Increased JNK Phosphorylation and Oxidative Stress in Response to Increased Glucose Flux through Increased GLUT1 Expression in Rat Retinal Endothelial Cells

Abstract

purpose. To investigate whether increased glucose flux through increased glucose transporter1 (GLUT1) expression results in increased oxidative stress and increased c-jun N-terminal kinase (JNK) phosphorylation.

methods. GLUT1-overexpressing cells were established using a rat retinal endothelial cell line. The intracellular reactive oxygen species was detected by the oxidation of 5- (and -6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate, acetyl ester (CM-H2-DCFDA). Western blot was performed to determine JNK phosphorylation and lipid peroxidation. Differentially expressed genes were detected by cDNA microarray analysis and confirmed by Northern blot analysis.

results. Clones overexpressing GLUT1 showed an approximate four- to eightfold increase in GLUT1 expression and a 44% increase in intracellular glucose concentrations. GLUT1-overexpressing cells had a 80% increase in DCF fluorescence and increased lipid peroxidation, as well as increased JNK phosphorylation. Analysis of differentially expressed genes in GLUT1-overexpressing cells showed increased expression of JNK interacting protein (JIP)-1, a scaffold protein necessary for JNK activation. Northern blot analysis confirmed upregulation of JIP-1. Immunoprecipitation showed that phosphorylated JNK, but not total JNK, coimmunoprecipitated with JIP-1 protein. At the cellular level, JIP-1 was predominantly localized in cytoplasm, especially in the perinuclear area in retinal endothelial cells.

conclusions. GLUT1 overexpression and increased glucose flux result in increased oxidative stress and JNK phosphorylation in immortalized rat retinal endothelial cells. Further studies are needed to understand molecular events after increased glucose flux in retinal endothelial cells and the relation between increased oxidative stress and JNK phosphorylation.

Diabetic retinopathy (DR), a microvascular complication of diabetes mellitus, is one of the leading causes of adult blindness in developed countries. Despite its prevalence and severity, the molecular mechanisms underlying DR have not been fully elucidated. Various mechanisms have been proposed, such as increased flux through polyol ¹ and hexosamine pathways, ^{2 3} nonenzymatic glycosylation, increased formation of advanced glycation end products (AGEs), ^{4 5} protein kinase C activation, ⁶ glucose-induced DNA damage, ⁷ and oxidative stress. ⁸ All these pathologic changes appear to be initiated by chronic exposure of the retinal microvasculature to increased blood glucose concentrations. Clinical studies have demonstrated a strong association between long-term glycemic control and the development and progression of diabetic retinopathy. ^{9 10 11 12} On a cellular level, prolonged hyperglycemia associated with diabetes mellitus is deleterious to the retinal microvasculature and results in endothelial cell and pericyte death; formation of microaneurysms and acellular capillaries; thickening of basement membranes; and, in severe cases, retinal neovascularization. ¹³  

Glucose is the major energy source for the retina, and its transport from the blood to the neuroretina is mediated by a facilitative, sodium-independent glucose transporter known as GLUT1. ^{14 15 16} The transport of glucose into the retina by GLUT1 exceeds its phosphorylation by hexokinase, the rate-limiting step in retinal glucose metabolism. ^{17 18} Consequently, measurable free glucose is available as a substrate for biochemical processes thought to be responsible for the development of DR. It is also possible, however, that the effects of elevated glucose concentrations may be mediated by the binding of AGEs to endothelial cell surface receptors (RAGE). ¹⁹ Binding of AGE to its receptors has been shown to result in activation of specific signaling pathways, ²⁰ as well as increased oxidative stress ²¹ and apoptosis ^{22 23} in vascular cells. 

