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

J. Biol. Chem., Vol. 275, Issue 33, 25207-25215, August 18, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/33/25207    most recent M002608200v1 Alert me when this article is cited 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 Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Musil, L. S. Articles by Roberts, L. M. Search for Related Content PubMed PubMed Citation Articles by Musil, L. S. Articles by Roberts, L. M. Social Bookmarking                      What's this?

Regulation of Connexin Degradation as a Mechanism to Increase Gap Junction Assembly and Function^*

Linda S. Musil, Anh-Chi N. Le, Judy K. VanSlyke, and Lori M. Roberts

From the Vollum Institute for Advanced Biomedical Research, Oregon Health Sciences University, Portland, Oregon 97201

Received for publication, March 27, 2000, and in revised form, May 10, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Connexins, the integral membrane protein constituents of gap junctions, are degraded at a rate (t_1/2 = 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.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Gap junctions are ordered arrays of intercellular plasma membrane channels that directly link the cytosols of two adjoining cells. By serving as low resistance pathways for the diffusional transfer of low molecular weight substances such as ions, amino acids, and second messengers, gap junctions play important roles in embryonic development, growth control, and the function of differentiated tissues throughout the body (1). In vertebrates, gap junctions are composed of members of a highly homologous family of tetraspanning integral membrane proteins known collectively as connexins. The first step in the assembly of gap junctions is the noncovalent oligomerization of six connexin monomers into an annular structure known as a connexon, or hemichannel (2). After transport to the cell surface, two connexons on apposing cell surfaces dock to form an intercellular channel. These channels then cluster at densities up to 10,000/µm² into paracrystallin arrays referred to as gap junctional "plaques," each of which can contain from ~10 to more than 10,000 intercellular channels. To date, no nonconnexin protein has been demonstrated to be a component of gap junctional plaques or to be essential for gap junction assembly.

One of the most unusual aspects of gap junction biosynthesis is the exceptional metabolic lability of connexins. As assessed from the degradation rate of total cell surface biotinylated proteins, the half-life of the great majority of plasma membrane proteins exceeds 24 h (3). In contrast, pulse-chase analysis has demonstrated that connexin family members turn over with a half-life of only 1.5-5 h, even after incorporation into gap junctional plaques. This rapid rate of degradation has been observed in a wide variety of mammalian systems including primary and established tissue culture cells (4, 5), whole organs (6), and intact animals (7). The instability of connexins is especially remarkable given the relatively long half-lives of components of the tight junction (8) and of desmosomes (9), and because intercellular communication can be rapidly down-regulated by gating of the gap junctional channel without the need for connexin degradation.

To date, two proteolytic pathways have been implicated in connexin turnover in intact cells. The first is the lysosome, which electron microscopy localization studies have shown mediates the degradation of internalized gap junctional plaques in at least some cell types (10). The second is the proteasome, the multicatalytic protease complex that degrades most fast turnover proteins in the cytosol and nucleus and that also plays a role in the degradation of proteins in the secretory pathway. Laing and colleagues (6, 11, 12) have reported that chemical inhibitors of the proteasome decrease the rate of turnover of connexin43 (Cx43)¹ in tissue culture cells (11, 12) and in intact heart (6) as determined by pulse-chase analysis. We have recently obtained similar results for another member of the connexin family, connexin32 (Cx32) (13). Experiments in which brefeldin A was used to block intracellular transport suggested that proteasome inhibitor-sensitive degradation of connexins occurs at the level of the ER (13) as well as after transport to the cell surface (12). Both sites have previously been implicated in proteasome-mediated degradation of other integral plasma membrane proteins (14). In none of the aforementioned studies were the consequences of proteasome inhibition on gap junction function assessed.

Although highly unusual for an integral plasma membrane protein, rapid turnover kinetics are a common characteristic of mature, fully folded forms of cytoplasmic and nuclear proteins involved in signal transduction. Before such proteins can be degraded by the proteasome, they must undergo a multistep targeting process. First, the substrate is usually distinguished from other wild-type molecules by posttranslational "tagging" events such as a change in phosphorylation and/or protein-protein interaction that facilitate the addition of polyubiquitin chains and mark it for destruction (15). The substrate must then be completely unfolded to gain entry into the proteolytic chamber of the 20S proteasome core, a process that for at least some proteins appears to require extraproteasomal components such as cytosolic molecular chaperones (16). Decreasing the efficiency with which cytosolic and nuclear proteins are targeted for proteasomal degradation is a physiologically important mechanism by which their signaling function is prolonged or enhanced (17). For example, regulated proteasomal degradation of the cytosolic enzyme ornithine decarboxylase requires association with a protein known as antizyme. Preventing antizyme synthesis with cycloheximide stabilizes ornithine decarboxylase and preserves its enzymatic function (18, 19). Although the mechanism is not yet clear, protein synthesis inhibitors also slow the turnover and thereby prolong the activity of certain other (but not all) cytoplasmic proteasome substrates, including pp60^v-src (20), estradiol-liganded estrogen receptors (21), and some forms of p53 (22).

The objective of the current study was to determine whether the signaling function of integral membrane connexins (intercellular communication) could be increased by slowing the rate of connexin turnover. We show here that reducing connexin degradation with inhibitors of the proteasome or of protein synthesis (but not with inhibitors of the lysosome) is associated with a striking increase in gap junctional plaque assembly and intercellular dye transfer in cells that are inefficient in both processes under basal conditions. Our findings support a model in which protein synthesis inhibitors interfere with events required for targeting of folded forms of connexins for proteasomal degradation, possibly by lowering the levels of a fast-turnover protein required for this process. Regulation of the rate of connexin degradation is a novel means of posttranslationally increasing intercellular communication that is distinct from gating of the gap junctional channel.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Culture-- Mouse sarcoma cells stably transfected with the adhesion molecule L-CAM (S180L cells) (23), NRK fibroblasts, and L929 fibroblasts were cultured in DMEM plus 10% FCS, penicillin G, and streptomycin as described previously (4). CHO-K1 cells were maintained in F-12 medium with the same additions and were refed with fresh medium 6-8 h prior to use. Three-day-old, newly confluent cultures were used for all experiments unless otherwise specified. Serum-starved S180L (ssS180L) cells were generated by rinsing 2-day-old S180L cultures three times with serum-free F-12 medium and culturing the cells in the same medium for 8-12 h prior to use. Clonal lines of PC12J cells (a subclone of rat pheochromocytoma cells devoid of endogenous gap junction activity) stably expressing either wild-type human Cx32 or a CMTX-linked Cx32 mutant were generated and maintained in RPMI 1640 medium supplemented with 10% horse serum, 5% FCS, penicillin G, streptomycin, and 400 µg/ml G418 as described by Deschenes et al. (24).

