High Glucose Alters Connexin 43 Expression and Gap Junction Intercellular Communication Activity in Retinal Pericytes

Abstract

purpose. To investigate the role of the gap junction protein, connexin-43 (Cx43) in the maintenance of retinal vascular homeostasis in diabetic retinopathy.

methods. In human retinal pericytes (HRPs) and bovine retinal pericytes (BRPs) grown for 7 days in normal (5 mM) or high (30 mM)-glucose medium, the Cx43 protein level was determined by Western blot analysis. Parallel experiments were performed in HRPs to determine the Cx43 mRNA level by RT-PCR, the distribution and localization of Cx43 protein by immunostaining, and gap junction intercellular communication (GJIC) activity by a scrape-loading dye transfer technique. Distribution and localization of Cx43 protein was also determined in pericyte–endothelial cell cocultures.

results. Western blot analysis of the Cx43 protein level in HRPs and BRPs indicated reduced Cx43 expression in the high-glucose condition (69.1% ± 17% of control, P = 0.004; 62.3% ± 19% of control, P = 0.001, respectively). The Cx43 mRNA level in HRPs grown in high-glucose medium also showed significant reduction (71.4% ± 16.8% of control, P = 0.02). The relative number of Cx43 plaques indicative of Cx43 localization at specific sites of contact between adjacent cells showed significant reduction in the high-glucose condition (61% ± 10% of control, P = 0.002); similarly, a significant reduction in the number of plaques was observed in cocultures grown in high-glucose medium compared with those in normal medium (59.4% ± 29% of control, P = 0.001). Cells with reduced Cx43 expression showed significantly reduced transfer of lucifer yellow (61% ± 13% of control, P = 0.001; r = 0.9).

conclusions. High-glucose–induced downregulation of Cx43 expression and inhibition of GJIC in retinal pericytes may play a role in the disruption of vascular homeostasis in diabetic retinopathy.

Gap junction intercellular communication (GJIC) or cell–cell coupling plays a critical role in the maintenance of vascular homeostasis. Gap junctions are aggregates of intercellular channels that connect the cytoplasms of adjoining cells and allow the passage of small molecules between adjacent cells. The presence of gap junctions between endothelial cells and pericytes have been detected in vitro. ¹ Although GJIC has been extensively studied in the endothelial cells, ^{2 3 4 5 6} its involvement in pericytes remains unclear. Studies indicate that Cx43 proteins, the major components of gap junctions, are required for maintenance of GJIC and normal vascular function, ⁷ including cellular interactions during growth and development in a variety of systems ^{8 9} thought to be conduits for growth-modulating signals. 

The retinal capillaries have two vascular cell types: endothelial cells and pericytes. Our recent studies have shown that, in endothelial cells, high glucose may disrupt vascular homeostasis by downregulating expression of Cx43 and reducing GJIC. Currently, it is unknown whether pericytes undergo similar disruption before eventually disappearing and turning into pericyte ghosts. ^{10 11 12 13 14} Degeneration and loss of retinal capillary pericytes and selective disruption of cell–cell contacts in the retinal vasculature are characteristic of diabetic retinopathy. ^{15 16 17 18 19 20}  

Hyperglycemia is known to play a critical role in the development and progression of diabetic retinopathy. ^{10 21 22 23 24} Studies have shown that retinal pericytes undergo several molecular and biochemical changes under high-glucose conditions eventually undergoing apoptosis and accelerated cell death. ^{24 25 26 27} In the retinal microvasculature, pericytes and endothelial cells share a common vascular basement membrane. The pericyte cell bodies and their long cytoplasmic processes cover the abluminal surface of the retinal vasculature as documented in most profiles of EM sections of retinal capillaries. Each pericyte possesses a prominent nucleus in its cell body with surrounding cytoplasm. ²⁸ Protruding from the cell body are extensions of cytoplasmic processes that parallel the long axis of the capillary and often overlap, forming one or two layers ²⁸ and making frequent close contacts between adjacent pericyte processes. ²⁹  

To our knowledge there have been no reports regarding the effect of hyperglycemia or diabetes on Cx43 expression and its effect on GJIC in retinal pericytes. In this study we examined the effect of high glucose on Cx43 expression and GJIC in retinal pericytes and its possible role in diabetic retinopathy. 

