Role of the Ubiquitin-Proteasome Pathway in Downregulation of the Gap-Junction Protein Connexin43 by TNF-α in Human Corneal Fibroblasts

Abstract

Purpose.: Fibroblasts in the corneal stroma communicate with each other through gap junctions and form a three-dimensional-network structure. The proinflammatory cytokine tumor necrosis factor-α (TNF-α) downregulates the gap-junction protein connexin43 (Cx43) and thereby inhibits gap-junctional intercellular communication (GJIC) in corneal fibroblasts. The authors examined the role of the ubiquitin-proteasome system in the TNF-α–induced degradation of Cx43 in these cells.

Methods.: Human corneal fibroblasts were cultured with TNF-α in the absence or presence of the proteasome inhibitor MG132. The expression of Cx43 was detected by immunofluorescence and immunoblot analyses. GJIC was monitored by observing the intercellular diffusion of the fluorescent dye Lucifer yellow. The ubiquitination of Cx43 was evaluated by immunoprecipitation and immunoblot analysis.

Results.: TNF-α induced a decrease both in the amount of Cx43 as detected by immunoblot analysis and in the extent of specific staining for this protein as revealed by immunofluorescence analysis in corneal fibroblasts. These effects of TNF-α were inhibited by MG132. MG132 also attenuated the TNF-α–induced inhibition of GJIC in these cells. In addition, TNF-α induced the ubiquitination of Cx43 in corneal fibroblasts.

Conclusions.: The ubiquitin-proteasome pathway contributes to the degradation of Cx43 and the inhibition of GJIC induced by TNF-α in corneal fibroblasts. The ubiquitin-proteasome system may thus play an important role in the disruption of corneal homeostasis associated with corneal inflammation.

The corneal stroma consists of corneal fibroblasts, or keratocytes, embedded in extracellular matrix. Corneal fibroblasts are connected to each other by gap junctions ¹ and form a three-dimensional-network structure. ² Gap junctions mediate the diffusion of ions and small molecules, including metabolites and second messengers such as cyclic AMP between neighboring cells, thereby allowing the synchronization of tissue differentiation and the function and maintenance of homeostasis. ³ Gap junctions are composed of two connexons, one in each of the opposing membranes and each containing six molecules of connexin. Connexins constitute a diverse family of proteins that are expressed in a cell type–specific manner, with connexin43 (Cx43) one of the most abundant members of this family. ⁴ Functional gap junctions have been detected in both the normal and the wounded rabbit corneal stroma and in the normal human corneal stroma. ⁵ Corneal fibroblasts and myofibroblasts have also been found to express Cx43 and to form functional gap junctions in culture. ^6–8  

Inflammation is a response of the body to infection or other insults. Both infiltrated cells and tissue-resident cells participate in inflammatory reactions, releasing cytokines, chemokines, and growth factors that affect neighboring cells by autocrine or paracrine mechanisms. Tumor necrosis factor-α (TNF-α) is a proinflammatory cytokine ⁹ that contributes to ocular inflammation associated with injury, infection, or allergy. ^10–13 We have previously shown that TNF-α downregulates the expression of Cx43 and inhibits gap-junctional intercellular communication (GJIC) in cultured human corneal fibroblasts. ⁸  

The ubiquitin-proteasome system mediates intracellular protein degradation associated with the regulation of cellular processes such as cell cycle, cell growth and differentiation, gene transcription, inflammation, and response to infection. ^14,15 The targeting of proteins for degradation is mediated by the sequential action of a ubiquitin-activating enzyme (E1), a ubiquitin-conjugating enzyme (E2), and a ubiquitin-protein ligase (E3), with the latter determining the substrate specificity of ubiquitination. Proteins conjugated to ubiquitin in this manner are then recognized and degraded by the 26S proteasome. Several families of E3 proteins, such as the Cbl family and the Nedd4 family, have been identified. ^16–18  

Inflammation of the corneal stroma is characterized by the infiltration of leukocytes and stromal edema and can result in a marked change in tissue structure. ¹⁹ To investigate the mechanism of corneal stromal inflammation, we examined the possible role of the ubiquitin-proteasome pathway in the TNF-α–induced downregulation of Cx43 expression and the inhibition of GJIC in cultured human corneal fibroblasts. 

