HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 281, Issue 12, 7994-8009, March 24, 2006

This Article Abstract Full Text (PDF) All Versions of this Article: 281/12/7994    most recent M506533200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Beahm, D. L. Articles by Sosinsky, G. E. Search for Related Content PubMed PubMed Citation Articles by Beahm, D. L. Articles by Sosinsky, G. E. Social Bookmarking                      What's this?

Mutation of a Conserved Threonine in the Third Transmembrane Helix of - and -Connexins Creates a Dominant-negative Closed Gap Junction Channel^*

Derek L. Beahm¹, Atsunori Oshima¹², Guido M. Gaietta, Galen M. Hand, Amy E. Smock, Shoshanna N. Zucker, Masoud M. Toloue³, Anjana Chandrasekhar³, Bruce J. Nicholson³, and Gina E. Sosinsky⁴

From the Department of Biological Sciences, State University of New York, Buffalo, New York 14260 and National Center for Microscopy and Imaging Research, Department of Neurosciences, University of California, San Diego, La Jolla, California 92093-0608

Received for publication, June 15, 2005 , and in revised form, December 15, 2005.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Single site mutations in connexins have provided insights about the influence specific amino acids have on gap junction synthesis, assembly, trafficking, and functionality. We have discovered a single point mutation that eliminates functionality without interfering with gap junction formation. The mutation occurs at a threonine residue located near the cytoplasmic end of the third transmembrane helix. This threonine is strictly conserved among members of the - and -connexin subgroups but not the -subgroup. In HeLa cells, connexin43 and connexin26 mutants are synthesized, traffic to the plasma membrane, and make gap junctions with the same overall appearance as wild type. We have isolated connexin26T135A gap junctions both from HeLa cells and baculovirus-infected insect Sf9 cells. By using cryoelectron microscopy and correlation averaging, difference images revealed a small but significant size change within the pore region and a slight rearrangement of the subunits between mutant and wild-type connexons expressed in Sf9 cells. Purified, detergent-solubilized mutant connexons contain both hexameric and partially disassembled structures, although wild-type connexons are almost all hexameric, suggesting that the three-dimensional mutant connexon is unstable. Mammalian cells expressing gap junction plaques composed of either connexin43T154A or connexin26T135A showed an absence of dye coupling. When expressed in Xenopus oocytes, these mutants, as well as a cysteine substitution mutant of connexin50 (connexin50T157C), failed to produce electrical coupling in homotypic and heteromeric pairings with wild type in a dominant-negative effect. This mutant may be useful as a tool for knocking down or knocking out connexin function in vitro or in vivo.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Intercellular communication is a fundamental feature of all multicellular organisms. Gap junctions are one means by which cells communicate with each other and arise as tissue cells grow and abut each other. The morphologically distinctive cell-cell junctional areas allow the exchange of ions, nutrients, and small metabolites between neighboring cells. Gap junction structures are found throughout vertebrates and invertebrates, although the primary sequences of constituent proteins are different from each other even though electron micrographs and physiological assays indicate similar quaternary structure and functionality. Gap junctions are composed of two oligomeric channel structures called connexons, with each cell supplying one connexon that docks with the other at their extracellular surfaces. The connexin family of proteins has a very conserved protein folding topology with highly conserved transmembrane and extracellular primary sequences, but contains variable regions of the cytoplasmic loop and C terminus that confer the individual physiological properties to each connexin. Each connexon is made up of a cyclic arrangement of six protein monomers, called connexins (abbreviated as Cx⁵ plus the molecular mass of the protein as predicted by the amino acid sequence, e.g. Cx32). At present, over 20 unique connexins have been sequenced (1), and over 90 sequences are found in GenBank^TM from various species ranging the evolutionary scale from fish to human.

The correct synthesis and assembly of gap junctions are critically important for development (2-4), signaling (5, 6), and homeostasis (7). Recently, the genetic characterizations of connexin diseases have led to studies focused on understanding how single amino acid changes in connexin sequences result in macroscopic symptoms seen in patients (reviewed in Refs. 8 and 9) and whether these mutations arise from defects in synthesis, trafficking, docking, or channel function. Mutations in connexin genes have been demonstrated to be the cause of several diseases such as X-linked Charcot-Marie-Tooth (CMTX) syndrome (defects in Cx32 causing a peripheral neuropathy (10, 11), non-syndromic hereditary deafness (Cx26 (12)), erythrokeratodermia variabilis (Cx31, a skin disorder (13)); cataracts (Cx46 and Cx50 (14, 15)); oculodentodigital dysplasia, a pleiotropic syndrome displaying craniofacial and limb dysmorphisms, spastic paraplegia, and neurodegeneration (Cx43 (16)); and a few cases of visceroatrial heterotaxia manifested as cardiac developmental disorders (17)). In each case, no systematic pattern has emerged as to the cause and effect of these mutations; however, these naturally occurring mutations are useful in trying to build three-dimensional models based on bioinformatics considerations (18).

Here we present the functional and structural characterization of a mutation of a threonine residue situated in the third transmembrane (TM3) helix. Except as part of a larger mutagenesis study (19), the position and amino acid make up at this site have not been investigated. The presence of this mutation does not impede the formation of channels, which in conventional EM displays the same morphology as wild-type (wt) channels. However, as we show in dye transfer and electrophysiological assays, the channels carrying this mutation are nonfunctional and may mimic a "closed pore" state. This residue is strictly conserved among the - and -connexins. Because the majority of the naturally occurring disease-related mutations are site-specific ones that result in mis-folding or mis-trafficking of connexins, this mutation is interesting because the connexin carrying this relatively minor substitution is correctly synthesized, trafficked, inserted into the plasma membrane, and capable of docking with a partner connexon, yet it is completely nonfunctional. We show electron microscopic data indicating small differences in the mutant from wild-type structure. We have mutated this residue at the equivalent positions in three connexin isoforms, and we found that function is eliminated in all three cases in intercellular channels in a dominant-negative fashion. We also show that a loss of function occurs in Cx50 hemichannels in homomeric and heteromeric combinations of wt and mutant connexins. Hence, in addition to being interesting from a structure/function relationship point of view, this mutation is potentially very useful as a tool for knocking out connexin functionality without having to delete connexin genes or for examining potential heteromeric interactions between different connexins.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Sequence Analysis of Connexins—Connexin sequences were obtained from the National Center for Biotechnology Information (NCBI) protein sequence search web site and collated into a file format readable by the ClustalW set of sequence alignment programs (European Bioinformatics Institute). These aligned sequences were used as input for dendrograms using the program TreeViewPPC and also used to generate the TM3 sequences shown in Table 1.

Transient transfections in HeLa cells were performed using FuGENE (Roche Diagnostics). Transductions were performed with recombinant virus generated in HEK293 cells. pCLNCX Cx43-GFP-4C was co-transfected with the Imgenex packaging vector pCL-Ampho. HeLa cell lines stably expressing the wt or mutant connexins were isolated after a 2-week selection in 100 µg/ml hygromycin in growth medium (Dulbecco's modified Eagle's medium, 10% fetal bovine serum, 1% penicillin/streptomycin).

Dye Injection Studies—Dye injection was performed by injecting Alexa 568 (Molecular Probes) into HeLa cells that were transiently transfected with connexins containing the GFP-TC tag. Images were recorded in epifluorescence mode on a BX50WI Olympus light microscope using a fluorescein isothiocyanate filter set before injection to image the Cx-GFP-TC expression and after injection using a rhodamine filter set to image the Alexa 568. Image acquisition was performed with a Nikon SLR digital camera. Only cells showing fluorescent gap junctions were injected. The images shown in Fig. 6 were obtained by superimposing before and after pictures.

ReAsH Labeling of TC Constructs, Light Microscopy, and Photooxidation for Electron Microscopy—Cells expressing protein with tetracysteine tags were labeled with FlAsH-EDT₂ or ReAsH-EDT₂ similarly to what we described previously (20). Briefly, FlAsH-EDT₂ or ReAsH-EDT₂ was used at final concentrations of 0.5 and 1.25 µM, respectively, in the presence of EDT (12.5 µM). The labeling was performed for 1 h at 37 °C in 1x Hanks' balanced salt solution (Invitrogen). Free and nonspecifically bound ligands were removed by washing with EDT/Hanks' balanced salt solution (250 µM EDT for TC-tagged connexins, 300 µM EDT for MP-tagged connexins, 600 µM EDT for 4C-tagged connexins). After washing, cells were either fixed in 4% paraformaldehyde (light microscopy imaging) or 2% glutaraldehyde (electron microscopy imaging) or imaged live (using a Bio-Rad MRC-1024 confocal microscope) in complete growth media (phenol red-free Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum). GFP and ReAsH were easily discriminated from one another using standard fluorescein and rhodamine settings, respectively.

Cx43-GFP-TC-expressing cells were photooxidized as described previously (24) except that the excitation was achieved using 490 nm light. At this wavelength, direct excitation of ReAsH (and thus photooxidation of diaminobenzidine by non-GFP-bound ReAsH) was minimal. Prior to photoconversion, cells were fixed with 2% glutaraldehyde in sodium cacodylate buffer (0.1 M, pH 7.4) for 20 min, rinsed in buffer, and treated for 30 min in KCN (20 mM), aminotriazole (5 mM), glycine (50 mM), and H₂O₂ (0.01%) in cacodylate buffer to reduce nonspecific background. Diaminobenzidine (1 mg/ml) in oxygenated sodium cacodylate (0.1 M) was added to the culture dish, and the cells were irradiated with 585 nm light from a xenon lamp for 10-15 min until a brownish reaction product appeared in place of the red fluorescence. After photoconversion, cells were washed in buffer and post-fixed in 1% osmium tetroxide for 30 min. Cells were rinsed in double distilled water and incubated in 2% uranyl acetate overnight at 4 °C and rinsed again in double distilled water dehydrated in ethanol, embedded in Durcupan ACM resin, and polymerized at 60 °C for 48 h. Sections or 7-8 nm thickness were cut with a Leica Ultracut UTC microtome (Leica, Bannockburn, IL).