Methods

Cell Culture

BRPs were isolated and cultured after Mandarino’s method with modification. ³⁰ Briefly, bovine retinas were dissected and washed with PBS followed by DMEM to eliminate residual retinal pigment epithelium. Retinas were individually cut into small pieces and homogenized with one stroke of a handheld homogenizer, and the homogenate was digested with 0.1% collagenase type 2 (Worthington Biochemical Corp., Lakewood, NJ) for 1 hour at 37°C. The digested material was pelleted by centrifugation, washed ×2 in DMEM, resuspended in complete medium (DMEM containing 5.5 mM glucose and supplemented with 10% fetal calf serum, antibiotics, and antimycotics), and plated in 60-mm culture dishes. On reaching subconfluency, pericytes were trypsinized and seeded in 35-mm culture dishes in normal (5.5 mM) or high (30 mM)-glucose medium. All experiments were performed with passage-1 through -3 cells. For osmotic control, pericytes were grown in normal medium containing 25 mM mannitol. The BRPs were identified as pericytes on the basis of several criteria: (1) irregular, stellate morphology; (2) lack of vWF antigen; (3) presence of desmin and vimentin ³¹ ; (4) presence of smooth muscle cell α-actin ³² ; and (5) presence of a surface ganglioside identified by monoclonal antibody 3G5 (gift of Ramesh C. Nayak, New England Medical Center, Boston, MA). ³³ Cultures in first and subsequent passages were routinely found to be free of contaminants, and therefore the purity of the cultures in subsequent passages was assessed on the basis of cell morphology. HRPs (Clonetics, San Diego, CA) were characterized in a manner similar to the BRPs before use in the study. The bovine retinal endothelial cells (BRECs), a gift from Thomas W. Gardner (Penn State University College of Medicine, Hershey, PA), used for cocultures, stained positive for von Willebrand factor (vWF). Cocultures of BRECs and BRPs were grown by plating endothelial cells and pericytes at a ratio of 30:1. The cell suspension was plated onto coverslips, and the cocultures were allowed to grow in either normal or high-glucose medium for 7 days. The coverslips were then subjected to Cx43 immunostaining for identifying Cx43 plaques between endothelial cells and pericytes. 

Western Blot Analysis

Cells in normal or high-glucose medium were washed with PBS and lysed with buffer containing 10 mM Tris (pH 7.4; Sigma-Aldrich, St. Louis, MO), 1 mM EDTA, and 0.1% Triton X-100 (Sigma-Aldrich). Protein content in cell lysate was determined by the bicinchoninic acid (BCA) protein assay method (Pierce, Rockford, IL). Fifteen micrograms of protein was electrophoresed on 8% SDS-PAGE. The samples were then transferred onto nitrocellulose membranes according to the procedure of Towbin et al. ³⁴ The membranes were blocked with 5% nonfat dry milk in Tris-buffered saline (TBS) for 2 hours and incubated with mouse anti-rat Cx43 antibody (Chemicon, Temecula, CA) solution (1:500) in 0.2% nonfat dry milk overnight at 4°C. After three washes in TBS containing 0.1% Tween-20, the membranes were incubated with goat anti-mouse IgG conjugated with alkaline phosphatase (Sigma-Aldrich) for 1 hour. The membranes were washed as just described, a chemiluminescence substrate was applied (Immun-Star; BioRad, Richmond, CA), and the membranes were exposed to x-ray film (Fuji, New York, NY). Similar procedures were followed for the detection of Cx37 and Cx40 except that rabbit anti-mouse antiserum for Cx37, Cx40 (Alpha Diagnostic International, San Antonio, TX), and goat anti-rabbit IgG were used. Densitometric signals from Western blots were analyzed with NIH Image software (available by ftp at zippy.nimh.nih.gov/ or at http://rsb.info.nih.gov/nih-image; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD). Western blot analysis for determining the β-actin protein level was performed after stripping membranes previously probed with connexin antibodies. After an extensive wash in TBS containing 0.1% Tween-20, followed by a wash in 62.5 mM Tris (pH 6.8), 2% SDS, and 100 mM 2-mercaptoethanol (Sigma-Aldrich) for 30 minutes at 70°C, the membranes were blocked with 5% nonfat dry milk in TBS for 2 hours and incubated with mouse anti-rat β-actin antibody (Abcam, Cambridge, UK) solution (1:2500) in 0.2% nonfat dry milk overnight at 4°C. After three washes in TBS containing 0.1% Tween-20, the membranes were incubated with goat anti-mouse IgG (1:10,000) conjugated with alkaline phosphatase (Sigma-Aldrich) for 1 hour. After three more washes, the chemiluminescence substrate was applied to the membranes (Immun-Star; Bio-Rad) and the membranes exposed to x-ray film. Densitometry was performed as described earlier. 

