Fibroblast growth factor reversal of the corneal myofibroblast phenotype.

PURPOSE: Keratocytes give rise to fibroblasts and myofibroblasts in wounded cornea. It is well established that treatment of fibroblasts with transforming growth factor (TGF) beta will induce myofibroblast differentiation. We investigated whether this differentiation could be reversed by the administration of fibroblast growth factor (FGF). METHODS: Cultured corneal myofibroblasts were plated at 200 cells/mm(2), and cells were grown in DMEM/F12 containing (1) 10% FBS or (2) 10% FBS with FGF and heparin or (3) 1% FBS or (4) 1% FBS with TGF-beta. As distinguished from the fibroblast phenotype, the myofibroblast phenotype was identified by the assembly of alpha-smooth muscle (SM) actin protein into the stress fiber cytoskeleton. To further characterize growth factor regulation of the two phenotypes, the phenotypic expression of TGF-beta receptor types I and II, cadherins, and connexin 43 by immunocytochemistry, Western blot analysis, and immunoprecipitation and of alpha-SM actin mRNA in Northern blot analysis were evaluated. RESULTS: Corneal myofibroblasts replated and grown in the presence of FGF-1 or FGF-2 (20 ng/ml) plus heparin (5 microg/ml) in 10% FBS medium had decreased expression of alpha-SM actin protein, TGF-beta receptors, and cadherins. Thus, FGF-heparin decreased the myofibroblast phenotype and promoted the fibroblast phenotype. Administration of either 20 ng/ml FGF or 5 microg/ml heparin alone was not effective. Addition of TGF-beta further enhanced the expression of alpha-SM actin mRNA and protein and cell surface expression of TGF-beta receptors in myofibroblast cultures. CONCLUSIONS: FGF-1 or -2 and heparin promoted the fibroblast phenotype and reversed the myofibroblast phenotype. This finding supports the idea that corneal myofibroblasts and fibroblasts are alternative phenotypes rather than terminally differentiated cell types.

Transforming growth factor-beta-stimulated connective tissue growth factor expression during corneal myofibroblast differentiation.

PURPOSE: Transforming growth factor beta1 (TGF-beta) stimulates the differentiation of myofibroblasts as indicated by the nascent expression of alpha-smooth muscle (alpha-SM) actin protein and its organization into stress fibers. Downstream messengers of TGF-beta in the conversion from the fibroblast to the myofibroblast phenotype were investigated. Whether TGF-beta increases the transcription of a second growth factor, connective tissue growth factor 1 (CTGF), which could mediate myofibroblast differentiation, was evaluated. CTGF, a newly identified growth factor, is highly expressed in dermal granulation tissue. METHODS: In this study, primary cultures of rabbit corneal fibroblasts were exposed to growth factors to investigate CTGF mRNA and protein expression during myofibroblast differentiation. Statistical analysis was used to evaluate the impact of growth factor treatment on myofibroblast differentiation. RESULTS: . TGF-beta treatment induced both CTGF mRNA and protein in rabbit corneal fibroblasts; in contrast, fibroblast growth factor-2 (FGF) and heparin led to a decrease in CTGF mRNA. Addition of recombinant CTGF to rabbit corneal fibroblast cultures did not significantly increase alpha-SM actin mRNA or protein nor did it appear to affect assembly of alpha-SM actin stress fibers. CONCLUSIONS: This is the first study to present evidence for the induction of CTGF by TGF-beta treatment of corneal fibroblasts. It is doubtful that CTGF is the TGF-beta mediator of the corneal fibroblast to myofibroblast transition because CTGF does not induce alpha-SM actin in subconfluent fibroblast cultures. CTGF may play a supporting role in myofibroblast differentiation.

TGF-beta receptor expression and smad2 localization are cell density dependent in fibroblasts.