Antibodies and Reagents-- Anti-Cx43 antibodies were affinity purified from a rabbit antiserum generated against a synthetic peptide encoding amino acids 252-271 of rat Cx43 (4) using a glutathione Sepharose-immobilized recombinant fusion protein consisting of glutathione-S-transferase linked to residues 229-382 of rat Cx43. Cx32 was immunoprecipitated with a crude antiserum from rabbits immunized with a synthetic peptide corresponding to amino acids 98-124 of rat Cx32 (kindly provided by D. Goodenough, Harvard Medical School, Boston, MA) (25). Unless otherwise noted, all other reagents were from Sigma except for carboxybenzyl-leucyl-leucyl-leucine-vinylsulfone (ZL₃VS), a gift of M. Bogyo (Harvard Medical School). Final concentrations of reagents in experiments were as follows: 100 µM N-acetyl-leu-leu-norleucinal (ALLN), 10 µM lactacystin, 20 µM ZL₃VS, 100 µg/ml leupeptin, 200 µM chloroquine, 2 mM dithiothreitol (DTT), and 20 µg/ml cycloheximide. At this concentration, cycloheximide reduced the synthesis of total cellular proteins (including connexins) by >96% within 10 min (not shown).

Immunofluorescence Microscopy-- Cells grown on glass coverslips were fixed in 2% paraformaldehyde in phosphate-buffered saline (PBS) (pH 7.5) for 30 min at room temperature, rinsed for 30 min with PBS, and then postfixed for 5 min with 20 °C acetone. After a second 30-min rinse with PBS, the cells were incubated for 30 min in PBS supplemented with 0.5% normal goat serum, 0.1% bovine serum albumin, 0.2% Triton X-100, and 0.02% sodium azide. Cells were sequentially incubated with anti-Cx43 antibodies and then with rhodamine-conjugated goat anti-rabbit IgG (Pierce) as described by Le and Musil (26). The coverslips were mounted onto glass slides with MOWIOL 4-88 mounting medium (Calbiochem) and immunofluorescence images captured using a Leica DM LD photomicrography system and Scion Image 1.60 software.

Scrape Loading/Dye Transfer Assay of Gap Junctional Intercellular Communication-- Cells grown on 5-mm-diameter glass coverslips were rinsed three times with Hanks' balanced salt solution containing 1% bovine serum albumin (HB). The HB was removed and 2.5 µl of PBS containing 1% Lucifer Yellow (LY) (Sigma) and 0.75% of the gap junction-impermeant compound rhodamine dextran (fluoro-ruby; molecular mass, 10 kDa) (Molecular Probes) was applied to the center of the glass coverslip. A 27 gauge needle was then used to create two longitudinal scratches through the cell monolayer. Cells were incubated in the dye mix for exactly 1 min and then quickly rinsed three times with HB prior to a 5-min incubation in HB at room temperature to allow dye transfer. The cells were rinsed three times with PBS and immediately fixed for 30 min at room temperature with 2% paraformaldehyde/PBS, pH 7.5. LY and rhodamine dextran were subsequently examined by fluorescence microscopy (Leitz DMR) using fluorescein and rhodamine filter sets, respectively. Only images taken with fluorescein optics are shown; in all cases, rhodamine dextran remained confined to a single row of cells immediately bordering the wound. In some experiments, 1% biocytin (Molecular Probes) was substituted for LY and visualized with avidin-FITC (Molecular Probes) as described by Le and Musil (26). The scrape loading/dye transfer assay has the advantage over microinjection techniques of allowing simultaneous monitoring of dye coupling in a large population of cells; its utility in the assessment of gap junction-mediated intercellular communication has been well documented (27, 28).

Metabolic Labeling and Immunoprecipitation-- Cells were starved for methionine for 30 min at 37 °C in DMEM lacking methionine and supplemented with 5% dialyzed FCS and 2 mM glutamine ("labeling" medium). The medium was then replaced with fresh labeling medium containing [³⁵S]methionine (Expre³⁵S³⁵S, NEN Life Science Products; 0.1 mCi/35-mm dish of cells, scaled proportionally for dishes of other sizes). The radioactive medium was removed after a 30-min pulse, and if desired, the cells were then rinsed three times with complete DMEM supplemented with 0.5 mM unlabeled methionine and chased in DMEM/0.5 mM unlabeled methionine/10% FCS at 37 °C. At the end of the chase period, cells were rinsed once with PBS at 4 °C and resuspended in lysis buffer (5 mM Tris base, 5 mM EDTA, 5 mM EGTA, 10 mM iodoacetamide, 2 mM phenylmethylsulfonyl fluoride, pH 8.0) supplemented with 0.6% SDS, 250 mg/ml soybean trypsin inhibitor, and 200 µM leupeptin. The cell lysates were then immunoprecipitated with the desired antibody as described by Le and Musil (26). Two modifications were made for immunoprecipitation of Cx32: first, cells were lysed in SDS at room temperature for 30 min instead of at 100 °C for 3 min to minimize aggregation of Cx32; and second, samples were precleared with protein A-Sepharose for 2 h at 4 °C prior to addition of anti-Cx32 antibody (29). Cx43 and Cx32 immunoprecipitates were analyzed on 10 or 11% SDS-polyacrylamide gel electrophoresis gels, respectively, after incubation of samples in SDS-polyacrylamide gel electrophoresis sample loading buffer for either 3 min at 100 °C (Cx43) or 30 min at room temperature (Cx32). The dried gels were quantitated on a PhosphorImager 445 SI (Molecular Dynamics, Inc.) utilizing IPLab Gel software (Signal Analytics Corp.).