The c-Jun N-terminal kinase pathway is important in modulating cellular responses to stress. Increased oxidative stress has been implicated in several animal models of diabetes through activation of the c-Jun N-terminal kinase, such as β-cell dysfunction and the formation of coronary atherosclerosis. ^{24 25} JNK is a stress-induced protein kinase and is involved in regulation of gene expression and stabilization through phosphorylation of Jun and other proteins. ²⁶ Only the phosphorylated JNK (phospho-JNK), which is translocated to the nucleus, activates c-Jun. ²⁷ The assembly of JNK and its upstream kinases (MAPKK and MAPKKK) requires molecular scaffold proteins, the JNK interacting proteins (JIPs). ^{28 29} Several studies have reported that JNK activity is increased in response to diabetes, and elevated JNK activity interferes with insulin action both in cell culture and in animal models. ^{30 31 32 33}  

To understand whether increased intracellular glucose concentrations potentiate oxidative stress and activate the JNK signaling pathway in retinal endothelial cells, we established lines of stable transfected GLUT1-overexpressing rat retinal endothelial cells. In this report, we demonstrate that increased GLUT1 expression results in an increase in intracellular glucose and elevated oxidative stress and JNK phosphorylation in this cell line. These results indicate that the JNK signaling pathway is activated in response to increased glucose flux. 

Methods

Antibodies

The anti-phosphoJNK and anti-total JNK antibodies were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA); the anti-malondialdehyde (MDA) antibody from Abcam, Inc. (Cambridge, MA); and the anti-rabbit and anti-mouse antibodies coupled to horseradish peroxidase from GE Healthcare (Piscataway, NJ). Staining of transferred proteins on the Western blot membranes, as well as an anti-GAPDH monoclonal antibody (Advanced Immunochemical, Inc., Long Beach, CA) were used as internal loading controls. A polyclonal antibody raised against a purified human erythrocyte glucose transporter (GLUT1) was the kind gift of Christin Carter-Su (University of Michigan) and the polyclonal anti-JIP-1 antibodies were kindly provided by Benjamin Margolis (University of Michigan). The anti-GLUT1 and anti-JIP-1 antibodies have been characterized previously. ^{34 35}  

Cell Culture

An immortalized rat retinal endothelial cell line TRiBRB, which was established from retinal capillaries isolated from transgenic rats carrying temperature-sensitive SV-40 large T antigen gene, was used in this study. ³⁶ Although experiments investigating the endothelial cell biology of diabetic microvascular complications have frequently been performed in primary endothelial cell cultures of bovine or human origin, the well-known difficulty in transfecting retinal endothelial cells with exogenous DNA prevented use of a primary retinal endothelial cell culture model in these experiments. ³⁷ The conditionally immortalized rat retinal endothelial cell line, TRiBRB, was therefore used to create stable retinal endothelial cell overexpression of GLUT1. TRiBRB cells are a well-described line that have typical characteristics of primary retinal endothelial cells in culture, including spindle-shaped morphology, expression of factor VIII, VEGF receptor-2, and p-glycoprotein, as well as uptake of acetylated LDL, and facilitated transport of glucose, oxidized vitamin C, and amino acids. ^{36 38 39 40 41} The cells were grown in DMEM (Invitrogen, Carlsbad, CA) supplemented with fetal bovine serum (10%), endothelial cell growth factor (15 μg/mL; Roche, Indianapolis, IN), heparin (100 μg/mL), and antibiotics and antimycotics (100 U/mL penicillin, 100 μg/mL streptomycin, and 250 ng/mL amphotericin B; Sigma-Aldrich, St. Louis, MO). To establish stable transfected cells, full-length human GLUT1 cDNA (a gift from Michael Mueckler, Washington University, St. Louis, MO) was subcloned into pcDNA3.1 at the BamHI site (BD-Clontech, Palo Alto, CA) and transfected into TRiBRB cells. A pcDNA3.1 was also transfected into TRiBRB cells as a vector control. Stable transfected cells were propagated through G418 selection at 32°C. Cells were then switched to 37°C for all experiments. It has been shown that SV40 expression is substantially reduced after 1 to 2 days of culture at a nonpermissive temperature. ³⁶ Expression of GLUT1 was determined by Western blot analysis. ⁴² All cells were collected at the same passages and same cell density. 