Materials and Methods

Materials

Eagle's minimum essential medium (MEM) and fetal bovine serum were obtained from Invitrogen-Gibco (Carlsbad, CA). Bovine serum albumin (fraction V) and a protease inhibitor cocktail were from Sigma-Aldrich (St. Louis, MO). Recombinant human TNF-α was from R&D Systems (Minneapolis, MN). Mouse monoclonal antibodies to Cx43 were obtained from Chemicon (Temecula, CA), those to ubiquitin were from Santa Cruz Biotechnology (Santa Cruz, CA), and rabbit polyclonal antibodies to Nedd4 were from Millipore (Billerica, MA). Lucifer yellow CH (Li⁺ salt), rhodamine-phalloidin, TOTO-3, and Alexa Fluor 488–labeled goat antibodies to mouse immunoglobulin G were from Invitrogen (Carlsbad, CA). Horseradish peroxidase–conjugated goat antibodies to mouse or rabbit immunoglobulin G, enhanced chemiluminescence (ECL) detection reagents, and protein G–Sepharose beads were from Amersham Biosciences GE Healthcare (Little Chalfont, UK). Carboxybenzoyl-leucyl-leucyl-leucinal (MG132), epoxomicin, and AG490 were obtained from Merck (Darmstadt, Germany). Plastic culture dishes (60 mm) were from Corning (Corning, NY), and glass-bottom dishes (35 mm) were from Iwaki (Chiba, Japan). 

Isolation and Culture of Human Corneal Fibroblasts

Human corneas were obtained for corneal transplantation surgery from NorthWest Lions Eye Bank (Seattle, WA). Human tissue was used in strict accordance with the tenets of the Declaration of Helsinki. Corneal stromal fibroblasts were prepared and cultured as described previously. ⁸ In brief, the endothelial layer of the rim of the cornea remaining after transplantation surgery was removed mechanically, and the tissue was then incubated with dispase (2 mg/mL in MEM) for 1 hour at 37°C. After mechanical removal of the epithelial sheet, the tissue was treated with collagenase (2 mg/mL in MEM) at 37°C until a single-cell suspension of corneal fibroblasts was obtained. Isolated corneal fibroblasts were maintained under a humidified atmosphere of 5% CO₂ at 37°C in MEM supplemented with 10% fetal bovine serum. They were used for experiments after four to seven passages. All cells were positive for vimentin and negative for cytokeratin, suggesting the absence of contamination of the cultures by epithelial cells (data not shown). For experiments, the cells were seeded on plastic or glass-bottom dishes and cultured for 3 days in complete medium. They were then deprived of serum for 24 hours before exposure to MG132, epoxomicin, AG490, or TNF-α in unsupplemented MEM, as indicated. MG132, epoxomicin, and AG490 were dissolved in dimethyl sulfoxide, and cells were exposed to 0.1% dimethyl sulfoxide as a vehicle control. 

Immunofluorescence Microscopy

Cells (1 × 10⁵) cultured in glass-bottom dishes were fixed for 15 minutes with ice-cold acetone, washed with Ca²⁺- and Mg²⁺-free phosphate-buffered saline (PBS−), and incubated for 1 hour at room temperature first with 1% bovine serum albumin (BSA) in PBS− and then with monoclonal antibodies to Cx43 (1:200 dilution) in PBS− containing 1% BSA. The cells were washed with PBS− and then incubated for 1 hour at room temperature with Alexa Fluor 488–conjugated goat antibodies to mouse immunoglobulin G (1:1000 dilution), rhodamine-phalloidin (1:200 dilution), and TOTO-3 (1:1000 dilution) in PBS− containing 1% BSA. Finally, they were examined with a laser confocal microscope (LSM5; Carl Zeiss, Hallbergmoos, Germany). 

Dye Coupling

Cells (3 × 10⁵) cultured in glass-bottom dishes were monitored for GJIC with Lucifer yellow, as described previously. ^20,21 In brief, the cells were washed with PBS−, and Lucifer yellow (10% in water) was injected into widely separated cells with the use of a microinjector (Micromanipulator and Transinjector; Eppendorf, Hamburg, Germany). The cells were observed with a fluorescence inverted microscope (Axioscope; Carl Zeiss) and photographed 1 minute after dye injection. The number of cells containing the dye was counted. GJIC activity was expressed as the mean number of cells coupled to each injected cell. 