Pulse-Chase Analysis of Cx43 in Newly Forming S180L Cell Monolayers-- S180L cells were grown in 100-mm tissue culture dishes in DMEM/10% FCS until ~80% confluent. After two rinses with PBS lacking Ca⁺² and Mg⁺² (PBS-), the cultures were incubated in PBS-, 5 mM EDTA, and 2% FCS for 10 min at room temperature on a orbital shaker. The cells were then removed from the plate and triturated 10 times with a 10-ml pipette to generate a single-cell suspension. The cells were pelleted at 150 × g for 7 min and resuspended in labeling medium (see above). After verification of cell viability (>95% Trypan blue-excluding) and the absence of cell clumps, 5 × 10⁵ cells in 500 µl of labeling medium were added to each well of a 12-well tissue culture plate. Label (60 µCi of Expre³⁵S³⁵S/well) was then immediately added, and the cells were incubated at 37 °C for 30 min, during which time the cells quantitatively adhered to the dish. The cultures were then rinsed three times with complete DMEM supplemented with 0.5 mM unlabeled methionine and chased for the desired period prior to cell lysis and immunoprecipitation of Cx43 as described above.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Cell Type-specific Differences in Gap Junction Assembly Competence-- A striking feature of gap junction biosynthesis both in vivo and in tissue culture is that cells differ greatly in their inherent ability to assemble connexins into gap junctional plaques. The four cell lines shown in Fig. 1 synthesize comparable quantities of endogenously expressed, wild-type Cx43, the coding sequence of which is very highly conserved among mammals and not subject to alternative mRNA splicing. We have previously shown that each cell type assembles this Cx43 into connexons and transports them to the cell surface (30, 31). However, under basal conditions, only NRK and S180L cells are able to efficiently assemble Cx43 into large gap junctional plaques, which appear as punctate or linear concentrations of anti-connexin immunoreactivity at cell-cell interfaces (Fig. 1, A and C). CHO cells have substantially fewer and generally smaller gap junctions (Fig. 1E; see also Figs. 5A and 7A), even if the level of Cx43 expression is increased by refeeding the cells in fresh culture medium for 6-8 h (data not shown). At the resolution of immunofluorescence microscopy, L929 fibroblasts lack detectable gap junctions (Fig. 1G). The number of immunodetectable gap junctions in NRK, S180L, CHO, and L929 cells correlates closely with their capacity to mediate intercellular communication as assessed by a scrape loading/dye transfer assay. As previously reported (4), NRK and S180L cells are highly communication-competent and transfer LY to several orders of cells distal to the wound edge (Fig. 1, B and D). Intercellular movement of LY is still detectable but much more restricted in CHO cells (Fig. 1F), whereas the dye remains confined to wounded cells in monolayers of L929 fibroblasts (Fig. 1H). Based on their capacity to assemble functional gap junctions under basal conditions, we herein refer to cells as either "assembly-efficient" (e.g. NRK and S180L cells), "assembly-inefficient" (e.g. CHO cells), or "assembly-incompetent" (e.g. L929 cells).

View larger version (48K): [in this window] [in a new window]   Fig. 1.   Three gap junction assembly phenotypes. Confluent monolayers of NRK, serum-maintained S180L, CHO, and L929 cells were either fixed and immunostained with anti-Cx43 antibodies (top row) or assessed for their ability to carry out gap junction-mediated intercellular transfer of LY as described under "Experimental Procedures" (bottom row).

Assembly and Turnover of Cx43 during Initial Establishment of Gap Junctional Plaques-- Despite their differences in gap junction phenotype, each of the four cell lines shown in Fig. 1 degrades Cx43 with a t_1/2 of only ~1.5-4 h as assessed by pulse-chase analysis of confluent cell monolayers (4, 11). These kinetics are in keeping with results obtained in other systems with high levels of preexisting intercellular contact (5-7, 32). One possibility not addressed by such experiments is that the first gap junctional plaques assembled at newly established cell-cell contacts might be metabolically stable and that there are a limited number of such stable gap junctions that can be accommodated at cell interfaces. If so, then any [³⁵S]methionine-labeled Cx43 synthesized in confluent cultures of gap junction-rich cells after saturation is reached would be turned over at a rate indistinguishable from that of connexins in less assembly-competent cells. Differential cell surface stability has been described for the nicotinic acetylcholine receptor, which has a t_1/2 of 8-10 days when localized to the neuromuscular junction but only 1 day at extrajunctional sites (33, 34). To test whether the extent of preexisting intercellular contact influences Cx43 assembly and/or turnover, pulse-chase studies were conducted in both established and newly forming monolayers of S180L cells. In 2-day-old, confluent cultures, pulse-labeled [³⁵S]methionine-Cx43 underwent the posttranslational changes in electrophoretic mobility characteristic of Cx43 in assembly-efficient cells (4) (Fig. 2A). Following synthesis as a 42-kDa species referred to as Cx43-NP, [³⁵S]methionine-Cx43 was progressively converted to two slower migrating species, Cx43-P₁ and Cx43-P₂, by phosphorylation on serine residues. Numerous studies in a wide variety of systems, including S180L cells, have established that conversion to the P₁ and P₂ forms occurs only after transport to the cell surface and that Cx43-P₂ is concentrated in gap junctional plaques (30, 32, 35). All forms of [³⁵S]methionine-Cx43 were largely turned over by 8 h of chase (Fig. 2A, lane 4). To examine Cx43 processing during de novo formation of cell-cell contacts, S180L cells were dissociated into a single-cell suspension under protease-free conditions and then replated at confluent density to rapidly reform a monolayer. [³⁵S]Methionine-Cx43 synthesized during the initial 30-min replating period was converted to the plaque-associated P₂ form at a rate that closely correlated with the appearance of gap junctions as assessed by immunofluorescence microscopy (data not shown). The labeled Cx43 was degraded over an 8-h period with kinetics nearly identical to those of Cx43 in two-day-old monolayers (Fig. 2B). Similar results were obtained when cells were pulsed 3 h after replating (not shown). The lack of a detectable pool of metabolically stable Cx43 indicates that the turnover rate of connexins is unaffected by the presence or absence of preexisting gap junctional plaques.