Intracellular Glucose Measurement

Intracellular glucose concentration was measured by gas chromatography (GC)-mass spectrometry. Briefly, vector control and GLUT1-overexpressing cells were cultured in complete medium, as stated earlier including 5 mM glucose and allowed to grow to 90% confluence in six-well plates. Cells were then washed three times in cold 1× phosphate-buffered saline. After they were freeze-thawed three times, aliquots (200 μL) of supernatants were treated with trichloroacetic acid (final concentration 5% trichloroacetic acid [TCA]) and then derivatized with hydroxylamine and 4-dimethylamino pyridine. ⁴³ A GC-mass spectrometer (5972 series; Hewlett Packard, Palo Alto, CA) was used to measure glucose concentrations. Aliquots were taken to determine total protein amount using bicinchoninic acid (BCA) protein assays, based on the manufacturer’s instructions (Pierce Biotechnology, Inc., Rockford, IL). Intracellular glucose concentrations were compared between GLUT1-overexpressing and vector control cells. 

Intracellular Reactive Oxygen Species Measurement

The intracellular reactive oxygen species was determined by a fluorescence detection method based on the oxidation of 5- (and -6)-chloromethyl-2′,7′- dichlorodihydrofluorescein diacetate, acetyl ester (CM-H₂DCFDA; Molecular Probes Inc., Eugene, OR). ⁸ GLUT1-overexpressing and vector control cells were grown in complete medium, as stated in the prior section, including 5 mM glucose, and allowed to grow in six-well plates until 90% confluence was achieved. CM-H₂DCFDA was added for 1 hour (20 μM) and fluorescence was analyzed by a cell sorter (Elite ESP; Beckman Coulter, Hialeah, FL) using excitation and emission wavelengths of 495 and 525 nm, respectively. 

RNA Extraction, cDNA Microarray, and Northern Blot Analyses

GLUT1-overexpressing and vector control cells were grown in 5 mM glucose for 5 days to 90% confluence. Vector control and GLUT1-overexpressing cells were homogenized in extraction reagent (TRIzol; Invitrogen, Carlsbad, CA). Total RNA was extracted and further purified using an RNA cleanup procedure (RNeasy; Qiagen, Valencia, CA). RNA yields were assessed by absorbance at 260 nm, and the quality was confirmed on agarose-formaldehyde gels. 

The cDNA microarray analysis was performed in conjunction with the Michigan NIDDK Biotechnology Core using a rat genome microarray (U34 chips; Affymetrix, Inc., Santa Clara, CA) containing gene expression data for >8000 known genes. Total RNA was extracted and purified as just described. Comparisons were made between GLUT1-overexpressing and vector control cells at normal glucose concentrations, to determine whether GLUT1 overexpression per se causes changes in gene expression profiles. Gene expression patterns detected by cDNA microarray analysis were confirmed in replicated experiments. Changes in gene expression that met a minimum of a twofold change, compared with control samples, were confirmed by Northern blot analysis. 

To perform Northern blot analysis, total RNA was fractionated on agarose-formaldehyde gel electrophoresis. RNAs were transferred to membranes (Nytran SuPerCharge membrane; Schleicher & Schuell, Keene, NH) and hybridized with rat JIP-1b cDNA fragment and a housekeeping gene acidic ribosomal phosphoprotein PO (ARP/36B4) cDNA probe labeled with [α-³²P]dCTP by random primer labeling (GE Healthcare). The intensity of each band was visualized and quantified (PhosphorImager and Quantity One software; Bio-Rad, Hercules, CA). 

Immunoblot and Immunoprecipitation

Vector control and GLUT1-overexpressing cells were washed three times in cold 1× phosphate-buffered saline and lysed in lysis buffer (1% SDS, 62.5 μM Tris [pH 6.8], 10% glycerol). As positive controls in MDA detection, both vector and GLUT1-overexpressing cells were treated with 80 μM H₂O₂ for 1 hour at 37°C. Cells were then rinsed twice with PBS. Proteins were resolved by electrophoresis on 10% SDS polyacrylamide gels, transferred to membranes (Hybond-P; GE Healthcare), and immunoblotted with antibodies against GLUT1, total JNK, phosphoJNK, and MDA, followed by secondary horseradish peroxidase-anti-rabbit antibodies according to methods described previously. ⁴² Blots were visualized by a chemiluminescence assay (ECL-Plus; GE Healthcare, Piscataway, NJ). To verify equal loading of proteins, Western blots were stained with ponceau-S after transfer. In addition, after immunoblot, each blot was stripped and reprobed with an anti-GAPDH antibody, as an internal control. 