Immunofluorescence Microscopy

To study the localization and distribution of Cx43 in retinal pericytes, immunofluorescence staining for Cx43 was performed in HRPs and BRPs grown in normal or high-glucose medium. Briefly, cells were plated on glass coverslips and fixed in ice-cold acetone for 15 minutes at −20°C. After PBS washes, the coverslips were blocked with 2% bovine serum albumin in PBS for 15 minutes. Then they were incubated with mouse anti-rat Cx43 antibody (Chemicon) solution (1:100) overnight at 4°C in a moist chamber. After three washes in PBS, they were incubated with goat anti-mouse IgG conjugated with fluorescein isothiocyanate (FITC) or rhodamine red (Jackson ImmunoResearch Laboratories, West Grove, PA) for detecting Cx43 protein in HRPs or BRPs, respectively. The coverslips were mounted in antifade medium (Slow-Fade; Molecular Probes, Eugene, OR) after washes in PBS. Negative control samples were processed in the same procedure, except that the primary antibody was omitted. The cells were viewed and photographed with a fluorescence microscope (Diaphot; Nikon, Tokyo, Japan, with Spot software; Diagnostic Instruments, Sterling Heights, MI). Cx43 appearing as fluorescent dots on cell processes of HRPs and BRPs was quantified in random areas. Similarly, immunofluorescence microscopy was performed to identify Cx43 in cocultures, together with rabbit anti-human vWF antibody (Sigma-Aldrich) to identify endothelial cells. FITC-conjugated goat anti-rabbit IgG and rhodamine-red–conjugated goat anti-mouse IgG (1:100; Jackson ImmunoResearch Laboratories) were used as secondary antibodies. Cx43 plaque numbers were obtained by counting plaques located on adjacent cell bodies and processes of BRECs and BRPs. 

Reverse Transcription–Polymerase Chain Reaction

Total RNA was isolated from cultured HRPs in normal or high-glucose medium with a total RNA isolation system (RNAgent; Promega, Madison, WI) according to the instructions of the manufacturer. RT reactions were performed in a 20-μL volume with 1 μg RNA at 42°C. At the end of RT, samples were treated with RNase-H (1 U) for 15 minutes at 37°C. PCR was designed to measure Cx43 mRNA level relative to the internal standard β-actin gene. The Cx43 and β-actin cDNA generated in the same RT reaction were amplified in separate tubes containing increasing volumes of the RT products (0.5, 1, and 2 μL) to document amplification in the linear phase of the reaction. The specificity of the PCR was enhanced by using the hot-start approach. ³⁵ PCR was performed in a volume of 50 μL with aliquot of RT product, 0.2 μM primers, Taq polymerase 2.5 U, MgCl₂ (1.9 mM) and PCR buffer, using the following cycle conditions: denaturation at 95°C for 1 minute, annealing at 55°C for 1 minute and extension at 72°C for 2 minutes for 29 cycles. For β-actin PCR, the cycle conditions were exactly similar to Cx43 except that the annealing temperature was 53°C, and the number of cycles was 21. The PCR primers for human Cx43 are 5′-TCAATGTGGACATGC-3′ and 5′-ATGATATTCAAGGCCA-3′; PCR primers for human β-actin are 5′-CAGATCATGTTTGAGAC-3′ and 5′-GTCACACTTCATGATG-3′. 

PCR products were resolved on the same gel (1% agarose) containing 0.05 μL/mL nucleic acid stain (Gelstar; BMA, Lowell, MA). The gels were photographed with positive/negative instant film (667; Polaroid, Cambridge, MA). Quantitative analysis of the PCR products was performed using NIH Image software. Densitometric units per microliter of RT products were averaged from the increasing volume of RT products (0.5, 1, and 2 μL) for both Cx43 and β-actin. Cx43 mRNA level was converted into the percentage of β-actin mRNA level in each sample. 