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Cx43 processing and turnover during de novo formation of cell-cell contacts. A, a confluent monolayer of S180L cells established over a 2-day period was pulsed for 30 min with [35S]methionine and then chased for 0, 1, 4, or 8 h. B, S180L cell cultures were dissociated into a single-cell suspension under protease-free conditions that inhibit the function, but preserve the structural integrity, of calcium-dependent cell-cell adhesion molecules. The cells were replated at confluent density in the presence of [35S]methionine and after a 30-min labeling period were chased for 0, 1, 4, or 8 h. After the chase period, all cells were lysed, and Cx43 was analyzed by immunoprecipitation and SDS-polyacrylamide gel electrophoresis. Exposure time was adjusted for optimal visualization of [35S]methionine-Cx43 at the 8-h time point. The percentage of pulse-labeled Cx43 remaining at this time was 10% in replated cells (B, lane 4) and 16% in established monolayers (A, lane 4). Cx43-NP, Cx43-P1, and Cx43-P2 refer to differentially phosphorylated forms of Cx43 (see text).

Effect of Inhibitors of the Proteasome and Lysosome on Cx43 Turnover-- As a first step toward determining whether a reduction in the rate of connexin degradation influences gap junction assembly and/or function, we assayed inhibitors of either the proteasome or the lysosome for their ability to decrease Cx43 turnover in NRK, S180L, CHO, and L929 cells (Fig. 3). As expected, only 9-25% of the Cx43 synthesized during a 30-min pulse was recovered after a 7-h chase in the absence of inhibitors in all four cell types. Although there was some variability between experiments in the magnitude of the effect, the proteasome inhibitor ALLN reproducibly (in 15 of 15 experiments) increased the amount of labeled Cx43 that survived the chase period between 2- and 7-fold in all cell types examined relative to untreated controls within the same trial. In a less extensive series of experiments, mechanistically distinct, highly selective inhibitors of the proteasome (ZL₃VS and lactacystin) (36) yielded similar results (data not shown). The lysosomotropic amine chloroquine (CLQ) significantly decreased Cx43 degradation in assembly-efficient NRK and S180L cells as well as in assembly-incompetent L929 cells. In contrast, CLQ only very weakly reduced Cx43 turnover in assembly-inefficient CHO cells despite its reported ability to block lysosome-mediated catabolism in this cell type (37). A mechanistically distinct inhibitor of lysosomal proteolysis, leupeptin, was also ineffective at concentrations (100 µg/ml) that reduce connexin degradation in other cell types (13) (data not shown). Insensitivity of Cx43 degradation to lysosomal inhibitors has been previously reported for a subclone of CHO cells by Laing and Beyer (11). The susceptibility of Cx43 to proteasomal or lysosomal inhibition therefore did not correlate with gap junction assembly competence in the cell types tested.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Effect of inhibitors of the proteasome, lysosome, and protein synthesis on the turnover of Cx43. A, NRK, S180L, CHO, and L929 cells were metabolically labeled for 30 min with [35S]methionine and then chased at 37 °C for 7 h in unsupplemented chase medium (0) or in the presence of an inhibitor of either the proteasome (100 µM ALLN, 10 µM lactacystin, or 20 µM ZL3VS), the lysosome (200 µM CLQ), or protein synthesis (20 µg/ml CHX). Due to interexperimental (but not intraexperimental) differences, the percentage of pulse-labeled [35S]methionine-Cx43 recovered after a 7-h chase without inhibitor ranged from 9 to 25% in each of the cell lines examined. To permit comparison of results obtained in experiments with different basal rates of Cx43 turnover, the data are presented as the fold increase in the percentage of recovery of Cx43 after a 7-h chase in the presence of inhibitor relative to the percentage of recovery of Cx43 in cells in the same experiment chased in the absence of inhibitor. B, plot of the percentage of pulse-labeled Cx43 remaining in CHO cells after 0, 1, 3, and 7 h of chase in either the absence (open circles) or presence of 100 µM ALLN (closed circle), 200 µM CLQ (open squares), or 20 µg/ml CHX (closed squares) in a typical experiment. Note that the effect of ALLN and CHX on Cx43 degradation becomes significant after 1 h of chase, possibly reflecting the time required for ALLN to maximally inhibit proteasome function in intact cells and for CHX to reduce the levels of the putative fast-turnover mediator of connexin turnover. An apparent break in the linearity of [35S]methionine-Cx43 turnover in untreated cells at later time points (after ~80% of pulse-labeled Cx43 has been degraded) has previously been observed in rodent heart (6).

Inhibition of Proteasomal, but not Lysosomal, Degradation of Cx43 Is Associated with Up-regulation of Gap Junction Formation in Assembly-inefficient Cells-- -We next tested whether the metabolic stabilization of Cx43 caused by inhibitors of protein degradation was associated with changes in the assembly and/or function of gap junctions. The results obtained were dependent on the capacity of the cells to form gap junctional plaques under basal conditions. NRK cells showed a small increase in the already very high number of gap junctional plaques after a 3-7-h treatment with any of the proteasome inhibitors tested (ALLN, ZL₃VS, MG132, and lactacystin; ALLN data shown in Fig. 4B). Comparable results were obtained with assembly-efficient S180L cells and primary chick lens epithelial cells (not shown), consistent with previous reports that these reagents modestly enhance plaques in other intrinsically gap junction-rich cell types (12, 6). The lysosomal inhibitor CLQ did not increase the number of immunodetectable junctions in NRK, S180L, or lens epithelial cells. Instead, Cx43 staining accumulated intracellularly in vesicles with the swollen appearance characteristic of endosomal/lysosomal compartments after neutralization of their lumenal pH by this lysosomotropic amine (Fig. 4C, arrowhead). None of the proteasomal (ALLN; Fig. 4E) (ZL₃VS, MG132) or lysosomal (CLQ; Fig. 4F) (leupeptin) inhibitors tested induced detectable gap junction formation or function in assembly-incompetent L929 cells.

