OSE) is a primitive epithelium made up by a single layer of mesothelial-type epithelial cells. When these cells get trapped in the ovarian stroma, expression of epithelial specific markers, such as E-cadherin, are induced. Most epithelial cells are also characterized by the ability to form tight junctions (TJ). Incomplete TJ have earlier been demonstrated in the OSE by electron microscopy studies. We have investigated expression and localization of the TJ proteins ZO-1, occludin, and claudin-1 in tissue biopsies from normal human ovaries and OSE in culture. The dynamics of TJ formation were studied in human OSE cultured on porous filters in culture inserts by measuring trans epithelial resistance (TER) including Ca(2+) switch experiments. Confluent OSE cells were also analyzed by electron microscopy. The results show that normal human OSE has expression of all three TJ proteins investigated. These proteins, ZO-1, occludin, and claudin-1, were localized to OSE cell borders both in ovarian biopsies and in cultured OSE. There was no difference in this regard between fertile and postmenopausal women. Cells in culture were polarized and presented junctional complexes seen by electron microscopy. In the Ca(2+) switch experiments, removing free Ca(2+) transiently, TER decreased significantly (P < 0.05) in the Ca(2+)-free group compared with nontreated OSE. TER was fully restored after 24 h. N-cadherin but not E-cadherin was expressed in the OSE and localized to the cell borders. We conclude that normal human OSE express and form functional TJ both in vivo and vitro. This report also describes a method to study the influence of ovarian-derived mediators on TJ in cultured OSE.     

Molecular composition of tight and adherens junctions in the rat olfactory epithelium and fila

Tight and adherens junctions (TJs, AJs) between neurons, epithelial and glial cells provide barrier and adhesion properties in the olfactory epithelium (OE), and subserve functions such as compartmentalization and axon growth in the fila olfactoria (FO). Immunofluorescence and immunoelectronmicroscopy were combined in sections of rat OE and FO to document the cellular and subcellular localization of TJ proteins occludin(Occl), claudins(Cl) 1-5 and zonula occludens(ZO) proteins 1-3, and of AJ proteins N-cadherin(cad), E-cad, and alpha-, beta- and p120-catenin(cat). With the exception of Cl2, all TJ proteins were colocalized in OE junctions. Differences in relative immunolabeling intensities were noted between neuronal and epithelial TJs. In the FO, Cl5-reactivity was localized in olfactory ensheathing cell (OEC) junctions, Cl1-reactivity in the FO periphery, with differential colocalization with ZOs. Supporting cells formed N-cad-immunoreactive (ir) AJs with olfactory sensory neurons, E-cad-ir junctions with microvillar and gland duct cells, and both N-cad and E-cad-ir junctions in homotypic contacts. Alpha, beta- and p120-cat were localized in all AJs of the OE. AJs were scarce in the globose basal cell layer. Immature and mature neurons formed numerous contacts. In the FO, AJs were documented between OECs, between OECs and axons, and between axons. Most AJs colocalized N-cad with catenins, occasionally E-cad-ir AJs were found in the FO periphery. Characteristics of molecular composition suggest differential properties of TJs formed by neuronal, epithelial and glial cells in the OE and FO. The presence and molecular composition of AJs are consistent with a role of AJ proteins in neuroplastic processes in the peripheral olfactory pathway.     

Permeability to water in a tight epithelium: possible modulating action of gap junctions