Assessment of GJIC

The scrape-loading dye transfer technique (SLDT) was used to assess GJIC. ^{2 36} Measurement of GJIC can be achieved by performing SLDT, a direct and rapid technique that uses scrape-loading to introduce lucifer yellow, a low molecular mass (457) dye into cells in culture followed by monitoring its transfer into contiguous cells. ^{2 36} Briefly, the confluent HRPs on coverslips were washed three times with PBS containing 0.01% Ca²⁺ and Mg²⁺. An aliquot of 1.5 mL of PBS containing 0.1% lucifer yellow CH (Molecular Probes) was added to cover the coverslips and several razor cuts were made in the monolayer on each coverslip. The cells were incubated for 10 minutes at 37°C and then rinsed three times with PBS containing 0.01% Ca²⁺ and Mg²⁺ to remove background fluorescence. The cells were fixed with 4% paraformaldehyde in PBS and photographed with a fluorescence microscope (Diaphot; Nikon). The dye-coupled cell layers on either sides of the scrape were counted to evaluate the GJIC. 

Statistical Analysis

All data are expressed as mean ± SD. Comparisons between groups were performed with Student’s t-test. A value of P < 0.05 was considered statistically significant. 

Results

Effect of High Glucose on Cx37, Cx40, and Cx43 Protein Levels in Retinal Pericytes

Human retinal pericytes and BRPs grown in high-glucose medium for 7 days exhibited a decreased Cx43 protein level compared with cells grown in normal medium. Connexin 43 protein expression was significantly reduced in both HRPs (69.1% ± 17% of control; P = 0.004, n = 7; Fig. 1 ) and BRPs (62.3% ± 19% of control; P = 0.001, n = 7; Fig. 2 ). The high-glucose–induced decrease in Cx43 expression was not due to fewer pericytes, because the actin level in the Western blot analysis was similar in both groups. No significant changes in the Cx37 (98% of control in HRPs, 96% of control in BRPs grown in high-glucose medium) or Cx40 (103% of control in HRPs, 87% of control in BRPs grown in high-glucose medium) protein levels was observed in either HRPs or BRPs grown in high-glucose medium (Figs. 1 2) . BRPs grown in normal medium supplemented with 25 mM mannitol, as an osmotic control, showed no change in the Cx43 protein level compared with cells grown in normal medium. 

Effect of High Glucose on Cx43 Distribution and Localization

Immunofluorescence microscopy was used to assess distribution and relative quantity of Cx43 protein. Cx43 immunostaining appeared as plaques on the cytoplasmic processes of HRPs (Fig. 3A) . A semiquantitative analysis of the number of fluorescent plaques in random but fixed areas showed HRPs grown in high-glucose medium had fewer plaques than those grown in normal medium (61% ± 10% of control, P = 0.002, n = 3; Fig. 3B ). In BRPs, the number of Cx43 plaques in cells grown in high-glucose medium was also decreased compared with those in cells in normal medium (51.6% ± 12% of control, n = 3); indicating reduced Cx43 protein on the cell membranes of retinal pericytes grown in high-glucose medium (Fig. 4) . In coculture, the endothelial cells were distinguishable from pericytes by vWF positive staining (Figs. 5A 5B 5D 5E) . The number of Cx43 plaques in the cocultures between BRECs and BRPs in high-glucose medium was significantly reduced compared with the cells grown in normal medium (59.4% ± 29% of control, n = 3; Figs. 5C 5F ). 

Effect of High Glucose on Cx43 mRNA Levels

Relative levels of Cx43 mRNA determined by RT-PCR were significantly downregulated in HRPs grown in high-glucose medium compared with cells grown in normal medium (71.4% ± 16.8% of control; P = 0.02, n = 6; Fig. 6 ). The β-actin mRNA level used as internal control was not affected by high glucose. The band intensity of the PCR products increased linearly with the increasing amount of RT reaction (0.5, 1, and 2 μL) for both Cx43 and β-actin, indicating that the comparison was made within the exponential phase of the PCR reaction (Fig. 6A) . 

Effect of High Glucose on GJIC in HRP

GJIC was determined by the SLDT technique, which measured the ability of HRPs to transfer lucifer yellow through gap junctions. A reduced number of dye-coupled cell layers was observed on either side of the scrape in HRPs grown in high-glucose medium compared with cells grown in normal medium. (3.8 ± 1.0 vs. 2.3 ± 0.4; P = 0.001, n = 3; Fig. 7 ). 