View larger version (39K): [in this window] [in a new window]   Fig. 4.   Effect of protein degradation inhibitors on gap junction formation in assembly-efficient and assembly-incompetent cells. Assembly-efficient NRK cells or assembly-incompetent L929 cells were stained for Cx43 after a 7-h incubation in unsupplemented medium (A and D) (0) or in the presence of 100 µM ALLN (B and E) or 200 µM chloroquine (C and F).

A fundamentally different result was obtained with assembly-inefficient cells. Incubation of CHO cells with any of the four proteasomal inhibitors tested (ALLN, ZL₃VS, lactacystin; MG132) induced a large increase in the number of immunocytochemically detectable gap junctions over a 3-6-h period (ALLN data shown in Fig. 5B). Up-regulation of plaques in CHO cells was associated with a correspondingly large increase in dye coupling as assessed by scrape loading/dye transfer analysis (Fig. 5E). Intercellular transfer of LY or biocytin was inhibited by pretreatment of cells with the gap junction blocker 18-glycyrrhetinic acid (GA) and was not observed for junction-impermeant rhodamine-dextran, demonstrating a bona fide increase in gap junction-mediated intercellular communication (data not shown). In contrast, the lysosomal inhibitors CLQ (Fig. 5, C and F) and leupeptin (not shown) induced neither gap junction assembly nor function.

View larger version (82K): [in this window] [in a new window]   Fig. 5.   Proteasome inhibition up-regulates gap junctions in assembly-inefficient cells. CHO cells and ssS180L cells were incubated for 7 h in new medium without additions (A, D, G, and J), with 100 µM ALLN (B, E, H, and K), or with 200 µM chloroquine (C, F, I, and L). The cells were then either fixed and immunostained with anti-Cx43 antibodies (A-C and G-I) or assessed for their ability to mediate intercellular transfer of the gap junction permeant LY (D-F and J-L).

Up-regulation of gap junctions in CHO cells in response to proteasome inhibitors could either be coincidentally or causally related to their assembly-inefficient phenotype under basal conditions. Evidence in support of the latter possibility was obtained in experiments in which the ability of S180L cells to assemble gap junctions was modulated by changing their culture conditions. We have empirically found that overnight incubation of S180L cells in the absence of serum greatly reduces the number of morphologically or functionally detectable gap junctions despite the continued synthesis of Cx43 and the cell-cell adhesion molecule LCAM. Treatment of such ssS180L cells with proteasome inhibitors for 3-6 h in either the absence or presence of new 10% FCS resulted in a large increase in functional gap junctions (Fig. 5, H and K) similar to that observed in CHO cells. Serum starvation therefore converted S180L cells from an assembly-efficient to an assembly-inefficient phenotype and concomitantly conferred the ability to up-regulate gap junctions in response to proteasome inhibitor treatment. Despite their efficacy in reducing the turnover of pulse-labeled Cx43 (Fig. 3A), lysosomal inhibitors (CLQ and, not shown, leupeptin) induced neither gap junction assembly nor activity in ssS180L cells (Fig. 5, I and L).

The Effect of Proteasome Inhibitors on Gap Junction Formation Is Mimicked by Inhibitors of Protein Synthesis-- Degradation of properly folded cytosolic or nuclear proteins by the ubiquitination/proteasome system usually requires that the substrate undergo one or more regulatable targeting events that distinguish it from other mature proteins (15, 19). To investigate whether such a mechanism might also control proteasome-mediated degradation of connexins, we tested agents known to block the targeting and therefore the turnover of other proteins for their effect on connexins. Among the compounds assayed, only protein synthesis inhibitors significantly and reproducibly (in 29 of 29 experiments) reduced the rate of pulse-labeled Cx43 degradation. As shown in Fig. 6, addition of 20 µg/ml cycloheximide (CHX) to the chase medium of NRK, S180L, CHO, or L929 cells increased the amount of [³⁵S]methionine-Cx43 recovered after a 7-h chase between 2.6- and 5.7-fold relative to untreated control cells within the same experiment (results of four independent experiments, quantitated in Fig. 3A). The ability of CHX to enhance the recovery of Cx43 was optimal at concentrations that inhibited protein synthesis >90% and was reversible by incubation in CHX-free medium. Moreover, synthesis-blocking quantities of the mechanistically and structurally distinct protein synthesis inhibitors puromycin and anisomycin had a similar effect (data not shown). Inhibition of protein synthesis rather than some other property of CHX therefore appears to account for its enhancement of connexin stability. There was no significant reproducible difference among the cell types tested in the extent to which CHX decreased Cx43 turnover.

View larger version (38K): [in this window] [in a new window]   Fig. 6.   Cycloheximide reduces the turnover of Cx43. CHO, L929, NRK, and serum-maintained S180L cells were pulse-labeled for 30 min with [35S]methionine and then chased for 7 h in either the absence (lanes 2, 6, 10, and 13) (cont, control) or presence (lanes 3, 7, 11, and 14) of 20 µg/ml CHX prior to immunoprecipitation of Cx43 and analysis by SDS-polyacrylamide gel electrophoresis. Cx43 immunoprecipitates from pulse-labeled CHO and L929 cells chased in the presence of 100 µM ALLN are included for comparison (lanes 4 and 8). The asterisk indicates the gap junctional plaque-associated P2 phosphoform of Cx43.

