Structure and closure of connexin gap junction channels

Connexin gap junctions comprise assembled channels penetrating two plasma membranes for which gating regulation is associated with a variety of factors, including voltage, pH, Ca²⁺, and phosphorylation. Functional studies have established that various parts of the connexin peptides are related to channel closure and electrophysiology studies have provided several working models for channel gating. The corresponding structural models supporting these findings, however, are not sufficient because only small numbers of closed connexin structures have been reported. To fully understand the gating mechanisms, the channels should be visualized in both the open and closed states. Electron crystallography and X-ray crystallography studies recently revealed three-dimensional structures of connexin channels in a couple of states in which the main difference is the conformation of the N-terminal domain, which have helped to clarify the structure in regard to channel closure. Here the closure models for connexin gap junction channels inferred from structural and functional studies are described in the context of each domain of the connexin protein associated with gating modulation.     

Voltage-dependent gap junction channels are formed by connexin32, the major gap junction protein of rat liver.

We report here experiments undertaken in pairs of hepatocytes that demonstrate a marked voltage sensivity of junctional conductance and, thus, contradict earlier findings reported by this laboratory (Spray, D.C., R.D.ginzberg, E.A., E. A. Morales, Z. Gatmaitan and I.M. Arias, 1986, J. Cell Biol. 101:135-144; Spray C.D. R.L. White, A.C. Campos de Carvalho, and M.V.L. Bennett. 1984. Biophys. J. 45:219-230) and by others (Dahl, G., T. Moller, D. Paul, R. Voellmy, and R. Werner. 1987. Science [Wash. DC] 236:1290-1293; Riverdin, E.C., and R. Weingart. 1988. Am. J. Physiol. 254:C226-C234). Expression in exogenous systems, lipid bilayers in which fragments of isolated gap junction membranes were incorporated (Young, J.D.-E., Z. Cohn, and N.B. Gilula. 1987. Cell. 48:733-743.) and noncommunicating cells transfected with connexin32 cDNA (Eghbali, B., J.A. Kessler, and D.C. Spray. 1990. Proc. Natl. Acad. Sci. USA. 87:1328-1331), support these findings and indicate that the voltage-dependent channel is composed of connexin32, the major gap junction protein of rat liver (Paul, D. 1986. J. Cell Biol. 103:123-134).     

Asymmetric configurations and N-terminal rearrangements in connexin26 gap junction channels

Gap junction channels are unique in that they possess multiple mechanisms for channel closure, several of which involve the N terminus as a key component in gating, and possibly assembly. Here, we present electron crystallographic structures of a mutant human connexin26 (Cx26M34A) and an N-terminal deletion of this mutant (Cx26M34Adel2-7) at 6-Å and 10-Å resolutions, respectively. The three-dimensional map of Cx26M34A was improved by data from 60° tilt images and revealed a breakdown of the hexagonal symmetry in a connexin hemichannel, particularly in the cytoplasmic domain regions at the ends of the transmembrane helices. The Cx26M34A structure contained an asymmetric density in the channel vestibule ("plug") that was decreased in the Cx26M34Adel2-7 structure, indicating that the N terminus significantly contributes to form this plug feature. Functional analysis of the Cx26M34A channels revealed that these channels are predominantly closed, with the residual electrical conductance showing normal voltage gating. N-terminal deletion mutants with and without the M34A mutation showed no electrical activity in paired Xenopus oocytes and significantly decreased dye permeability in HeLa cells. Comparing this closed structure with the recently published X-ray structure of wild-type Cx26, which is proposed to be in an open state, revealed a radial outward shift in the transmembrane helices in the closed state, presumably to accommodate the N-terminal plug occluding the pore. Because both Cx26del2-7 and Cx26M34Adel2-7 channels are closed, the N terminus appears to have a prominent role in stabilizing the open configuration.     

