An adherens junction protein is a member of the family of lactose-binding lectins

We previously described a pig junction protein of M(r) 37,000 found in oral epithelium but not in epidermis, limited to suprabasal cells, and colocalizing by immunofluorescence with adherens junction proteins. A 1.1-kilobase pair cDNA of the 37-kDa protein yielded an open reading frame encoding a 323-amino acid protein of 35,852 Da, and Northern analysis demonstrated a band of 1.2 kilobases in tongue RNA. Secondary structure predictions indicate that the 37% identical 16-17-kDa N- and C-terminal domains from beta-sheet-rich barrels linked by a compact proline-rich segment. The protein is 72% identical in amino acid sequence and shares symmetrical two-domain structure with L-36, a lectin of unknown function from rat intestine, indicating that the 37-kDa protein is the porcine form of L-36. Of the homologous lactose binding lectins known, two others, invertebrate lectins, share this symmetrical structure. Expression of the C-terminal domain of the pig lectin in bacteria yields a lectin which binds lactosyl-Sepharose, and binding is inhibited by lactose. The expressed protein binds a glycoprotein of 120 kDa from pig tongue epithelium on Western blots, and this is also inhibited by lactose. The findings suggest that the lectin function may be involved in the assembly of adherens junctions.

Filipin vs enzymatic localization of cholesterol in guinea pig, mink, and mallard duck testicular cells

To test the validity of filipin cytochemistry for localization of cholesterol in testicular cells, we compared the results obtained by this technique with those obtained by a two-step enzymatic method involving cholesterol esterase and cholesterol oxidase. In all the animals models tested (guinea pig, mink, and mallard duck) the disappearance of subsurface filaments along Sertoli cell junctional membranes was accompanied by a significant increase in the number of filipin-cholesterol complexes/microns 2 in these membranes. Enzyme histochemistry allowed localization of free cholesterol in the limiting membrane of multivesicular bodies, in membranes within lysosomes, in mitochondrial membranes, and in junctional membranes, with or without subsurface filaments. The method also permitted selective visualization of cholesterol esters in lipid droplets. We conclude that filipin mapping of cholesterol induces false-negative cytochemical results. The enzymatic method is superior to filipin because it allows localization of free cholesterol in junctional membranes and of cholesterol esters in lipid droplets. This compartmentalization of the compounds may represent the basis of a system that helps to maintain constant free cholesterol levels in the testis.

Plasmodesmata: composition, structure and trafficking

Plasmodesmata are highly specialized gatable trans-wall channels that interconnect contiguous cells and function in direct cytoplasm-to-cytoplasm intercellular transport. Computer-enhanced digital imaging analysis of electron micrographs of plasmodesmata has provided new information on plasmodesmatal fine structure. It is now becoming clear that plasmodesmata are dynamic quasi-organelles whose conductivity can be regulated by environmental and developmental signals. New findings suggest that signalling mechanisms exist which allow the plasmodesmatal pore to dilate to allow macromolecular transport. Plant viruses spread from cell to cell via plasmodesmata. Two distinct movement mechanisms have been elucidated. One movement mechanism involves the movement of the complete virus particle along virus-induced tubular structures within a modified plasmodesma. Apparently two virus-coded movement proteins are involved. A second movement mechanism involves the movement of a non-virion form through existing plasmodesmata. In this mechanism, the viral movement protein causes a rapid dilation of existing plasmodesmata to facilitate protein and nucleic acid movement. Techniques for the isolation of plasmodesmata have been developed and information on plasmodesma-associated proteins is now becoming available. New evidence is reviewed which suggests that plasmodesmatal composition and regulation may differ in different cells and tissues.

Differential localization and sequence analysis of capping protein beta-subunit isoforms of vertebrates

Capping protein nucleates the assembly of actin filaments and stabilizes actin filaments by binding to their barbed ends. We describe here a novel isoform of the beta subunit of chicken capping protein, the beta 2 isoform, which arises by alternative splicing. The chicken beta 1 isoform and the beta 2 isoform are identical in their amino acid sequence except for a short region at the COOH terminus; this region of the beta subunit has been implicated in binding actin. Human and mouse cDNAs of the beta 1 and beta 2 isoforms also were isolated and among these vertebrates, the COOH-terminal region of each isoform is highly conserved. In contrast, comparison of the sequences of the vertebrate beta subunit COOH-termini to those of lower eukaryotes shows no similarities. The beta 2 isoform is the predominant isoform of nonmuscle tissues and the beta 1 isoform, which was first characterized in studies of capping protein from chicken muscle, is the predominant isoform of muscle tissues, as shown by immunoblots probed with isoform-specific antibodies and by RNAse protection analysis of mRNAs. The beta 2 isoform also is a component of dynactin complex from brain, which contains the actin-related protein Arp1. Both beta-subunit isoforms are expressed in cardiac muscle but they have non-overlapping subcellular distributions. The beta 1 isoform is at Z-discs of myofibrils, and the beta 2 isoform is enriched at intercalated discs; in cardiac myocytes grown in culture, the beta 2 isoform also is a component of cell-cell junctions and at sites where myofibrils contact the sarcolemma. The biochemical basis for the differential distribution of capping protein isoforms is likely due to interaction with specific proteins at Z-discs and cell-cell junctions, or to preferential association with different actin isoforms. Thus, vertebrates have developed isoforms of capping protein that associate with distinct actin-filament arrays.

