Posttranslational, Reversible O-Glycosylation Is Stimulated by High Glucose and Mediates Plasminogen Activator Inhibitor-1 Gene Ex
http://www.100md.com
《内分泌学杂志》
Department of Medicine (H.J.G., C.I.W., I.G.F.), Mount Sinai Hospital and University Health Network, Toronto, Ontario, Canada M5G 1X5
Banting and Best Diabetes Centre (H.J.G., C.I.W., I.G.F.), Toronto, Ontario, Canada M5G 2C4
Department of Physiology (I.G.F.), University of Toronto, Toronto, Ontario, Canada M5S 1A8
Department of Biological Chemistry (G.W.H.), Johns Hopkins University School of Medicine, Baltimore, Maryland 21205-2185
Abstract
Metabolic flux through the hexosamine biosynthetic pathway (HBP) is increased in the presence of high glucose (HG) and potentially stimulates the expression of genes associated with the development of diabetic nephropathy. A number of synthetic processes are coupled to the HBP, including enzymatic intracellular O-glycosylation (O-GlcNAcylation), the addition of single O-linked N-acetylglucosamine monosaccharides to serine or threonine residues. Despite much data linking flow through the HBP and gene expression, the exact contribution of O-GlcNAcylation to HG-stimulated gene expression remains unclear. In glomerular mesangial cells, HG-stimulated plasminogen activator inhibitor-1 (PAI-1) gene expression requires the HBP and the transcription factor, Sp1. In this study, the specific role of O-GlcNAcylation in HG-induced PAI-1 expression was tested by limiting this modification with a dominant-negative O-linked N-acetylglucosamine transferase, by overexpression of neutral -N-acetylglucosaminidase, and by knockdown of O-linked -N-acetylglucosamine transferase expression by RNA interference. Decreasing O-GlcNAcylation by these means inhibited the ability of HG to increase endogenous PAI-1 mRNA and protein levels, the activity of a PAI-1 promoter-luciferase reporter gene, and Sp1 transcriptional activation. Conversely, treatment with the -N-acetylglucosaminidase inhibitor, O-(2-acetamido-2-deoxy-D-glucopyranosylidene)amino-N-phenylcarbamate, in the presence of normal glucose increased Sp1 O-GlcNAcylation and PAI-1 mRNA and protein levels. These findings demonstrate for the first time that among the pathways served by the HBP, O-GlcNAcylation, is obligatory for HG-induced PAI-1 gene expression and Sp1 transcriptional activation in mesangial cells.
Introduction
PROLONGED HYPERGLYCEMIA plays an essential role in the development of diabetic nephropathy and other microvascular complications of diabetes (1, 2, 3). The adverse effects of high glucose (HG) have been proposed to be due to a number of biochemical alterations, including increased reactive oxygen species, increased polyol pathway activity, activation of protein kinase C, formation of advanced glycation end-products, and, potentially, increased flux through the hexosamine biosynthetic pathway (HBP) (2, 3). Approximately 2–5% of glucose normally enters the HBP, of which the first and rate-limiting enzyme, glutamine:fructose-6-phosphate amidotransferase, catalyzes the formation of glucosamine-6-phosphate from fructose-6-phosphate and glutamine (4, 5). Further modifications generate the end-product, UDP-N-acetylglucosamine, a substrate for a number of processes, including O-linked glycosylation of intracellular proteins, complex N- and O-linked protein glycosylation of membrane and secreted proteins, and the synthesis of glycolipids, proteoglycans, and glycosylphosphatidylinositol anchors (6). HG increases flux through the HBP in vitro and in vivo in both insulin-responsive cells and in noninsulin-responsive cells, such as mesangial, endothelial, and smooth muscle cells (7, 8, 9, 10, 11, 12). The HBP was initially proposed to be a mediator of insulin resistance and was found to contribute to the HG-induced expression of leptin, plasminogen activator inhibitor-1 (PAI-1), and other genes in muscle and adipose tissue (5, 13, 14, 15, 16, 17, 18). More recently, HG-stimulated HBP flux has been postulated to be involved in the development of diabetic nephropathy (2, 7, 9, 10, 19, 20, 21, 22, 23). This hypothesis is based on the ability of HBP flux to stimulate the expression of prosclerotic genes in mesangial cells in vitro, such as transforming growth factor-1, PAI-1, fibronectin, and laminin (7, 9, 10, 19, 20, 21, 22, 23). Such changes would serve to accelerate glomerular extracellular matrix accumulation and mesangial expansion in vivo, which are characteristic of diabetic nephropathy (24).
Enzymatic intracellular O-glycosylation (O-GlcNAcylation), the addition of single O-linked -N-acetylglucosamine (O-GlcNAc) monosaccharides to serine or threonine residues on cytosolic and nuclear proteins, is one of the processes coupled to the HBP (4). HG increases protein O-GlcNAcylation in a variety of cell types, including mesangial cells (7, 8, 10, 11, 12). O-GlcNAcylation is concentrated in the nucleus, where it has been identified on transcription factors, such as Sp1, c-myc, YY1, and CRE-binding protein, as well as other nuclear proteins (4). Certain cytosolic proteins have also been reported to be subject to O-GlcNAcylation, including endothelial nitric oxide synthase (8), insulin receptor substrate-1 (16), glycogen synthase (16), the Rpt2 ATPase subunit of the 26S proteasome (25), and heat shock protein 70 (4). Unlike traditional N-linked glycosylation of extracellular proteins, O-GlcNAcylation is dynamic, reversible, and responsive to extracellular stimuli (4). O-GlcNAc transferase (OGT), which catalyzes O-GlcNAcylation, has been cloned and determined to consist of 11.5 amino-terminal tetratricopeptide repeats (TPRs) and a carboxyl-terminal catalytic domain (26, 27). Conversely, the O-GlcNAc-selective deglycosylating enzyme, neutral cytosolic -N acetylglucosaminidase (O-GlcNAcase), removes O-GlcNAc from proteins (28).