Baculovirus Expression of Cx26 T135A Mutant with Insect Cells—Cx26T135A-His₆ cDNA was subcloned into the vector Bac-N-Blue (Invitrogen) at EcoRI and HindIII sites and was used to generate recombinant baculovirus. A plaque assay was performed for purifying recombinant viruses. Insect Sf9 cells were cultured in the Sf900II-SFM media supplemented with 2% fetal bovine serum and 0.1% antibiotic/antimycotic (Invitrogen) at 27 °C. Cells were infected at a density of 1.5 x 10⁶ cells/ml for expression and then recovered 65 h after infection.

Gap Junction and Hemichannel Purification—Gap junctions were purified from HeLa cells grown to confluency by following a procedure published previously (25) or isolated from baculovirus-infected Sf9 cells according to published methods (26). The baculovirus-expressed connexins contained a His₆ tag for affinity purification (26). Solubilized connexons were prepared from isolated Sf9 membranes by incubation in 2% dodecyl maltoside in HEPES buffer (10 mM HEPES, 200 mM NaCl), and the protein was purified using the C-terminal His₆ tag and nickel-nitrilotriacetic acid-agarose. Samples were negatively stained with 2% uranyl acetate for EM observation.

Native gels were carried out according to Schagger (27) and visualized with enhanced chemiluminescence Western blotting procedures. Electrophoresis was performed using a 4-20% gradient gel (NuPAGE Novex and Invitrogen).

Electron Microscopy and Image Processing Procedures—Conventional EM images were recorded either on a JEOL 1200 EX operating at 80 kV or a JEOL 2000 operating at 200 keV. Cryoelectron microscopy (cryo-EM) was performed using a JEOL 2000 at 200 keV equipped with a Tietz Video and Image Processing System low dose tomography package and Tietz Video and Image Processing System TemCam-F224 2k x 2k CCD camera. Images were acquired at 40,000 magnification with a 1.6 post-column magnification and had an effective pixel step size of 3.76 Å/pixel on the specimen. For correlation averaging, sub-areas were chosen based on a correlational search using a reference image central gap junction channel surrounded by six neighbors obtained by local Fourier filtering. The program QINDEX in the EMAN package of three-dimensional reconstruction single particle programs (28) performs both the local averaging and correlation search. Images were first high pass filtered to eliminate or dampen any correlations with inconsistent background values. A threshold of 160-300 of the highest correlating areas was chosen, and the x and y coordinates for each of these areas were output to a file for use with the automated boxing program BOXER. These were then averaged for each image analyzed. Similar averages were obtained using smaller references as well. Image averages for each of the two samples (mutant and wild-type) were then translationally and rotationally aligned and averaged using the rotational filtering algorithms (29) in ROBEM to produce the images in Fig. 4. The mutant average was translationally and rotationally aligned to the wt average. A difference map was calculated by subtracting the mutant from the wt average.

Preparation of Connexin Mutants and cRNA for Oocyte Injection Studies—Cx26, Cx43, and mouse Cx50 were subcloned into the pGEM 7Zf(+) (Cx26 and Cx43) and SP64T (Cx50) transcription vectors. Cx26T135A, Cx43T154A, and Cx50T157C mutants were generated by site-directed mutagenesis using overlap extension by PCR as described previously (30). Sequence analysis across the entire coding region confirmed the success of single point mutagenesis.

Constructs were linearized with restriction endonucleases, and capped mRNAs were transcribed in vitro with the appropriate RNA polymerase using the mMessage mMachine kit (Ambion Inc., Austin, TX) according to the manufacturer's instructions. RNA concentrations were determined by spectroscopy, and master stocks of 1 µg/µl RNA were stored at -80 °C. In some cases, relative RNA concentrations were assayed by nondenaturing gel electrophoresis combined with ethidium bromide staining. Working stocks were prepared by serial dilution of the master stock into RNase-free water and also stored at -80 °C.

Oocyte Preparation—Ovarian lobe tissue was surgically removed from adult female Xenopus laevis anesthetized in 0.2% tricaine chilled to 4-6 °C. The tissue was teased apart into smaller clumps containing 6-12 oocytes and incubated on a rocking platform at 22 °C for 1 h in OR2 solution containing 1.5 mg/ml collagenase. The OR2 buffer contains 83 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 5 mM HEPES, pH 7.4, with NaOH. After washing with OR2, stage V and VI oocytes were selected from the population, manually defolliculated if necessary, and incubated at 17 °C for 24 h in half-strength L-15 media (Sigma) supplemented with 0.05 mg/ml gentamicin.

Stage V and VI oocytes were injected with 23-46 nl of a mixture of connexin cRNA and 3-5 ng of Cx38 antisense RNA to suppress endogenous gap junction expression (31-33) and incubated as above for 1-2 days; see "Results" for the various amounts of connexin cRNA injected. For Cx43 experiments, oocytes were also pre-injected with Cx38 antisense 2 days prior to injection with the cRNA mixtures.

For experiments involving the assay of gap junction currents, oocytes were transferred to Petri dishes coated with 2% agarose 24-72 h after cRNA injections for removal of the vitelline layer prior to pairing. The vitelline layer was separated from the plasma membrane by first incubating the oocytes in half-strength L-15 media made slightly hypertonic by addition of 20-50 mM NaCl and then manually removing the vitelline layer with number 5 Dumont forceps. After removing the vitelline membrane, oocytes were incubated for 20-30 min in half-strength L-15 and then paired in agar wells in the same media.

For experiments involving the assay of hemichannel currents, oocytes with their vitelline layers intact were perfused with ND96 buffer containing either 2 mM CaCl₂ or no added calcium. ND96 buffer with no added calcium contains 96 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 10 mM HEPES, pH 7.4, with NaOH.

Electrophysiology (Oocytes)—Voltage clamp experiments began 18-48 h after cRNA injections. Voltage clamp recordings were obtained using a two-electrode voltage clamp technique for single oocytes or a dual two-electrode voltage clamp technique for oocyte pairs. GeneClamp 500B voltage clamps (Axon Instruments, Inc., Foster City, CA) were used in one or two oocyte configurations using a 1x L head stage for voltage recording and a 10x MG head stage for passing current. The bath potential was actively clamped to 0 mV using a 100x VG head stage. Voltage recording and current passing electrodes were pulled from borosilicate glass on a horizontal puller (Flaming-Brown P-87, Sutter Instruments, Novato, CA). Electrodes had resistances between 0.5 and 3 megohms using an internal pipette solution of 150 mM KCl, 10 mM EGTA, 10 mM HEPES, pH 7.2. The amplifiers were interfaced to a personal computer through a Digidata 1320 A/D converter, and data were acquired and analyzed using Pclamp 8 software (Axon Instruments, Inc.). Currents were filtered at 50-200 Hz and acquired directly from the hard drive.

To assay hemichannel currents, individual oocytes were subjected to a two-pulse voltage protocol before and after exchanging the bath media from ND96 containing 2 mM CaCl₂ to ND96 with no added calcium. The voltage protocol consisted of a 500-ms -10-mV prepulse followed by a 5-s voltage step ranging from -120 to 60 mV in 10-20-mV increments from a holding potential of -20 mV. Whole-cell hemichannel conductance was determined as the current elicited by the pre-pulse divided by 10 mV. To assay gap junctional currents, both oocytes in a pair were clamped to -20 mV, and one oocyte was subjected to a similar voltage pulse protocol as above but using a series of voltage steps ranging from -120 to 80 mV to generate transjunctional potentials of -100 to 100 mV. In gap junction experiments, the pre-pulse current recorded in the nondriven oocyte was divided by the transjunctional voltage of 10 mV to obtain a measurement of the junctional conductance.

Fluorescent Imaging (Oocytes)—For fluorescent measurement of dye transfer, oocytes were prepared as for electrophysiological recording. After recording the initial junctional conductance, the donor oocyte was injected with 41.4 nl of 10 mM hydrazide sodium salt of Alexa 488 (Molecular Probes). At 1, 3, and 6 h after injection, the oocyte pair was imaged with a Zeiss Axiovert S100 fluorescent microscope using a CCD camera (model RTE/CCD-1300-Y, Princeton Instruments, Trenton, NJ). D450-490 and LP520 (Chroma Technology) excitation and emission filter sets were used for Alexa 488. To quantify the diffusion of dye, uniformly sized boxes were centered over both donor (D) and acceptor (A) oocytes of a pair, and their pixel intensity ratio (acceptor:donor) was measured using Metamorph 6.2R6 software (Universal Imaging Corp., West Chester, PA). For visualization, the software also assigned color based on fluorescent intensity over a range of 0-4095 pixels, with the blue end of the spectrum being the lower and red the higher intensities. 0 was assigned to black and 4095 to white.

View larger version (17K): [in this window] [in a new window]   FIGURE 1. Sequence analysis of threonine mutation. In the connexin family, this conserved mutation sits near the top of the third transmembrane helix at the cytoplasmic side. A, the approximate location (as indicated by the star symbol) is illustrated on a canonical connexin topology diagram. B, bar graph histogram analysis comparing all connexin sequences in the NCBI protein data base was performed. The majority of connexins has threonine at this position. Deviations of this amino acid are shown in the table at the right-hand side of B. C, a radial dendrogram of the mouse connexin showing the familial relationships between isoforms contain a clustering of non-threonine positions outside of the - and -subgroups (names shaded in gray).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Site-specific mutants are commonly used as a tool for exploring structure/function relationships. In particular, there has been great interest in finding connexin mutants that form gap junction structures but fail to establish open channels, because such mutants would allow us to distinguish functions of connexins associated with the protein or the structures per se, and those associated with their function in intercellular communication. It would also provide us a way of stably isolating closed pores. Strictly or absolutely conserved amino acids often indicate critical positions that can eliminate functionality if mutated or perturbed. Here we became interested in a threonine amino acid located at the cytoplasmic end of TM3. This mutation was fortuitously discovered in some Cx43 sequences we had generated, and we chose to study it further because of the conservation within the connexin family, its loss of function when mutated, and its three-dimensional location within the connexin polypeptide.