Induction of polarized cell-cell association and retardation of growth by activation of the E-cadherin-catenin adhesion system in a dispersed carcinoma line

PC9 lung carcinoma cells cannot tightly associate with one another, and therefore grow singly, despite their expression of E-cadherin, because of their lack of alpha-catenin, a cadherin-associated protein. However, when the E-cadherin is activated by transfection with alpha-catenin cDNA, they form spherical aggregates, each consisting of an enclosed monolayer cell sheet. In the present work, we examined whether the alpha-catenin-transfected cell layers expressed epithelial phenotypes, by determining the distribution of various cell adhesion molecules on their surfaces, including E-cadherin, ZO-1, desmoplakin, integrins, and laminin. In untransfected PC9 cells, all these molecules were randomly distributed on their cell surface. In the transfected cells, however, each of them was redistributed into a characteristic polarized pattern without a change in the amount of expression. Electron microscopic study demonstrated that the alpha-catenin-transfected cell layers acquired apical-basal polarity typical of simple epithelia; they formed microvilli only on the outer surface of the aggregates, and a junctional complex composed of tight junction adherens junction, and desmosome arranged in this order. These results indicate that the activation of E-cadherin triggered the formation of the junctional complex and the polarized distribution of cell surface proteins and structures. We also found that, in untransfected PC9 cells, ZO-1 formed condensed clusters and colocalized with E-cadherin, but that other adhesion molecules rarely showed such colocalization with E-cadherin, suggesting that there is some specific interaction between ZO-1 and E-cadherin even in the absence of cell-cell contacts. In addition, we found that the activation of E-cadherin caused a retardation of PC9 cell growth. Thus, we concluded that the E-cadherin-catenin adhesion system is essential not only for structural organization of epithelial cells but also for the control of their growth.

Complexus adhaerentes, a new group of desmoplakin-containing junctions in endothelial cells: II. Different types of lymphatic vessels

In diverse mammalian species, including (man, cow and rat) the very flat endothelial cells of lymphatic vessels of various organs, including the retothelial meshwork of sinus of lymph nodes, are connected by zonula-like plaque-bearing junctions which differ from the similarly structured junctions of blood vessel endothelia by the presence of desmoplakin or an as yet unknown but closely related plaque protein. These extended junctions, which also contain plakoglobin but none of the presently known desmogleins and desmocollins, are therefore different from the spot-like desmosomes (maculae adhaerentes) present in epithelia, myocardium and dendritic reticulum cells of lymphatic follicles, and are collectively subsumed under the new category of complexus adhaerentes, including the 'syndesmos' connecting the processes of the retothelial cells. The lymphatic endothelial cells possessing these special desmoplakin-containing junctions also contain the calcium-dependent transmembrane glycoproteins, V-cadherin and cadherin 5, of which the latter has also been partly localized to regions with desmoplakin-positive junctions. Possible functional reasons for the formation and maintenance of complexus adhaerentes are discussed as well as the potential value of reagents which allow their identification in relation to physiology and pathology.

Spatial and temporal relationships between cadherins and PECAM-1 in cell-cell junctions of human endothelial cells

The integrity of the endothelial layer, which lines the entire cavity of the vascular system, depends on tight adhesion of the cells to the underlying basement membrane as well as to each other. It has been previously shown that such interactions occur via membrane receptors that determine the specificity, topology, and mechanical properties of the surface adhesion. Cell-cell junctions between endothelial cells, in culture and in situ, involve both Ca(2+)-dependent and -independent mechanisms that are mediated by distinct adhesion molecules. Ca(2+)-dependent cell-cell adhesion occurs mostly via members of the cadherin family, which locally anchor the microfilament system to the plasma membrane, in adherens junctions. Ca(2+)-independent adhesions were reported to mainly involve members of the Ig superfamily. In this study, we performed three-dimensional microscopic analysis of the relative subcellular distributions of these two endothelial intercellular adhesion systems. We show that cadherins are located at adjacent (usually more apical), yet clearly distinct domains of the lateral plasma membrane, compared to PECAM-1. Moreover, cadherins were first organized in adherens junctions within 2 h after seeding of endothelial cells, forming multiple lateral patches which developed into an extensive belt-like structure over a period of 24 h. PECAM-1 became associated with surface adhesions significantly later and became progressively associated with the cadherin-containing adhesions. Cadherins and PECAM-1 also differed in their detergent extractability, reflecting differences in their mode of association with the cytoskeleton. Moreover, the two adhesion systems could be differentially modulated since short treatment with the Ca2+ chelator EGTA, disrupted the cadherin junctions leaving PECAM-1 apparently intact. These results confirm that endothelial cells possess distinct intercellular contact mechanisms that differ in their spatial and temporal organization as well as in their functional properties.

