Caco-2 cells, originated in a human colonic cancer, are currently used as model systems to study transepithelial transports. To further characterize their water permeability properties, clone P1 Caco-2 cells were cultured on permeable supports. At confluence, the transepithelial net water movement (JW), mannitol permeability (Ps), and electrical resistance (R) were simultaneously measured. The observed results were correlated with transmission and freeze-fracture electron microscopy studies and compared with those obtained, in similar experimental conditions, in a typical mammalian epithelial barrier: the rabbit rectum. When the serosal solution was made hypertonic (50 mm polyethylene glycol-PEG), the spontaneously observed secretory Jw rapidly reversed, became absorptive and then stabilized. Simultaneously, the R values dropped and Ps went up. In the case of the rabbit rectal epithelium, a similar treatment did not elicit significant changes in the water permeability during the first 20 min following the osmotic challenge while there was a significant increase in the transepithelial resistance. After exposure to serosal hypertonicity, several morphological modifications developed in the Caco-2 cells: Localized dilations in the intercellular spaces and vacuoles in the cytoplasm appeared. Nevertheless, most cells remained in contact and no evidence of cell shrinking was observed. Simultaneously, the tight-junction structure was more or less disorganized. The filament network lost its sharpness and “omega” figures appeared, bordering the intercellular spaces. In some cases the tight-junction network was completely disrupted. In the case of the rabbit rectum the structural modifications were completely different: Serosal hypertonicity rapidly induced cell shrinking and the opening of the intercellular spaces, with no noticeable change in the tight-junction structure. These results suggest that Caco-2-P1 cell membranes, contrary to the case of the basolateral membrane of rabbit rectal cells, have no water channels and that a paracellular route could play a central role in the water movements across this epithelial barrier.

Aquaporin CHIP, a 28 kDa channel forming protein, has been proposed to function as water channel in both erythrocyte and kidney proximal tubule. Recently, we have reported that in frog urinary bladder, a model of the kidney collecting tubule, polyclonal antibodies against human erythrocyte CHIP recognize and immunoprecipitate a 30 kDa protein from the epithelial cell homogenate. In the present work confocal fluorescence microscopy was used to determine the cellular and subcellular localization of CHIP28-like proteins in the urinary epithelium. A clear labeling of the apical border was found after Triton X-100 permeabilization. The labeling was distributed throughout the apical domain and not restricted to specific domains of the membrane. The staining was also present in the deeper confocal sections where the fluorescence seems to be localized at the cellular contour. No difference in the labeling patterns was observed between resting and ADH-treated bladder. Specificity of the staining was confirmed by the absence of the labeling pattern when antiserum was preadsorbed on CHIP28 protein immobilized on Immobilon P stripes. Our results suggest that CHIP-like proteins are not proteins inserted in the apical membrane during the antidiuretic response. Moreover, we do not know whether the labeling was due to the presence of CHIP28 itself or an as-yet-unidentified protein sharing immunological analogies with aquaporin CHIP.

The urinary bladder of the aquatic toad Xenopus laevis is known to exhibit a low permeability to water and a poor sensitivity to antidiuretic hormone. In order to precise the characteristics and the specific cellular mechanisms of this reduced hydroosmotic response we used a sensitive volumetric technique to monitor net water flow and studied the correlation between the anti-diuretic hormone (ADH)-induced net water flow and the fine ultrastructural appearence of the urinary bladder epithelium. Transmural net water flow was entirely dependent on the osmotic gradient across the preparation and not on the hydrostatic pressure difference. We observed the existence of a low but significant hydro-osmotic response to arginine vasopressin. Freeze-fracture electron microscopy demonstrated the presence of typical aggrephores in the subapical cytoplasm. The response to the hormone was accompanied by the appearance of typical intramembrane aggregates into the apical plasma membrane. Water permeability increase and apical aggregate insertion were both slowly but fully reversible. Except for the multilayered structure of the epithelium and the particularly low response to antidiuretic hormone, all the studied permeability and ultrastructural characteristics of the bladder were thus very similar to those observed in other sensitive epithelia such as the amphibian bladder and skin and the mammalian collecting duct which exhibit a high hydro-osmotic response to the hormone.

1. In salt-acclimated frogs, water and urea bladder permeabilities are markedly higher than in tap water-acclimated animals. 2. Intra-membrane particle aggregates (IMPA) cover an unusually large surface area of the salt-acclimated frog bladder apical plasma membrane. 3. In saline-adapted animals, proteins extracted from the apical plasma membrane contain additional species of 19, 26, 31 and 53-61 kDa. These proteins might be related to the water channels contained by IMPA.

It is clearly established that the changes induced by antidiuretic hormone (ADH) in its target epithelial cells result from the insertion in the apical membrane of new components that contain channels for water. We have already undertaken an initial study of these channels by raising polyclonal antibodies against Triton X-100 apical extracts from ADH-treated bladders and approached their purification by different adsorption steps. In the present study, we used the label-fracture technique to investigate the localization of the binding sites of the obtained polyclonal antibodies on the apical membrane of ADH-stimulated frog urinary bladder. The results obtained clearly demonstrated a preferential labeling by the selected antibodies of morphological structures such as the groove arrays and the fusion images that are generally accepted as being involved in ADH-induced changes in water permeability of the apical membrane.

Antidiuretic hormone increases the water permeability of its target epithelial tissues by triggering the insertion into the apical cell membrane of aggregated intramembrane particles that contain channels specific for water. Little is known about the chemical composition of these membrane particles and of the water channel components. Present work describes a procedure for obtaining selected antibodies that specifically recognize ADH-induced components of the apical membrane in the amphibian urinary bladder epithelial cells.