Selective effect of PDGF on connexin43 versus connexin40 comprised gap junction channels

The goals of the current study were to determine whether the conductance of Cx40 and Cx40-Cx43 mixed composition junctions was regulated by platelet-derived growth factor (PDGF)-activated signaling cascades, to ascertain the minimum number of Cx43 subunits/connexon required to confer PDGF sensitivity, and to identify specific residues in Cx43 required for this regulation. Junctional and channel conductances (g(j) and gamma(j), respectively) were determined for Cx40/Cx40, Cx43/Cx43, Cx40/Cx43, and Cx40-Cx43/Cx40-Cx43 mixed composition channels. PDGF had no effect on g(j) in Cx40/Cx40 pairs, but decreased g(j) in the remaining combinations by 53% (Cx43/Cx43), 48% (Cx40/Cx43), 41% (4:1 Cx40:Cx43 expression ratio) and 24% (10:1 Cx40:Cx43 expression ratio). Based on the predicted connexin composition of channels in cells expressing Cx40 and Cx43 at either 4:1 or 10:1 ratios, these decreases in g(j) suggest that a single subunit of Cx43 is sufficient to confer PDGF sensitivity. The effect of PDGF on g(j) involved a decrease in both gamma(j) and Po and required serine 368 in the C-terminus. These data implicate protein kinase C as the mediator of the PDGF effect and strongly suggest that acute regulation of gap junction function by PDGF-activated signaling cascades is conferred by low levels of expression of a sensitive connexin in cells that otherwise express insensitive connexins.     

Evidence for heteromeric gap junction channels formed from rat connexin43 and human connexin37

Homomeric gap junction channels are composed solely of oneconnexin type, whereas heterotypic forms contain two homomeric hemichannels but the six identical connexins of each are different fromeach other. A heteromeric gap junction channel is one that containsdifferent connexins within either or both hemichannels. The existenceof heteromeric forms has been suggested, and many cell types are knownto coexpress connexins. To determine if coexpressed connexins wouldform heteromers, we cotransfected rat connexin43 (rCx43) and humanconnexin37 (hCx37) into a cell line normally devoid of any connexinexpression and used dual whole cell patch clamp to compare the observedgap junction channel activity with that seen in cells transfected onlywith rCx43 or hCx37. We also cocultured cells transfected with hCx37 orrCx43, in which one population was tagged with a fluorescent marker tomonitor heterotypic channel activity. The cotransfected cells possessedchannel types unlike the homotypic forms of rCx43 or hCx37 or theheterotypic forms. In addition, the noninstantaneous transjunctionalconductance-transjunctional voltage(G_j/V_j)relationship for cotransfected cell pairs showed a large range ofvariability that was unlike that of the homotypic or heterotypic form.The heterotypic cell pairs displayed asymmetric voltage dependence. Theresults from the heteromeric cell pairs are inconsistent with summedbehavior of two independent homotypic populations or mixed populationsof homotypic and heterotypic channels types. TheG_j/V_jdata imply that the connexin-to-connexin interactions are significantlyaltered in cotransfected cell pairs relative to the homotypic andheterotypic forms. Heteromeric channels are a population of channelswhose characteristics could well impact differently from theirhomotypic counterparts with regard to multicellular coordinatedresponses.

     

Chemical gating of gap junction channels; roles of calcium, pH and calmodulin

Both Ca(2+) and H(+) play a role in chemical gating of gap junction channels, but, with the possible exception of Cx46 hemichannels, neither of them is likely to induce gating by a direct interaction with connexins. Some evidence suggests that low pH(i) affects gating via an increase in [Ca(2+)](i); in turn, Ca(2+) is likely to induce gating by activation of CaM, which may act directly as a gating particle. The effective concentrations of both Ca(2+) and H(+) vary depending on cell type, type of connexin expressed and procedure employed to increase their cytosolic concentrations; however, pH(i) as high as 7.2 and [Ca(2+)](i) as low as 150 nM or lower have been reported to be effective in some cells. Some data suggest that Ca(2+) and H(+) affect gating by acting synergistically, but other data do not support synergism. Chemical gating follows the activation of a slow gate distinct from the fast V(j)-sensitive gate, and there is evidence that the chemical/slow gate is V(j)-sensitive. At the single channel level, the chemical/slow gate closes the channels slowly and completely, whereas the fast V(j) gate closes the channels rapidly and incompletely. At least three molecular models of channel gating have been proposed, but all of them are mostly based on circumstantial evidence.     