Sequence Analysis of the TM3 Helix and Conservation of the Threonine Position
We examined the sequence of the third transmembrane helix and found a threonine that was strictly conserved among the - and -connexins but not the -connexins. The diagram in Fig. 1A shows the approximate location of this amino acid on the connexin topology diagram. This amino acid sits on the N-terminal side of a strictly conserved tyrosine that is found in 100% of the connexin sequences and was mapped in Cx32 as the most cytoplasmic pore-lining residue (19). The amino acid on the other side varies among the connexins (arginine, asparagine, glycine, and tryptophan being the most common substitutions). Next, we searched the NCBI data base for all connexin sequences and aligned the sequences using the ClustalW set of programs. The bar graph of the aligned sequences shown in Fig. 1B shows that this threonine occurs in 71% of the connexins. Deviations from having Thr in this position only occurred among highly divergent species and isoforms. For example, nonconservation of this amino acid occurred in all Cx45 (Ile) and Cx36 (Phe) homologues. Using ClustalW and TreeViewPPC, we plotted the connexin sequences from mouse to see the familial relationships (Fig. 1C). When viewed as a radial plot, connexins with deviations from Thr at this position cluster in a manner that is separate from the - and -subgroups.

Expression and Trafficking of the Threonine Mutant in HeLa Cells
When the Cx26T135A mutant construct is transiently or stably expressed in HeLa cells, "normal" gap junction patterns are seen in confocal micrographs (Fig. 2A). We also generated a Cx43T154A-GFP-4C construct and transiently expressed it in HeLa cells to elucidate whether the protein trafficked in a native manner. This particular construct was generated to perform correlated light and electron microscopy; however, the confocal microscopy reinforces the result seen with Cx26T135A that this Cx43 mutant forms normal appearing gap junctions. Fig. 2B shows immunofluorescence from the GFP and from ReAsH labeling demonstrating not only the large gap junction plaque but also intracellular labeling indicative of trafficking vesicles and lysosomes. When GFP is excited, it transfers energy to the ReAsH molecule in a fluorescence resonance energy transfer-dependent manner (22, 23). The confocal image shown in Fig. 2B is of an area prior to photoconversion. The green signal is from the GFP, and the red is ReAsH labeling. Note that in the large gap junction the GFP and ReAsH signal overlap. These large gap junctions are typical of tagged Cx43 (20, 36, 37).

View larger version (121K): [in this window] [in a new window]   FIGURE 2. Light and EM of threonine mutant expressed in HeLa cells. A, when stably expressed in HeLa cells, immunolabeled Cx26T135A forms native appearing punctate gap junctions as illustrated in this confocal micrograph. B, this confocal image of a transiently transfected Cx43T154A-GFP-4C construct contains a large junction; the green fluorescence is from the GFP excitation, and the red fluorescence is excitation for the ReAsH from fluorescence resonance energy transfer of the GFP. The ReAsH is used for generating singlet oxygen for photooxidation by O2 and diaminobenzidine for increased osmium deposition during preparation for EM. C, this EM from a different area than in B shows a large gap junction (GJ) between two HeLa cells that has been selectively stained (denoted by arrow). Lysosomes in the process of degrading the Cx43T154A are indicated by the white L and arrows. D, a higher magnification area from another area shows gap junctions plus areas of deposition corresponding to hemichannel plaque (HP) and small trafficking vesicles (TV) on their way to the plasma membrane. The inset in D is from a higher magnification view illustrating the classic pentalaminar appearance of the gap junction. Note the large diffuse cytoplasmic staining because of the bulky addition of the GFP-4C tag. Scale bar is 20 nm.

Connexon Structure by Negative Staining and Cryo-EM
To study the structural appearance of gap junction plaques between mutant and wt, we isolated Cx26T135A gap junctions from both HeLa cells (Fig. 3, A and B) and from baculovirus-infected Sf9 cells (Fig. 3C). We chose to perform our structural studies with Cx26 because of our previous successes in the isolation and analysis of gap junctions composed of this isoform (25, 41). Gap junctions isolated from both HeLa cells and Sf9 cells showed the classic doughnut appearance with plaques ranging from larger and less ordered (Fig. 3, A and C) to occasional plaques that were relatively small but had better lattice order. Under these low resolution and high irradiation conditions, the channels appear similar to wild-type Cx26 (Fig. 3D).

We applied cryo-EM and image analysis methods to isolated Sf9 gap junction membranes to evaluate structural changes in the averaged projection maps between mutant and wt connexons. Because of the inherent lattice disorder in these samples, standard Fourier averaging could not be used effectively, and instead a correlation average approach was used (42). Lattice filtered subregions of images for the wt (Fig. 4A) and Cx26T135 mutant (Fig. 4B) show small differences where the ice density within the pore in the wt is more diffuse than in the mutant as indicated by its black appearance. We used correlation averaging of frozen-hydrated wt (Fig. 4C) and mutant gap junctions (Fig. 4D) and computed a difference map from the subtraction of wt minus mutant (Fig. 4E). The wt average contains about 400 unit cells, although the mutant image was computed from 600 unit cells. The average and difference maps have been rotationally filtered to enhance the 6-fold symmetry of the connexon. The difference map (Fig. 4E) has been displayed on a step gray scale where the largest differences are either black (negative differences) or white (positive differences). The zero difference level would appear as a medium gray. We find that the pore size is larger in the wild-type average than in the mutant, and there is some rearrangement of features within the lobes the connexons. There is an 17% reduction in the pore size from wt to mutant (27 to 22.5 Å), although it is not possible in these projection images to distinguish whether this reduction occurs at the cytoplasmic surface or at the extracellular surface, or if a larger constriction could occur at a local point within the pore. Single site mutations often result in instability of the hexamer when released from the membrane (26) and indicate a potential role in maintaining intra-molecular or inter-molecular contacts. To test the stability of the connexon, Cx26T135A-His₆ and wt-His₆ gap junctions were detergent-solubilized in 2% dodecyl maltoside and purified with a nickel-nitrilotriacetic acid resin. We found that Cx26T135A-His₆ connexons became unstable as judged by their appearance using conventional, negative-stained EM (Fig. 5A). Although wt-His₆ maintained a uniform "doughnut" appearance (Fig. 5B, circles), images of Cx26T135A-His₆ contained aggregates of smaller substructures (arrows) as well as some round shaped particles (circles). Native gel PAGE analysis was performed with these purified samples (Fig. 5C). The Cx26T135A-His₆ mutant sample (Fig. 5C, left lane) has strong bands at the dimer, trimer, tetramer, and hexamer positions, whereas the wt-His₆ (Fig. 5C, right lane) shows a strong band at the hexamer molecular weight (156 daltons). The relative amount of each species in the mutant was variable in three preparations examined by native gel analysis (data not shown); however, in the mutant lane shown in this figure the ratio of wt:tetramer:trimer: dimer is 1:0.8:0.65:0.35. Because both wt and mutant connexons have a His₆ tag, the difference in appearance of the oligomers can only be explained by the difference between the connexin isotypes. These results suggest that mutant protein is in a slightly altered conformation from the wt because it is still retained as hexamers but also dissociates to smaller oligomers. A highly disruptive mutation would cause disruption into monomers and aggregates or fail to assemble in vivo.

View larger version (149K): [in this window] [in a new window]   FIGURE 3. Isolated gap junctions of Cx26T135A and wt gap junctions show very similar structures. A and B, conventional EM of a negatively stained Cx26T135A gap junction from HeLa cells shows wt appearance at this resolution and under high irradiation conditions (scale bar for A and B the same). C, negatively stained Cx26T135A-His6 gap junction obtained from expressing the mutant connexin in Sf9 cells with baculovirus expression system. D, negatively stained EM image of a Cx26wt-His6 gap junction is shown here for comparison. Under these imaging conditions, the mutant and wild-type connexin channels are indistinguishable.

View larger version (112K): [in this window] [in a new window]   FIGURE 4. Cryo-EM analysis of wt and mutant Cx26T135A-His6 gap junctions. Frozen-hydrated imaging of wt versus mutant structures allows us to directly image differences in mass density. Portions of lattice-filtered images are shown for wt (A) and Cx26T135A (B) and show differences in density in the pore region but are noisy within the subunit structure. Correlation averages were obtained from cryo-EM images of frozen-hydrated wt (C) and mutant gap junctions (D) and the difference map from the subtraction of wt minus mutant (E). The averages shown here have been rotationally filtered to enhance the 6-fold symmetry. The two averages are displayed on a 15 gray level step scale, and the difference map is displayed using 9 gray levels. In the difference map, the pore size is larger in the wt average than in the mutant, and there is a shift in the hexagonal symmetry in the lobes of the connexons.

View larger version (52K): [in this window] [in a new window]   FIGURE 5. Analysis of purified Cx26T135A-His6 disassembled connexons. Isolated Sf9 membranes for both mutant-His6 and wt-His6 were solubilized in 2% dodecyl maltoside, and the protein was purified using C-terminal His6 tag and nickel-nitrilotriacetic acid-agarose. A, the mutant protein has heterogeneous appearance in the detergent solution, suggesting some structural degradation. Examples of hexamers are indicated by with a circle, and the arrows point to smaller oligomers. B, the purified wt sample shows typical hemichannel structure, which has a doughnut shape with a hole in the center (three examples are indicated by the circles). This suggests that wt hexameric structure is maintained in the detergent solution. C, native gel PAGE was performed for purified protein. Cx26 was visualized by ECL Western blotting. The Cx26T135A-His6 mutant (T135A, left lane) has strong bands at the dimer and hexamer positions, whereas the wt-His6 (wt, right lane) shows strong bands at molecular weights of the trimer, tetramer, and hexamer. This suggests that mutant protein is in a different conformational form from the wt.