HBP-regulated gene expression has been documented by treating cells with glucosamine or glutamine and by overexpression of glutamine:fructose-6-phosphate amidotransferase, which increase cellular UDP-N-acetylglucosamine pools in the absence of HG (9, 10, 11, 14, 17, 18, 19, 20, 21, 22, 23). Furthermore, reversal of HG-induced gene expression with antisense RNA or antisense oligonucleotides specific for glutamine:fructose-6-phosphate amidotransferase or through the use of pharmacological inhibitors of glutamine:fructose-6-phosphate amidotransferase, such as 6-diazo-5-oxonorleucine and azaserine, support the link between the HBP and increased gene expression (7, 9, 11, 19, 20, 21, 22, 23, 29). Yet, the molecular mechanisms governing HBP-induced gene expression are poorly understood. To elucidate these mechanisms, it is essential to first establish which aspect of the HBP regulates gene expression in response to altered HBP flux. O-GlcNAcylation has been correlated with HBP-stimulated transcription and is an attractive candidate to mediate this effect of the HBP (4, 7, 11, 23, 29). However, the reagents used in the studies of HBP-induced gene expression affect overall HBP flux and are not entirely specific for O-GlcNAcylation. Therefore, these would have modulated other consequences of increased HBP flux, such as the development of oxidative stress (30, 31), N-glycosylation (32), glucosamine-6-phosphate accumulation (31, 33), or ganglioside synthesis (34), which have the potential to alter gene expression through changes in intracellular signaling. For example, an enzyme participating in N-glycosylation, 1,6 N-acetylglucosaminyltransferase V, was found to be rate limiting for cytokine signaling (32). Synthesis of the glycolipids, GM1 and GM2 ganglioside, in response to glucosamine has been linked to mesangial cell hypertrophy and cell cycle arrest (34). High levels of glucosamine-6-phosphate generated in response to glucosamine inhibit glucose-6-phosphate dehydrogenase, thereby depleting NADPH and enhancing oxidative stress (30, 31, 33). Thus, a causal relationship between O-GlcNAcylation and increased gene expression has not been established.
Elevated PAI-1 gene expression is involved in the development of diabetic nephropathy and atherosclerosis associated with diabetes (35, 36). Using the PAI-1 gene as a model for HBP-mediated gene expression, we previously found that Sp1 binding sites in the PAI-1 promoter were targets of HG- and glucosamine-induced HBP flux in mesangial cells (19). Increased HBP flux was accompanied by elevated Sp1 transcriptional activation, measured with an Sp1-GAL4 fusion protein (22). To address the role of O-GlcNAcylation in these effects of HG, several approaches were used in this study to specifically target this posttranslational modification. Thus, we demonstrate that O-GlcNAcylation is indispensable for HG-induced PAI-1 gene expression, PAI-1 promoter activation, and Sp1 transcriptional activation.
Materials and Methods
Cell culture
Primary rat mesangial cells were isolated from Sprague-Dawley rat kidney glomeruli, characterized, and grown as described previously (19, 22). For HG experiments, mesangial cells were grown in DMEM Base (Sigma, St. Louis, MO) with 2 mM L-glutamine, and either 1 (control) or 15 (HG) mM glucose for 96 h. Mesangial cells were generally used between passages 15–20, with additional studies performed in passage 8–12 cells comparing 5 with 25 mM glucose. O-(2-acetamido-2-deoxy-D-glucopyranosylidene)amino-N-phenylcarbamate (PUGNAc) was purchased from Toronto Research Chemicals (Toronto, Canada).
Plasmids
Plasmid expression vectors for human O-GlcNAcase and Sp1-GAL4 (amino acids 83–778 of Sp1 fused to Gal4 amino acids 1–147), and the plasmids, pCMV-Luc and pPAI-Luc (which contain nucleotides –740 to +44 of the human PAI-1 promoter fused to a luciferase reporter gene), were described previously (19, 22, 28). To produce a dominant-negative OGT (DNOGT) mutant, nucleotides 170-2212 of human nucleocytoplasmic OGT (GenBank accession no. BC014434), consisting of the TPR domain without the catalytic domain, were amplified from a cDNA template (Invitrogen, Carlsbad, CA) by PCR with pfu turbo DNA polymerase (Stratagene, La Jolla, CA), ligated between the BamHI and HindIII sites of pcDNA3MycHis (Invitrogen), and verified by sequencing. A luciferase reporter gene, pFR-Luc, driven by five GAL4 binding sites was obtained from Stratagene. The expression vector for green fluorescent protein (GFP), pEGFP-N1, was from BD Biosciences Clontech (Palo Alto, CA).
Transient expression of DNOGT and O-GlcNAcase
Mesangial cells were grown in 10-cm dishes, and after 24 h, the cells were transfected at approximately 40% confluence with 1.1 μg plasmid DNA expression vector for DNOGT or O-GlcNAcase, or the empty vector pcDNA3, using Effectene (Qiagen, Valencia, CA) as specified by the manufacturer.