Immunoblot Analysis

Cells (3 × 10⁵) cultured in 60-mm dishes were lysed on ice in 0.5 mL solution containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 5 mM NaF, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM Na₃VO₄, and 1% protease inhibitor cocktail. The lysates were centrifuged at 15,000g for 15 minutes at 4°C, and the resultant supernatants were subjected to SDS-polyacrylamide gel electrophoresis on a 10% gel. The separated proteins were transferred to a nitrocellulose membrane, which was then exposed to 5% skim milk for 1 hour at room temperature before incubation for 1 hour at room temperature with primary antibodies at a dilution of 1:1000 in washing buffer (20 mM Tris-HCl [pH 7.4], 5% skim milk, 0.1% Tween 20). The membrane was washed in washing buffer, incubated for 1 hour at room temperature with horseradish peroxidase–conjugated secondary antibodies (1:1000 dilution in washing buffer), washed again, incubated with ECL detection reagents for 5 minutes, and exposed to film. 

Immunoprecipitation

Cells (3 × 10⁵) cultured in 60-mm culture dishes were lysed on ice in 0.3 mL of a solution containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 5 mM NaF, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM Na₃VO_4, and 1% protease inhibitor cocktail. The lysates were centrifuged at 15,000g for 10 minutes at 4°C, and portions of each supernatant (100 μg protein) were incubated for 16 hours at 4°C in a final volume of 200 μL with antibodies to Cx43 (1:100 dilution) and 20 μL protein G–Sepharose beads. The beads were isolated by centrifugation and washed twice with cell lysis solution, and the bound proteins were subjected to immunoblot analysis, as described. 

Statistical Analysis

Quantitative data are presented as mean ± SE and were analyzed with Dunnett's test. P < 0.05 was considered statistically significant. 

Results

We first examined the effect of TNF-α on the distribution of Cx43 in cultured human corneal fibroblasts. The cells were deprived of serum for 24 hours, exposed to TNF-α (1 ng/mL) for 24 hours, fixed, and stained with antibodies to Cx43 and phalloidin (for the detection of F-actin). Fluorescence microscopy revealed a dotlike pattern of Cx43 immunoreactivity that did not colocalize with F-actin staining in control cells. In contrast, the number of dots corresponding to Cx43 immunoreactivity was greatly reduced in cells treated with TNF-α (Fig. 1A). Immunoblot analysis of corneal fibroblasts exposed to various concentrations (0–10 ng/mL) of TNF-α for 24 hours revealed that the cytokine induced a concentration-dependent decrease in the amount of Cx43, with the maximal effect apparent at a TNF-α concentration of 1 ng/mL (Fig. 1B). 

Figure 1.

 

TNF-α–induced downregulation of Cx43 in cultured human corneal fibroblasts. (A) Cells deprived of serum for 24 hours were incubated in the absence (Control) or presence of TNF-α (1 ng/mL) for 24 hours. They were then fixed and subjected to immunofluorescence staining with antibodies to Cx43 (green). Cells were also stained with rhodamine-phalloidin to detect F-actin (red) and with TOTO-3 to reveal nuclei (blue). Scale bar, 10 μm. (B) Serum-deprived cells were incubated with the indicated concentrations of TNF-α for 24 hours, lysed, and subjected to immunoblot analysis with antibodies to Cx43. All data are representative of three independent experiments.

We next examined whether the ubiquitin-proteasome pathway contributed to the TNF-α–induced decrease in Cx43 expression in human corneal fibroblasts. The cells were incubated for 1 hour with the proteasome inhibitor MG132 (10 μM) or epoxomicin (1 μM) or with AG490 (25 μM), an inhibitor of the JAK (Janus kinase)–STAT (signal transducer and activator of transcription) signaling pathway activated by TNF-α. They were then incubated for 24 hours in the additional absence or presence of TNF-α (1 ng/mL). Immunoblot analysis revealed that MG132 and epoxomicin each inhibited the TNF-α–induced downregulation of Cx43 (Fig. 2). In contrast, AG490 had no such effect. This inhibitory action of MG132 was concentration dependent, with the maximal effect at 24 hours apparent with an MG132 concentration of 1 μM (Fig. 3A). The TNF-α–induced decrease in the amount of Cx43 was time dependent, with the effect first apparent at 12 hours and maximal at 24 hours; the effect was not observed at either time point in the presence of 10 μM MG132 (Fig. 3B). Immunofluorescence analysis also revealed that MG132 (10 μM) inhibited the effect of TNF-α (1 ng/mL) on the pattern of Cx43 immunoreactivity in human corneal fibroblasts (Fig. 4). 