Conductance and permeability of the residual state of connexin43 gap junction channels

We used cell lines expressing wild-type connexin43 and connexin43 fused with the enhanced green fluorescent protein (Cx43-EGFP) to examine conductance and perm-selectivity of the residual state of Cx43 homotypic and Cx43/Cx43-EGFP heterotypic gap junction channels. Each hemichannel in Cx43 cell-cell channel possesses two gates: a fast gate that closes channels to the residual state and a slow gate that fully closes channels; the transjunctional voltage (V(j)) closes the fast gate in the hemichannel that is on the relatively negative side. Here, we demonstrate macroscopically and at the single-channel level that the I-V relationship of the residual state rectifies, exhibiting higher conductance at higher V(j)s that are negative on the side of gated hemichannel. The degree of rectification increases when Cl(-) is replaced by Asp(-) and decreases when K(+) is replaced by TEA(+). These data are consistent with an increased anionic selectivity of the residual state. The V(j)-gated channel is not permeable to monovalent positively and negatively charged dyes, which are readily permeable through the fully open channel. These data indicate that a narrowing of the channel pore accompanies gating to the residual state. We suggest that the fast gate operates through a conformational change that introduces positive charge at the cytoplasmic vestibule of the gated hemichannel, thereby producing current rectification, increased anionic selectivity, and a narrowing of channel pore that is largely responsible for reducing channel conductance and restricting dye transfer. Consequently, the fast V(j)-sensitive gating mechanism can serve as a selectivity filter, which allows electrical coupling but limits metabolic communication.     

Gating properties of gap junction channels assembled from connexin43 and connexin43 fused with green fluorescent protein.

We used cell lines expressing wild-type connexin43 (Cx43) and Cx43 fused with enhanced green fluorescent protein (Cx43-EGFP) to examine mechanisms of gap junction channel gating. Previously it was suggested that each hemichannel in a cell-cell channel possesses two gates, a fast gate that closes channels to a nonzero conductance or residual state via fast (< approximately 2 ms) transitions and a slow gate that fully closes channels via slow transitions (> approximately 10 ms). Here we demonstrate that transjunctional voltage (V(j)) regulates both gates and that they are operating in series and in a contingent manner in which the state of one gate affects gating of the other. Cx43-EGFP channels lack fast V(j) gating to a residual state but show slow V(j) gating. Both Cx43 and Cx43-EGFP channels exhibit slow gating by chemical uncouplers such as CO(2) and alkanols. Chemical uncouplers do not induce obvious changes in Cx43-EGFP junctional plaques, indicating that uncoupling is not caused by dispersion or internalization of junctional plaques. Similarity of gating transitions during chemical gating and slow V(j) gating suggests that both gating mechanisms share common structural elements. Cx43/Cx43-EGFP heterotypic channels showed asymmetrical V(j) gating with fast transitions between open and residual states only when the Cx43 side was relatively negative. This result indicates that the fast V(j) gate of Cx43 hemichannels closes for relative negativity at its cytoplasmic end.     

Gating of gap junction channels.

Gap junctional conductance ( gj ) in various species is gated by voltage and intracellular pH (pHi). In amphibian embryos, gj is reduced to half by a 14 mV transjunctional voltage ( Vj ), a change that in fish embryo requires approximately 28 mV. Crayfish septate axon and pairs of dissociated rat myocytes show no voltage dependence of gj over a range of Vj greater than +/- 50 mV. In fish and amphibian blastomeres , gj is steeply decreased by decrease in pHi (n, Hill coefficient: 4.5) and the apparent pKH (7.3) is in the physiological range. In crayfish septate axon the pKH is lower (6.7) and the curve is less steep (n = 2.7). Rises in cytoplasmic Ca can also decrease gj but much higher concentrations are required (greater than 0.1 mM in fish blastomeres). Voltage and pH gates on gap junctions in amphibian embryos appear independent. In squid blastomeres pH gates exhibit some sensitivity to potential, both transjunctional and between inside and outside. A pharmacology of gap junctions is being developed: certain agents block gj directly (aldehydes, alcohols, NEM in crayfish); others block by decreasing pHi (esters that are hydrolyzed by intrinsic esterases, NEM in vertebrates, and, as in the experiments demonstrating the effect of pHi, weak acids). Certain agents block pH sensitivity without affecting voltage dependence (retinoic acid, glutaraldehyde, EEDQ), further indicating separateness of pH and voltage gates. These studies demonstrate a dynamics of gap junctional conductance and variability in gating in a series of possibly homologous membrane channels.     