PURPOSE: In nonconfluent cultures, TGF-beta induces differentiation of corneal fibroblasts to myofibroblasts. However, in confluent cultures, few fibroblasts differentiate to myofibroblasts after TGF-beta1 addition. This study investigated the hypothesis that functional TGF-beta receptor expression is greater in low-density cultures and is decreased in confluent cultures. METHODS: Northern and western blot analyses were used to detect smooth muscle (SM) a-actin message and protein. 125I-labeled TGF-beta1 was used in a radioreceptor-binding assay as an index of functional receptors on the cell surface of rabbit corneal fibroblast cultures prepared either at high density (cell-cell contact) or low density (absence of contact). Cell lysates were analyzed by SDS-PAGE and autoradiography. Total TGF-beta receptor expression was evaluated in western blot analysis. Smad2, a downstream effector of TGF-beta receptor activation, was immunodetected. RESULTS: Low-density cultures expressed more SM alpha-actin mRNA and protein than high-density cultures, indicating that the low-density cells were differentiating into myofibroblasts. When 125I-TGF-beta1 was added to low- and high-density fibroblasts, fibroblasts plated at low density bound more than fibroblasts in high density, confluent cultures. Furthermore, after the cells differentiated into myofibroblasts, they continued to bind 125I-TGF-beta1. Specificity of 125I-TGF-beta1 binding was demonstrated by complete inhibition by excess nonradioactive TGF-beta1. Localization of Smad2 was correlated with SM alpha-actin induction: Smad was nuclear in low-density cells and cytoplasmic in high-density cells. After TGF-beta1 treatment, Smad2 remained cytoplasmic in high-density cells but was localized to nuclei in cells that were nonconfluent. CONCLUSIONS: Low cell density is correlated with increased functional expression of TGF-beta receptors and promotion of signal transmission from these receptors. Thus, conditions that decrease cell density such as wounding favor myofibroblast differentiation in response to TGF-beta.

Matrix adhesion characteristics of corneal myofibroblasts.

PURPOSE: To investigate the adhesion characteristics of corneal myofibroblasts in a cell culture model. METHODS: Immunocytochemistry, immunoprecipitation, western blot analysis, and attachment assays were used to evaluate matrix adhesion characteristics of myofibroblasts. RESULTS: Myofibroblasts, defined by their expression of the smooth muscle isoform of alpha-actin, were evaluated and compared with fibroblasts. Myofibroblasts had larger vinculin-containing focal adhesions and expressed more of the classic fibronectin receptor (FNR) alpha5beta1 per cell. However, myofibroblasts had less surface expression of the higher molecular weight alpha4 subunit of another FNR, alpha4beta1, than did fibroblasts. Myofibroblasts adhered more avidly in an integrin-dependent manner to fibronectin than did fibroblasts. The attachment to fibronectin was actin-dependent for both phenotypes, but the myofibroblasts' adhesion was more resistant to disruption by cytochalasin than were fibroblasts'. In addition to the previously described expression of a 135-kDa classic cadherin, myofibroblasts also expressed a 115-kDa mesenchymal cadherin, cadherin-11. CONCLUSIONS: Differentiation of corneal fibroblasts into myofibroblasts is associated with characteristics that would indicate that the latter have a special role in wound closure. The increase in focal and cell adhesion molecules that accompanies smooth muscle-specific actin expression provides the basis for the myofibroblasts' enhanced cell-fibronectin and cell-cell adhesion.

Functional gap junctions in corneal fibroblasts and myofibroblasts.