Immunofluorescence
Mesangial cells were seeded on coverslips in six-well plates. For the PAI-1 studies, the cells were transiently transfected after 24 h with an expression vector for GFP, pEGFP-N1, and either O-GlcNAcase or the empty vector, pcDNA3, as described above. The cells were then incubated for an additional 72 h in either 5 or 25 mM glucose and transferred to 0.1% serum during the final 48 h. In additional experiments, the cells were stimulated with 10% serum for 24 h. For the Sp1 studies, the mesangial cells were incubated in 1 or 15 mM glucose for 96 h and transferred to 0.1% serum during the final 48 h. In both cases, the cells were washed with PBS, fixed with 4% formaldehyde for 10 min, permeabilized in 100% methanol at –20 C for 10 min, and blocked in PBS containing 1% goat serum and 0.1% BSA for Sp1 or in 2.5% BSA for PAI-1 for 45 min. The cells were incubated for 1.5 h with either polyclonal Sp1 antibodies, 1/100 (sc-59, Santa Cruz Biochemicals, Santa Cruz, CA), or for the PAI-1 studies, monoclonal GFP antibodies, 1/100 (sc-9996, Santa Cruz Biochemicals) and rat PAI-1 polyclonal antibodies, 1/100 (no. 1062, American Diagnostica, Greenwich, CT). After washing, the cells were incubated with secondary antibodies for 1 h, 1/300 goat antirabbit IgG-Texas Red (Jackson ImmunoResearch, West Grove, PA) for the Sp1 studies or for the PAI-1 studies, and 1/300 goat antirabbit IgG-Texas Red and 1/300 goat antimouse IgG-fluorescein isothiocyanate (Santa Cruz Biochemicals). The coverslips were then mounted onto glass slides with fluorescence mounting media (DakoCytomation, Mississauga, Ontario, Canada). In some experiments, the nuclei were stained with 1 μg/ml 4',6'-diamidino-2-phenylindole for 15 min. Images were captured with a Nikon Eclipse microscope (Nikon, Melville, NY) equipped for epifluorescence and with a monochrome charge-coupled device camera, using SimplePCI software (Compix, Cranberry Township, PA). The images were colorized with SimplePCI and quantified with Scion Image software (Scion Corp., Frederick, MD), which is a version of NIH Image for the Windows operating system. PAI-1 and Sp1 staining were assessed in the cytosol and nucleus, respectively. No staining was observed when boiled Sp1 or PAI antibodies were used (data not shown).
Sp1 immunoprecipitation and Western blotting
To analyze Sp1 O-GlcNAcylation, mesangial cells were lysed in a buffer consisting of 50 mM HEPES (pH 7.9), 150 mM NaCl, 2 mM EDTA, 10% glycerol, 0.1% SDS, 1% NP-40, 0.5% sodium deoxycholate, and protease inhibitor mixture (Roche Applied Science, Indianapolis, IN). Sp1 was immunoprecipitated by rotating cell extracts containing equal amounts of protein (300 μg) with 2.4 μg polyclonal Sp1 antibodies and 90 μl protein G PLUS-agarose beads (Santa Cruz Biochemicals). Western blotting was carried out as previously described (22) using RL-2 O-GlcNAc antibodies, 1/1000 (Affinity BioReagents, Golden, CO), or Sp1 antibodies, 1/1000 (sc-59, Santa Cruz Biochemicals). The resulting x-ray films were scanned and quantified using Scion Image software.
Semiquantitative RT-PCR
Total RNA was isolated with TRIzol Reagent (Invitrogen), and 1.5 μg RNA was reverse transcribed into cDNA with a First Strand cDNA Synthesis kit (Fermentas, Hanover, MD) using random hexamers according to the manufacturer’s protocol. Subsequently, a 2-μl aliquot of cDNA was amplified by PCR in a total volume of 50 μl containing 10 mM Tris (pH 8), 50 mM KCl, 0.08% NP-40, 0.2 mM dNTP, 1.5 mM MgCl2, 8 ng of each primer, and 5 U recombinant Taq DNA polymerase (Fermentas). For PAI-1 and OGT, parallel control PCRs were carried out with primers for 18S rRNA. PCR was performed with cycles consisting of: denaturation at 94 C for 0.5 min, annealing at the temperature indicated below for 1 min, and primer extension at 72 C for 2 min. All cycle numbers were in the linear range. The primers used were: rat PAI-1, sense, 5'-ACCCTCAGCATGTTCATTGC-3' and antisense, 5'-CTCGTTCACCTCGATCTTGAC-3'; 18S rRNA, sense, 5'-CGGCTACCACATCCAAGGAA-3' and antisense, 5'-GCTGGAATTACCGCGGCT-3'; and rat OGT 3' region, sense, 5'-GCCTGTTGAAGTCACTGAGTCAGCCT-3' and antisense, 5'-ATCACCCATCTGCAACACAGTACTATAC-3'. DNA oligonucleotides were synthesized by The Centre for Applied Genomics (The Hospital for Sick Children, Toronto, Canada). The PCR parameters (number of cycles, annealing temperature, and product size) were: rat PAI-1, 29 cycles, 61 C, and 297 bp; 18S rRNA, 21 cycles, 60 C, and 187 bp; and rat OGT 3' region, 30 cycles, 60 C, and 145 bp. Both the PAI-1 and OGT primers crossed exon boundaries. The PCR products were fractionated on a 1.3% agarose gel, stained with ethidium bromide, and visualized with UV light and a charge-coupled device camera using a Gel Doc 2000 instrument (Bio-Rad, Hercules, CA). The resulting TIFF images were quantified with Scion Image software, and the data were normalized to 18S rRNA levels.
Reporter gene analysis
Luciferase reporter gene assays were performed as described previously (19, 22). To take into account transfection efficiency differences between plasmids (such as between O-GlcNAcase and pcDNA3), relative luciferase activity was adjusted based on the luciferase activity obtained when these plasmids were cotransfected in parallel with a luciferase reporter gene driven by a cytomegalovirus promoter, pCMV-Luc. All transfections were repeated at least three times and with different DNA preparations.
EMSA
Mesangial cells were exposed to 5 or 25 mM glucose for 4 d. Nuclear extracts were prepared and EMSAs performed as previously described (19) using the human PAI-1 promoter –85 to –63 Sp1 binding site (sense, 5'-CAGTGAGTGGGTGGGGCTGGAAC-3'; antisense, 5'-GTTCCAGCCCCACCCACTCACTG-3'). All buffers included 50 mM N-acetylglucosamine and 100 nM okadaic acid to prevent deglycosylation and dephosphorylation, respectively. After drying, the gels were exposed to x-ray film, scanned, and quantified with Scion Image software.