Microtubules and microfilaments were localized by an immunocytochemical method in the granular cells of the frog bladder after fixation and isolation. An extensive array of microtubules was observed in the granular cells with an orientation towards the luminal plasma membrane in the supranuclear zone. Actin filaments formed a continuous bundle that underlined the cellular membrane. After incubation in the presence of colchicine, nocodazole, or tubulozole, the microtubular network appeared fragmented but did not disappear completely. These observations are related to the role of the cytoskeleton in the permeability response of the frog bladder epithelium to vasopressin.

3,3'-diallyldiethylstilbestrol (DADES), a blocker of the facilitated diffusion of glucose, was found to interfere markedly with the hydrosmotic response to antidiuretic hormone and its related agonists. Frog urinary bladders were isolated and monitored for transmural net water flow. DADES was added either to the serosal or to the apical medium at concentrations ranging from 10(-4) M to 10(-6) M. Pretreatment for 30 min with apical 10(-4) M DADES drastically reduced the subsequent hydrosmotic response: (a) to oxytocin (4.4 x 10(-8) M) by 91.7 +/- 17.6% versus 6.2 +/- 7.8 in control; (b) to 8-bromo 3',5'-cyclic AMP by 93.5 +/- 19.4% versus 19.4 +/- 11.4%; (c) to serosal hyperosmolarity (mannitol 220 mOsm) by 99.3 +/- 0.5% versus 12.3 +/- 18.2%. This effect was dose-dependent. Inhibitory action of DADES was more effective on the apical side than on the serosal side (97.0 +/- 1.5 versus 45.8 +/- 10.8). Freeze-fracture studies revealed a modified distribution of the particles and unusual endocytotic pits and vesicles in the apical membrane of both granular and mitochondria-rich epithelial cells. These observations point to multiple and complex effects of the drug. Thus, it seems that DADES has numerous effects on urinary epithelium, which makes it a nonspecific inhibitor of water permeation. Conclusions on its use should therefore be drawn with suitable caution.

We have developed a techique for recovering apical membranous sheets from aphibian urinary bladders by gelatin stripping. The tissue is mounted on a lucite support and the apical surface is first stuck onto a gelatin-coated glass slide at 30°C. This sandwich is then chilled on ice and the bladder is pulled away from the slide. Preliminary results indicate that this simple technique could be used to remove membranous apical sheets of various sizes, almost devoid of cytoplasmic contamination and without significant damage to the underlying cell structures. The method could also be adapted to prepare perforated cells and to study the cohesive forces between the different layers of the tissue.

In the amphibian urinary bladder, water permeability is correlated with the insertion of intramembrane particle aggregates (IMPAs) into the apical plasma membrane (AM) of the granular cells. These aggregates are believed to contain water channels. Characterization of the IMPAs by comparing AM fractions of antidiuretic hormone (ADH)-treated and resting epithelia requires isolation and purification of AM-rich material, free of other cytoplasmic aggregate-containing organelles, in both cases. A technique derived from freeze-fracture was chosen to isolate large sheets of apical membrane material from frog (Rana esculenta) urinary bladder epithelium. The apical side was plated on a polylysine-coated glass slide, frozen with liquid nitrogen, and fractured. A nylon mesh was inserted between the glass slide and the bladder, in order to bring the fracture plane back to the AM periodically. Fluorescent markers were used to characterize the material having fractured with the glass slide. Samples were observed by fluorescence and phase contrast microscopy. We obtained evidence that numerous patches of fractured AM remained on the glass surface without nuclei. A phase contrast picture was obtained only at a high magnification, indicating a low thickness of the recovered material. Further characterization was made with SDS-PAGE. Protein contents of samples were extracted under various experimental conditions and the patterns of ADH-treated, resting AM samples, or whole epithelial cell crude homogenates, were compared. Staining of some bands increased under certain conditions, whereas many others disappeared. Both morphological and biochemical approaches demonstrate that the recovered material was apical in origin.

The functional and structural changes induced by apical wheat germ agglutinin (WGA) 100 μg/ml exposure on frog urinary bladder have been investigated and the possible correlation between these effects discussed. Bladders, apically exposed to WGA for 30 min to 3 hr exhibit a marked reduction of their response to antidiuretic hormone (ADH) challenge and of their hydrosmotic reactivity. Structural changes triggered by WGA trreatmentare: 1. apical invaginations of the plasma membrane, interpreted as endocytotic in nature, taking into account the results of carbohydrate cytochemical detection and horseradish peroxidase (HRP) exposure: 2. cytoskeleton disorganization and microvilli collapse. These phenomena do not interfere with cortical granule traffic and are independent of ADH challenge: they occur in ADH-stimulated bladders as well as in bladders at rest. These findings could be interpreted as follows: binding of the divalent lectin WGA to its coat specific receptors would induce changes in the apical membrane structure which in turn could provoke disorganization and disruption of apical cytoskeletal elements associated with plasma membrane. Reduction of bladder response to ADH challenge could result from a reduced recyling of aggrephores, as they are associated with cytoskeletal elements in the subapical cytoplasm. Collapse of microvilli and endocytotic events also could result from apical cytoskeleton disruption, as microvilli are sustained by bundles of actin filaments interconnected with apical cytoskeleton filaments and as plasma membrane is associated with apical cytoskeleton. However, these two last events evidently occur in ADH-challenged or non-challenged bladders.