Translational diffusion measurements of a fluorescent phospholipid between MDCK-I cells support the lipid model of the tight junctions

The diffusion of a lipid fluorescent analogue across tight junctions in MDCK-I cells was studied by FRAP measurements. Cells were labelled with C6-NBD-PC at the apical cell membrane. Transepithelial resistance of MDCK-I cells was found to be higher than 3000 omega x cm2, which indicates the presence of a very tight epithelium. After bleaching of large areas, fluorescence recovery was detected by confocal laser scanning microscopy, and diffusion coefficients were calculated. The lipid probes were able to diffuse to neighbouring cells as long as tight junctions between cells were intact. The average diffusion coefficient of the lipid was determined to be 5 x 10(-9) +/- 3.5 x 10(-9) cm2/s (n = 12) at room temperature. In controls, where the tight junctions were opened by complexing Ca2+ ions with EDTA, no recovery was observed. Redistribution of C6-NBD-PC into the bleached apical membrane did not occur in experiments at temperatures below 10 degrees C. We conclude that the tight junction structure can be described by the lipid model, assuming the formation of an intermembranous hexagonal cylinder of lipids for the cell contact. In this model the exoplasmatic leaflets of the plasma membranes are continuous, thus allowing free diffusion of suitable lipids from one cell to another.

Different effects of protein kinase inhibitors on the localization of junctional proteins at cell-cell contact sites

The protein kinase inhibitor H-7 prevents the assembly of tight junctions in cultured Madin Darby Canine Kidney (MDCK) epithelial cells (Balda et al. (1991) J. Membr. Biol. 122, 193–202; Nigam et al. (1991) Biochem. Biophys. Res. Commun. 181, 548–553); however, its mechanism of action is unknown. To understand the basis of the activity of H-7 and other inhibitors we compared the effect of H-7 on the localization of proteins belonging to tight junctions and adherens-type junctions (zonula adhaerens and desmosome), and on the organization of actin microfilaments. Junction assembly was induced in MDCK cells either by the ‘Ca2+ switch’ procedure or by incubating trypsinized cells at normal extracellular Ca2+, and the cells were then immunofluorescently labeled with antibodies against cingulin, ZO-1, E-cadherin and desmoplakin, and with FITC-phalloidin. Here we show by measuring the transepithelial resistance that, in addition to H-7, H-8 and staurosporine can also significantly block the assembly of tight junctions, whereas HA1004 is poorly active. H-7 inhibited the accumulation of cingulin and ZO-1 in junctional areas most effectively when added during assembly at normal extracellular Ca2+. On the other hand, H-7 did not have major effects on the accumulation of E-cadherin and desmoplakin in the regions of cell-cell contact using either assembly protocol. Electron microscopy confirmed that H-7 does not abolish the formation of adherens-type junctions, suggesting that phosphorylation plays a different role in the assembly of tight junctions versus adherens-type junctions. Finally, in both protocols of junction assembly H-7 caused a major disorganization of actin microfilaments, suggesting that H-7 may prevent TJ assembly through its effect on the cytoskeleton.

Molecular characterization and tissue distribution of ZO-2, a tight junction protein homologous to ZO-1 and the Drosophila discs-large tumor suppressor protein