Characterization of the gap junction protein,connexin45

Connexin45 is a gap junction protein which forms channels with unique characteristics. RNA blots demonstrated that connexin45 is expressed in a number of cell lines including WB, SK Hepl, BHK, A7r5, CLEM, and BWEM cells. Connexin45 was further studied in BWEM cells using specific affinity-purified antibodies directed against a synthetic peptide representing amino acids 285–298 of its sequence. Immunofluorescence experiments demonstrated that the BWEM cells expressed both connexin43 and connexin45 and that these connexins colocalized. Connexin45 polypeptide, immunoprecipitated from BWEM cells metabolically labeled with [³⁵S]-methionine, consisted of a predominant 48 kD polypeptide. Connexin45 and connexin43 contained radioactive phosphate when immunoprecipitated from BWEM cells metabolically labeled with [³²P]-orthophosphoric acid. This phosphate label was removed from connexin45 by alkaline phosphatase digestion. Treatment of BWEM cells with the tumor promoting agent 12-O-tetradecanoylphorbol-13-acetate (TPA) inhibited intercellular passage of microinjected Lucifer yellow. While TPA treatment induced phosphorylation of connexin43 in these cells, it reduced the expression of connexin45. Furthermore, the connexin45 expressed after TPA treatment was not phosphorylated. These results suggest that treatments which alter protein phosphorylation may regulate connexin43 and connexin45 in BWEM cells by different mechanisms.These studies were supported by National Institutes of Health grants HL45466 and EY08368. J.G.L. is supported by a fellowship from the Lucille P. Markey Foundation. E.C.B. is an Established Investigator of the American Heart Association.     

Cx23, a connexin with only four extracellular-loop cysteines, forms functional gap junction channels and hemichannels

Gap junction channels may be comprised of either connexin or pannexin proteins (innexins and pannexins). Membrane topologies of both families are similar, but sequence similarity is lacking. Recently, connexin-like sequences have been identified in mammalian and zebrafish genomes that have only four conserved cysteines in the extracellular domains (Cx23), a feature of the pannexins. Phylogenetic analyses of the non-canonical "C4" connexins reveal that these sequences are indeed connexins. Functional assays reveal that the Cx23 gap junctions are capable of sharing neurobiotin, and further, that Cx23 connexins form hemichannels in vitro.     

Regulation of connexin degradation as a mechanism to increase gap junction assembly and function

Connexins, the integral membrane protein constituents of gap junctions, are degraded at a rate (t(12) = 1.5-5 h) much faster than most other cell surface proteins. Although the turnover of connexins has been shown to be sensitive to inhibitors of either the lysosome or of the proteasome, how connexins are targeted for degradation and whether this process can be regulated to affect intercellular communication is unknown. We show here that reducing connexin degradation with inhibitors of the proteasome (but not with lysosomal blockers) is associated with a striking increase in gap junction assembly and intercellular dye transfer in cells inefficient in both processes under basal conditions. The effect of proteasome inhibitors on wild-type connexin stability, assembly, and function was mimicked by treatment of assembly-inefficient cells with inhibitors of protein synthesis such as cycloheximide. Sensitivity of connexin degradation to cycloheximide, but not to proteasome inhibitors, was abolished when connexins were rendered structurally abnormal by perturbation of essential disulfide bonds or by mutation. Our findings provide the first evidence that intercellular communication can be up-regulated at the level of connexin turnover and that a short-lived protein may be required for conformationally mature connexins to become substrates of proteasomal degradation.     