PURPOSE: Within the corneal stroma, keratocytes communicate through gap junctions. These plasma membrane channels, which connect the cytoplasm of adjacent cells, are composed of connexins. In a cell culture model, an investigation was conducted to determine whether connexin-based gap junction intercellular communication is present in fibroblasts and myofibroblasts, both of which replace keratocytes after wounding. METHODS: Fibroblasts and myofibroblasts were grown according to preestablished methods. Phenotype was determined by immunocytochemistry. A gap junction-permeant dye, Lucifer yellow or Cascade blue, and nonpermeant 10-kDa Texas red-dextran were used. Tracer fluorescent dyes were introduced by scrape-loading or by microinjection, and their diffusion into adjacent cells was recorded photographically. Inhibition of gap junction dye transfer was elicited by treatment with 18-alpha-glycyrrhetinic acid (AGA). RESULTS: In confluent fibroblast or myofibroblast cultures, the scrape-loaded dextran probe remained within wounded cells, whereas the Lucifer yellow or Cascade blue dye diffused into adjacent intact cells. Similarly, in nonconfluent fibroblast and myofibroblast cultures, microinjected Lucifer yellow rapidly diffused from the microinjected cell to adjacent cells. Treatment with 2 microM AGA, an uncoupling agent, blocked the spread of Lucifer yellow in fibroblast and myofibroblast cultures. CONCLUSIONS: Cultured fibroblasts and myofibroblasts have functional gap junctions as has previously been demonstrated for keratocytes in vivo. Thus, fibroblasts and myofibroblasts have the ability to establish and maintain intercellular communication with themselves and with nonactivated keratocytes. This property may be critical in the wound-healing process, especially in the avascular corneal environment.

Tumor-specific cell surface expression of the-KDEL containing, endoplasmic reticular heat shock protein gp96.

Heat shock protein (HSP) gp96/grp94 contains a signal peptide at the amino terminus and a -KDEL sequence at the carboxy terminus and is a major component of the lumen of the mammalian endoplasmic reticulum (ER). We show, by a number of immunolocalization methods using light and electron microscopy, that a significant proportion of intact gp96 molecules is also expressed on the cell surface. Surface gp96 molecules truly represent surface expression and do not result from adventitious deposition of gp96 released by dead cells on to the live cells in culture. Cell surface expression of gp96 is enhanced by heat shock and exposure to reducing agents. Gp96 molecules are not released from plasma membranes by repeated salt washes, and gp96 is not an integral membrane protein. Our observations suggest that gp96 and perhaps other HSPs are anchored to the cell surface as part of larger molecular complexes, which also transport them to the cell surface.

Immunodetection of connexins and cadherins in corneal fibroblasts and myofibroblasts.

PURPOSE: In normal cornea, stromal fibroblasts (keratocytes) interact with one another by gap junctions. After corneal wounding, the remaining corneal stroma cells are phenotypically fibroblasts and myofibroblasts. For insight into the respective roles of fibroblasts and myofibroblasts in wound healing, the authors have investigated the molecular basis of cell-cell interaction in cultures of corneal fibroblasts and corneal myofibroblasts. METHODS: Using Western blot analysis and immunofluorescent microscopy, the authors determined the relative expression and localization of junction proteins-connexins, cadherins, and cadherin-associated proteins (catenins)-in cultured fibroblasts and myofibroblasts. RESULTS: In cultured corneal fibroblasts, the gap junction protein, connexin 43, was highly expressed and was localized to dense maculae; cadherins were not detected in cell-cell contacts. Cultured myofibroblasts showed the opposite pattern: Cadherins were highly expressed and localized at the cell-cell contacts, whereas myofibroblast connexin 43 was primarily intracellular. Myofibroblast cadherin was identified by a pan-cadherin antibody as a molecule of 135 kDa that reacted weakly with an N-cadherin monoclonal antibody. In addition, cadherin-associated cytoplasmic proteins, alpha- and beta-catenins, co-localized with cadherin at the cell-cell borders of the myofibroblasts. CONCLUSIONS: The presence of connexin 43 at the cell-cell borders of corneal fibroblasts is consistent with a primary communication role of junctions in confluent corneal fibroblasts. In contrast, the presence of cadherin at the cell-cel borders of myofibroblasts may provide a site for insertion of actin filaments. A cadherin-actin association could support actin-based force generation for effective wound closure.

Myofibroblasts differentiate from fibroblasts when plated at low density.