Discussion

In this study, we evaluated the effect of high glucose on Cx37, Cx40, and Cx43 expression in HRPs and BRPs. Cx43 expression was significantly reduced in both pericyte strains grown in high-glucose medium compared with those grown in normal medium. The expression of Cx37 and Cx40 showed no change in both HRPs and BRPs grown in high-glucose medium. Immunofluorescence microscopy revealed a reduction in the number of Cx43 plaques in pericytes grown in high-glucose medium compared with cells grown in normal medium. A similar reduction in Cx43 expression was observed in pericyte-endothelial cell cocultures grown in high-glucose medium. In addition, we observed that the high-glucose condition impairs not only connexin expression in retinal pericytes, but also gap junction activity. These findings are strikingly similar to our previous observation in microvascular endothelial cells grown in high-glucose medium showing reduced Cx43 expression and GJIC. ² Our observation is consistent with other studies reporting inhibition of gap junction activity by high glucose in aortic smooth muscle cells and aortic endothelial cells. ³⁷ Taken together, these results indicate that a high-glucose condition downregulates Cx43 expression and inhibits GJIC in retinal vascular cells. 

Our finding that high glucose reduces GJIC in retinal pericytes is in agreement with the findings in a study performed in retinal microvessels of diabetic rats. ³⁸ In that study, a marked reduction in cell-to-cell coupling was observed in pericyte-containing retinal microvessels soon after diabetes was induced. Currently, it is unknown why vascular cells downregulate Cx43 expression in a high-glucose condition. Because pericytes and endothelial cells communicate through gap junctions, ¹ express Cx43, ^{2 37 39} and have a high frequency of junctional contacts, ^{29 40 41 42} it is likely that the high-glucose–induced reduction of Cx43 expression and impairment of GJIC contributes to breakdown and disruption of vascular homeostasis in diabetic retinopathy. Reduced GJIC can impair diffusible transport of small molecules, such as calcium ions, potassium ions, and sodium ions, necessary for the upkeep and proper maintenance of cell proliferation, cell differentiation, and cellular homeostasis. Studies have shown that reduced Cx43 expression accelerates the progression of apoptosis, ⁴³ a mechanism associated with vascular endothelial cell and pericyte loss in diabetic retinopathy. ^{26 27 44 45}  

Despite progress in understanding the structure–function relationship of pericytes, a link between dysfunction of pericytes and the development of early vascular lesions in diabetic retinopathy remains undefined. Direct cell–cell interactions through GJIC appear to be particularly important in the maintenance of vascular homeostasis in retinal capillaries. Studies have determined that direct cell–cell interactions through gap junctional communication is influenced by cytokines such as TNF-α, which is increased in diabetes ⁴⁶ and can induce translocation and activation of nuclear factor (NF)-κB. ⁴⁷ A recent study has shown that activation of NF-κB, which occurs selectively in retinal pericytes in response to hyperglycemia and diabetes, may trigger apoptosis and result in loss of pericytes. ²⁴ It is thus possible that one of the early steps toward commencement of vascular cell apoptosis in a hyperglycemic environment may be inhibition of Cx43 expression and GJIC. 

In a high-glucose condition or diabetes, pericytes exhibit altered biochemical and molecular changes. Pericytes upregulate synthesis of basement membrane components, ^{30 48} reduce structural support of vascular walls, ^{49 50 51} and inhibit contractile action that can modulate blood flow. ^{52 53} High glucose may inhibit pericyte contractility ⁵² by regulating Na⁺/K⁺-adenosine triphosphatase (ATPase) activity. Investigators have shown that vascular smooth muscle cells (VSMCs) grown in high glucose exhibit decreased Na⁺ and K⁺ transport and alter membrane permeability to cations, possibly leading to changes in contractility and proliferation. ⁵⁴ Regulation of Na⁺ and K⁺ levels is to a large extent controlled by GJIC. For example, gap junctions between glial cells allow intercellular exchange of Na⁺ and equalize intracellular concentrations of Na⁺. ⁵⁵ The association between GJIC and Na⁺/K⁺-ATPase activity has been reported in other cell types. ⁵⁶ Pericyte contractility has been shown to be dependent on the presence of adenosine triphosphate (ATP), ^{57 58} which is necessary for ATPase activity. Recent studies indicate that not only does Na⁺ and K⁺ ions pass through gap junctions but also the transfer of ATP between gap junction channels. Previous studies have shown that labeled nucleotides transfer through gap junctions of pericytes ¹ and that ATP transfer to adjacent cells is 300 times greater through channels formed by Cx43 than through channels formed by other connexins. ⁵⁹ Moreover, studies have indicated altered levels of retinal Na⁺, K⁺-ATPase in streptozotocin-induced diabetic rats, ⁶⁰ which may be associated with altered GJIC in diabetes. Overall, high glucose-induced reduction in GJIC may have a profound effect on pericyte contractility and other physiologic functions. Further studies are needed, to understand the role of connexin expression in the maintenance of retinal vascular homeostasis. A challenge for future studies is to explore the pathophysiological consequences of defective cell–cell communication in diabetic retinopathy. 