TC-tagged connexins were expressed in the paired Xenopus oocyte assay system and assessed for changes in electrophysiological behavior from wt. An example of the junctional currents elicited in oocyte pairs expressing either Cx26wt or Cx26-TC is presented in Fig. 6A. The average normalized initial and steady-state conductance-voltage relationships obtained from three different pairs indicates that tagged and Cx26wt gap junctions behave very similarly (Fig. 6B). These experiments were performed in different batches of oocytes, where junctional currents were elicited by imposing transjunctional potentials of -120 to +120 mV in 20-mV increments for the Cx26-TC oocyte pairs and -130 to +130 mV in 20-mV increments for the Cx26wt pairs. Fig. 6C show examples of Cx43wt and Cx43-TC gap junctional currents elicited by imposing transjunctional potentials of -130 to +130 mV in 20-mV increments. The average normalized initial and steady-state conductance-voltage relationships obtained from three pairs each are presented in Fig. 6D. As had been the case for Cx26, the Cx43T154A channels behaved nearly identical to wt channels.

View larger version (55K): [in this window] [in a new window]   FIGURE 6. Tetracysteine constructs show similar electrophysiological profiles to wt. A, examples of the junctional currents elicited by imposing transjunctional potentials of -120 to +120 mV in 20-mV increments in oocyte pairs expressing either Cx26wt or Cx26-TC reveal similar kinetic profiles (B). The average normalized initial (circles) and steady-state (squares) conductance-voltage relationships obtained from three different pairs of Cx26wt (open symbols) and Cx26-TC (closed symbols) expressing oocytes largely superimpose. Both curves show the asymmetry about Vj = 0 that has been consistently reported in the literature for Cx26 (33). C, examples of Cx43t and Cx43-TC gap junctional currents elicited by imposing transjunctional potentials of -130 to +130 mV in 20-mV increments show essentially identical profiles. D, steady-state and instantaneous conductance was plotted against Vj from three separate oocyte pairs of Cx43wt and Cx43-TC as shown in B. As for Cx26, these traces indicate that the tag did not interfere with the bulk characteristics of voltage gating.

View larger version (19K): [in this window] [in a new window]   FIGURE 7. Analysis of threonine mutant Cx43, Cx26, and Cx50 connexins in homotypic and heterotypic combinations. A, mean junctional conductances normalized to the average percentage of wt conductances for several oocyte pairs show that no significant coupling is found with the Cx26T135A construct, either in homotypic pairings or in heterotypic pairings with Cx26wt. This is in contrast to Cx26wt construct that produces robust homotypic coupling. However, if one of the wt cells is co-injected with equal levels of Cx26T135A cRNA, there is a drastic reduction of coupling that goes to background levels when both oocytes are co-injected with wt and mutant RNA. Similar results are also seen in Cx43T154A (B) and Cx50T157C (C). These results indicate that this mutant acts as a fairly complete dominant-negative suppressor of function.

Oocytes injected with either 10 or 20 ng of Cx26T135A cRNA developed no detectable gap junctional coupling beyond that observed in oligonucleotide-injected control oocyte pairs over the course of 2 days, compared with similar levels of wt Cx26 cRNA that produced 40-62 µS of conductance. The lack of junctional coupling suggests that the mutation rendered the connexin nonfunctional but does not indicate whether the protein was synthesized or correctly processed and trafficked to the plasma membrane, although this seemed likely from the microscopy results shown in Fig. 2. One way of assessing this was to determine whether Cx26T135A had a dominant-negative effect on wt junctional conductance. Oocytes were injected with either wt (20 ng), mutant (20 ng), or an equal mix of mutant and wt cRNA (20 ng each) and were subsequently paired with either the same type of oocyte or an oocyte injected with only wt cRNA. All oocytes also received 5 ng of an antisense oligonucleotide to endogenous Cx38 to suppress contamination of results by endogenous coupling. Because oocytes received the same amount of wt cRNA, if the mutant were unable to interact with wt connexin, then the junctional conductance levels between oocytes injected with a combination of wt and mutant cRNA should be comparable with the conductance levels between oocytes injected with only wt cRNA. We had shown previously in oocytes that at these levels of RNA, there were minimal effects because of competition for translational machinery (46), although this was also confirmed in these experiments (see below). As shown in Fig. 7A, the mutant Cx26T135A clearly inhibited the formation of functional gap junction channels when co-expressed with Cx26wt, reducing the conductance in two experiments from 40.1 ± 7.5 µS (n = 11) and 61.8 ± 17.0 µS (n = 9) for wt Cx26 to 8.7 ± 2.0 µS (n = 11) and 11.2 ± 1.6 µS (n = 15), respectively when Cx26wt and Cx26T135A were co-expressed in only one of the pairs. The properties of the gap junctional currents (kinetics and voltage dependence) measured in the Cx26wt-(Cx26wt + Cx26T135A) pairs were similar to those of Cx26wt-Cx26wt pairs (data not shown). When mutant and wt Cx26 were co-expressed in both cells, the conductance was reduced even further so as not to be significantly different to pairs injected only with antisense oligonucleotides to the endogenous connexin.

View larger version (17K): [in this window] [in a new window]   FIGURE 8. Dominant-negative effect of mutation on Cx50 hemichannels. A, representative whole-cell currents from oocytes, injected with the cRNAs indicated on the left before and after changing the bath solution from ND96 with 2 mM CaCl2 (left column) to ND96 with 0 added calcium (right column). Oocytes were held at -20 mV and subjected to a series of voltage steps ranging from -120 to +60 mV (30-mV increments shown above). The scale bar represents 5 µA (vertical) and 2 s (horizontal). The hemichannel currents characteristic of Cx50wt are not seen in the mutant and only minimally detected if wt and mutant are co-expressed in the same oocyte. B, histograms of the average whole-cell conductance after removing calcium from the medium show that Cx50T157C does not produce functional hemichannels alone and suppresses the ability of wt to make hemichannels. The control consists of injection of the same solution as in the other conditions (Cx38 mRNA only) but lacks any exogenous Cx50 connexin mRNA.

Combining data from experiments in four different batches of oocytes, Cx26T135A never induced coupling above background in the system (Fig. 7A). Furthermore, co-injection of the mutant with Cx26wt reduced conductance of the Cx26wt-Cx26wt pairs to 15.4 ± 2.3% when paired with Cx26wt cells and to near background levels when two co-injected cells were paired (6.2 ± 1.8% of Cx26wt-Cx26wt pairs versus 4.6 ± 1.1% and 5.1 ± 1.1% of wt for Cx26T135A-Cx26wt pairs and pairs receiving only antisense oligonucleotide, respectively).

Theoretically, the binomial theorem can predict the subunit composition of hemichannels formed in oocytes injected with different ratios of mutant and wt cRNA. The binomial theorem predicts that equal amounts of mutant and wt RNA will generate mainly heteromeric hemichannels (>95%), assuming that the two cRNAs are translated with equal efficiency and assuming that the connexin polypeptides are homogeneously mixed in the ER and can assemble into a hemichannel with equal probability. For example, if there are equal numbers of mutant and wt subunits, the probability of having a homomeric wt hemichannel would be (1/2)⁶ or 1.5% (i.e. the probability of any one subunit being wt raised to the power of the number of subunits). Similar probabilities can be calculated for any heteromeric mix by multiplying this probability by the number of possible combinations of the particular kind of heteromeric channel.

Because of the complexity of accurately quantifying the levels of plasma membrane resident Cx26 and Cx26T135 within the same cell, we did not attempt to determine the minimum number of mutant subunits required to render a hemichannel nonfunctional by manipulating subunit stoichiometry using various ratios of wt and T135A cRNA. However, if the conditions of the binomial theorem were met, the absence of detectable junctional coupling between wt oocytes and oocytes co-expressing wt and mutant in the third round of experiments suggests that a single mutant subunit may be capable of rendering the channel nonfunctional (i.e. based on the example above, only 1.5% of all channels would be functional, homomeric wt if the mutant and wt proteins were equally expressed).

The lower level of dominant-negative effects at the higher RNA concentration used in the first two experiments described above might suggest a higher stoichiometry of mutant channels is needed to ablate function of wt Cx26. However, the high levels of conductance in wt pairs is likely to lead to a significant underestimate, i.e. 40% (47) of the actual junctional conductance, once access resistance is taken into account. The inhibition by co-injection of mutant would then be closer to the 95% predicted by the Poisson distribution, assuming equal affinity of the two types of subunits.

The Dominant-negative Effect of the Mutation in Cx43 Gap Junctions—To determine whether the specific mutation at the equivalent position of Cx26T135 affected other types of connexins, experiments were performed using oocytes injected with Cx43wt (4 ng), Cx43T154A (4 ng), or a combination of Cx43wt (4 ng) + Cx43T154A (4 ng). In these experiments, oocytes were pre-injected with 5 ng of Cx38 antisense oligonucleotide 2 days prior to injecting the connexin cRNAs to reduce the potential interactions between Cx43 and Cx38. The results of various pairing configurations are shown in Fig. 7B and resemble the Cx26 experiments with the Cx43T154A mutation showing a clear dominant-negative effect on the formation of functional Cx43 gap junction channels, reducing the conductance from 12.4 ± 2.6 µS for Cx43wt (n = 7) to 2.4 ± 0.9 µS for Cx43wt-(Cx43wt + Cx43T154A) pairs (n = 9) and to background levels 0.3 ± 0.2 µS(n = 4) when mutant and wt were mixed in both oocytes. The 80% reduction seen in the heterotypic pairing is somewhat less than that seen with Cx26 or predicted by the Poisson distribution, as discussed above. This could also reflect a less efficient suppression in the case of Cx43 or differential translation of the mutant and wt Cx43 constructs.