Myofibroblasts, defined by their expression of smooth muscle alpha-actin, appear at corneal and dermal incisions and promote wound contraction. We report here that cultured fibroblasts differentiate into myofibroblasts by a cell density-dependent mechanism. Fibroblasts seeded at low density (5 cells per mm2) produced a cell culture population consisting of 70-80% myofibroblasts, 5-7 days after seeding. In contrast, fibroblasts seeded at high density (500 cells per mm2) produced cultures with only 5-10% myofibroblasts. When the myofibroblast-enriched cultures were subsequently passaged at high density, the smooth muscle alpha-actin phenotype was lost within 3 days. Furthermore, initially 60% of the low density-cultured cells incorporated BrdUrd compared to 30% of cells passaged at high density. Media from myofibroblast-enriched cultures had more latent and active transforming growth factor beta (TGF-beta) than did media from fibroblast-enriched cultures. Although there was a trend towards increased numbers of myofibroblasts after addition of exogenous TGF-beta, the results did not reach statistical significance. We conclude that myofibroblast differentiation can be induced in fibroblasts by plating at low density. We propose a cell density-dependent model of myofibroblast differentiation during wounding and healing in which at least two factors interact: loss of cell contact and the presence of TGF-beta.

Integrin-dependent tyrosine phosphorylation in corneal fibroblasts.

PURPOSE: A major pathway for intracellular signaling from cell surface receptors, such as integrins, involves intracellular phosphorylation. In corneal fibroblasts, the authors have investigated the role of tyrosine phosphorylation in integrin-dependent cell adhesion to extracellular matrix. METHODS: Antibodies were used to detect phosphotyrosine-containing proteins, including focal adhesion kinase in lysates and immunoprecipitates of corneal fibroblasts. The authors used anti-phosphotyrosine antibodies to localize phosphotyrosines in fixed cultured corneal fibroblasts. Similarly, immunocytochemical detection of vinculin was used to identify focal adhesions, the subcellular structures in which integrins organize attachment to matrix extracellularly and to cytoskeletal components intracellularly. RESULTS: Suspension of corneal fibroblasts produced a dramatic decrease in detectable phosphotyrosines. During integrin-dependent fibroblast attachment to exogenously supplied fibronectin, the cytoplasmic phosphotyrosine kinase, focal adhesion kinase (FAK), pp125FAK, became tyrosine phosphorylated. However, FAK was not phosphorylated during fibroblast attachment to vitronectin or polylysine or when cells were kept in suspension. In addition, the treatment of suspended cells with antibody to the extracellular domain of fibronectin receptor caused FAK phosphorylation. Phosphotyrosine was colocalized with vinculin in newly formed focal adhesions. Focal adhesion formation was prevented by herbimycin A, an inhibitor of tyrosine kinases. CONCLUSIONS: In corneal fibroblasts, fibronectin receptor-specific signal transduction from extracellular matrix during the formation of focal adhesions requires tyrosine kinase activation, including phosphorylation of FAK. This underscores a role for the fibronectin receptor in signaling from the extracellular matrix in corneal fibroblasts.

Induction of cell migration by pro-urokinase binding to its receptor: possible mechanism for signal transduction in human epithelial cells.