PAI-1 ELISA
For determination of PAI-1 protein content in conditioned media, mesangial cells were seeded in 60-mm dishes in DMEM base, 14% fetal bovine serum, 2 mM L-glutamine, and the indicated concentration of glucose for 24 h. This was replaced with the same medium supplemented with 5% fetal bovine serum for 24 h, 0.1% fetal bovine serum for 24 h, and serum-free medium supplemented with 20 mM HEPES (pH 7.4), 2 mM L-glutamine, and 24 mM NaHCO3 for the final 24 h. After centrifugation at 1000 rpm to remove debris, PAI-1 protein was detected in the conditioned medium supernatant with a rat PAI-1 sandwich ELISA kit (American Diagnostica). Results were normalized to the total cellular protein content of each dish.
RNA interference (RNAi)
To reduce OGT expression, a small interfering RNA (siRNA) was targeted to the sequence, 5'-AAAGGGAACTAGATAACATGC-3', corresponding to nucleotides 3479–3499 of the 3' region of rat OGT (GenBank accession no. U76557). This siRNA was prepared by in vitro transcription with a Silencer siRNA synthesis kit (Ambion, Austin, TX) according to the manufacturer’s instructions. The sequence of the DNA oligonucleotide used against OGT was sense, 5'-AAAGGGAACTAGATAACATGCCCTGTCTC-3'. For the control scrambled siRNA, the sequence was sense, 5'-AATCAAGATAGAGCGAACATGCCTGTCTC-3'. A BLAST search of GenBank was carried out to avoid exact matches with other known sequences. The siRNAs were transfected at a final concentration of 20 nM with 48 μl GeneEraser siRNA transfection reagent (Stratagene) according to the manufacturer’s instructions. After 72 h, the cells were used to prepare RNA or cell lysates. OGT antibody, used at a 1/4000 dilution in Western blots, was previously described (26).
Statistics
Results are presented as mean ± SE of at least three independent experiments. Differences between groups were assessed by Student’s t test, with P < 0.05 considered to be statistically significant. Where indicated, results were analyzed by ANOVA with Newman-Keuls post hoc comparisons of means. Statistics were determined with Statistica software (Statsoft, Tulsa, OK).
Results
Dominant-negative OGT interferes with HG-induced Sp1 O-GlcNAcylation and PAI-1 gene expression in mesangial cells
The role of O-GlcNAcylation in HG-stimulated PAI-1 gene expression was examined in rat glomerular mesangial cells. We compared the effect of exposure to 15 mM glucose (HG) for 96 h to 1 mM glucose (control) because the glucose dose-response curve is shifted to lower concentrations in the mesangial cells studied, as previously demonstrated (19). This phenomenon, which has been described in other cultured cell lines (12, 37), is due to high-level glucose transporter 1 expression (data not shown) and/or altered hexokinase activity that results in high basal glucose uptake. Data were also obtained in earlier passage mesangial cells using 5 mM glucose in the control group (see below).
To inhibit O-GlcNAcylation, a nucleocytoplasmic OGT (38) mutant that included the amino-terminal TPRs, but lacked most of the carboxy-terminal catalytic domain, was transiently transfected. This construct was designed to impair OGT function in a dominant-negative manner by competing with wild-type OGT for trimerization, targeting, and substrate recognition (4). An analogous construct containing the TPRs from the Arabidopsis OGT homologue, SPINDLY, exerts dominant-negative effects in Arabidopsis (39). Preliminary experiments with GFP revealed a transient transfection efficiency of approximately 65% in the mesangial cells using Effectene (data not shown). Expression of DNOGT was documented by immunoblotting with an antibody directed against the OGT myc-epitope tag. A band with an approximate molecular mass of 68 kDa was identified in DNOGT-transfected cells but not in cells transfected with pcDNA3 (Fig. 1A). Sp1 O-GlcNAcylation, detected by immunoprecipitating Sp1 and immunoblotting with monoclonal RL-2 O-GlcNAc antibodies, was used as a marker for cellular O-GlcNAcylation. This and other Western blots presented below were quantified by scanning the x-ray films and using Scion Image software, giving a semiquantitative assessment of protein expression. As depicted in Fig. 1B, mesangial cells incubated in HG displayed increased Sp1 O-GlcNAcylation, 2.9 ± 0.8-fold above control (P < 0.05, n = 3), and this was diminished to 83 ± 42% of basal levels (P < 0.05 compared with HG, n = 3) by transfected DNOGT.
To determine whether endogenous PAI-1 mRNA levels are regulated by O-GlcNAcylation, PAI-1 mRNA levels were measured by semiquantitative RT-PCR using 18S rRNA amplified in parallel as an internal control. In pcDNA3-transfected cells, PAI-1 mRNA levels were elevated 2.1-fold by a 96-h treatment with HG (Fig. 1C, P < 0.001). Transient expression of DNOGT did not reduce baseline PAI-1 mRNA levels significantly but inhibited the increase in response to HG (Fig. 1C).
Overexpression of O-GlcNAcase blocks HG-induced increases in PAI-1 gene expression, promoter activity, and protein levels, and Sp1 transcriptional activation in mesangial cells
The inhibitory effects of DNOGT suggested that O-GlcNAcylation was required for stimulation of PAI-1 expression by HG. However, another possibility could have been that OGT interfered with gene expression by protein-protein interactions via the overexpressed TPR domain (4). Thus, O-GlcNAcase, which has previously been demonstrated to diminish O-GlcNAcylation when overexpressed, was also transiently transfected (28). As shown in Fig. 2A, in the presence of HG Sp1 O-GlcNAcylation was reduced by O-GlcNAcase to 40 ± 13% of the levels in pcDNA3-transfected cells (P < 0.01, n = 3). Transfected O-GlcNAcase did not significantly reduce baseline PAI-1 mRNA levels (Fig. 2B). At the same time, HG failed to increase PAI-1 mRNA levels when O-GlcNAcase was overexpressed (Fig. 2B).