View larger version (55K): [in this window] [in a new window]   FIGURE 9. Effects of mutants on dye coupling in Xenopus oocytes. A-F, oocytes injected with cRNA encoding the relevant connexin were paired, the conductance measured as in Fig. 6, and the left-hand member of the pair injected with Alexa 488 dye. Images that were captured after 6 h indicated that pairs injected with Cx43 (A) and Cx26 (B) showed clear dye transfer between cells, whereas pairs injected with Cx43T154A (C), Cx26T135A (D), or co-injected with equal amounts of Cx26wt and Cx26T135A (F) showed no dye transfer above the antisense oligonucleotide-injected controls (E). Images are shown in pseudocolor that represents the intensity of fluorescence according to the scale bar shown in G. The compilation of data from all pairs studied is presented in the histogram (H), where the ratio of fluorescence intensity in the acceptor over the donor cell is compared 6 h after injection. The average conductance of Cx43 and Cx26 pairs was 37.6 and 37 µS, respectively, whereas all other pairs showed conductances indistinguishable from the antisense controls. The average conductances and their standard deviations are displayed in this graph above the histogram bar.

View larger version (53K): [in this window] [in a new window]   FIGURE 10. Functional imaging studies of HeLa cells expressing mutant connexins show no dye transfer as compared with wt. HeLa cells were transiently transfected with chimeric Cx26T135A-GFP-TC DNA and monitored for gap junction plaque formation with the GFP fluorescence. The TC tag was the 4C peptide and was used for the photooxidation experiments in Fig. 2. Alexa 564 was injected into cells presenting gap junctions, and dye transfer was monitored. Cells expressing Cx26T135A-GFP-TC did not show dye transfer (A), although dye transfer occurs in wt cells (B). Similarly, Cx43T154A-GFP-TC-expressing cells did not transfer Alexa 568 (C), whereas Alexa 568 diffused through wt connexin channels expressing cells (D).

View larger version (55K): [in this window] [in a new window]   FIGURE 11. Keratinocytes infected with Cx43T154A show a reduction in dye transfer compared with Cx43wt. Keratinocytes that had been retrovirally infected with a GFP expression plasmid (pBMNIGFP) alone (top panel) or with the vector containing the insert for Cx43wt (middle panel) or the mutant Cx43T154A (bottom panel) were microinjected with Alexa 350 dye, and the degree of dye transfer was recorded after 15 s (A, D, and G, respectively). B, E, and H, phase contrast image showing transfected and nontransfected cells. C, F, and I, GFP fluorescence from pBMNIGFP expression depicts the cell expressing GFP that was injected (denoted by arrow).

Effect of Cx26T135A and Cx43T154A on Dye Transfer between Oocytes—To complement the electrophysiological analysis, the dye permeability of these dominant-negative mutants was also tested in the paired Xenopus oocyte expression system. Fig. 9, A and B, shows robust transfer of Alexa 488 (MW570) between cells expressing Cx43 and Cx26 with an average conductance of 35 and 36 µS, respectively. In contrast, oocytes expressing Cx43T154A (Fig. 9C) and Cx26T135A (Fig. 9D) show no transfer of dye over 6 h. The slight "halo" of fluorescence that can be seen in the acceptor cell is similar to that seen in the antisense oligonucleotide-injected controls (Fig. 9E) and appears to be an optical artifact of the system. Expression of the mutant protein was confirmed by demonstrating its dominant-negative effect on the wt connexin, as seen by the lack of dye coupling in the Cx26 + Cx26T135A co-injected pairs (Fig. 9F). Note that the dye intensity in these images is expressed in pseudocolor, according to the scale shown in Fig. 9G.

These observations were quantified by placing boxes of equivalent area within the donor and acceptor cells and measuring pixel intensity using Metamorph software. The resulting ratio of intensity in the acceptor to donor box was then measured at 1-, 3- (data not shown), and 6-h time points (Fig. 8). The latter is shown in Fig. 9H. Cx43 and Cx26 passed dye efficiently. Although the average conductances of the pairs tested were the same for each connexin (37.6 and 37 µS, respectively), there were more Cx43 channels connecting the cells, as this connexin has a lower single channel conductance than Cx26 (90 and 130 pS, respectively). Cx43 channels also have a higher permeability for this dye than Cx26 channels (2-fold) (40). Together, this would explain the higher acceptor:donor ratio that is observed for Cx43. Most importantly, all pairs expressing the mutant form of either protein, including those where it was co-injected with wt (in the case of Cx26), showed no transfer of dye above that seen in the antisense oligonucleotide-injected negative controls. Hence, dye injection studies reflect the identical pattern to that seen in the electrophysiological analyses of these channels, demonstrating that these mutants are equally impermeable to ions or larger molecules.

Functional Analysis of Constructs in Mammalian Cells
Dye Transfer Assays in Mammalian Cells—In addition to oocytes, we also tested the functionality of these mutants through dye transfer between mammalian calls. Initially, we employed dye coupling of HeLa transfectants that had been shown to have no detectable expression of endogenous connexins. Alexa 564, a dye that should pass through the channel pore (47) and whose emission does not significantly overlap with GFP, was employed in these experiments. Gap junctions were monitored using a construct that was a chimera of GFP with a 4C tetracysteine peptide appended to the C terminus of the GFP. These constructs were used in combination with fluorescence photooxidation for obtaining the EM images in Fig. 2 and allow us to monitor the protein and gap junction expression (green color) and to unambiguously select only cell pairs containing gap junctions. Typical images of mutant versus wt dye transfer experiments are shown in Fig. 10. Images are presented for Cx26T135A mutant (Fig. 10A) and wt (Fig. 10B) and Cx43T154A mutant (Fig. 10C) and wt (Fig. 10D). In both isoforms, Alexa 564 (red color) was transferred in the wt but not in the mutant, indicating a loss of function in the mutant.

The dominant-negative effects of the Cx43 mutant was also tested using primary human keratinocytes. These cells were transfected with either the pBMNIGFP vector alone or with Cx43wt or T154A mutant inserted into this vector. Transfected cells were microinjected with Alexa 350 dye, which has also been shown to pass through Cx43 channels (40), and photographed after 15 s (Fig. 11). The results show that the expression of Cx43wt enhances dye transfer, although the expression of Cx43T154A significantly reduces the dye transfer. However, when the time course was allowed to progress beyond 15 s, dye transfer was detected even with the mutant.⁸ This might be explained if the mutant were expressed at lower levels than the endogenous protein, leading to incomplete "knockdown" of endogenous Cx43. However, the Alexa 350 dye is transported by all of the connexins surveyed to date (47), and there are a total of seven connexins expressed in human skin (48). Although Cx43T154A may have dominant-negative effects on connexins other than Cx43, it seems likely that it may not affect all of the endogenous connexins. In future experiments, use of a combination of larger and smaller Alexa dyes (e.g. Alexa 350 and Alexa 564) may help to discriminate between different isoform channels in cells expressing multiple connexins.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The specific mechanisms for connexin channel gating have yet to be identified despite examinations of gating in response to calcium (49) or electrophysiological dissection of different voltage gates (for a review see Ref. 50). The absence of a widely accepted structural model for gating is partially due to the fact that although closed and open states have been characterized via functional methods such as electrophysiology and dye transfer, the isolation or preparation of distinct and homogenous open and closed states for structural analysis has been problematic. Atomic force microscopy investigations have shown opening and closing of the channel pore (41, 51); however, because these images are surface views, they cannot provide an insight into internal protein changes. Perturbation analysis studies from potassium channels have shown that single site amino acid substitutions can be useful in defining inter-helical associations (52). One way to achieve this goal is the isolation of mutant connexons or channels whereby the mutation causes the hexamer or dodecamer to be locked into a permanent open or closed state. Here we have shown by using correlative structural and functional studies that a dominant-negative nonfunctional mutant is closed, and we observed small structural differences in projection maps between the mutant connexon and wt. Given that this is a conservative amino acid substitution, it is important to understand what effect the mutation has on neighboring amino acids, the structure of TM3, and its three-dimensional spatial inter-helical and intra-helical interactions.

Comparative Sequence Analysis of the TM3 at the Threonine or Equivalent Position—Conservation of residues can provide insights into the importance of certain amino acid positions. A search of the connexin sequences in GenBank^TM revealed that this threonine is conserved among a subset of connexins that group into - and -subfamilies. However, the tyrosine that is next to the threonine (e.g. Tyr-136 in Cx26 and Tyr-155 in Cx43), one amino acid down into the transmembrane region, is absolutely conserved throughout the connexin family. In addition, an analysis of hereditary mutations leading to connexin diseases reveals that the 8 connexins linked to diseases are all and ; therefore, all contain this Thr. Table 1 contains the TM3 sequences of naturally occurring mutations found in connexin diseases with mutations highlighted in a large boldface and underlined font. There are no published mutations within the TM3 sequences of Cx37-, Cx46-, and Cx50-caused diseases, although this is most likely due to the small number (1-10) of mutations for each of these connexin diseases that have been documented as compared with Cx26 and Cx32, where over 100 and 250 distinct mutations, respectively, have been identified and published (Human Gene Mutation Data base at the Institute of Medical Genetics in Cardiff; archive.uwcm.ac.uk/uwcm/mg/hgmd0.html). The mutation of Cx26Y136 into a termination codon as been shown to be one cause of nonsyndromic hereditary deafness (53), whereas a mutation of Cx32Y135, the neighbor of Thr-134, to a Cys creates manifestations of CMTX (54). In addition, mutation of Trp-133, the other Thr-134 neighboring amino acid in Cx32, to an Arg-55, Cys-56, or stop codon (57) also resulted in patients displaying CMTX. It is interesting to note that in CMTX, although mutations span the entire sequence of TM3 with the four upstream residues and one downstream amino acid from this Thr, this threonine residue has yet to be identified as a disease mutation or in Cx26-caused diseases. There are three reasons why this might be so: 1) cases have not yet been identified; 2) mutations at this position cause no change in the expression and functionality of the connexin, or 3) this mutation would be embryonic lethal. Given the functional data presented in this paper showing an almost complete knock-out of function in a dominant-negative fashion, we believe that the third hypothesis, lethality of this mutation in vivo, is the most likely explanation (58). Although most mutations would only ablate the function of the connexin mutated, these mutants would be expected to cause widespread effects on co-expressed connexins, eliminating any "redundancy" protections in tissues. The biological consequences of such a mutation are likely to be more severe. This being said, it is still likely that mutations in connexins with limited expression (e.g. Cx46 and Cx50 in the eye lens) would not be expected to be lethal.