A human epithelial cell line, WISH, and a mouse cell line, LB6-uPAR, transfected with the human urokinase receptor (uPAR), both expressed high affinity uPAR but undetectable levels of urokinase (uPA). In two independent assays, binding of exogenous pro-uPA produced an up to threefold enhancement of migration. The migration was time and concentration dependent and did not involve extracellular proteolysis. This biologic response suggested that uPAR can trigger an intracellular signal. Since this receptor is a glycosyl-phosphatidylinositol-linked protein, we postulated that it must do so by interacting with other proteins, among which, by analogy to other systems, would be a kinase. To test this hypothesis, we carried out a solid phase capture of uPAR from WISH cell lysates using either antibodies against uPAR or pro-uPA adsorbed to plastic wells, followed by in vitro phosphorylation of the immobilized proteins. SDS-PAGE and autoradiography revealed two phosphorylated protein bands of 47 and 55 kD. Both proteins were phosphorylated on serine residues. Partial sequence of the two proteins showed a 100% homology to cytokeratin 18 (CK18) and 8 (CK8), respectively. A similar pattern of phosphorylation was obtained with lysates from A459 cells, a lung carcinoma, but not HL60, LB6-uPAR or HEp3 cell lysates, suggesting that the identified multiprotein uPAR-complex may be specific for simple epithelia. Moreover, immunocapture with antibody to another glycosyl-phosphatidylinositol-linked protein, CD55, which is highly expressed in WISH cells, was ineffective. The kinase was tentatively identified as protein kinase C, because it was inhibited by an analogue of staurosporine more specific for PKC and not by a PKA or tyrosine kinase inhibitors. The kinase was tentatively identified as PKC epsilon because of its resistance to PMA down-modulation, independence of Ca2+ for activity, and reaction with a specific anti-PKC epsilon antibody in Western blots. Cell fractionation into cytosolic and particulate fractions revealed that all four proteins, the kinase, uPAR, CK18, and CK8, were present in the particulate fraction. In vivo, CK8, and to a lesser degree CK18, were found to be phosphorylated on serine residues. Occupation of uPAR elicited a time-dependent increase in the phosphorylation intensity of CK8, a cell shape change and a redistribution of the cytokeratin filaments. These results strongly suggest that uPAR serves not only as an anchor for uPA but participates in a signal transduction pathway resulting in a pronounced biological response.

Isolation and immunolocalization of a rat renal cortical membrane urate transporter.

Two modalities of urate transport have been reported in rat kidney, a urate/anion exchanger and a potential sensitive, uricase-like uniporter. As an initial attempt to isolate and characterize the responsible transport protein(s), rat renal cortical membranes were harvested, solubilized, and subjected to affinity chromatography with urate or xanthine as the affinity ligand. Pig liver peroxisomal uricase was purified with the same system, and the enzymatically active protein was used to generate polyclonal antibodies in rabbit. Silver stain of SDS-polyacrylamide gel electrophoresis gels of the eluted fraction containing the affinity-purified renal membrane protein(s) demonstrated bands at 25, 32, 36, and 41 kDa. On Western blot, two of these bands (32 and 36 kDa) were immunoreactive to the polyclonal antibody to pig liver uricase. In 6 of 10 studies, the affinity-purified renal membrane protein(s) also oxidized urate. Anti-pig liver uricase produced a selective and dose-dependent inhibition of the uricase-like urate uniporter in renal membrane vesicles, but did not affect the urate/anion exchanger or the sodium-dependent glucose transporter. Immunocytochemical studies of rat renal cortex with the same antibody indicated that the immunoreactivity was localized to proximal tubules. These studies demonstrate that the renal cortical plasma membranes contain urate-binding proteins, which have some functional and immunological homology to the hepatic peroxisomal core protein, uricase. Within the renal cortex, these proteins are localized to proximal tubules, the site of urate transport. Since the antibody that reacts with the affinity-purified urate-binding proteins on Western blot selectively inhibits urate transport in intact membrane vesicles, it is concluded that at least one of the affinity-purified urate-binding proteins is a uricase-like urate transporter.

Identification of integrins in cultured corneal fibroblasts and in isolated keratocytes.

PURPOSE: The integrins are a family of transmembrane glycoproteins that function in attachment of cells to one another and to the extracellular matrix. When cell--cell and cell--matrix interactions are altered, the population of integrins may change. In particular, removing cells from their normal environment may be used as a model of wounding. The current study reports the identification of the integrins expressed at the cell surface of noncultured keratocytes and of cultured corneal fibroblasts, which are derived from keratocytes grown in primary culture. METHODS: For integrin identification, the surface proteins of keratocytes and cultured corneal fibroblasts were labeled with biotin, and the integrins were immunoprecipitated using anti-integrin antibodies. Attachment assays determined (1) the extracellular matrix preference of the cultured corneal fibroblasts and (2) the effects of function-perturbing antibodies against the fibronectin receptor (alpha 5 beta 1) or against other beta 1-containing integrins. RESULTS: The integrins of noncultured keratocytes were present as heterodimeric alpha, beta surface proteins that were immunoprecipitated by anti-beta 1, anti-alpha v, anti-alpha 6, anti-alpha 3, anti-alpha 1, and anti-beta 3. Furthermore, when the keratocytes were placed in culture, the integrin pattern changed. The classic fibronectin receptor, alpha 5 beta 1, is then expressed along with additional integrins that bind to fibronectin. Using attachment assays, we determined that the cultured corneal fibroblasts prefer fibronectin to collagen, vitronectin, or laminin as extracellular matrix substrate. In addition, function-perturbing antibodies against the fibronectin receptor (alpha 5 beta 1) or against beta 1 inhibit attachment of cultured corneal fibroblasts to fibronectin. CONCLUSIONS: Receptors for fibronectin and other extracellular matrix molecules are expressed at the cell surface in cultured corneal fibroblasts, and are in position to play a significant functional role as seen in attachment to extracellular matrix.