ZO-1 is a 210-225-kD peripheral membrane protein associated with cytoplasmic surfaces of the zonula occludens or tight junction. A 160-kD polypeptide, designated ZO-2, was found to coimmunoprecipitate with ZO-1 from MDCK cell extracts prepared under conditions which preserve protein associations (Gumbiner, B., T. Lowenkopf, and D. Apatira. 1991. Proc. Natl. Acad. Sci. USA. 88: 3460-3464). We have isolated ZO-2 from MDCK cell monolayers by bulk coimmunoprecipitation with ZO-1 followed by electroelution from preparative SDS-PAGE gel slices. Amino acid sequence information obtained from a ZO-2 tryptic fragment was used to isolate a partial cDNA clone from an MDCK library. The deduced amino acid sequence revealed that canine ZO-2 contains a region that is very similar to sequences in human and mouse ZO-1. This region includes both a 90-amino acid repeat domain of unknown function and guanylate kinase-like domains which are shared among members of the family of proteins that includes ZO-1, erythrocyte p55, the product of the lethal(1)discs-large-1 (dlg) gene of Drosophila, and a synapse-associated protein from rat brain, PSD-95/SAP90. The dlg gene product has been shown to act as a tumor suppressor in the imaginal disc of the Drosophila larva, although the functions of other family members have not yet been defined. A polyclonal antiserum was raised against a unique region of ZO-2 and found to exclusively label the cytoplasmic surfaces of tight junctions in MDCK plasma membrane preparations, indicating that ZO-2 is a tight junction-associated protein. Immunohistochemical staining of frozen sections of whole tissue demonstrated that ZO-2 localized to the region of the tight junction in a number of epithelia, including liver, intestine, kidney, testis, and arterial endothelium, suggesting that this protein is a ubiquitous component of the tight junction. Double-label immunofluorescence microscopy performed on cryosections of heart, a nonepithelial tissue, revealed the presence of ZO-1 but no ZO-2 staining at the fascia adherens, a specialized junction of cardiac myocytes which has previously been shown to contain ZO-1 (Itoh, M., S. Yonemura, A. Nagafuchi, S. Tsukita, and Sh. Tsukita. 1991. J. Cell Biol. 115:1449-1462). Thus it appears that ZO-2 is not a component of the fascia adherens, and that unlike ZO-1, this protein is restricted to the epithelial tight junction.

Cuprolinic blue visualization of cytosolic and membrane-associated glycosaminoglycans in the rat junctional epithelium and gingival epithelia

The gingiva of rat molars was studied at the light microscope level using glutaraldehyde as fixative, Cuprolinic Blue for visualizing polyanionic glycosaminoglycans and the autometallographic technique for enhancing the copper signal of the cationic dye. The polyanions were located inside the epithelial cells in the junctional epithelium, whereas a network located along either the plasma membrane or the intercellular spaces, or both, of the gingival oral epithelium and sulcular oral epithelium was evident with autometallography. In these cases, positive staining was limited to the basal and spinous layers, the granular and keratinized layers being unstained. With the transmission electron microscope, electron-dense aggregates were seen in the gingival lamina propria, in the basement membrane and along the plasma membrane of the keratinocytes of the basal and spinous layers of the gingival and sulcular oral epithelia. In the junctional epithelium, Cuprolinic Blue-positive granules, 25 nm in diameter, were seen in the cytoplasm. Together with some vesicles containing electron-dense material, they may account for the staining process noted after autometallography. When the ultra-thin sections were digested with bovine testicular hyaluronidase, the staining was abolished. This indicates that glycosaminoglycans were primarily responsible for the staining pattern visualized with these methods. In the junctional epithelium, the cytosolic location of the 25 nm granules reflects either transcellular transfer between the plasma membrane and the nucleus or accumulation of glycosaminoglycans in this group of keratinocytes. The glycoconjugates located inside vesicles or vacuoles are related to endocytosis and lysosomal degradation. Interstitial glycosaminoglycans seen in the two types of oral epithelium may play a role in the diffusion of water and nutriments.

Polysaccharide distribution in the cellular junctions of immature fibre cells of flax seedlings

The characteristic features of the pectins present in the walls of immature fibre cells of the hypocotyl of flax seedlings have been studied by a combination of three subtractive methods (treatment with boiling water, calcium chelator, and free endopolygalacturonase), three staining reactions (periodic acid-thiocarbohydrazide-silver, Ruthenium Red, and ferric hydroxylamine) and labelling with an endopolygalacturonase-gold probe. The primary wall and the periphery of the tricellular junctions were shown to contain pectic molecules made of blocks either with free acidic functions or methyl-esterified, these molecules being removed from the wall by splitting alpha (1-4) linkages. On the contrary, the pectic molecules in the core of the tricellular junctions were mainly with free acidic groups, but with an appreciable acetylesterification of their hydroxyl groups; and they were linked with one another chiefly by calcium bonds. This unexpected constitution of the core of the tricellular junctions may be considered to be an early marker of the cells destined to give rise to the fibre bundles of the mature plant.

Effects of modulated and continuous microwave irradiation on pyroantimonate precipitable calcium content in junctional complex of mouse small intestine

The pyroantimonate precipitable calcium content of intestinal epithelial cells was investigated in mice following total body irradiation with 2450 MHz continuous and low frequency (16 Hz) square modulated waves. In the control animals the reaction products appeared in the intercellular space of adjacent cells including intermediate junctions and desmosomes and were absent in the area of tight junctions. Immediately after low frequency modulated microwave irradiation at 0.5 and 1mW/cm2 power densities, a rapid distribution of pyroantimonate precipitable calcium content was observed. The pyroantimonate deposits were located on the cytoplasmic side of lateral membrane, in the area of junctional complex, including tight junction, and in other parts of lateral plasma membrane. These changes were reversible and 24 hours after the irradiation the distribution of pyroantimonate deposits was similar to the control. Continuous waves with same energy not altered the distribution of precipitable calcium. We conclude the low frequency modulated microwave irradiation can modify the calcium distribution without heat effects.