Regulation of connexin biosynthesis, assembly, gap junction formation, and removal

Gap junctions (GJs) are the only known cellular structures that allow a direct transfer of signaling molecules from cell-to-cell by forming hydrophilic channels that bridge the opposing membranes of neighboring cells. The crucial role of GJ-mediated intercellular communication (GJIC) for coordination of development, tissue function, and cell homeostasis is now well documented. In addition, recent findings have fueled the novel concepts that connexins, although redundant, have unique and specific functions, that GJIC may play a significant role in unstable, transient cell-cell contacts, and that GJ hemi-channels by themselves may function in intra-/extracellular signaling. Assembly of these channels is a complicated, highly regulated process that includes biosynthesis of the connexin subunit proteins on endoplasmic reticulum membranes, oligomerization of compatible subunits into hexameric hemi-channels (connexons), delivery of the connexons to the plasma membrane, head-on docking of compatible connexons in the extracellular space at distinct locations, arrangement of channels into dynamic, spatially and temporally organized GJ channel aggregates (so-called plaques), and coordinated removal of channels into the cytoplasm followed by their degradation. Here we review the current knowledge of the processes that lead to GJ biosynthesis and degradation, draw comparisons to other membrane proteins, highlight novel findings, point out contradictory observations, and provide some provocative suggestive solutions.     

Asparagine 175 of connexin32 is a critical residue for docking and forming functional heterotypic gap junction channels with connexin26

The gap junction channel is formed by proper docking of two hemichannels. Depending on the connexin(s) in the hemichannels, homotypic and heterotypic gap junction channels can be formed. Previous studies suggest that the extracellular loop 2 (E2) is an important molecular domain for heterotypic compatibility. Based on the crystal structure of the Cx26 gap junction channel and homology models of heterotypic channels, we analyzed docking selectivity for several hemichannel pairs and found that the hydrogen bonds between E2 domains are conserved in a group of heterotypically compatible hemichannels, including Cx26 and Cx32 hemichannels. According to our model analysis, Cx32N175Y mutant destroys three hydrogen bonds in the E2-E2 interactions due to steric hindrance at the heterotypic docking interface, which makes it unlikely to dock with the Cx26 hemichannel properly. Our experimental data showed that Cx26-red fluorescent protein (RFP) and Cx32-GFP were able to traffic to cell-cell interfaces forming gap junction plaques and functional channels in transfected HeLa/N2A cells. However, Cx32N175Y-GFP exhibited mostly intracellular distribution and was occasionally observed in cell-cell junctions. Double patch clamp analysis demonstrated that Cx32N175Y did not form functional homotypic channels, and dye uptake assay indicated that Cx32N175Y could form hemichannels on the cell surface similar to wild-type Cx32. When Cx32N175Y-GFP- and Cx26-RFP-transfected cells were co-cultured, no colocalization was found at the cell-cell junctions between Cx32N175Y-GFP- and Cx26-RFP-expressing cells; also, no functional Cx32N175Y-GFP/Cx26-RFP heterotypic channels were identified. Both our modeling and experimental data suggest that Asn(175) of Cx32 is a critical residue for heterotypic docking and functional gap junction channel formation between the Cx32 and Cx26 hemichannels.     

Role of intramolecular interaction in connexin50: mediating the Ca2+-dependent binding of calmodulin to gap junction

Gap junction channels formed by connexin50 (Cx50) are critical for maintenance of eye lens transparency. Cleavage of the carboxyl terminus (CT) of Cx50 to produce truncated Cx50 (Cx50trunc) occurred naturally during maturation of lens fiber cells. The mechanism of its altered properties is under confirmation. It has been suggested that calmodulin (CaM) participates in gating some kinds of gap junction. Here, we performed confocal colocalization and co-immunoprecipitation experiments to study the relationships between Cx50 and CaM. Results exhibited that the CaM could colocalize Ca2+ dependently with CT in the linear area of cell-to-cell contact formed by Cx50trunc, while it could not localize in the linear area without expression of CT. Further study indicated that the CT could interact Ca2+ independently with the cytoplasmic loop (CL) of Cx50. These data put forward the importance of Ca2+-independent intramolecular interaction between CT and CL of Cx50, which mediate the Ca2+-dependent binding of CaM to Cx50. These intra- and intermolecular interactions may further improve our understanding of biological significance of the Cx50 in the eye lens.     