IgM anti-Fc gamma R autoantibodies trigger neutrophil degranulation.

Anti-Fc gamma R IgM monoclonal antibodies (mAbs) isolated from lipopolysaccharide-stimulated spleen cells from tightskin (TSK) mice were found to be polyspecific, reacting with a wide variety of molecules, including double-stranded DNA, topoisomerase, RNA polymerase, and different collagen types. Approximately 60% of the polyspecific IgM mAbs have anti-Fc gamma R specificity. These anti-Fc gamma R mAbs induce the release of hydrolases from both azurophil and specific granules of human neutrophils. 25-45% of the total cellular content (determined in Nonidet P-40 lysates) of neutrophil elastase, 10-25% of beta-glucuronidase, and 30-50% of alkaline phosphatase was released after incubation with the mAbs. The degranulation process was accompanied by dramatic morphological changes shown by scanning and transmission electron microscopy. The release of hydrolytic enzymes stimulated by the IgM anti-Fc gamma R mAbs was inhibited by preincubation of neutrophils with Fab fragments of either anti-human Fc gamma RII (IV.3) or anti-human Fc gamma RIII (3G8) mAbs. The binding of the anti-Fc gamma R TSK mAbs to human neutrophils was inhibited by Fab fragments of mAb 3G8. However, we found that the TSK anti-Fc gamma R mAbs do not bind to human Fc gamma RII expressed in either CHO cells or the P388D1 mouse macrophage cell line. Since the enzyme release could be inhibited by Fab fragments of mAb IV.3, we suggest that the signal transduction may require Fc gamma RII activation subsequent to crosslinking of the glycan phosphatidyl inositol-anchored Fc gamma RIII-1. These data demonstrate for the first time that polyspecific autoantibodies with Fc gamma R specificity can trigger neutrophil enzyme release via human Fc gamma RIII-1 in vitro and indicate a possible role for such autoantibodies in autoimmune inflammatory processes.

Reversible inhibition of endocytosis in cultured neurons from the Drosophila temperature-sensitive mutant shibirets1.

The Drosophila mutant, shibirets1 (shits1), is paralyzed at restrictive temperatures (greater than 29 degrees C) by a reversible block in synaptic transmission. Heat pulses deplete synaptic vesicles in nerve terminals and inhibit endocytic internalization of plasma membrane in garland cells and oocytes. In dissociated cultures of larval central nervous system (CNS), a temperature-sensitive defect is also expressed in shits1 neurons: at 30 degrees C, growth cone formation is retarded and neurite outgrowth is arrested. We now report that we have examined constitutive endocytosis in Drosophila CNS culture and have demonstrated directly an endocytic defect in shits1 neurons. At the permissive temperature, 20-22 degrees C, both shits1 and wild-type neurons actively endocytosed fluorescein-labelled dextran (40 KD, 5%) or horseradish peroxidase (HRP, 1%). Within 5 min, HRP was seen in vesicles, cup-shaped bodies, tubules and multivesicular bodies in neurites and cell bodies. In contrast, endocytosis was inhibited in cultures derived from the temperature-sensitive paralytic shits1 by a 15 min heat pulse (30 degrees C). Even after 30 min of HRP exposure at 30 degrees C, HRP-containing membranes were absent from almost all shits1 neurites; a minority of cell bodies had a few HRP-containing vesicles. The temperature-dependent block in endocytosis was readily reversed at 20 degrees C. Interestingly, the block was overcome by high concentration of external cations: shits1 neurons in culture actively took up HRP in numerous vesicles at 30 degrees C if 18 mM Ca2+ or Mg2+ was added to the medium. Our results support the notion that membrane recycling plays a critical role in regulating neurite outgrowth. This study also provides baseline information for further mutational analysis of the mechanism underlying the membrane cycling process in cultured neurons.