To perform immunoprecipitation, vector control and GLUT1-overexpressing cells were washed in 1× phosphate-buffered saline and scraped with a cell scraper in lysis buffer (50 mM HEPES [pH 7.5], 10% glycerol, 150 mM NaCl, 1% Triton X-100, 1.5 mM MgCl₂ ,1 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 100 mM NaF, 200 mM sodium orthovanadate, 10 mM tetrasodium pyrophosphate, 10 mg aprotinin/mL and 10 mg leupeptin/mL). ³⁵ Lysates (4.5 mg) were incubated with a anti-JIP-1 antibody (antiserum no. 152) and protein A agarose (RepliGen Corp., Waltham, MA) overnight at 4°C. Pellets were washed three times in washing buffer (20 mM HEPES [pH 7.5], 150 mM NaCl, 10% glycerol, and 0.1% Triton X-100) and boiled in 1× sample buffer. Supernatants were resolved by 8% SDS polyacrylamide gel electrophoresis, transferred to a membrane (Hybond-P; GE Healthcare), and immunoblotted with antibodies against JIP-1, GLUT1, total JNK, and phosphoJNK. Blots were visualized by chemiluminescence as above (ECL Plus; GE Healthcare). 

Immunohistochemistry

Vector control and GLUT1-overexpressing cells were grown on chamber slides to 90% confluence. Cells were fixed with 4% paraformaldehyde for 30 minutes at room temperature and permeabilized with 0.1% Triton X-100 in 1× phosphate-buffered saline for 10 minutes. The slides were blocked with 1% goat serum for 1 hour and incubated with either anti-GLUT1 (1:500 dilution) or anti-JIP-1 (1:250 dilution) overnight. After washing, the slides were incubated with goat anti-rabbit-FITC (1:200 dilution; Molecular Probes) for 1 hour, and mounted (Prolong; Molecular Probes) and viewed with a laser scanning confocal microscope (model OZ; Noran Instruments, Middleton, WI). 

Statistical Analysis

Comparisons between vector control and GLUT1-overexpressing cells were made with Student’s t-test. P < 0.05 was considered significant. 

Results

Increased Glucose Flux in GLUT1-Overexpressing Cells

Under normal culture conditions, immortalized rat retinal endothelial cells overexpressing GLUT1 showed an approximate four- to eightfold increase in GLUT1 expression compared with vector control cells (Fig. 1A) . Such levels of expression were confirmed by Western blot in every experiment performed. Increased GLUT1 expression resulted in increased intracellular glucose, as determined by GC-mass spectrometry (Fig. 1B) . There was an average 40% increase in intracellular glucose in GLUT1-overexpressing cells compared with vector control cells, and the increase was statistically significant (P = 0.0003; Fig. 1B ). Immunohistochemistry confirmed membrane expression of GLUT1 protein in both vector control and GLUT1-overexpressing cells (Fig 1C) . Increased GLUT1 expression was also visible in GLUT1-overexpressing cells, with much of GLUT1 protein localized in the cytoplasmic membrane. 

Increased Oxidative Stress in GLUT1-Overexpressing Cells

The presence of free glucose in the retinal endothelial cells may provide a substrate for a variety of biochemical processes. We examined the possibility that GLUT1-overexpression induces oxidative stress in TRiBRB cells. GLUT1-overexpressing and vector control cells were incubated with CM-H₂DCFDA (20 μM) for 1 hour, and the fluorescence was measured by flow cytometry (FACScan; BD Biosciences, Franklin Lakes, NJ). As shown in Figure 2A , there was approximately an 80% increase in generation of DCF-sensitive ROS in GLUT1-overexpressing compared with vector control cells, a statistically significant increase (P < 0.0001). To determine whether lipid peroxidation products were increased with GLUT1-overexpression, GLUT1-overexpressing and vector control cells were grown in 5 mM glucose, and Western blot analysis was performed using antibody against MDA-protein addicts. Identical cultures were treated with 80 μM H₂O₂ for 1 hour as positive controls. As shown in Figure 2B , there was increased lipid peroxidation in GLUT1-overexpressing cells as determined by increased MDA-modified protein addicts in Western blot analysis (Fig. 2B) . The most prominent MDA-modified protein adducts were of higher molecular weight, approximately 200 and 80 kDa, in GLUT1-overexpressing cells compared to vector control cells. 