The Third Transmembrane Helix Tertiary Structure—Although such sequence comparisons are useful, these bioinformatics approaches are not enough to explain observed differences in functionality and trafficking between isoforms even in such highly conserved regions as TM3. Lagree et al. (59) reported that replacement of the residues at 152 and 153 of Cx43 with those occurring in Cx32 (L152W and R153W, respectively) resulted in formation of gap junctions, but the L152W had markedly reduced dye transfer (59). In addition, the neighboring residue to Cx43T154, Cx43R153, was shown to be critical in preventing oligomerization of Cx43 in the ER when an ER retention sequence was genetically added to the C terminus (60). Kumari et al. (61) investigated a valine to aspartic acid substitution at position 156 in Cx37 (two residues to the N-terminal side of the Thr under study here) expressed in N2A cells and found a variety of conductive states with rapid transitions. Most interestingly, Cx40 contains an Asp at this position, but its conductance is different from Cx37V156D, indicating that tertiary structure plays a large role in determining the functional importance of specific amino acids.

The four membrane spanning segments (TM1, TM2, TM3, and TM4) form a four -helix bundle, so that when the hexamer is formed, a closed cylindrical structure is obtained (62) with one major helix (and possibly part of a second) from each subunit contributing to the pore lining of the channel. A three-dimensional 7-Å cryo-EM structure contains four transmembrane rods corresponding to -helices (63).

The third transmembrane helix, TM3, has been a particular region of interest within the connexin structure since the first publication of the primary sequence (64, 65). This membrane-bound stretch of amino acids contains polar or charged residues at every 4th position, and by analogy with other membrane channels it was speculated that this was the pore-lining helix because the polar or charged residues would form one face of the helix that could be exposed to an aqueous environment (66). Largely consistent with this, in the 7-Å structure determined by electron crystallography (63) the pore is formed by one major pore-lining and one minor pore-lining helix. Unequivocal identification of the helices could not be determined because of the limited resolution, so alternative techniques such as SCAM analysis have been used to probe accessibility of specific amino acids to aqueous reagents that block channel functionality. A SCAM analysis to map accessible residues in this section of Cx32 (19) has indicated that the tyrosine residue (Tyr-135) to the immediate C-terminal side of this threonine is located within the pore, and that mutation of the threonine to cysteine (Cx32T134C) for SCAM studies resulted in an elimination of function, just as reported here for Cx50 and the alanine mutations at this site on Cx43 and Cx26. Amino acids implicated as Cx32 TM3 "pore liners" were Tyr-135, Val-139, Phe-141, Leu-144, Ala-147, Phe-149, and Tyr-151. This pattern of reactivity reported for pore-lining residues would suggest that Thr-134 would likely lie at the cytoplasmic mouth of the pore but approximately one-third of a turn away from the pore lumen itself. This would be consistent with it interacting with one of the adjacent tightly packed helices.

A new model of the arrangement of the -carbon atoms of the transmembrane -helices has been proposed by Fleishman et al. (18). This model was based on a three-dimensional map, 5.7 Å in plane and 19.8 Å perpendicular to the membrane plane, which was obtained by data analysis refinement of the original 7 Å data of Unger et al. (63). The four transmembrane segments in the oleamide-treated Cx43 truncation mutant structure were traced as canonical -helices, and the -carbons were assigned to specific amino acid residues in the protein sequence. The criteria for these assignments were evolutionary conservation of residues, hydrophobicity of the amino acid residues, biochemical and phylogenic data from mutagenesis studies, and analysis of naturally occurring disease mutations in Cx32. Because Cx32 mutations in CMTX are the most highly characterized, with over 250 published mutations that extend over the entire length of the sequence, the amino acid sequence of Cx32 was used for assigning the -carbons to the canonical -helices. In this model, it is assumed that highly conserved domains will share a common architecture between isoforms that highly conserved amino acid residues will preferentially pack at helix-helix interfaces and perturbations of the helix-helix packing will manifest as CMTX mutations. Using this computational approach, Fleishman et al. (18) suggest that the major pore-lining helix is TM3, and the minor pore-lining helix is TM1. The TM2 and TM4 helices face the lipid environment with TM2 closely apposed to TM1 and TM4 closely apposed to TM3 in this model. It is important to point out that this is not a consensus model, because the errors in assignment of the helix orientations are within 40°, and the helical register could vary by 1 turn of a helix because of the decreased vertical resolution in the EM map. More importantly, mutagenesis studies often differ greatly with this interpretation (19, 67, 68) with portions of TM1, TM2, and/or TM3 all being identified as contributing to the pore lining. However, this model provides a first approximation for its spatial location and insights into why it plays a critical functional role.

Within the Fleishman et al. (18) model, the Thr-134 position of Cx32 is in a region of the TM3 helix that is buried on all sides by other helices at the cytoplasmic end of the bilayer. This is consistent with the deductions from the SCAM studies (19) described above where Thr-134 might be expected to interact with adjacent helices. The cytoplasmic end of TM3 is the region of the connexon structure that shows the tightest packing of helices in the electron crystallographic model (61). Thus, Thr-134 could form a contact with the TM4 helix by possibly interacting with the Glu-208 residue and therefore may be important for determining inter-helix contacts. Whether the Thr OH group forms hydrogen bonds with amino acids nearby in three-dimensional space, but in a different part of the sequence, remains to be determined. Certainly, a test of this theory is whether replacements of the Thr with a Val (same size but no OH group) or a Ser (maintains the OH group but is smaller) would maintain function of the channels. This would indicate that hydrogen bonding at this position plays a role in inter-helical packing or whether it is the specific size of the Thr that is important in maintaining the helix integrity or both are a factor. Notably, cysteine substitutions for this threonine in both Cx32 and Cx50 fail to preserve function, despite retaining some H-bonding potential in the introduced SH group, albeit reduced compared with the endogenous OH group. However, the cysteine does not have the -branched structure of Thr and is likely to present the SH group in a distinct orientation.

Although the Fleishman model might suggest an interaction of Thr-134 with Glu-208 in TM4 as described above, the mapping of Tyr-135 as being within the pore by Skerrett et al. (19) would place Thr-134 on the opposite face to TM4 based on the models presented in both Refs. 19 and 18. This would suggest an interaction with either TM1 or TM2, depending on the model. This is consistent with the demonstration that Arg-142 (8 residues away and hence on the opposite side of the helix) interacts with TM4, based on the R142C mutant forming a spontaneous disulfide with Cys-201 in TM4 that locks the channel in the closed state (19). Comparison of the models deduced by Fleishman et al. (18) and Skerrett et al. (19) is further complicated by the fact that the SCAM analysis was based on accessibility in open channels, whereas the Fleishman et al. (18) model was constructed from an oleamide-treated sample, where the channel has been speculated to be in a closed or partially closed state.⁹ Hence, should TM3 undergo rotation during gating, Thr-134 could interact with different helices in the open and closed states.

Consequences of Destabilizing Structure—Our EM data indicate that substitution at the Thr position causes only minor alterations in the structure as seen in our projection images but causes profound changes on the physiology of the mutant connexin channel. In particular, in all of our electrophysiological and dye transfer assays, the pore is essentially closed. Whether this reflects a true closed state or an artificial one is a subject for further investigation. Nonetheless, to our knowledge, this is the first time that a direct structure/function relationship has been shown for a mutant connexon at the resolutions that EM provides. Specifically, we see a smaller pore size consistent with the idea that the pore is closed in this mutant and a rearrangement of the subunit structure that would indicate a shifting of subunit substructure between wt and mutant, perhaps reflecting a destabilizing effect of this substitution on this structure. The reduction in measured pore diameter is 17%, although in projection images, it is difficult to assign that reduction to specific areas of the structure. The expectation is that the relative decrease in the diameter at the extracellular surface (also called the extracellular "loop" gate (69) would be greater than the cytoplasmic gate based on atomic force microscopy measurements of the extracellular surface pore diameter before and after calcium-induced closure (41).

Our EM data argue for the role of Thr in maintaining either inter-helical or intra-helical stability that is consistent with both models discussed above, because detergent solubilization of the channels results in breakdown of the hexamer. The normal appearance of mutant gap junction plaques signifies that this mutant could form the complex within the membrane. Because the substitution is a relatively modest one, when held together in the membrane, the lipid environment helps to stabilize the connexin channel. As the complex is solubilized and released from the membrane, one see a series of structures in the EM and oligomers in the native gel, further enforcing the idea that the substitution causes a modest destabilization in the structure. If the modification had caused a dramatic change in structure, there should have been only been aggregates in the detergent and no hexameric forms. This contrasts to studies by Oshima et al. (26), which showed that Cx26M34T or Cx26M34A expressed and purified from baculovirus-infected Sf9 insect cells remained intact when solubilized in dodecyl maltoside, but have greatly reduced dye transfer permeability arguing that the structural change affected the pore but left inter-subunit contacts intact. Other residues have also been implicated in the stability of subunit interactions, including Arg-75 in E1 and Arg-153 and Gln-173 (58, 59) in TM3 of Cx43.