Layer-by-layer desquamation of corneal epithelium and maturation of tear-facing membranes.

A method to devitalize single layers of apically exposed rabbit corneal epithelial cells through the use of digitonin is described. Devitalized cells exfoliate spontaneously as loosely cohesive, trypan-blue-stained layers, exposing underlying viable cells. Repeated application of this devitalization-exfoliation methodology results in the gradual elimination of each of the epithelial cells. The generation of corneal surfaces composed of the tear-facing membranes of all intraepithelial cell types--subsurface, wing, and basal--is thus attainable. Exposed surfaces were studied with respect to microanatomy, the binding of lectins, and the adherence of Pseudomonas aeruginosa. Microprojections (microvilli or microplicae) were absent in the basal cells but were present in all suprabasal layers, and increased gradually in density as cells approached the surface position. Wheat germ agglutinin and concanavalin A were found to bind to the tear-facing membranes of all suprabasal cell layers. The tear-facing membrane of the basal cells, in contrast, was not labeled. Within each labeled layer, the magnitude of lectin binding differed markedly from cell to cell; lectin binding decreased as the cellular area exposed to the tear surface increased. Pseudomonas were found exclusively at microprojection-free cellular areas, suggesting that inhibition of attachment is linked to the ontogeny of these microprojections.

Very low osmotic water permeability and membrane fluidity in isolated toad bladder granules.

Osmotic water permeability of the apical membrane of toad urinary epithelium is increased greatly by vasopressin (VP) and is associated with exocytic addition of granules and aggrephores at the apical surface. To determine the physiological role of granule exocytosis, we measured the osmotic water permeability and membrane fluidity of isolated granules, surface membranes and microsomes prepared from toad bladder in the presence and absence of VP. Pf was measured by stopped-flow light scattering and membrane fluidity was examined by diphenylhexatriene (DPH) fluorescence anisotropy. In response to a 75 mM inward sucrose gradient, granule size decreased with a single exponential time constant of 2.3 +/- 0.1 sec (SEM, seven preparations, 23 degrees C), corresponding to a Pf of 5 x 10(-4) cm/sec; the activation energy (Ea) for Pf was 17.6 +/- 0.8 kcal/mole. Under the same conditions, the volume of surface membrane vesicles decreased biexponentially with time constants of 0.13 and 1.9 sec; the fast component comprised approximately 70% of the signal. Granule, surface membrane and microsome time constants were unaffected by VP. However, in surface membranes, there was a small decrease (6 +/- 2%) in the fraction of surface membranes with fast time constant. DPH anisotropies were 0.253 (granules), 0.224 (surface membranes) and 0.190 (microsomes), and were unaffected by VP. We conclude: (1) granules have among the lowest water permeabilities of biological membranes, (2) granule water permeability is not altered by bladder pretreatment with VP, (3) granule membrane fluidity is remarkably lower than that of surface and microsomal membranes, and (4) rapid water transport occurs in surface membrane vesicles. The unique physical properties of the granule suggests that apical exocytic addition of granule membrane may be responsible for the low water permeability of the unstimulated apical membrane.

Identification of the plasma membrane proteolipid protein as a constituent of brain coated vesicles and synaptic plasma membrane.