As shown in Fig. 3, the effect of CHX on Cx43 turnover was comparable to that of the proteasomal inhibitor ALLN. Moreover, both compounds shared the ability to induce phosphorylation of Cx43 to the Cx43-P₂ species in CHO cells but not in assembly-incompetent L929 fibroblasts (Fig. 6). Given the close correlation between modification of Cx43 to the P₂ form and assembly of functional gap junctional plaques (30, 32, 35), we tested whether protein synthesis inhibitors also selectively up-regulated gap junctions in assembly-inefficient cells (Fig. 7). Incubation of CHO cells with 20 µg/ml CHX resulted in an increase in the number of immunologically detectable gap junctional plaques over a 4-7-h period despite the lack of new connexin synthesis (Fig. 7C). Importantly, the increase in plaques was accompanied by a corresponding up-regulation of gap junction-mediated intercellular dye coupling (Fig. 7D). Protein synthesis-blocking concentrations of CHX or (not shown) anisomycin also increased the level of functional gap junctions in assembly-inefficient ssS180L cells (Fig. 7, G and H). As was the case with proteasome inhibitors, CHX and other protein synthesis inhibitors failed to induce detectable gap junction formation or function in assembly-incompetent L929 cells after 2-8 h of treatment (data not shown). With regard to assembly-efficient cells, CHX did not notably up-regulate gap junctions in NRK cells between 3-10 h of treatment but instead modestly decreased both gap junction number and intercellular dye coupling after extended (6-10 h) incubations (Fig. 7, K and L). Similar results were obtained with assembly-efficient primary lens epithelial cells and serum-maintained S180L cells. Potential reasons for the difference in the effect of proteasome inhibitors and protein synthesis blockers on assembly-efficient cells are addressed under "Discussion."

View larger version (46K): [in this window] [in a new window]   Fig. 7.   Cycloheximide up-regulates gap junction formation and function in assembly-inefficient, but not assembly-efficient, cells. CHO cells, ssS180L cells, and NRK fibroblasts were incubated for 7 h in new medium with or without 20 µg/ml CHX. A, C, E, G, I, and K, anti-Cx43 immunostaining. B, D, F, H, J, and L, scrape loading analysis of gap junctional intercellular communication.

Protein Synthesis Inhibitors Do Not Affect the Degradation of Misfolded Connexins-- The results presented above are consistent with the possibility that inhibition of protein synthesis reduces targeting (tagging and/or unfolding) of connexins for destruction by the proteasome (possibility 1), but they do not rule out the possibility that CHX acts instead by somehow interfering with proteasome-mediated hydrolysis of connexins after their unfolding (possibility 2) or by inhibiting a degradation pathway that mimics but is not causally related to the proteasome (possibility 3). To distinguish between these three alternatives, we compared the effect of inhibitors of the proteasome or of protein synthesis (ALLN and CHX, respectively) on the turnover of nonnative forms of connexins. Our rationale was that the tagging of a protein for proteasomal destruction as well as the ease with which it is unfolded is influenced by its conformational state, whereas its hydrolysis after linearization is not. A change in the sensitivity of degradation of misfolded connexins to CHX but not to ALLN would therefore be consistent with the first, but not the second, possibility; a loss in the ability of ALLN, but not of CHX, to inhibit the turnover of misfolded connexins would indicate that CHX is affecting a process independent of the proteasome (possibility 3). The degradation of two types of improperly folded connexin molecules was examined (Fig. 8). First, Cx43 was synthesized and chased in CHO cells in the continuous presence of 2 mM DTT, which abolishes the formation of the highly conserved intramolecular disulfide bonds that are essential for connexin function (38). Although degradation of DTT-unfolded Cx43 remained sensitive to proteasome inhibitors, CHX had virtually no effect on this process whereas turnover of Cx43 in control (no DTT) cells was effectively reduced by both ALLN and CHX (Fig. 8A). Given the multiple effects of DTT, it was, however, conceivable that the reducing agent affected Cx43 turnover indirectly instead of by altering connexin conformation. To control for this possibility and to extend our findings to another member of the connexin family, we also examined the effect of inhibitors of protein synthesis and of the proteasome on the degradation of wild-type and mutant forms of Cx32 expressed in stable PC12 cell transfectants (24, 13). Wild-type Cx32 in PC12 cells was degraded at the rapid rate characteristic of the connexin family in a process sensitive to proteasome inhibitors as well as to CHX. In the typical experiment shown in Fig. 8B, CHX increased the amounts of pulse-labeled wild-type Cx32 and ALLN that survived a 6-h chase by 4.8- and 3.8-fold, respectively, over the untreated control in lane 2. We have previously reported that three nonfunctional point mutants of Cx32 (E208K, R142W, and E186K Cx32) linked to the human peripheral neuropathy Charcot-Marie-Tooth X disease are unable to assemble into connexons and have defects in intracellular transport indicative of conformational abnormalities (13). Degradation of each of these three Cx32 mutants was sensitive to proteasome inhibitors (Fig. 8B, lanes 8, 12, and 16) but was not reduced by CHX (lanes 7, 11, and 15). CHX also caused the loss of all immunofluorescently detectable E186K, E208K, and R142W Cx32 (but not of wild-type Cx32) within a 6-h period, indicating that sensitivity to CHX is a property of the total cellular pool of mutant Cx32 and not just the fraction that incorporates [³⁵S]methionine during a 30-min pulse (data not shown). Given that each mutation is in a different domain of Cx32 and confers a distinct intracellular localization, the lack of effect of CHX on mutant turnover cannot be due to the loss of a specific amino acid residue or to retention in a particular subcellular compartment. Instead, misfolding of connexin molecules, induced by either DTT or mutation, appears to account for the loss of sensitivity to CHX. These findings support a model in which proteasomal degradation of native, but not of misfolded, connexins involves a CHX-sensitive step that targets them for proteasomal destruction. The fact that Cx43 and Cx32 show comparable responses to proteasomal and protein synthesis inhibitors despite their low sequence similarity (relative to other connexins) makes it likely that this model will apply to additional members of the connexin family.