 Supported by grants from the American Diabetes Association, and Grant EY11990-01A1 from the National Eye Institute; and in part by departmental grants from the Massachusetts Lions Eye Research Foundation, Inc., and Research to Prevent Blindness, Inc.

 Submitted for publication April 8, 2003; revised July 5, 2003; accepted August 4, 2003.

 Disclosure: A. Li, None; T. Sato, None; R. Haimovici, None; T. Okamoto, None; S. Roy, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

(A) Western blot analysis of Cx37, Cx40, Cx43, and β-actin protein levels in HRPs grown in normal (N) or high-glucose (HG) medium. (B) Histogram shows that the Cx43 protein level was reduced in cells grown in high-glucose medium compared with those grown in normal medium, whereas, Cx37 and Cx40 protein levels were not altered by high glucose. Data are expressed as the mean ± SD; *P < 0.05; n = 7.

Figure 2.

 

(A) Western blot analysis of Cx37, Cx40, and Cx43 protein levels in BRPs grown in normal (N) or high-glucose (HG) medium. (B) Histogram shows that the Cx43 protein level was significantly reduced in cells grown in high-glucose medium compared with cells grown in normal medium, whereas Cx37 and Cx40 protein levels were slightly reduced by high glucose. Data are expressed as the mean ± SD; *P < 0.05; n = 7.

Figure 3.

 

Effect of high glucose on Cx43 immunoreactivity in HRPs. Representative photomicrographs of (A) cells grown in normal (N) or high-glucose (HG) medium. Note the dotlike Cx43 immunofluorescence on pericyte processes. (B) Histogram showing counts of Cx43 plaques in pericytes grown in normal or high-glucose medium. Data are expressed as the mean ± SD; *P < 0.05; n = 3. Magnification, ×35.

Figure 4.

 

Effect of high glucose on Cx43 immunoreactivity in BRPs. Representative photomicrographs of (A) cells grown in normal (N) medium or high-glucose (HG) medium. (B) Histogram shows counts from Cx43 plaques in BRPs grown in normal or high-glucose medium. Data are expressed as the mean ± SD; *P < 0.05; n = 3. Magnification, ×80.

Figure 5.

 

Effect of high glucose on Cx43 immunoreactivity in a coculture of BRPs and BRECs. Representative phase-contrast micrographs (A, D), immunofluorescent micrographs of vWF staining (B, E), and micrographs of Cx43 immunolabeling in (C, F) cocultures grown in normal (A–C) and high-glucose (D–F) medium. Magnification. ×100. (G) Histogram shows Cx43 plaque counts in BRECs and BRPs in cocultures grown in normal or high-glucose medium. Data are expressed as the mean ± SD; *P < 0.05; n = 3.

Figure 6.

 

Effect of high glucose on Cx43 mRNA level in HRPs. (A) RT-PCR analysis of Cx43 mRNA level in cells exposed to normal (N) or high glucose (HG) medium. PCR amplification products for Cx43 and β-actin as detected with nucleic acid stain on agarose gel with 0.5-, 1-, and 2-μL samples of reverse transcribed cDNA. PCR with the Cx43 primer pair generated a single 316-bp product; PCR with the β-actin primer pair generated a single 498-bp product. (B) Histogram shows that the Cx43 mRNA level was significantly reduced in cells grown in high-glucose medium compared with those grown in normal medium. Data are expressed as the mean ± SD; *P < 0.05; n = 6.

Figure 7.

 

Analysis of GJIC in retinal pericytes using the SLDT technique. Transfer of lucifer yellow into contiguous cells was detected after scrape-loading. Histogram shows counts from dye-coupled cell layers in normal or high-glucose medium. Data are expressed as the mean ± SD; *P < 0.05; n = 3.