Next, the influence of O-GlcNAcylation on PAI-1 gene promoter activity was assessed by transiently transfecting a luciferase reporter gene driven by a (–740 to +44) PAI-1 promoter (19, 22). O-GlcNAcase had no effect on baseline PAI-1 promoter activity. A 1.8-fold increase in PAI-1 promoter activity was observed in HG-treated cells cotransfected with pcDNA3 (P < 0.001, Fig. 2C), whereas PAI-1 promoter activity was refractory to stimulation by HG in cells cotransfected with O-GlcNAcase. PAI-1 promoter activation by HG was specific because luciferase activity driven by a cytomegalovirus promoter was not affected by HG (1 ± 0.1-fold of control, data not shown). Sp1 transcriptional activity was evaluated by cotransfecting Sp1-GAL4, an expression vector for Sp1 (amino acids 83–778) fused to the DNA-binding domain of the yeast transcription factor, GAL4, and a luciferase reporter gene driven by multiple GAL4 binding sites, pFR-Luc (22). HG-stimulated Sp1-GAL4 activity was substantially greater (2.5-fold, P < 0.01) than control in pcDNA3-transfected cells (Fig. 2D). Significantly, cotransfected O-GlcNAcase diminished baseline Sp1-GAL4 activity and rendered Sp1-GAL4 activation unresponsive to HG (Fig. 2D). Similar results were obtained with DNOGT (data not shown).
To confirm that HG-stimulated PAI-1 gene expression is dependent on O-GlcNAcylation in mesangial cells cultured in a physiologic baseline glucose concentration, the effect of 25 mM glucose was compared with that of 5 mM glucose in passage 8–12 rat mesangial cells. These cells were transiently transfected with GFP to mark the transfected cells and with either O-GlcNAcase or pcDNA3. PAI-1 and GFP were detected by immunofluorescence staining of mesangial cells with specific antibodies, and PAI-1 levels were quantified in the cytosol of GFP-positive cells. The percentage of GFP-positive cells, 3%, indicated a lower transfection efficiency in these slower growing, lower passage mesangial cells. Accordingly, PAI-1 protein expression was examined in individual transfected cells. Figure 3 shows that 25 mM glucose triggered a 2.03- ± 0.09-fold (P < 0.001) increase in PAI-1 protein staining compared with 5 mM glucose in cells transfected with pcDNA3, whereas O-GlcNAcase suppressed this increase (1.04- ± 0.05-fold, n = 4) without any effect on baseline PAI-1 protein levels. In contrast, PAI-1 protein staining in mesangial cells exposed to 10% serum was not reduced by transiently transfected O-GlcNAcase (Fig. 4).
Silencing OGT expression with RNAi reduces PAI-1 mRNA levels in mesangial cells
Additional evidence that O-GlcNAcylation regulates PAI-1 gene expression in mesangial cells was obtained using RNAi to diminish endogenous OGT expression. A siRNA targeted to OGT was transfected into mesangial cells at a low concentration, 20 nM, to minimize off-target gene regulation (40). After 72 h, OGT siRNA significantly reduced OGT mRNA levels to 37 ± 0.07% of control (P < 0.001, n = 3) (Fig. 5A), decreased OGT protein expression (Fig. 5B), and concomitantly decreased Sp1 O-GlcNAcylation (Fig. 5C) compared with cells transfected with a scrambled siRNA. Consistent with a positive role for O-GlcNAc in PAI-1 gene expression, reduction of OGT levels by RNAi diminished PAI-1 mRNA abundance in mesangial cells grown under HG conditions to 41 ± 9% (P < 0.01, n = 3) of the levels in cells transfected with a scrambled siRNA (Fig. 5D).
HG does not increase nuclear Sp1 levels or DNA binding in mesangial cells
Previous studies of insulin-stimulated calmodulin gene expression and glutamine-stimulated argininosuccinate synthetase gene expression revealed that Sp1 O-GlcNAcylation correlated with and may have induced Sp1 nuclear translocation (41, 42). To ascertain whether HG affects Sp1 nuclear levels in mesangial cells, Sp1 immunofluorescence staining was performed. As illustrated in Fig. 6A and confirmed by 4',6'-diamidino-2-phenylindole staining (data not shown), Sp1 was primarily localized to the nucleus. The intensity of Sp1 nuclear or cytosolic staining was not elevated in HG-treated mesangial cells compared with control cells in three experiments (Fig. 6B). Similar results were obtained by immunoblotting for Sp1 in nuclear extracts (data not shown).
Based on reported in vitro data, O-GlcNAcylation may up-regulate Sp1 DNA binding activity (43). Therefore, EMSAs were carried out to assess Sp1 binding to an oligonucleotide containing one of the PAI-1 promoter Sp1 sites (–85 to –63). A previous study confirmed Sp1 binding to this PAI-1 promoter site in mesangial cells by supershifting of the Sp1 band with Sp1-specific antibodies (19). Binding to this PAI-1 oligonucleotide was specific because it was eliminated by a 50-fold excess of unlabeled oligonucleotide (Fig. 7B). However, the amount of Sp1-DNA complexes formed were similar in nuclear extracts prepared from mesangial cells incubated with either 25 or 5 mM glucose (Fig. 7, A and C).