Osmotic water flow (Jw) across tight distal nephron epithelial membranes increases upon exposure to vasopressin: following binding of the hormone to its receptors, intracellular cyclic AMP concentration increases, leading to insertion of aquaporins in the apical membrane. The involvement of intercellular communication in the process, however, has not been adequately explored. Octanol, 1.2 x 10(-3) M, a gap junction inhibitor, significantly reduced Jw (expressed as mg.20 min(-1)) in isolated toad urinary bladders (a model of the distal nephron) subjected to a transepithelial osmotic gradient and exposed to agents mimicking the vasopressin-triggered mechanism: oxytocin, 50 mIU.mL(-1) (from 185.3 +/- 28.0, P < 0.001, to 69.0 +/- 23.6, P < 0.05; Pdiff < 0.01, n = 6), and cyclic AMP, 2.5 x 10(-3) M (from 98.0 +/- 32.6, P < 0.02, to 31.0 +/- 13.9, NS; Pdiff < 0.05, n = 12), without altering the effect of nystatin, 450 U.mL(-1), which increases Jw via a mechanism unrelated to apical aquaporin insertion (163.2 +/- 16.3, P < 0.001, in controls vs. 150.3 +/- 10.4, P < 0.001, in octanol-treated bladders; Pdiff: NS, n = 6). Another gap junction blocker, carbenoxolone, 2.0 x 10(-4) M (CBX), exerted similar effects on the responses to oxytocin, 100 mIU.mL(-1), reducing the response from 256.7 +/- 33.6, P < 0.001, to 102.7 +/- 10.4, P < 0.001; Pdiff < 0.01, n = 6) and nystatin, which was unaffected (95.0 +/- 20.9, P < 0.01, vs. 132.0 +/- 27.0, P < 0.01; Pdiff: NS, n = 6). Our results suggest that either gap junctions or, alternatively, unopposed gap junction hemichannels, may be important in the regulation of Jw in the isolated toad bladder, by modulating a step in the physiological process leading to increased apical membrane permeability.     

The structure of tight junctions in the tracheal epithelium may not correlate with permeability

Summary To test the hypothesis that cigarette smoke produces changes in the morphology of tight junctions guinea pigs were exposed to cigarette smoke or air in a previously standardized fashion (Simani et al. 1974). Permeability is greatest one half hour following exposure to cigarette smoke (Hulbert et al. 1981). The animals were sacrificed at that time. The tracheal epithelium was studied using both thin-section and freeze-fracture techniques. A quantitative analysis of the organization and integrity of junctional complexes was performed for each animal. Organization was assessed by measuring and comparing areas delimited by PF fibers and EF furrows. PF fiber integrity was assessed by measuring uninterrupted lengths of fibers and furrows from freeze-fracture replicas. This assessment did not demonstrate a change in tight-junction morphology following exposure to cigarette smoke.     

Expression and function of tight junctions in the crypt epithelium of human palatine tonsils.

The human palatine tonsils have surface and crypt stratified epithelium and may be initiated via the epithelium to mount immune responses to various presenting antigens. Here we investigated the expression and function of tight junctions in the epithelium of human palatine tonsils from patients with tonsillar hypertrophy or recurrent tonsillitis. Occludin, ZO-1, JAM-1, and claudin-1, -3, -4, -7, -8, and -14 mRNAs were detected in tonsillar hypertrophy. Occludin and claudin-14 were expressed in the uppermost layer of the tonsil surface epithelium, whereas ZO-1, JAM-1, and claudin-1, -4, and -7 were found throughout the epithelium. In the crypt epithelium, claudin-4 was preferentially expressed in the upper layers. In freeze-fracture replicas, short fragments of continuous tight junction strands were observed but never formed networks. In the crypt epithelium of recurrent tonsillitis, the tracer was leaked from the surface regions where occludin and claudin-4 disappeared. Occludin, ZO-1, JAM-1, and claudin-1, -3, -4, and -14, but not claudin-7, mRNAs were decreased in recurrent tonsillitis compared with those of tonsillar hypertrophy. These studies suggest unique expression of tight junctions in human palatine tonsillar epithelium, and the crypt epithelium may possess an epithelial barrier different from that of the surface epithelium.     