Potential Use of This Mutant for Knocking Out or Down Connexin Functionality—Beyond the interest in this mutant for the role it plays in gap junction structure, we propose that this mutant is also useful as a tool for the gap junction community. As shown in Fig. 11, transfection of keratinocytes with Cx43T154A significantly reduced initial Alexa 350 dye transfer, presumably by making heteromeric combinations with Cx43wt but also with any other connexins that are compatible for heteromeric combinations with Cx43. Because keratinocytes contain multiple connexins, the eventual dye transfer is probably due to transfer through other connexin channels found in smaller amounts than Cx43. We believe it is easier to transfect cells with this construct to knock down function rather than either generating knock-out animals or using small interfering RNA interference, which would be specific to a single connexin isotype. Our studies indicate a broad dominant-negative phenotype among both -class (Cx43 and 50) and -class (Cx26 and possibly Cx32) connexins.

Conclusions—Mutation of the Thr residue at the cytoplasmic end of TM3 in connexins of the - and -subgroups does not interfere with synthesis, trafficking, and docking, yet it leads to the formation of nonfunctional channels with dramatically reduced dye and ion transfer properties, a smaller pore size, and structural instability of the connexin oligomer.

The Thr mutation in this location also has a dominant-negative effect, as we show in homotypic, heterotypic, and heteromeric combinations of Cx43, Cx26, and Cx50. This mutation may be useful as a tool for knocking down or knocking out connexin function in endogenously expressing connexin cell lines.

	FOOTNOTES

 
^* This work was supported by National Institutes of Health Grants GM065937 (to G. E. S.), GM072881 (to G. E. S.), GM048773 (to B. J. N.), and CA048049 (to B. J. N.) and National Science Foundation Grant MCB-0131425 (to G. E. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

¹ Both authors contributed equally to this work.

² Present address: Dept. of Biophysics, Kyoto University, Kyoto 606-8502, Japan.

³ Present address: Dept. of Biochemistry, University of Texas Health Sciences Center, San Antonio, TX 78229-3900.

⁴ To whom correspondence should be addressed: University of California, San Diego, 1070 Basic Science Bldg. MC 0608, 9500 Gilman Dr., La Jolla, CA 92093-0608. Tel.: 858-534-0128; Fax: 858-534-7497; E-mail: gsosinsky{at}ucsd.edu.

⁵ The abbreviations used are: Cx, connexin; wt, wild type; TC, tetracysteine; GFP, green fluorescent protein; SCAM, substituted cysteine accessibility method; CMTX, X-linked Charcot-Marie-Tooth; TM, transmembrane; MP, MPCCPGCCGS; ER, endoplasmic reticulum; EDT, 1,2-ethanedithiol.

⁶ The tetracysteine constructs used in this paper are as follows: TC, AEAAAREACCRECCARA (20); TC-PG, AEAAAREACCPGCCARA (21); MP, MPCCPGCCGS (B. R. Martin, B. N. Giepmans, S. R. Adams, and R. Y. Tsien, unpublished results); and 4C, FLNCCPGCCMEP (22, 23). These are listed in order of oldest to newest published sequences (lowest affinity to highest affinity) developed by Tsien and co-workers.

⁷ D. L. Beahm, M. M. Toloue, and B. J. Nicholson, unpublished observations.

⁸ S. N. Zucker, A. Chandrasekhar, and B. J. Nicholson, unpublished data.

⁹ M. Yeager, personal communication.