View larger version (31K): [in this window] [in a new window]   Fig. 8.   Degradation of misfolded connexins is sensitive to inhibitors of proteasomal degradation but not to protein synthesis inhibitors. A, CHO cells endogenously expressing wild-type Cx43 were pulse-labeled for 30 min with [35S]methionine and then chased at 37 °C for 6 h in unsupplemented chase medium or in the presence of the proteasome inhibitor ALLN (100 µM) or 20 µg/ml CHX. In +DTT cells, both the pulse and chase media contained 2 mM DTT. The data are presented as the fold increase in the recovery of [35S]methionine-Cx43 after a 6-h chase in the presence of either ALLN or CHX relative to the percentage of recovery of [35S]methionine-Cx43 in the absence of inhibitor (17% -DTT, 6% +DTT). Shown is a representative data set from three independent experiments. B, stable PC12 cell transfectants expressing either wild-type Cx32 or the R142W, E186K, or E208K Cx32 mutant were labeled for 30 min with [35S]methionine and then chased for 6 h in unsupplemented chase medium (lanes 2, 6, 10, and 14) or in the presence of 20 µg/ml CHX (lanes 3, 7, 11, and 15) or 100 µM ALLN (lanes 4, 8, 12, and 16).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In vivo as in tissue culture, cells range from assembly-efficient (e.g. hepatocytes and cardiac myocytes) to assembly-incompetent (e.g. some tumor cells) (39) in their ability to oligomerize endogenously expressed, wild-type connexins into gap junctional plaques. Some cell types with relatively low levels of gap junction plaques can be operationally defined as assembly-inefficient and are able to dramatically up-regulate gap junction formation and function on a rapid time scale. This latter category includes precompaction eight-cell stage blastomeres (40) and day 21 pregnant rat myometrial smooth muscle cells (41), in which increased intercellular communication is thought to play an important role in embryonic development and parturition, respectively. A molecular understanding of up-regulation of gap junction formation and function in assembly-inefficient cells is therefore of relevance to a variety of physiological processes. In this study, we have investigated whether decreasing the remarkably rapid turnover rate of connexins (t_1/2 = 1.5-5 h) can serve as a mechanism to increase gap junction formation and function. We found that lysosomal inhibitors (CLQ and leupeptin) slowed the degradation of Cx43 in a cell type-specific manner but did not up-regulate gap junction assembly or dye coupling in any of the lines tested, as expected given that these compounds prevent proteolysis within, but not transport to, the endosome/lysosome system (37). In contrast, we found that inhibitors of the proteasome strongly decreased connexin degradation in all of the cell types examined. Most significantly, proteasome inhibition resulted in a striking up-regulation of Cx43 phosphorylation, gap junction formation, and intercellular dye coupling in previously assembly-inefficient cell types. These findings provide the first evidence that cell-cell communication can be up-regulated at the level of connexin turnover. Moreover, they distinguish connexins from several other types of integral plasma membrane proteins (including the cystic fibrosis transmembrane conductance regulator), the degradation of which is reduced by proteasome inhibitors without increasing the functional pool of the protein on the cell surface (42, 43).

Physiologically, degradation of cytosolic and nuclear proteins is down-regulated by preventing tagging events that serve to facilitate recognition of the substrate by the ubiquitination/proteasomal machinery (15, 17). For ornithine decarboxylase, as well as several other cytosolic regulatory molecules, an essential step in the targeting process is physical association with another protein. Polyamine-stimulated degradation of ornithine decarboxylase is blocked if synthesis of its tagging protein (antizyme) is prevented with CHX (18, 19). In an analogous process, CHX is thought to inhibit the proteasome-mediated degradation of iron regulatory protein 2 in iron-rich hepatoma cells by reducing the levels of an as yet unknown short-lived tagging protein (44). We show here that CHX and other general inhibitors of protein synthesis also reduce the degradation of wild-type Cx43 and Cx32 to an extent comparable to that obtained with direct inhibitors of the proteasome. Moreover, both classes of inhibitors up-regulate gap junction formation and function in assembly-inefficient (but not in assembly-incompetent) cell types. Proteasome inhibitors do not block protein synthesis and, conversely, protein synthesis inhibitors are not general perturbants of proteasomal function as evidenced by the lack of effect of CHX on the degradation of several well characterized proteasomal substrates (45-47). The similarity in the effect of inhibitors of protein synthesis and of the proteasome on connexins must therefore either be a mechanistically unrelated coincidence or reflect a specific requirement for ongoing protein synthesis in the proteasomal degradation of wild-type connexins. Our data support the second interpretation for the following reasons. First, the effects of proteasome inhibitors and CHX on gap junctions are not additive (data not shown), consistent with the possibility that they act on the same pathway rather than on two independent processes. Second, although protein synthesis inhibitors, including CHX, have been reported to interfere with lysosomal function by an unknown mechanism (48), such an activity cannot account for their effect on connexin turnover given that standard lysosomal inhibitors only very weakly reduced Cx43 degradation in CHO cells despite the sensitivity of this process to CHX. Third, we are not aware of any nonproteasomal, nonlysosomal proteolytic system that might degrade connexins that requires ongoing protein synthesis. Fourth, rendering connexins conformationally abnormal by either reduction of disulfide bonds or by mutation eliminates the effect of CHX (but not of proteasome inhibitors) on connexin turnover, in keeping with the known sensitivity of proteasomal targeting processes to substrate conformation. One possibility is that the fast-turnover protein binds directly to mature connexins and, analogous to antizyme in ornithine decarboxylase degradation, targets connexins for proteasome-mediated, but ubiquitin-independent, degradation. Given the role of ubiquitin in the turnover of many proteasome substrates (49), an alternative possibility not addressed in this study is that the short-lived protein is involved in polyubiquitination, either of connexins themselves or of proteins that (when modified) facilitate targeting to the proteasome (50). It is also conceivable that CHX reduces the amount of a protein involved in unfolding connexins for insertion into the catalytic chamber of the proteasome and that conformationally abnormal, unaggregated (13) connexins require less of this activity than mature, stably folded connexin species. This scenario would be in keeping with the finding that unfolding can be the rate-limiting step in the proteasomal degradation of mature proteins and that a major cytosolic chaperone implicated in this process (HSP70) is a fast-turnover protein (16). Implicit in either model is that the putative short-lived targeting protein can distinguish between different conformational states of the connexin molecule. Precedence for this type of interaction is provided by the proteasome-mediated degradation of the estrogen receptor, which is sensitive to CHX or puromycin when it is liganded to estradiol but not when it adopts a different conformation upon binding the antiestrogen RU 58668 (51).