Increased JNK Phosphorylation in GLUT1-Overexpressing Cells

Increased JNK activity in response to high glucose has been reported in human umbilical vein endothelial cells. ⁴⁴ To assess whether increased glucose flux through increased GLUT1 expression also results in the phosphorylation of JNK in retinal endothelial cells, we performed Western blot analysis and probed the blots with anti-phosphoJNK and total JNK. As shown in Figure 3 , there was a basal level of JNK phosphorylation in vector control cells. Increased GLUT1 expression and increased glucose flux resulted in an increase in JNK phosphorylation. There was no detectable change in total JNK (Fig. 3) . Western blot with an anti-p38 antibody revealed no change in p38 protein expression (data not shown). 

Increased JIP-1 Expression in GLUT1-Overexpressing Cells

To identify additional molecular changes in GLUT1-overexpressing cells, cDNA microarray analysis was performed using a rat genome microarray (Set U34; Affymetrix, Inc.). We found increased JIP-1 gene expression among the upregulated genes. JIP-1 expression was increased more than twofold in two replicates (Fig. 4A) . To confirm the observed changes in JIP-1 expression, we performed Northern blot analysis. As shown in Figure 4B , there was an approximate 80% increase in JIP-1 expression in GLUT1-overexpressing cells compared with that in vector control cells. The expression of transcripts of enzymes involved in glycolytic, polyol, and hexosamine pathways were unchanged in the overexpressing cells compared with the vector control (data not shown). 

Association of JIP-1 with PhosphoJNK

Endogenous JIP-1 protein has been detected mainly in pancreatic islet and neuronal cells. ^{35 45} To determine whether JIP-1 protein is expressed in retinal endothelial cells, we performed immunoprecipitation using whole-cell lysates (4.5 mg). ³⁵ In both vector control and GLUT1-overexpressing cells, a prominent band corresponding to 90 kDa was detected by antibodies against both JIP-1 carboxyl-terminal peptide (antiserum no. 176) and GST-JIP-1 Src homology 3 domain fusion protein (antiserum no. 152), ³⁵ which was consistent with JIP-1 detected in neuronal cells (Fig. 5A) . ^{35 45} As a JNK scaffold protein, JIP-1 binds to three kinases, MLK3, MAP2K7, and JNK, to form a signaling complex. ^{28 46} In both vector control and GLUT1-overexpressing cells, phosphorylated JNK was found to coimmunoprecipitate with JIP-1 (Fig. 5B) , which is consistent with the role of JIP-1 as a scaffold protein essential in JNK activation. Immunoprecipitation also showed a basal level of JNK phosphorylation associated with JIP-1 in vector control cells. JIP-1 has been reported to be in direct contact with cell surface receptors, guanine exchange factor, and motor kinesin, ^{35 47 48 49} however, GLUT1 did not coimmunoprecipitate with JIP-1 (data not shown). 

Subcellular Localization of JIP-1 in Retinal Endothelial Cells

To determine the localization of endogenous JIP-1 in retinal endothelial cells, both vector control and GLUT1-overexpressing cells were stained by indirect immunofluorescence. JIP-1 was found to be predominantly localized in the cytoplasm, especially in the perinuclear area. Confocal microscopy revealed that a small amount of JIP-1 was localized in the nucleus (Fig. 6) . The pattern of JIP-1 staining was similar in both vector control and GLUT1-overexpressing cells. 

Discussion

In this study, we report changes in oxidative stress and JNK phosphorylation in stable transfected GLUT1-overexpressing rat retinal endothelial cells in response to increased intracellular glucose. Our results demonstrate that increased oxidative stress and JNK phosphorylation resulted from increased glucose flux through increased GLUT1 expression. The increased JNK phosphorylation probably occurs through upregulation of JIP-1, as evidenced by physical association of JIP-1 with phosphorylated JNK protein. 