Preservation of the structural integrity of tight junctions during the freeze-fracture of ciliary epithelium

It has recently been reported that losses of tight junction material could result from the freeze-fracture process. To verify this assumption, we tried to increase the possibility, if any, of losses of junction material, by inducing an important fragmentation of junctional fibrils by bathing ciliary epithelium in a 0.5M sucrose solution before glutaraldehyde fixation and freeze-fracturing at −160 °C. In spite of a significant redistribution of junctional material on both fracture faces, careful examination of complementary replicas and measurements of junction elements and interruption lengths showed that no loss of junctional material occurred in this tissue. The influence of physical parameters (i.e. temperature) on the preservation of the structural integrity of the tight junction during fracturing is now a problem to be considered.     

A mechanism for isotonic fluid flow through the tight junctions of Necturus gallbladder epithelium

During isotonic fluid flow, Necturus gallbladder epithelium mediates net fluxes of paracellular probes by a convective process. We show here that the paracellular system is modeled by permeation through three populations of channels: (i) convective parallel-sided ones of width 7.7 nm (ii) small diffusive ones of radius 0.6 nm, and (ii) large diffusive ones of radius exceeding 50 nm. The reflexion coefficient of the convective channels is very low and the calculated osmotic flow rate is close to zero when compared with the observed fluid absorptive rate of 2 x 10^–6 cm/sec. Analysis reveals that the convective channels behave as though closed to back-diffusion of probes; if this is due to solvent drag then very high fluid velocities are required, acting through minute areas. There are no transjunctional gradients that could drive the flow, and so the fluid must be propelled through the channel by components of the junction.We propose a mechanism based upon an active junctional peristalsis which allows discrimination on the basis of molecular size, in which the channels are always occluded at some point and so back-diffusion cannot occur. There is no local gradient of salt distal to the junctions and therefore the osmotic permeability of the membranes is irrelevant. High fluid velocities are not required, and the flow can occur over a substantial fraction of the junction. The mechanism must involve motile and contractile elements associated with the junction for which there is already considerable evidence.Symbols A _i filtration area of channel i;i=b (big), s (small) and c (convectional) - B constant for streamline flow - C _i concentration of probe at i - D diffusion coefficient - D ^o diffusion coefficient in free solution - d width of junction - F _i diffusive drag factor in channel i - g ionic conductivity - G _i convective drag factor in channel i - J _ij probe flux from i to j - J _net net probe flux - J _v volume flow per cm² of epithelium - l linear extent of junction per cm² epithelial plane - L length of junctional channel - L _p hydraulic conductivity - N Avogadro's number - q available filtration area fraction of channel - r _s probe molecular radius - r _c channel radius or half-width - S _i steric factor in channel i - V _w,s partial molar volume of water or salt - v _i fluid velocity in channel i - _w dynamic viscosity of water - specific conductivity - ratio of solute radius to channel radius or half-width - diffusive/pressure-driven flow ratio - reflexion coefficient     

Multifunctional strands in tight junctions

Tight junctions are one mode of cell-cell adhesion in epithelial and endothelial cellular sheets. They act as a primary barrier to the diffusion of solutes through the intercellular space, create a boundary between the apical and the basolateral plasma membrane domains, and recruit various cytoskeletal as well as signalling molecules at their cytoplasmic surface. New insights into the molecular architecture of tight junctions allow us to now discuss the structure and functions of this unique cell-cell adhesion apparatus in molecular terms.     