We have analyzed brain coated vesicles and synaptic plasma membrane for the presence of the plasma membrane proteolipid protein. Coated vesicles were isolated from calf brain gray matter with a final purification on Sephacryl S-1000 and reisolated twice by chromatography to ensure homogeneity. Fractions were analyzed by gel electrophoresis, immunoblotting for clathrin heavy chain, and by electron microscopy. Using an immunoblotting assay we were able to demonstrate the presence of the plasma membrane proteolipid protein in these coated vesicles at a significant level (i.e., approximately 1% of the bilayer protein of these vesicles). Reisolation of coated vesicles did not diminish the concentration of the protein in this fraction. Removal of the clathrin coat proteins or exposure of the coated vesicles to 0.1 M Na2CO3 showed that the plasma membrane proteolipid protein is not removed during uncoating and lysis but is intrinsic to the membrane bilayer of these vesicles. These studies demonstrate that plasma membrane proteolipid protein represents a significant amount of the bilayer protein of coated vesicles, suggesting that these vesicles may be a transport vehicle for the intracellular movement of the plasma membrane proteolipid protein. Isolation of synaptic plasma membranes proteolipid adult rat brain and estimation of the plasma membrane proteolipid protein content using the immunoblotting method confirmed earlier studies that show this protein is present in this membrane fraction at high levels as well (approximately 1-2%). The level of this protein in the synaptic plasma membrane suggests that the synaptic plasma membrane is one major site to which these vesicles may be targeted or from which the protein is being retrieved.

ADH and phorbol ester increase immunolabeling of the toad bladder apical membrane by antibodies made to granules.

Polyclonal antibodies were raised to isolated toad bladder granules. On immunoblots, the anti-granule antiserum specifically stained components of isolated granules. Immunocytochemically, the anti-granule antiserum labeled the apical surface of the bladder. Immunolabeling increased at the apical surface when the bladder was exposed to antidiuretic hormone (ADH) serosally or phorbol ester (PMA) mucosally--conditions which stimulate apical granule exocytosis. The increase in granule epitopes on the apical surface was sixfold greater than the net increase in surface area.

Endosomal compartment of toad bladder epithelium.

Apical exocytosis and increased permeability are induced by antidiuretic hormone (ADH). After this, endocytosis is also induced by ADH and is associated with the decline in ADH-induced water permeability at the apical surface of the toad urinary bladder (9, 19, 20). During this process, horseradish peroxidase (HRP), a fluid phase marker, is taken up from the mucosal solution into endocytic tubules and multivesicular bodies. We now report that we can introduce from the apical (mucosal) side, a viral transmembrane protein (the G-protein of VSV) and that this protein can be retrieved as an integral membrane protein in endocytic membranes. This was demonstrated by immunoisolation of endosomal vesicles loaded with HRP using a monoclonal antibody against the cytoplasmic domain of the G-protein.

Isolation and characterization of granules of the toad bladder.

The electron-dense granules that lie just below the apical plasma membrane of granular epithelial cells of toad urinary bladder contribute glycoproteins to that apical membrane. Also, exocytosis of granules (and tubules) elicited by antidiuretic hormone potentially doubles that apical surface, during the same period the transport changes characteristic of the hormonal response occur. Granules separated from other membrane systems of the cells provide the material to assess the importance of the granules as glycocalyx precursors and in hormone action. We used isosmotic media to effect preliminary separations by differential centrifugation. Then granules were isolated by centrifugation on self-forming gradients of Percoll of decreasing hypertonicity. We find qualitative and quantitative changes in protein composition and enzymic activities in the isolated fractions. The primary criterion for granule purification was electron microscopic morphology. In addition, polypeptide species found in the granule fraction are limited in number and quantity. The granules are enzymically and morphologically not lysosomal in nature. Granules may provide the glycoproteins of the apical glycocalyx but they differ from the isolated plasma membrane fraction enzymically, in protein composition and in proportion of esterified cholesterol. We conclude that the granules are not "average" plasma membrane precursors. Their role in the membrane properties of the toad urinary bladder may now be evaluated by characterizing permeability and other properties of the isolated organelles.