Increased glucose has been shown to generate increased reactive oxygen species (ROS) in cell types susceptible to hyperglycemic damage, such as endothelial cells and mesangial cells. ^{50 51} In our study, we found an approximate 80% increase in generation of cytosolic ROS, which was detected by DCF in GLUT1-overexpressing compared with vector control cells. It has been suggested that high-glucose–induced ROS production mainly consists of hydrogen peroxide. ⁵¹ In mesangial cells, H₂O₂ was produced through increased glucose uptake and glucose metabolism rather than glucose auto-oxidation and can be effectively inhibited by catalase. ⁵¹ Although the ROS species generated in retinal endothelial cells remains to be determined, our initial experiments did not find significant changes in catalase activity. Our study also detected increased MDA-modified protein addicts at higher molecular weight (approximately 80 and 200 kDa), which suggested increased lipid peroxidation—a marker of oxidative stress-in GLUT1-overexpressing cells. Previous studies in diabetic subjects have shown that lipid peroxidation in both plasma and the vitreous body is linked to the development and progression of diabetic retinopathy. The reduction of lipid peroxidation in patients with rigorously controlled diabetes conferred a 50% lower risk of development of proliferative retinal changes, compared with standard glucose control subjects. ⁵²  

Recent studies found that the JNK pathway plays an important role in the development and progression of diabetes and its complications. ^{30 31 32 33 44 53 54 55} Increased ambient glucose results in growth retardation in several primary endothelial cell cultures, such as human umbilical vein endothelial cells (HUVECs), bovine pulmonary artery endothelial cells (PAECs), and cultured human endothelial cells. ^{56 57 58} The inhibition of cell growth has been suggested to be associated with JNK activation. ⁵³ JNK activation induced by increased glucose was detected both in neurons and endothelial cells. ^{53 54} Our data also showed increased JNK phosphorylation to be associated with increased glucose flux and GLUT1 overexpression in rat retinal endothelial cells. 

The identification of JIP-1 upregulation and increased JNK phosphorylation in the present study suggests that JIP-1 potentiates JNK activation in response to increased glucose flux through GLUT1 overexpression. This raises intriguing questions regarding the molecular events that connect increased glucose flux and increased JNK activity. JIP-1/IB1 has been reported as a nuclear protein that binds to the GLUT2 promoter and regulates GLUT2 expression in islet cells. ⁴⁵ In neurons, JIP-1 is located in cytoplasm. On neuronal differentiation, JIP-1 concentrates at the extending end of neurites. ³⁵ In our study, JIP-1 was mainly localized in the cytoplasm, especially the perinuclear area. JIP-1 has been shown to interact with several membrane receptors to form a signaling complex. ^{47 48 59} JIP-1 has also been detected to interact with an exchange factor for the small GTPase RhoA and molecular motor kinesin to regulate cytoskeleton rearrangement in neuronal cells. ^{35 49} Our study did not find a physical association between JIP-1 and GLUT1. 

Increased oxidative stress has been linked to the activation of JNK pathway in many cell types. ^{60 61} Hydrogen peroxide produced during oxidative stress has been shown to be a potent activator that induces JNK activation in skeletal muscle fibers. ⁶⁰ Studies in bovine lung microvascular endothelial cells have indicated that increased glutathione reductase activity inhibits JNK activation. ⁶¹ Although we did not determine whether increased lipid peroxidation is directly linked to increased JNK phosphorylation in the present study, given the above experimental observations conducted in other cell types, we speculate that there is a mechanistic link between increased lipid peroxidation and JNK phosphorylation in response to increased glucose flux through increased GLUT1 expression in rat retinal endothelial cells. The pathologic consequences associated with increased oxidative stress and JNK phosphorylation in retinal endothelial cells remain to be determined. 

 Supported by an ARVO/Novartis Research Fellowship Grant (JZ) and the JDRF Center for Complications in Diabetes (AKK) and in part by National Institutes of Health Grant RPO60DK-20572, which supports the Michigan Diabetes Research and Training Center.

 Submitted for publication September 8, 2004; revised November 15, 2004, and March 22, 2005; accepted April 11, 2005.

 Disclosure: J. Zhou, None; B.K. Deo, None; K. Hosoya, None; T. Terasaki, None; I.G. Obrosova, None; F.C. Brosius, III, None; A.K. Kumagai, 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.