Figure 2.

 

Attenuation of the TNF-α–induced decrease in the expression of Cx43 in cultured human corneal fibroblasts by proteasome inhibitors. Serum-deprived cells were incubated first for 1 hour in the absence or presence of MG132 (MG, 10 μM), epoxomicin (Epox, 1 μM), or AG490 (AG, 25 μM) and for an additional 24 hours in the absence or presence of TNF-α (1 ng/mL). Cells were then lysed and subjected to immunoblot analysis with antibodies to Cx43 and to actin (loading control). Data are representative of three independent experiments.

Figure 3.

 

Concentration and time dependence of the inhibitory effect of MG132 on the TNF-α–induced downregulation of Cx43 in cultured human corneal fibroblasts. (A) Serum-deprived cells were incubated for 1 hour with the indicated concentrations of MG132 and then for an additional 24 hours in the absence or presence of TNF-α (1 ng/mL). Cells were then lysed and subjected to immunoblot analysis with antibodies to Cx43 and to actin. (B) Serum-deprived cells were incubated first for 1 hour in the absence or presence of MG132 (10 μM) and then for the indicated times in the additional presence of TNF-α (1 ng/mL). They were then lysed and analyzed as in (A). All data are representative of three independent experiments.

Figure 4.

 

Inhibition by MG132 of the effect of TNF-α on the pattern of Cx43 immunostaining in cultured human corneal fibroblasts. Serum-deprived cells were incubated for 1 hour in the absence or presence of MG132 (10 μM) and for an additional 24 hours in the absence or presence of TNF-α (1 ng/mL). Cells were then fixed and stained as in Figure 1. Scale bar, 10 μm. Data are representative of three independent experiments.

We examined the effect of MG132 on GJIC in human corneal fibroblasts exposed to TNF-α by measuring dye coupling. Cells were incubated with or without TNF-α (1 ng/mL) or MG132 (10 μM) for 6 hours, after which Lucifer yellow was injected into widely separated cells. The number of neighboring cells to which the dye had been transferred was determined 1 minute after dye injection. The number of dye-positive neighboring cells was 7.7 ± 0.9 for each injected control cell, confirming the existence of functional gap junctions, but it was only 1.0 ± 0.0 for cells exposed to TNF-α (Fig. 5). This inhibition of GJIC by TNF-α was attenuated by MG132 (4.7 ± 0.9 cells). 

Figure 5.

 

Attenuation by MG132 of the TNF-α–induced inhibition of GJIC in cultured human corneal fibroblasts. Serum-deprived cells were incubated in the absence or presence of TNF-α (1 ng/mL) and MG132 (10 μM) for 6 hours, after which individual cells were microinjected with Lucifer yellow. The dye was allowed to spread for 1 minute, after which the number of coupled cells was determined. Data are mean ± SE of values obtained from 10 injected cells in a representative experiment. *P < 0.05 (Dunnett's test). Data are representative of three independent experiments.

Finally, we investigated the effect of TNF-α on ubiquitination of Cx43 in human corneal fibroblasts. Cells incubated with TNF-α (1 ng/mL) for various times were subjected to immunoprecipitation with antibodies to Cx43, and the resultant precipitates were subjected to immunoblot analysis with antibodies to ubiquitin, to Nedd4, and to Cx43 (Fig. 6). Equivalent amounts of connexin43 were present in such immunoprecipitates from cells incubated with TNF-α (1 ng/mL) for various times. The ubiquitination of Cx43 was increased after exposure of the cells to TNF-α for 5 minutes. Nedd4 was detected in the Cx43 immunoprecipitates, but the amount of this E3 ubiquitin ligase in the precipitates was not affected by exposure of the cells to TNF-α. 

Figure 6.