On the mechanisms of cell uncoupling induced by a tumor promoter phorbol ester in clone 9 cells, a rat liver epithelial cell line

It is known that in Clone 9 (C9) cells, intercellular gap junctional communication (IGJC) is rapidly blocked by the tumor promoter phorbol ester, 12-O-tetradecanoylphorbol-13-acetate (TPA), but it recovers spontaneously a few hours later and becomes refractory to TPA (Yada et al., J. Membr. Biol. 88, 217-232 (1985)). We now report that gap junctions between C9 cells contain at least two junctional proteins, connexin26 (Cx26) and connexin43 (Cx43), and that the TPA-induced changes in IGJC correlate temporally to changes in the state of phosphorylation of Cx43. The latter changes were prevented by inhibition of protein kinase C. Phosphoamino acid analysis and two-dimensional tryptic peptide maps of 32P-labeled Cx43 showed that during the TPA-induced phosphorylation at least two of the phosphorylated forms of Cx43 were differentially phosphorylated in seryl residues as compared to control. TPA induced a drastic reduction in junctional conductance as well as a redistribution of unitary gap junction channel event sizes seen in control cells. These changes were associated with retrieval of Cxs from the plasma membrane. Reappearance of gap junctions formed by Cx43 but not by Cx26 accounted for the spontaneous recovery in IGJC. It is proposed that gap junctions between C9 cells contain two types of channels each formed by Cx43 or Cx26 and that they are differentially affected during the action of TPA.

Concurrent analysis of intracellular glutathione content and gap junctional intercellular communication

The potential for performing dual analysis of intracellular glutathione levels and assessment of gap junctional intercellular communication with thiol-specific fluorescent probes in anchored cells was evaluated. Gap junction-mediated diffusion of monochlorobimane and 5-chloromethylfluorescein diacetate following intracellular loading and conjugation with glutathione was compared with 5-carboxyfluorescein diacetate (which is routinely used in laser cytometry to monitor intercellular communication) by means of fluorescence recovery after photobleaching using a variety of communication-competent and communication-incompetent cells. The rate of diffusion of fluorescence among communication-competent cells was inversely proportional to the size of the fluorescent probe employed. The thiol-specific probes were also employed to monitor depletion and synthesis of glutathione following treatments to inhibit glutathione synthesis or consume glutathione by adduct formation. Analysis of gap junctional intercellular communication following glutathione depletion revealed a direct correlation between glutathione levels and intercellular communication. These studies support the utility of the thiol-specific probes to monitor the respective role of cellular glutathione and intercellular communication in the mechanisms of cellular injury.

Reduced content of connexin43 gap junctions in ventricular myocardium from hypertrophied and ischemic human hearts

BACKGROUND

Gap junctions are a determinant of myocardial conduction. Disturbances of gap-junctional content may account for abnormalities of impulse propagation, contributing to the arrhythmic tendency and mechanical inefficiency of ischemic and hypertrophied myocardium. The aim of this study was to characterize gap junction organization in normal human ventricular myocardium and to establish whether abnormalities exist in myocardium of chronically ischemic and hypertrophied hearts.

METHODS AND RESULTS

Cardiac gap-junctional connexin43 antibodies and confocal microscopy were used in a quantitative immunohistochemical study of surgical myocardial samples to explore the structural basis of electromechanical ventricular dysfunction in chronic ischemic and hypertrophic heart diseases. Normal adult human left ventricular myocardium had a gap-junctional surface area of 0.0051 micron2/micron3 myocyte volume; gap junctions were confined to intercalated disks, of which there was a mean of 11.6 per cell. The right ventricle showed similar gap junction surface area. Left ventricular myocardium from ischemic hearts (distant from any fibrotic scarring), despite normal numbers of intercalated disks per cell, had a reduced gap junction surface area (0.0027 micron2/micron3; P = .02), as did hypertrophied myocardium (0.0031 micron2/micron3; P = .05). The cardiac myocytes in the pathological tissues were larger than normal, and estimated gap-junctional content per cell was reduced in ischemic ventricle (P = .02) compared with normal.

CONCLUSIONS

Gap junctions in normal adult human working ventricular myocardium occupy an area of 0.0051 micron2/micron3 myocyte volume. This surface area is reduced in ventricular myocardium from hearts subject to chronic hypertrophy and ischemia, despite a normal number of intercellular abutments, and this alteration may contribute to abnormal impulse propagation in these hearts.