 

Vector control and GLUT-overexpressing cells[b]. (A) Western blot analysis for GLUT1 in vector and GLUT-overexpressing cells from two independent sets of clones (clone 1 and clone 2). Each lane contains 10 μg protein from whole-cell lysate. The membrane was first probed with anti-GLUT1 antibody, and then it was stripped and reprobed with anti-GAPDH antibody to serve as an internal loading control. (B) Intracellular glucose measurements in vector and GLUT1-overexpressing cells, by GC-mass spectrometry. Cells were cultured in medium containing 5 mM glucose. Data are the mean ± SE (n = 6). There was a statistically significant difference between the vector and GLUT1-overexpressing cells (P = 0.0003, Student’s t-test). (C) Immunocytochemistry for GLUT1 in vector control and GLUT1-overexpressing cells. GLUT1 expression was detected by confocal microscopy. Magnification, ×600.

Figure 2.

 

Oxidative stress measurements in vector and GLUT1-overexpressing cells. (A) DCF measurement in vector and GLUT1-overexpressing cells. Cells were grown in complete medium containing 5 mM glucose and allowed to grow in six-well plates until 90% confluence was achieved. CM-H₂DCFDA was added for 1 hour (20 μM) and analyzed by flow cytometry, using excitation and emission wavelengths of 495 and 525 nm, respectively. (B) Western blot for MDA-modified protein adducts in vector and GLUT1-overexpressing cells. Each lane contains 30 μg protein from whole-cell lysate. Vector control and GLUT1-overexpressing cells treated with 80 μM H₂O₂ were used as a positive control for detecting MDA-modified protein adducts. After probing with anti-MDA antibody, the membrane was stripped and reprobed with GAPDH as the loading control.

Figure 3.

 

Western blot of phosphorylated JNK and total JNK in vector control and GLUT1-overexpressing cells. Vector control and GLUT1-overexpressing cells from two independent clone sets were cultured in medium containing 5 mM glucose. Each lane contains 20 μg protein from whole-cell lysate, except for the UV control lane (5 μg/lane). Human embryonic kidney 293 cells were treated with 400 μJ of UV light to induce JNK phosphorylation and served as positive control for phosphorylated JNK. The membrane was first probed with anti-phosphorylated JNK. The membrane was then stripped and reprobed with anti-total JNK antibody. After the second stripping, the membrane was probed with anti-GAPDH antibody, as an internal loading control.

Figure 4.

 

Upregulation of the JIP-1 gene in GLUT1-overexpressing cells compared with vector control cells, detected by cDNA microarray analysis (A). GLUT1-overexpressing and vector control cells were cultured in medium containing 5 mM glucose for 5 days. Results are from two independent cDNA microarray analyses. JIP-1 was upregulated by more than twofold in both experiments. (B) Northern blot of JIP-1 in vector control and GLUT1-overexpressing cells. Total RNA was extracted from vector control and GLUT1-overexpressing cells. Each lane contained 25 μg total RNA. The membrane was probed with a ³²P-labeled JIP-1 fragment and ARP/36B4.

Figure 5.

 

Coimmunoprecipitation of JIP-1 and phosphorylated JNK in both vector and GLUT1-overexpressing cells. (A) Lysates from vector and GLUT1-overexpressing cells were immunoprecipitated with anti-JIP-1 antibody (antiserum no. 152) and immunoblotted with anti-JIP-1 (antiserum no. 176). Mouse neuroblastoma N1E-115 cells were used as a positive control for JIP-1. Two sets of independent clones were used in the experiments (marked #1 and #2). (B) After probing with anti-JIP-1 antibody (antiserum no. 176), the membrane was stripped and reprobed with anti-phosphorylated JNK antibody to detect the presence of phosphorylated JNK.

Figure 6.

 

Immunocytochemistry for JIP-1 in vector control and GLUT1-overexpressing cells grown on chambered slides. JIP-1 protein was detected with anti-JIP-1 antibody (antiserum no. 152). JIP-1 expression was visualized by confocal microscopy. Magnification, ×600.

The authors thank Kelli Sullivan, Stephen Lentz, and the staff of the JDRF Morphology Core for assistance with morphologic analysis and Martin J. Stevens and Judy Grossi for help with the GC-Mass spectrometry measurements.