 

Effect of TNF-α on the ubiquitination of Cx43 and its association with Nedd4 in cultured human corneal fibroblasts. Serum-deprived cells were incubated with TNF-α (1 ng/mL) for the indicated times, lysed, and subjected to immunoprecipitation (IP) with antibodies to Cx43. The resultant precipitates were then subjected to immunoblot analysis (IB) with antibodies to ubiquitin (Ub), to Nedd4, and to Cx43. The lysates before immunoprecipitation (Input) and the immunoprecipitate supernatants (Sup) were then subjected to immunoblot analysis with antibodies to Cx43. Data are representative of three independent experiments.

Discussion

We have shown that the proteasome inhibitor MG132 blocked the TNF-α–induced downregulation of Cx43 in human corneal fibroblasts in a concentration- and time-dependent manner. Moreover, immunofluorescence analysis revealed that MG132 largely prevented the TNF-α–induced disappearance of Cx43 immunoreactivity in these cells. Consistent with these findings, the TNF-α–induced inhibition of GJIC was also attenuated by MG132. Immunoprecipitation and immunoblot analysis revealed that Cx43 was constitutively associated with the E3 ubiquitin ligase Nedd4 and that TNF-α induced the ubiquitination of Cx43 in human corneal fibroblasts. These results suggest that TNF-α disrupts gap junctions in human corneal fibroblasts in a manner dependent on the degradation of Cx43 by the ubiquitin-proteasome pathway. 

Our finding that the TNF-α–induced degradation of Cx43 in human corneal fibroblasts is mediated by the ubiquitin-proteasome pathway is consistent with previous observations in lens epithelial cells, liver epithelial cells, and cardiomyocytes. ^22–24 MG132 is a cell-permeable peptide that effectively blocks the proteolytic activity of the 26S proteasome. ²⁵ It has been used as a selective proteasome inhibitor to clarify the roles of the ubiquitin-proteasome pathway in various cellular processes. ^26–28 Our finding that MG132 blocked the decrease in GJIC activity induced by TNF-α in human corneal fibroblasts is also consistent with previous results showing that inhibition of the ubiquitin-proteasome pathway increased the level of GJIC in lens epithelial cells and sarcoma cells. ^29,30  

TNF-α is a proinflammatory cytokine that has been detected in the corneas of patients with keratitis and in the tear fluid of those with ocular allergy. ^31–33 Moreover, TNF-α is implicated in the pathogenesis of autoimmune uveitis. ^34,35 Gap junctions play an important role in maintaining both tissue structure and function. ^36,37 We previously showed that keratocytes interact with each other through gap junctions and form a three-dimensional-network structure. ^1,2 Inflammation of the corneal stroma is associated with the infiltration of leukocytes into the stroma and changes in stromal structure. ¹⁹ These observations suggest that the degradation of Cx43 in response to inflammation in the cornea may affect not only GJIC among stromal fibroblasts but also the architecture of the stroma. 

The TNF-α–induced downregulation of Cx43 in human corneal fibroblasts was inhibited by the proteasome inhibitors MG132 and epoxomicin. Moreover, TNF-α induced the ubiquitination of Cx43, and Cx43 was associated with Nedd4 in these cells. Nedd4 has been shown to bind to Cx43 in a manner dependent on the phosphorylation state of Cx43. ³⁸ Most connexin family proteins turn over with a half-life between 1 and 5 hours. ^39,40 TNF-α may thus accelerate the turnover of Cx43 by the ubiquitin-proteasome pathway in human corneal fibroblasts during ocular inflammation. We have previously shown that TNF-α induces the release of matrix metalloproteinases, including MMP-2 and MMP-9, from human corneal fibroblasts ⁴¹ ; these enzymes are responsible for the degradation of collagen in the corneal stroma. TNF-α thus activates both intracellular and extracellular pathways of protein degradation in the corneal stroma, effects that likely contribute to the promotion of inflammation in this tissue. Further elucidation of the mechanism for regulation of GJIC through Cx43 degradation in the corneal stroma may provide the basis for the development of new treatments to prevent or ameliorate corneal inflammation associated with infection, allergy, and injury. 

Footnotes

 Supported in part by Grant 19791271 from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

Footnotes

 Disclosure: K. Kimura, None; T. Nishida, None

The authors thank Yasumiko Akamatsu and the staff of the Yamaguchi University Center for Gene Research for technical assistance.