Selective distributions of proteoglycans and their ligands in pericellular matrix of cultured fibroblasts. Implications for their roles in cell-substratum adhesion

We showed previously that a large chondroitin sulfate proteoglycan, PG-M (also known as versican), inhibits cell-substratum adhesion, while basement membrane heparan sulfate proteoglycan (recently named perlecan) does not (Yamagata et al. (1989) J. Biol. Chem. 264, 8012–8018). To extend our understanding of the adhesive function of these proteoglycans, we examined the pericellular localization of the proteoglycans and their ligands and also that of some matrix receptors and cytoskeletal molecules in various fibroblast culture systems. PG-M was abundant in the subcellular space of fibroblasts, but was excluded selectively from focal contacts where vinculin, integrins and fibronectin were localized. Hyaluronan, CD44 and tenascin were distributed similarly as PG-M. In contrast, perlecan was associated with fibronectin and was included in focal contacts. Syndecan-1, a membrane heparan sulfate/chondroitin sulfate proteoglycan, was associated with fibronectin at the cell surface, partly at focal contacts and in association with stress fibers. Thus, complexes of PG-M with hyaluronan, tenascin and CD44, are not involved in focal contacts. On the other hand, perlecan and syndecan-1 together with fibronectin may participate in focal contacts. The difference in localization between these proteoglycans may be related to their glycosaminoglycan content and to their distinctive roles in cell-substratum adhesion.

Immunohistochemical delineation of the conduction system. II: The atrioventricular node and Purkinje fibers

Using an antibody that reacts specifically with the myocytes of the conduction system of the bovine heart, we have studied the atrioventricular node and the spatial distribution of the Purkinje fibers in the bovine heart. This study was complemented by studying the distribution of the gap junction protein connexin43 in these areas in the bovine heart and in the human heart. The large Purkinje fibers in the bovine heart are arranged in a two-dimensional network underneath the endocardium. At discrete sites, these fibers branch to the Purkinje fibers situated between the muscle bundles of the ventricular mass. These intramural Purkinje fibers are arranged in sheets that form a complex three-dimensional network of lamellas. Contacts with the ventricular myocytes are found throughout the myocardial wall, with the exception of a subepicardial layer of 2-mm thickness, ie, 10% to 15% of the wall thickness. The spatial arrangement of the Purkinje fibers correlates well with data on electrophysiology. Connexin43 was not detected in the myocytes of the atrioventricular node, whereas in the Purkinje fibers of the atrioventricular bundle and of the bundle branches, abundant expression of connexin43 was found in both humans and cows. In the bovine Purkinje fibers, a remarkable subcellular distribution of connexin43 is found: it occupies the entire plasma membrane facing other Purkinje cells but not that facing the surrounding connective tissue. The structural differences in architecture of the ventricular conduction system in humans and cows seems not to result in substantial differences in conduction velocities. However, the Purkinje fiber network in the bovine heart may explain the efficient ventricular excitation, as reflected by the relatively short QRS complex compared with that in the human heart, where intramural Purkinje fibers are not found.

Immunohistochemical delineation of the conduction system. I: The sinoatrial node

We have raised a mouse monoclonal antibody that reacts specifically with the myocytes of the sinoatrial node of the bovine heart. By use of this antibody (445-6E10) and antibodies against the gap junction protein connexin43, the periphery of the sinoatrial node and the distribution of gap junctions in the nodal region were studied. The reaction patterns of 445-6E10 and anti-connexin43 are exactly complementary; ie, connexin43 was not detected in the nodal myocytes but was clearly present in the atrial myocytes. Both reaction patterns demonstrate that nodal myocytes and atrial myocytes can unambiguously be distinguished by their characteristic molecular phenotype. The transitional nodal myocytes at the periphery of the node that have intermediate morphological and electrophysiological characteristics could now clearly be defined as nodal by our immunohistochemical criteria. The center of the node is surrounded by a region of interdigitating nodal and atrial bundles. Nodal bundles, coming from the center of the node, penetrate the atrial myocardium aligned at atrial bundles, forming histological connections between nodal and atrial myocytes at regular distances. This interdigitating arrangement of bundles of connexin43-negative nodal and connexin43-positive atrial myocytes is also found in the human and rat heart. We hypothesize that the architecture of the periphery of the node is important to prevent silencing of the pacemaking nodal myocytes by the atrium while ensuring a sufficient source loading of the nodal myocytes.