Expression of gap junction protein connexin43 in the adult rat cochlea: comparison with connexin26.

To elucidate whether the two different gap junction proteins connexin43 (Cx43) and connexin26 (Cx26) are expressed and localized in a similar manner in the adult rat cochlea, we performed three-dimensional confocal microscopy using cryosections and surface preparations. In the cochlear lateral wall, Cx43-positive spots were localized mainly in the stria vascularis and only a few spots were present in the spiral ligament, whereas Cx26-positive spots were detected in both the stria vascularis and the spiral ligament. In the spiral limbus, Cx43 was widely distributed, whereas Cx26 was more concentrated on the side facing the scala vestibuli and in the basal portion. In the organ of Corti, Cx43-positive spots were present between the supporting cells but they were fewer and much smaller than those of Cx26. These data demonstrated distinct differences between Cx43 and Cx26 in expression and localization in the cochlea. In addition, the area of overlap of zonula occludens-1 (ZO-1) immunolabeling with Cx43-positive spots was small, whereas it was fairly large with Cx26-positive spots in the cochlear lateral wall, suggesting that the differences are not associated with the structural difference between carboxyl terminals, i.e., those of Cx43 possess sequences for binding to ZO-1, whereas those of Cx26 lack these binding sequences.     

Inhibition of gap junction and adherens junction assembly by connexin and A-CAM antibodies

We examined the roles of the extracellular domains of a gap junction protein and a cell adhesion molecule in gap junction and adherens junction formation by altering cell interactions with antibody Fab fragments. Using immunoblotting and immunocytochemistry we demonstrated that Novikoff cells contained the gap junction protein, connexin43 (Cx43), and the cell adhesion molecule, A-CAM (N-cadherin). Cells were dissociated in EDTA, allowed to recover, and reaggregated for 60 min in media containing Fab fragments prepared from a number of antibodies. We observed no cell-cell dye transfer 4 min after microinjection in 90% of the cell pairs treated with Fab fragments of antibodies for the first or second extracellular domain of Cx43, the second extracellular domain of connexin32 (Cx32) or A-CAM. Cell-cell dye transfer was detected within 30 s in cell pairs treated with control Fab fragments (pre-immune serum, antibodies to the rat major histocompatibility complex or the amino or carboxyl termii of Cx43). We observed no gap junctions by freeze-fracture EM and no adherens junctions by thin section EM between cells treated with the Fab fragments that blocked cell-cell dye transfer. Gap junctions were found on approximately 50% of the cells in control samples using freeze-fracture EM. We demonstrated with reaggregated Novikoff cells that: (a) functional interactions of the extracellular domains of the connexins were necessary for the formation of gap junction channels; (b) cell interactions mediated by A-CAM were required for gap junction assembly; and (c) Fab fragments of antibodies for A-CAM or connexin extracellular domains blocked adherens junction formation.     

Isolation and purification of gap junction channels

This paper reports methods we have developed to solubilize gap junction channels, or connexons, from isolated gap junctions and to purify them in milligram quantities. Two sources of material are used: rat liver gap junctions and gap junctions produced by infecting insect cells with a baculovirus containing the cDNA for human liver beta 1 protein (connexin 32). Complete solubilization is obtained with long chain detergents (lauryl dimethyl amineoxide, dodecyl maltoside) and requires high ionic strength and high pH as well as reducing conditions. The purification involves chromatography on hydroxylapatite and gel filtration on Superose 6. A homogeneous product is indicated by a single band on a silver-stained gel and a homogeneous population of doughnut-shaped particles under the electron microscope. These particles have hexameric symmetry. The purified connexons have a tendency to form aggregates: filaments and sheets. The filaments grow by end-to-end association of connexons and are nonpolar, suggesting that the connexons are paired as in the cell-to-cell channel. The sheets grow by lateral association of the filaments.