Structure and permeability of goblet cell tight junctions in rat small intestine

Summary Two major cell types, goblet and absorptive cells, dominate the epithelial lining of small intestinal villi. We used freezefracture replicas of rat ileal mucosa to examine the possibility that tight junction structure, known to relate to transepithelial resistance, might vary with cell type. Tight junctions between absorptive cells were uniform in structure while those associated with villus goblet cells displayed structural variability. In 23% of villus goblet cell tight junctions the strand count was less than 4 and in 30% the depth was less than 200 nm. In contrast, only 4% of absorptive cell tight junctions had less than 4 strands and only 9% had depth measurements less than 200 nm. Other structural features commonly associated with villus goblet cell tight junctions but less commonly with absorptive cell tight junctions were: deficient strand cross-linking, free-ending abluminal strands, and highly fragmented strands. Bothin vivo ileal segments and everted loops were exposed to ionic lanthanum. Dense lanthanum precipitates in tight junctions and paracellular spaces were restricted to a subpopulation of villus goblet cells and were not found between villus absorptive cells. After exposure of prefixed ileal loops to lanthanum for 1 hour, faint precipitates of lanthanum were found in 14% of tight junctions and paracellular spaces between absorptive cells compared to 42% of tight junctions and paracellular spaces adjacent to villus goblet cells. When tested in Ussing chambers, the methods used for lanthanum exposure did not lower transepithelial resistance. Everted loops exposed to ionic barium and examined by light microscopy showed dense barium precipitates in the junctional zone and region of the paracellular space of villus goblet cells but not in these regions between absorptive cells. However, the macromolecular tracers, microperoxidase, cytochromec and horseradish peroxidase, were excluded from both villus goblet cell and absorptive cell paracellular spaces inin vivo segments. These findings suggest that a subpopulation of villus goblet cells may serve as focal sites of high ionic permeability and contribute to the relatively low resistance to ionic flow which characterizes the small intestinal epithelium.     

The tight junctional protein occludin is found in the uterine epithelium of squamate reptiles

Occludin, an integral protein associated with the mammalian tight junction, has for the first time been identified in the uterus of squamate reptiles. The tight junction is made up of anastamosing strands and forms a selective barrier that regulates paracellular diffusion of solutes across uterine epithelium. Occludin exclusively labels tight junctional strands and is an excellent marker for tight junction permeability. Using western blotting and immunohistochemistry, occludin expression was examined in the uterine epithelium of five species of Australian skinks at different stages of gestation. More occludin was detected during late stage pregnancy/gravidity compared to the lower levels of occludin detected in vitellogenic and post-parturient females in three of the five species. We conclude that the paracellular permeability of the squamate uterine epithelium decreases as gestation progresses. As placental transport of ions and solutes to the embryo is highest during the last third of pregnancy in viviparous squamates, it is likely that a decrease in paracellular permeability is compensated by an upregulation of other transporting mechanisms such as histotrophy.     

Freeze-etch appearance of the tight junctions in the epithelium of small and large intestine of mice

Summary The ultrastructure of the central layer and the contributing plasma membranes of tight junctions has been studied in epithelia of the jejunum and colon of mice.Examination of freeze-etched plasma membranes of epithelial cells has revealed that they consist of a central layer, with fracturing characteristics similar to bimolecular lipid leaflets, which is covered on both sides with a layer of particles.The fusion of the outer membrane surfaces of adjacent cells in the region of the tight junction leads to the formation of a new common structure consisting of a meshwork of fibrils embedded in a matrix substance. The fibrils probably contain protein. They have a diameter of 65 ± 10 Å and are linked together so that they form around the distal end of each cell a continuous belt-like meshwork which is extended proximally at the joints where three cells meet. As the fibrillar mesh appears to be strongly attached to the central lipid layer of the two adjoining membranes, in contrast to the weakly bound surrounding matrix, it is believed that the fibrils forming the continuous meshwork could be the mechanical coupling and the sealing elements of the tight junction. Their arrangement in the form of a concertinalike mesh would make the whole structure very flexible. In the region of the junction the membranes are constricted along the lines of attachment to the fibrils and bulge outwards,i.e. towards the cytoplasm, in the areas of the matrix material. In the resulting grooves on the cytoplasmic side of the plasma membranes regularly spaced particles with a diameter of 90 ± 10 Å can be detected. Various observations suggest that these particles could be connected through the central layer of the membranes to the fibrils on the other side. This would offer a possible explanation for the known abhesion properties of tight junctions. The described structures are also evaluated in terms of current theories of cell communication.