Inhibition of O-GlcNAcase with PUGNAc up-regulates PAI-1 mRNA and protein levels
To determine whether elevated O-GlcNAc levels increase PAI-1 gene expression in the absence of HG, mesangial cells were treated with an inhibitor of O-GlcNAcase, PUGNAc. This inhibitor causes a generalized increase in steady-state O-GlcNAc-modified protein levels, reflecting the dynamic nature of O-GlcNAcylation (15, 44). Exposing mesangial cells to 100 μM PUGNAc for 24 h resulted in elevated Sp1 O-GlcNAcylation (Fig. 8A) and a 2.5-fold increase in PAI-1 mRNA levels (P < 0.01, Fig. 8B). There was no further increase in PAI-1 mRNA levels with the combination of PUGNAc and HG. To establish the impact of PUGNAc on PAI-1 protein levels, PAI-1 ELISAs were performed in conditioned media. HG and phorbol 12-myristate 13-acetate, used as positive controls, both increased PAI-1 protein concentration in conditioned media 2.1-fold above control, whereas 100 μM PUGNAc (for 24 h) elicited a similar, 2.3-fold increase (P < 0.05, Fig. 8C). Absolute PUGNAc-stimulated PAI-1 levels were 13.8 ± 1.7 ng/ml.
Discussion
In this study, we investigated the role of O-GlcNAcylation, one of the consequences of enhanced HBP flux, in HG-induced PAI-1 gene expression in mesangial cells. Suppression of O-GlcNAcylation by overexpression of DNOGT or O-GlcNAcase prevented HG-induced increases in PAI-1 mRNA levels and PAI-1 promoter activity. Reduced PAI-1 mRNA levels in the presence of HG was also demonstrated when OGT was depleted by RNAi. Sp1 transcriptional activation by HG, previously shown to be associated with PAI-1 promoter activation (22) and assessed with Sp1-GAL4 and pFR-Luc, was also attenuated by O-GlcNAcase. Furthermore, the ability of O-GlcNAcase to block HG-induced PAI-1 protein expression was demonstrated in early passage mesangial cells that were incubated in 5 and 25 mM glucose. Finally, treatment of mesangial cells with the O-GlcNAcase inhibitor, PUGNAc, led to increased Sp1 O-GlcNAcylation, PAI-1 mRNA levels, and PAI-1 protein levels in the presence of normal (basal) glucose concentrations. Taken together, these observations demonstrate for the first time that one aspect of the HBP, namely O-GlcNAcylation, is essential for HG-stimulated PAI-1 gene expression and Sp1 transcriptional activation in mesangial cells.
Overexpression of DNOGT decreased HG-stimulated Sp1 O-GlcNAcylation to a small extent below that detected in control cells not treated with HG (P = not significant) (Fig. 1). This probably reflects some reduction of basal O-GlcNAcylation, possibly in the context of greater transfection efficiency than determined in preliminary studies. It should also be noted that baseline O-GlcNAcylation is determined in large part by the rate of glucose uptake through glucose transporter 1 in each particular set of mesangial cells studied. Consequently, some decrease in basal Sp1 O-GlcNAcylation upon overexpression of DNOGT or O-GlcNAcase would not be unexpected in higher passage mesangial cells that tend to have high basal glucose uptakes. Nonetheless, basal PAI-1 mRNA levels did not decrease significantly as a consequence of DNOGT or O-GlcNAcase overexpression (Figs. 1 and 2). Similarly, no change in baseline PAI-1 protein expression was seen in the immunofluorescence study when O-GlcNAcase was overexpressed (Fig. 3). Furthermore, O-GlcNAcase did not impair serum-stimulated PAI-1 protein expression (Fig. 4). Thus, the ability of O-GlcNAcase to block HG-induced PAI-1 expression does not appear to be due to a nonspecific inhibition of basal cellular function or of PAI-1 response. A positive role of O-GlcNAcylation in the stimulation of PAI-1 expression by HG is also supported by the finding that PAI-1 protein and mRNA expression were increased in the presence of PUGNAc, which acts by raising O-GlcNAc levels above baseline. HG-stimulated Sp1-GAL4 activity was completely abolished by O-GlcNAcase overexpression, and even basal activity was somewhat suppressed (Fig. 2). These effects likely represent the high sensitivity of the artificial chimeric transcription factor, Sp1-GAL4, to Spl function, and support the notion that Sp1 is positively regulated by cellular O-GlcNAc levels in mesangial cells.
In addition to the positive effects of O-GlcNAcylation on gene expression found in the present experiments, O-GlcNAcylation can exert negative effects on Sp1-dependent transcription. Previous reports indicate that O-GlcNAcylation of Ser 484 in the glutamine-rich Sp1 B transcriptional activation domain disrupts its interaction with the transcription cofactor, Drosophila TATA-binding protein-associated factor II 110 (dTAFII110, human equivalent hTAFII130) (45). A GAL4 construct fused to this Sp1 domain was less active in an in vitro transcription assay when O-glycosylated (45). Overexpression of OGT inhibited the activity of a reporter gene driven by multiple Sp1 sites, in part due to the recruitment of the transcriptional repressor, mSin3A, by OGT (46). In contrast, we have found that O-GlcNAcylation is required for HG-induced PAI-1 gene expression and Sp1 transcriptional activation in mesangial cells. A number of differences between the present experiments and the previous reports may account for these different results. First, the present experiments argue that an overall increase in mesangial cell O-GlcNAcylation is necessary for HG-induced PAI-1 expression, but, it is important to note, do not demonstrate directly the role played by O-GlcNAcylation of Sp1 itself. Indeed, the O-GlcNAcylation status of Sp1 Ser 484 in the current experiments remains unknown. It is possible that in mesangial cells HG-induced O-GlcNAcylation of other Spl sites, or more likely other substrates, overcomes any negative effects of Sp1 Ser 484 O-GlcNAcylation. Second, we used various approaches to modify HG-induced O-GlcNAcylation but not overexpression of OGT as in the earlier reports. In preliminary experiments, we found that overexpression of full-length OGT did not increase Sp1 O-GlcNAcylation in mesangial cells (data not shown), perhaps because O-GlcNAcylation is limited primarily by substrate flux. The ability of full-length OGT to form complexes with other nuclear proteins, for example mSin3A (46), may cause OGT overexpression to affect transcription differently than HG and potentially in a manner independent of O-GlcNAcylation. Third, we assessed endogenous PAI-1 mRNA levels and the activity of the native PAI-1 promoter linked to a luciferase reporter rather than a reporter gene driven by a multimerized Sp1-binding site. Finally, O-GlcNAcylation or phosphorylation of Sp1 and its associated transcription cofactors may differ in mesangial cells compared with the previously used cell types, thereby modifying the overall impact of O-GlcNAcylation on transcription.