Disposition and orientation of ductin (DCCD-reactive vacuolar H(+)-ATPase subunit) in mammalian membrane complexes

The disposition and orientation of mouse ductin (the subunit c of the vacuolar H(+)-ATPase) in gap junctions has been examined. Like the Nephrops norvegicus (arthropod) form, mouse ductin in the intact junctional structure is resistant to high levels of nonspecific proteinase, suggesting that it is for the most part buried in the bilayer. Antisera to an octapeptide near the N-terminus cross-react with ductins in gap junction preparations from four different mouse tissues, from chicken and Xenopus laevis liver, and from N. norvegicus hepatopancreas. The antisera and antibodies, affinity purified against the octapeptide, agglutinate isolated gap junctions, suggesting that the N-terminus is located on the exposed surface, equivalent to the cytoplasmic face of an intercellular gap junction. The antibodies also block dye coupling when injected into cells in culture, confirming the cytoplasmic location of the epitope. The lipophylic reagent dicylohexyl carbodiimide (DCCD), which targets carboxyl groups within the membrane and selectively reacts with ductin in N. norvegicus gap junction preparations, rapidly inhibits junctional communication. Bafilomycin A1, which inhibits V-ATPase and stops vacuolar acidification, does not affect dye coupling, showing that the inhibition seen with antibodies and DCCD is not an indirect consequence of their action on the ductin of V-ATPase. Consistent with this interpretation the anti-peptide antibodies do not bind to intact chromaffin granules or inhibit their V-ATPase activity, but do bind to osmotically disrupted granule membrane. This suggests that ductin has an orientation (N-terminus pointing away from the cytoplasm) in the vacuolar membrane opposite to that in the gap junction membrane.

Gap junctional intercellular communication in mouse lung epithelial cell lines: effects of cell transformation and tumor promoters

Gap junctional intercellular communication (GJIC) is reduced by neoplastic transformation and treatment with tumor promoters in many types of cells but few data exist for the lung. GJIC was therefore evaluated in non-transformed (C10) and transformed (E9, 82-132, and PCC4) mouse lung epithelial cell lines and in C10 cells treated with tumor promoters. GJIC was assessed by fluorescent dye microinjection (dye-coupling). Dye-coupling levels were highest in C10 cells (85-90% communicating cells) followed by 82-132 cells (40-50%), E9 cells (15-20%), and PCC4 cells (3-10%). Indirect immunofluorescent staining with anti-gap junction protein (connexin) antibodies revealed that C10 cells expressed gap junctions comprised of connexin43, but not connexin32 or connexin26. The tumor promoters, butylated hydroxytoluene (BHT), 12-O-tetradecanoylphorbol-13-acetate (TPA), and p,p'-dichlorodiphenyltrichloroethane (DDT), inhibited dye-coupling in C10 cells but phenobarbital (PB) did not. BHT promotes mouse lung tumor formation, PB does not, while the effects of TPA and DDT on lung tumor development have not been reported. These data indicate that cell transformation and certain tumor promoters reduce GJIC in mouse lung epithelial cells and demonstrate correlations between the in vitro inhibition of GJIC and lung tumor promotion.

A bacterial protease perturbs the paracellular barrier function of transporting epithelial monolayers in culture

Tight junctions between cells and adhesion to the substratum maintain the barrier function of epithelia throughout the body. Damage to the epithelial barrier by microbial products allows penetration of bacteria and promotion of infection. We studied the effects of Pseudomonas elastase (PE) on the barrier function of epithelia by using Madin-Darby canine kidney (MDCK) epithelial cells; these cells form tight junctions (zonula occludens [ZO]) in vitro. PE decreased electrical resistance across the monolayers in a concentration- and time-dependent manner. Immunostaining of selected proteins of the ZO and zonula adherens was used to explore the effects of PE on junctional proteins. PE-treated monolayers of MDCK cells had markedly decreased immunostaining of ZO-1, a protein of the ZO, but light microscopy of PE-treated cells revealed no obvious morphologic changes. A chromium release assay indicated that, even with marked changes in transmonolayer electrical resistance, the permeability defect was not due to membrane disruption. Fluorescence staining of F-actin indicated diminution of cellular microfilaments in PE-treated cells, but E cadherin (uvomorulin), a protein of the zonula adherens, was unaffected by the enzyme. Elastases from porcine pancreas and human leukocytes with similar enzymatic activity (6 U/ml) did not decrease transmonolayer electrical resistance or degrade ZO-1. These results suggest that PE disturbs the barrier function of epithelial monolayers, in part, by changing the cell architecture and altering at least one protein of the ZO.