	ACKNOWLEDGMENTS

 
We thank Dr. Steve Ludtke and Robert Ashmore for their image processing programs, development of the software for this study, and aid in implementing them. We thank Brent Martin for first identification of this mutant and pointing this out to us and Dr. Jean Jiang for advice with the native gels, Dr. Stelios Andreadis for provision of the human keratinocytes and help with infection studies, and Dr. Martha Skerrett for help with the electrophysiology. We also thank Dr. D. Miller for the gift of the PA317 packaging cells. The EMAN development is supported by National Institutes of Health Grant NCRR P41RR02250. Some of the work included here was conducted at the National Center for Microscopy and Imaging Research at San Diego, which is supported by National Institutes of Health Grant RR04050 awarded to Dr. Mark Ellisman.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Willecke, K., Eiberger, J., Degen, J., Eckardt, D., Romualdi, A., Guldenagel, M., Deutsch, U., and Sohl, G. (2002) Biol. Chem. 383, 725-737[CrossRef][Medline] [Order article via Infotrieve]
2. Reaume, A. G., de Souza, P. A., Kulkarni, S., Langille, G. L., Zhu, D., Davies, T. C., Junega, S. C., Kidder, G. M., and Rossant, J. (1995) Science 267, 1831-1834[Abstract/Free Full Text]
3. Nelles, E., Butzler, C., Jung, D., Temme, A., Gabriel, H. D., Dahl, U., Traub, O., Stumpel, F., Jungermann, K., Zielasek, J., Toyka, K. V., Dermetziel, R., and Willecke, K. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 9565-9570[Abstract/Free Full Text]
4. Gourdie, R. G., and Lo, C. W. (2000) in Gap Junctions: Molecular Basis of Cell Communication in Health and Disease (Peracchia, C., ed) Vol. 49, pp. 581-602, Academic Press, San Diego
5. Goldberg, G. S., Lampe, P. D., and Nicholson, B. J. (1999) Nat. Cell Biol. 1, 457-459[CrossRef][Medline] [Order article via Infotrieve]
6. Beltramello, M., Piazza, V., Bukauskas, F. F., Pozzan, T., and Mammano, F. (2005) Nat. Cell Biol. 7, 63-69[CrossRef][Medline] [Order article via Infotrieve]
7. Gilula, N. B., Reeves, O. R., and Steinbach, A. (1972) Nature 235, 262-265[CrossRef][Medline] [Order article via Infotrieve]
8. Bennett, M. V. L., and Abrams, C. K. (2000) in Gap Junctions: Molecular Basis of Cell Communication in Health and Disease (Peracchia, C., ed) Vol. 49, pp. 423-459, Academic Press, San Diego
9. Krutovskikh, V., and Yamasaki, H. (2000) Mutat. Res. 462, 197-207[CrossRef][Medline] [Order article via Infotrieve]
10. Patel, P. I., and Lupski, J. R. (1994) Trends Genet. 10, 128-133[CrossRef][Medline] [Order article via Infotrieve]
11. Scherer, S. S., Bone, L. J., Deschênes, S. M., Abel, A., Balice-Gordon, R. J., and Fischbeck, K. H. (1999) Novartis Found. Symp. 219, 175-187[Medline] [Order article via Infotrieve]
12. Kelsell, D. P., Dunlop, J., Stevens, H. P., Lench, N. J., Liang, J. N., Parry, G., Mueller, R. F., and Leigh, I. M. (1997) Nature 387, 80-83[CrossRef][Medline] [Order article via Infotrieve]
13. Richard, G., Smith, L. E., Bailey, R. A., Itin, R., Hohl, D., Epstein, E. H. J., DiGovanna, J. J., Compton, J. G., and Bale, S. J. (1998) Nat. Genet. 20, 366-369[CrossRef][Medline] [Order article via Infotrieve]
14. Mackay, D., Ionides, A., Kibar, Z., Rouleau, G., Berry, V., Moore, A., Shiels, A., and Bhattacharya, S. (1999) Am. J. Hum. Genet. 64, 1357-1364[CrossRef][Medline] [Order article via Infotrieve]
15. Shiels, A., Mackay, D., Ionides, A., Berry, V., Moore, A., and Bhattacharya, S. (1998) Am. J. Hum. Genet. 62, 526-532[CrossRef][Medline] [Order article via Infotrieve]
16. Paznekas, W. A., Boyadjiev, S. A., Shapiro, R. E., Daniels, O., Wollnik, B., Keegan, C. E., Innis, J. W., Dinulos, M. B., Christian, C., Hannibal, M. C., and Jabs, E. W. (2003) Am. J. Hum. Genet. 72, 408-418[CrossRef][Medline] [Order article via Infotrieve]
17. Britz-Cunningham, S. H., Shah, M. M., Zuppan, C. W., and Fletcher, W. H. (1995) N. Engl. J. Med. 332, 1323-1329[Abstract/Free Full Text]
18. Fleishman, S. J., Unger, V. M., Yeager, M., and Ben-Tal, N. (2004) Mol. Cell 15, 879-888[CrossRef][Medline] [Order article via Infotrieve]
19. Skerrett, I. M., Aronowitz, J., Shin, J. H., Cymes, G., Kasperek, E., Cao, F. L., and Nicholson, B. J. (2002) J. Cell Biol. 159, 349-360[Abstract/Free Full Text]
20. Gaietta, G., Deerinck, T. J., Adams, S. R., Bouwer, J., Tour, O., Laird, D. W., Sosinsky, G. E., Tsien, R. Y., and Ellisman, M. H. (2002) Science 296, 503-507[Abstract/Free Full Text]
21. Adams, S. R., Campbell, R. E., Gross, L. A., Martin, B. R., Walkup, G. K., Yao, Y., Llopis, J., and Tsien, R. Y. (2002) J. Am. Chem. Soc. 124, 6063-6076[CrossRef][Medline] [Order article via Infotrieve]
22. Tsien, R. Y. (2005) FEBS Lett. 579, 927-932[CrossRef][Medline] [Order article via Infotrieve]
23. Martin, B. R., Giepmans, B. N., Adams, S. R., and Tsien, R. Y. (2005) Nat. Biotechnol. 23, 1308-1314[CrossRef][Medline] [Order article via Infotrieve]
24. Deerinck, T. J., Martone, M. E., Lev-Ram, V., Green, D. P. L., Tsien, R. Y., Spector, D. L., Huang, S., and Ellisman, M. H. (1994) J. Cell Biol. 126, 910-910
25. Hand, G. M., Muller, D. J., Nicholson, B. J., Engel, A., and Sosinsky, G. E. (2002) J. Mol. Biol. 315, 587-600[CrossRef][Medline] [Order article via Infotrieve]
26. Oshima, A., Doi, T., Mitsuoka, K., Maeda, S., and Fujiyoshi, Y. (2003) J. Biol. Chem. 278, 1807-1816[Abstract/Free Full Text]
27. Schagger, H. (1995) Methods Enzymol. 260, 190-202[Medline] [Order article via Infotrieve]
28. Ludtke, S. J., Baldwin, P. R., and Chiu, W. (1999) J. Struct. Biol. 128, 82-97[CrossRef][Medline] [Order article via Infotrieve]
29. Crowther, R. A., and Amos, L. A. (1971) J. Mol. Biol. 184, 81-98
30. Ho, S. N., Hunt, H. D., Horton, R. M., Pullen, J. K., and Pease, L. R. (1989) Gene (Amst.) 77, 51-59[CrossRef][Medline] [Order article via Infotrieve]
31. Barrio, L. C., Suchyna, T., Bargiello, T., Xu, L. X., Roginski, R. S., Bennett, M. V. L., and Nicholson, B. J. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 8410-8414[Abstract/Free Full Text]
32. Hennemann, H., Suchyna, T., Lichtenberg-Frate, H., Jungbluth, S., Dahl, E., Schwarz, J., Nicholson, B. J., and Willecke, K. (1992) J. Cell Biol. 117, 1299-1310[Abstract/Free Full Text]
33. Bruzzone, R., Haefliger, J. A., Gimlich, R. L., and Paul, D. L. (1993) Mol. Biol. Cell 4, 7-20[Medline] [Order article via Infotrieve]
34. Miller, A. D., and Buttimore, C. (1986) Mol. Cell. Biol. 6, 2895-2902[Abstract/Free Full Text]
35. Andreadis, S. T., Hamoen, K. E., Yarmush, M. L., and Morgan, J. R. (2001) FASEB J. 15, 898-906[Abstract/Free Full Text]
36. Hunter, A. W., Jourdan, J., and Gourdie, R. G. (2003) Cell Commun. Adhes. 10, 211-214[Medline] [Order article via Infotrieve]
37. Falk, M. (2000) J. Cell Sci. 113, 4109-4120[Abstract]
38. Lal, R., John, S. A., Laird, D. W., and Arnsdorf, M. F. (1995) Am. J. Physiol. 268, C968-C977[Medline] [Order article via Infotrieve]
39. Zampighi, G. A., Loo, D. D., Kreman, M., Eskandari, S., and Wright, E. M. (1999) J. Gen. Physiol. 113, 507-524[Abstract/Free Full Text]
40. Lauf, U., Giepmans, B. N., Lopez, P., Braconnot, S., Chen, S. C., and Falk, M. M. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 10446-10451[Abstract/Free Full Text]
41. Muller, D. J., Hand, G. M., Engel, A., and Sosinsky, G. E. (2002) EMBO J. 21, 3598-3607[CrossRef][Medline] [Order article via Infotrieve]
42. Sosinsky, G. E., Baker, T. S., Caspar, D. L. D., and Goodenough, D. A. (1990) Biophys. J. 58, 1213-1226[Medline] [Order article via Infotrieve]
43. Jongsma, H. J., Wilders, R., Van Ginneken, A. C. G., and Rook, M. B. (1991) in Biophysics of Gap Junction Channels (Peracchia, C., ed), pp. 163-172, CRC Press, Inc, Boca Raton, FL
44. Wilders, R., and Jongsma, H. J. (1992) Biophys. J. 63, 942-953[Medline] [Order article via Infotrieve]
45. Verselis, V. K., Ginter, C. S., and Bargiello, T. A. (1994) Nature 368, 348-351[CrossRef][Medline] [Order article via Infotrieve]
46. Skerrett, I. M., Di, W. L., Kasperek, E. M., Kelsell, D. P., and Nicholson, B. J. (2004) FASEB J. 18, 860-862[Abstract/Free Full Text]
47. Weber, P. A., Chang, H. C., Spaeth, K. E., Nitsche, J. M., and Nicholson, B. J. (2004) Biophys. J. 87, 958-973[CrossRef][Medline] [Order article via Infotrieve]
48. Di, W. L., Rugg, E. L., Leigh, I. M., and Kelsell, D. P. (2001) J. Investig. Dermatol. 117, 958-964[CrossRef][Medline] [Order article via Infotrieve]
49. Unwin, P. N. T., and Ennis, P. D. (1984) Nature 307, 609-613[CrossRef][Medline] [Order article via Infotrieve]
50. Bukauskas, F. F., and Verselis, V. K. (2004) Biochim. Biophys. Acta 1662, 42-60[Medline] [Order article via Infotrieve]
51. Thimm, J., Mechler, A., Lin, H., Rhee, S., and Lal, R. (2005) J. Biol. Chem. 280, 10646-10654[Abstract/Free Full Text]
52. Schonherr, R., Mannuzzu, L. M., Isacoff, E. Y., and Heinemann, S. H. (2002) Neuron 35, 935-949[CrossRef][Medline] [Order article via Infotrieve]
53. Fuse, Y., Doi, K., Hasegawa, T., Sugii, A., Hibino, H., and Kubo, T. (1999) Neuroreport 10, 1853-1857[Medline] [Order article via Infotrieve]
54. Nelis, E., Haites, N., and Van Broeckhoven, C. (1999) Hum. Mutat. 13, 11-28[CrossRef][Medline] [Order article via Infotrieve]
55. Bone, L., Dahl, N., Lensch, M., Chance, P., Kelly, T., Le Guern, E., Magi, S., Parry, G., Shapiro, H., Wang, S., and Fischbeck, K. H. (1995) Neurology 45, 1863-1866[Abstract/Free Full Text]
56. Sorour, E., and Upadhyaya, M. (1998) Hum. Mutat. Suppl. 1, S242-S247
57. Ressot, C., Latour, P., Blanquet-Grossard, F., Sturtz, F., Duthel, S., Battin, J., Corbillon, E., Ollagnon, E., Serville, F., Vandenberghe, A., Dautigny, A., and Pham-Dinh, D. (1996) Hum. Genet. 98, 172-175[CrossRef][Medline] [Order article via Infotrieve]
58. Gabriel, H.-D., Jung, D., Bützler, C., Temme, A., Traub, O., Winterhager, E., and Willecke, K. (1998) J. Cell Biol. 140, 1453-1461[Abstract/Free Full Text]
59. Lagree, V., Brunschwig, K., Lopez, P., Gilula, N. B., Richard, G., and Falk, M. M. (2003) J. Cell Sci. 116, 3189-3201[Abstract/Free Full Text]
60. Maza, J., Das Sarma, J., and Koval, M. (2005) J. Biol. Chem. 280, 21115-21121[Abstract/Free Full Text]
61. Kumari, S. S., Varadaraj, K., Valiunas, V., and Brink, P. R. (2001) Biochem. Biophys. Res. Commun. 280, 440-447[CrossRef][Medline] [Order article via Infotrieve]
62. Unwin, N. (1986) Nature 323, 12-13[CrossRef][Medline] [Order article via Infotrieve]
63. Unger, V. M., Kumar, N. M., Gilula, N. B., and Yeager, M. (1999) Science 283, 1176-1180[Abstract/Free Full Text]
64. Paul, D. L. (1986) J. Cell Biol. 103, 123-134[Abstract/Free Full Text]
65. Kumar, N. M., and Gilula, N. B. (1986) J. Cell Biol. 103, 767-776[Abstract/Free Full Text]
66. Milks, L. C., Kumar, N. M., Houghten, R., Unwin, N., and Gilula, N. B. (1988) EMBO J. 7, 2967-2975[Medline] [Order article via Infotrieve]
67. Kronengold, J., Trexler, E. B., Bukauskas, F. F., Bargiello, T. A., and Verselis, V. K. (2003) J. Gen. Physiol. 122, 389-405[Abstract/Free Full Text]
68. Zhou, X.-W., Pfahnl, A., Werner, R., Hudder, A., Lianes, A., Luebke, A., and Dahl, G. (1997) Biophys. J. 72, 1946-1953[Medline] [Order article via Infotrieve]
69. Trexler, E. B., Bennett, M. V. L., Bargiello, T. A., and Verselis, V. K. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 5836-5841[Abstract/Free Full Text]
70. Frei, K., Lucas, T., Ramsebner, R., Schofer, C., Baumgartner, W. D., Weipoltshammer, K., Erginel-Unaltuna, N., Wachtler, F. J., and Kirschhofer, K. (2004) Audiol. Neurootol. 9, 47-50[CrossRef][Medline] [Order article via Infotrieve]
71. Cryns, K., Orzan, E., Murgia, A., Huygen, P. L., Moreno, F., del Castillo, I., Chamberlin, G. P., Azaiez, H., Prasad, S., Cucci, R. A., Leonardi, E., Snoeckx, R. L., Govaerts, P. J., Van de Heyning, P. H., Van de Heyning, C. M., Smith, R. J., and Van Camp, G. (2004) J. Med. Genet. 41, 147-154[Abstract/Free Full Text]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				D. A. Goodenough and D. L. Paul Gap Junctions Cold Spring Harb. Perspect. Biol., July 1, 2009; 1(1): a002576 - a002576. [Abstract] [Full Text] [PDF]

				A. L. Harris Gating on the outside J. Gen. Physiol., June 1, 2009; 133(6): 549 - 553. [Full Text] [PDF]

				K. Otsu, S. Das, S. D. Houser, S. K. Quadri, S. Bhattacharya, and J. Bhattacharya Concentration-dependent inhibition of angiogenesis by mesenchymal stem cells Blood, April 30, 2009; 113(18): 4197 - 4205. [Abstract] [Full Text] [PDF]

				G. Mese, V. Valiunas, P. R. Brink, and T. W. White Connexin26 deafness associated mutations show altered permeability to large cationic molecules Am J Physiol Cell Physiol, October 1, 2008; 295(4): C966 - C974. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 281/12/7994    most recent M506533200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Beahm, D. L. Articles by Sosinsky, G. E. Search for Related Content PubMed PubMed Citation Articles by Beahm, D. L. Articles by Sosinsky, G. E. Social Bookmarking                      What's this?