A number of mechanisms may explain the linkage of O-GlcNAcylation to PAI-1 gene transcription. Because HG did not increase Sp1 nuclear protein levels (Fig. 6), the present results are unlikely to be consequences of either O-GlcNAc-stimulated Sp1 nuclear translocation or O-GlcNAc-mediated inhibition of the proteasome (25), leading to Sp1 accumulation. Possibly, these latter effects only require basal levels of O-GlcNAcylation. The high basal levels of Sp1 observed in the nucleus in the present experiments are consistent with this concept and would have masked any effects of O-GlcNAc on Sp1 nuclear translocation. Similarly, the lack of effect of HG on Sp1 DNA binding in EMSAs (Fig. 7) argues against O-GlcNAc-induced Sp1 DNA binding to account for the observed effects of O-GlcNAcylation on transcription.
Because PAI-1 expression is known to be stimulated by transforming growth factor-, and the latter is up-regulated by increased HBP flux (10, 23, 47), it was possible that the O-GlcNAc effects could be indirect. However, our previous data argue against a role for transforming growth factor- in HBP-stimulated PAI-1 promoter activity under the current experimental conditions (19, 21).
The observation that HG-stimulated Sp1-GAL4 activity was blocked by cotransfection of O-GlcNAcase points to Sp1 and/or its coactivators as potential functional targets of O-GlcNAcylation. Theoretically, O-GlcNAcylation may augment Sp1 function by altering protein-protein interactions between Sp1 and its cofactors. A significant precedent is that O-GlcNAcylation of the transcription factor, Stat5a, increases its interaction with its cofactor, CRE-binding protein (48). Further work is necessary to investigate the modulation of Sp1 protein interactions by O-GlcNAcylation.
Finally, several studies have reported that flux through the HBP stimulates a plethora of signal transduction pathways, including protein kinase C and reactive oxygen species formation (8, 10, 15, 20, 22, 30, 31, 49, 50), although the role of O-GlcNAc in most of these effects is not clear. In this regard, it was previously demonstrated that overall HBP flux increased protein kinase C- and - kinase activity; however, protein kinase C O-GlcNAcylation was not detected (22).
In summary, these results demonstrate that O-GlcNAcylation plays an essential role in HG-induced PAI-1 gene expression and Sp1 transcriptional activation in mesangial cells. These findings indicate that the search for the relevant mechanism by which HG-mediated increased HBP flux stimulates gene expression should focus on O-GlcNAc-modified proteins. Furthermore, these data motivate in vivo experiments designed to establish the role of O-GlcNAcylation as the major effector of HBP flux in the regulation of PAI-1 and the expression of other genes in the diabetic kidney. These data support the concept that HG-stimulated O-GlcNAcylation acts in concert with other pathways, such as reactive oxygen species, protein kinase C, transforming growth factor-, and advanced glycosylation end-products, to augment prosclerotic gene expression in mesangial cells, leading to diabetic nephropathy.
Footnotes
This work was supported by grants from the Juvenile Diabetes Foundation International and by the Canadian Institutes of Health Research (Grant 88811).
First Published Online September 22, 2005
Abbreviations: DMSO, Dimethyl sulfoxide; DNOGT, dominant-negative OGT; GFP, green fluorescent protein; HBP, hexosamine biosynthetic pathway; HG, high glucose; O-GlcNAc, O-linked -N-acetylglucosamine; O-GlcNAcase, neutral cytosolic -N acetylglucosaminidase; O-GlcNAcylation, O-glycosylation; OGT, O-GlcNAc transferase; PAI, plasminogen activator inhibitor; PUGNAc, O-(2-acetamido-2-deoxy-D-glucopyranosylidene)amino-N-phenylcarbamate; RNAi, RNA interference; siRNA, small interfering RNA; TPR, tetratricopeptide repeat.
Accepted for publication September 14, 2005.