How could inhibition of connexin degradation by the proteasome elicit the cell type-specific differences in gap junction formation that we have observed? NRK, S180L, CHO, and L929 cells all synthesize wild-type Cx43 of comparable metabolic stability that is assembled into connexons and transported to the cell surface. The difference in the number of gap junctions expressed by these cells under basal conditions therefore appears to reflect dissimilarities in their intrinsic ability to assemble connexons into gap junctional plaques. We have shown that serum-maintained S180L cells form gap junctions shortly after being replated at confluent density (Fig. 2), consistent with the observations of Rook et al. (52) that cultured cardiac myocytes establish functional gap junctional channels within 2-20 min of initiating cell-cell contact. We propose that the rate and efficiency with which cell surface Cx43 is incorporated into plaques in these and other assembly-efficient cells is such that almost all of the cell-cell contact area available for gap junctions becomes occupied under basal conditions. Increasing the total amount of connexin protein with proteasome inhibitors therefore induces only a minor enhancement in the number of gap junctional plaques, whereas blocking the production of new connexin molecules with protein synthesis inhibitors leads to a gradual decline in the number of junctions. The rate of gap junction loss in the latter case is, however, considerably slower than would be expected from the half-life of connexins in untreated cells because of the inhibitory effect of protein synthesis blockers on connexin turnover (see below). In assembly-inefficient cells, the probability that a given connexin molecule will be assembled into a plaque before it is degraded is relatively low. We propose that decreasing the rate of connexin turnover with either proteasome inhibitors or protein synthesis blockers leads to a net increase in the number of connexins available for assembly. Given the cooperative nature of insertion of functional intercellular channels into developing gap junctional plaques (53), even a modest elevation in cell surface connexin levels could result in a rapid increase in gap junction size. It is also conceivable that proteasome inhibitors and/or protein synthesis inhibitors exert additional, less direct effects on connexins in assembly-inefficient cells, such as activating signaling events that promote gap junction formation. Despite increased Cx43 stability, assembly-incompetent L929 cells did not up-regulate gap junctional plaques in response to either proteasome or protein synthesis inhibitors. Although the basis for this behavior is not known, it is possible that these cells are unable to form cell-cell contacts permissive for gap junction assembly under the conditions tested or are defective in some other aspect of junction establishment or maintenance. A similar phenotype has been described for the human colon tumor cell lines HT-29 and HCT-8, both of which remain assembly-incompetent even after exogenous overexpression of wild-type Cx43 (54).

Although pulse-chase analyses in a variety of systems have supported the concept that connexins turn over rapidly, studies utilizing protein synthesis inhibitors have lead to a very different view of connexin stability. For example, Flagg-Newton et al. (55) have reported that the number of gap junctional plaques detectable by freeze-fracture electron microscopy in Balb-3T3 fibroblasts after a 24-h treatment with 10 µg/ml CHX was ~25% of that in untreated controls, which would correspond to a t_1/2 of 12 h. Moreover, the gap junctions remaining were exceptionally large, leading to a less than 40% reduction in the gap junction area/µm² of total cell surface. Similarly, Epstein et al. (56) demonstrated that the capacity of Novikof hepatoma cells to assemble endogenously expressed Cx43 into functional gap junctions in a dissociation-reaggregation assay was only minimally affected by a 12-h pretreatment with 100 µg/ml CHX. Our demonstration that protein synthesis inhibitors decrease the rate of connexin degradation provides an explanation for these previously puzzling observations. Up-regulation of connexin stability and assembly is also likely to have contributed to the unexpectedly long persistence of gap junctional plaques in CHX-treated BICR-M1R_k cells reported by Laird et al. (32). Such effects must be taken into account in the interpretation of experiments conducted in the presence of agents that inhibit protein synthesis as either their primary activity or as a nonspecific side effect.

An important issue is the physiological significance of the ability of inhibitors of protein synthesis and of the proteasome to up-regulate gap junctions. Protein synthesis is greatly reduced in vivo by starvation, viral infection, or exposure to certain metabolic poisons, such as ricin and diphtheria toxin (57). Stabilization of connexins under such conditions could facilitate cell-to-cell transfer of metabolites and signaling molecules and might also serve to reduce apoptosis of intercellular contact-dependent cells. A more likely possibility is that inhibition of protein synthesis is an experimental means to reduce the activity of the putative fast-turnover mediator of connexin degradation and that the function of this protein is normally controlled at the posttranslational level. A major goal of future studies is to identify other effectors of connexin stability that regulate intercellular communication in vivo.

	ACKNOWLEDGEMENTS

We thank D. Goodenough (Harvard Medical School) for providing anti-Cx32 antibodies and M. Bogyo (Harvard Medical School) and D. Koop (Oregon Health Sciences University, Portland, OR) for proteasome inhibitors.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Vollum Institute for Advanced Biomedical Research L474, Oregon Health Sciences University, 3181 Southwest Sam Jackson Park Rd., Portland, OR 97201. Tel.: 503-494-1300; Fax: 503-494-8230; E-mail: Musill@ohsu.edu.

Published, JBC Papers in Press, June 2, 2000, DOI 10.1074/jbc.M002608200

	ABBREVIATIONS

The abbreviations used are: Cx43, connexin43; Cx32, connexin32; NRK, normal rat kidney; ssS180L, serum-starved S180L; ZL₃VS, carboxybenzyl-leucyl-leucyl-leucine vinylsulfone; ALLN, N-acetyl-leu-leu-norleucinal; DTT, dithiothreitol; LY, Lucifer Yellow; CLQ, chloroquine; CHX, cycloheximide; CHO, Chinese hamster ovary; DMEM, Dulbecco's modified Eagle's medium; FCS, fetal calf serum; PBS, phosphate-buffered saline; HB, Hanks' balanced salt solution containing 1% bovine serum albumin.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				X. Li, V. Su, W. E. Kurata, C. Jin, and A. F. Lau A Novel Connexin43-interacting Protein, CIP75, Which Belongs to the UbL-UBA Protein Family, Regulates the Turnover of Connexin43 J. Biol. Chem., February 29, 2008; 283(9): 5748 - 5759. [Abstract] [Full Text] [PDF]