Polycation-induced enhancement of epithelial paracellular permeability is independent of tight junctional characteristics

The nature of polycation-induced change in transepithelial permeability was investigated in strains I (tight) and II (leaky) MDCK epithelial monolayers. Apical exposure to poly(L-lysine) (PLL, mol. wt. (MW) approximately 20,000) induced a dose-dependent increase in transepithelial conductance (GT) in both strains which correlated with increasing transepithelial flux of extracellular markers (thiourea/inulin) indicating that PLL enhanced paracellular permeability in these epithelia. Coincident with the increase in GT, PLL also induced an inward short circuit current (Isc) which was associated with the early phase of the increase in GT and may be responsible for part of it. Morphological studies showed that immunofluorescent staining of the tight junction protein, ZO-1, was abolished following PLL exposure. In addition, F-actin staining in monolayers challenged with PLL demonstrated breaks in the zonulae occludentes at the apical surface. PLL had similar effects on monolayers of T84 and HCT-8 human intestinal cells indicating that polycation action may be general for a range of epithelial types. We conclude that epithelial exposure to polycations results in opening of the paracellular route by mechanisms which are independent of tight junction characteristics.

Ultrastructural analysis of gap junctions in C6 glioma cells transfected with connexin43 cDNA

In the past, we transfected C6 glioma cells with connexin43 cDNA, resulting in a significant increase in connexin43 mRNA and protein, as well as reduced proliferation and tumorigenesis. To investigate the morphological aspects of increased connexin43 expression in these cells, we have used a combination of immunocytochemistry, cytochemistry, and electron microscopy. By confocal immunofluorescence microscopy, connexin43 protein was localized to the plasma membrane of transfected cells and extensive intracellular accumulations of connexin43 were also demonstrated. Freeze fracture preparations showed large aggregates of particles typical of mature gap junction plaques in the plasma membrane of these cells. Ultrastructural immunogold labeling with anti-connexin43 serum revealed that connexin43 protein was present in gap junctions in the plasma membrane, some of which were found in proximity to clathrin-coated pits. In addition, various intracellular membranous profiles were immunoreactive for connexin43, including annular profiles, some with fuzzy coats and some associated with lysosome-like structures. Enzyme cytochemistry revealed that these annular gap junction profiles were often associated with acid phosphatase-positive lysosomes. These studies on the intracellular localization of gap junction protein in connexin43-transfected cells are consistent with the functional expression of the transfected connexin43 cDNA and provide a useful model to study the pathways of gap junction assembly and degradation.

A structural basis for the unequal sensitivity of the major cardiac and liver gap junctions to intracellular acidification: the carboxyl tail length

The regulation of junctional conductance (Gi) of the major cardiac (connexin43; Cx43) and liver (connexin32; Cx32) gap junction proteins by intracellular hydrogen ion concentration (pH; pHi), as well as well as that of a truncation mutant of Cx43 (M257) with 125 amino acids deleted from the COOH terminus, was characterized in pairs of Xenopus laevis oocytes expressing homologous channels. Oocytes were injected with 40 nl mRNAs (2 micrograms/microliters) encoding the respective proteins; subsequently, cells were stripped, paired, and incubated for 20-24 h. Gj was measured in oocyte pairs using the dual electrode voltage-clamp technique, while pHi was recorded simultaneously in the unstimulated cell by means of a proton-selective microelectrode. Because initial experiments showed that the pH-sensitive microelectrode responded more appropriately to acetate than to CO2 acidification, oocytes expressing Cx32 and wild type and mutant Cx43 were exposed to a sodium acetate saline, which was balanced to various levels of pH using NaOH and HCl. pH was changed in a stepwise manner, and quasi-steady-state Gj -pHi relationships were constructed from data collected at each step after both Gj and pHi had reached their respective asymptotic values. A moderate but significant increase of Gj was observed in Cx43 pairs as pHi decreased from 7.2 to 6.8. In both Cx32 and M257 pairs, Gj increased significantly over a wider pH range (i.e., between 7.2 and 6.3). Further acidification reversibly reduced Gj to zero in all oocyte pairs. Pooled data for the individual connexins obtained during uncoupling were fitted by the Hill equation; apparent 50%-maximum (pK;pKa) values were 6.6 and 6.1 for Cx43 and Cx32, respectively, and Hill coefficients were 4.2 for Cx43 and 6.2 for Cx32. Like Cx32, M257 had a more acidic pKa (6.1) and steeper Hill coefficient (6.0) than wild type Cx43. The pKa and Hill coefficient of M257 were very similar to those of Cx32. These experiments provide the first direct comparison of the effects of acidification on Gj in oocyte pairs expressing Cx43 or Cx32. The results indicate that structural differences in the connexins are the basis for their unequal sensitivity to intracellular acidification in vivo. The data further suggest that a common pH gating mechanism may exist between amino acid residues 1 and 256 in both Cx32 and Cx43. However, the longer carboxyl tail of Cx43 relative to Cx32 or M257 provides additional means to facilitate acidification-induced gating; its presence shifts the pKa from 6.1 (Cx32 and M257) to 6.6 (Cx43) in the conductance of these channels.