References
The Diabetes Control and Complications Trial Research Group 1993 The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 329:977–986
Brownlee M 2001 Biochemistry and molecular cell biology of diabetic complications. Nature 414:813–820
Sheetz MJ, King GL 2002 Molecular understanding of hyperglycemia’s adverse effects for diabetic complications. JAMA 288:2579–2588
Zachara NE, Hart GW 2004 O-GlcNAc a sensor of cellular state: the role of nucleocytoplasmic glycosylation in modulating cellular function in response to nutrition and stress. Biochim Biophys Acta 1673:13–28
McClain DA 2002 Hexosamines as mediators of nutrient sensing and regulation in diabetes. J Diabetic Compl 16:72–80
Boehmelt G, Wakeham A, Elia A, Sasaki T, Plyte S, Potter J, Yang Y, Tsang E, Ruland J, Iscove NN, Dennis JW, Mak TW 2000 Decreased UDP-GlcNAc levels abrogate proliferation control in EMeg32-deficient cells. EMBO J 19:5092–5104
Du XL, Edelstein D, Rossetti L, Fantus IG, Goldberg H, Ziyadeh F, Wu J, Brownlee M 2000 Hyperglycemia-induced mitochondrial superoxide overproduction activates the hexosamine pathway and induces plasminogen activator inhibitor-1 expression by increasing Sp1 glycosylation. Proc Natl Acad Sci USA 97:12222–12226
Du XL, Edelstein D, Dimmeler S, Ju O, Sui C, Brownlee M 2001 Hyperglycemia inhibits endothelial nitric oxide synthase activity by posttranslational modification at the Akt site. J Clin Invest 108:1341–1348
Kolm-Litty V, Sauer U, Nerlich A, Lehmann R, Schleicher ED 1998 High glucose-induced transforming growth factor 1 production is mediated by the hexosamine pathway in porcine glomerular mesangial cells. J Clin Invest 101:160–169
Burt DJ, Gruden G, Thomas SM, Tutt P, Dell’Anna C, Viberti GC, Gnudi L 2003 P38 mitogen-activated protein kinase mediates hexosamine-induced TGF1 mRNA expression in human mesangial cells. Diabetologia 46:531–537
Sayeski PP, Kudlow JE 1996 Glucose metabolism to glucosamine is necessary for glucose stimulation of transforming growth factor- gene transcription. J Biol Chem 271:15237–15243
Han I, Oh ES, Kudlow JE 2000 Responsiveness of the state of O-linked N-acetylglucosamine modification of nuclear pore protein p62 to the extracellular glucose concentration. Biochem J 350:109–114
Marshall S, Bacote V, Traxinger RR 1991 Discovery of a metabolic pathway mediating glucose-induced densenitization of the glucose transport system. Role of hexosamine biosynthesis in the induction of insulin resistance. J Biol Chem 266:4706–4712
Wang J, Liu R, Hawkins M, Barzilai N, Rossetti L 1998 A nutrient-sensing pathway regulates leptin gene expression in muscle and fat. Nature 393:684–688
Vosseller K, Wells L, Lane MD, Hart, GW 2002 Elevated nucleocytoplasmic glycosylation by O-GlcNAc results in insulin resistance associated with defects in Akt activation in 3T3–L1 adipocytes. Proc Natl Acad Sci USA 99:5313–5318
McClain DA, Lubas WA, Cooksey RC, Hazel M, Parker GJ, Love DC, Hanover JA 2002 Altered glycan-dependent signaling induces insulin resistance and hyperleptinemia. Proc Natl Acad Sci USA 99:10695–10699
Obici S, Wang J, Chowdury R, Feng Z, Siddhanta U, Morgan K, Rossetti L 2002 Identification of a biochemical link between energy intake and energy expenditure. J Clin Invest 109:1599–1605
Einstein FH, Atzmon G, Yang XM, Ma XH, Rincon M, Rudin E, Muzumdar R, Barzilai N 2005 Differential responses of visceral and subcutaneous fat depots to nutrients. Diabetes 54:672–678
Goldberg HJ, Scholey J, Fantus IG 2000 Glucosamine activates the plasminogen activator-inhibitor 1 gene promoter through Sp1 DNA binding sites in glomerular mesangial cells. Diabetes 49:863–871
Singh LP, Crook ED 2000 Hexosamine regulation of glucose-mediated laminin synthesis in mesangial cells involves protein kinases A and C. Am J Physiol 279:F646–F654
James LR, Fantus IG, Goldberg H, Ly H, Scholey JW 2000 Overexpression of glutamine:fructose-6-phosphate amidotransferase (GFAT) activates PAI promoter. Am J Physiol 279:F718–F727
Goldberg HJ, Whiteside CI, Fantus IG 2002 The hexosamine pathway regulates the plasminogen activator inhibitor-1 gene promoter and Sp1 transcriptional activation through protein kinase C- I and -. J Biol Chem 277:33833–33841
Weigert C, Brodbeck K, Sawadogo M, Haring HU, Schleicher ED 2004 Upstream stimulatory factor (USF) proteins induce human TGF-1 gene activation via the glucose-response element-1013/-1002 in mesangial cells: up-regulation of USF activity by the hexosamine biosynthetic pathway. J Biol Chem 279:15908–15915
Steffes MW, Bilous RW, Sutherland DE, Mauer SM 1992 Cell and matrix components of the glomerular mesangium in type I diabetes. Diabetes 41:679–684
Zhang F, Su K, Yang X, Bowe DB, Paterson AJ, Kudlow JE 2003 O-GlcNAc modification is an endogenous inhibitor of the proteasome. Cell 115:715–725
Kreppel LK, Blomberg MA, Hart GW 1997 Dynamic glycosylation of nuclear and cytosolic proteins. Cloning and characterization of a unique O-GlcNAc transferase with multiple tetratricopeptide repeats. J Biol Chem 272:9308–9315
Lubas WA, Frank DW, Krause M, Hanover JA 1997 O-linked GlcNac transferase is a conserved nucleocytoplasmic protein containing tetratricopeptide repeats. J Biol Chem 272:9316–9324
Wells L, Gao Y, Mahoney JA, Vosseller K, Chen C, Rosen A, Hart GW 2002 Dynamic O-glycosylation of nuclear and cytosolic proteins: further characterization of the nucleocytoplasmic -N-acetylglucosaminidase, O-GlcNAcase. J Biol Chem 277:1755–1761
James LR, Tang D, Ingram A, Ly H, Thai K, Cai L, Scholey JW 2002 Flux through the hexosamine pathway is a determinant of nuclear factor B-dependent promoter activation. Diabetes 51:1146–1155
Kaneto H, Xu G, Song KH, Suzuma K, Bonner-Weir S, Sharma A, Weir GC 2001 Activation of the hexosamine pathway leads to deterioration of pancreatic -cell function through the induction of oxidative stress. J Biol Chem 276:31099–31104
Wu G, Haynes TE, Li H, Yan, Meininger, CJ 2001 Glutamine metabolism to glucosamine is necessary for glutamine inhibition of endothelial nitric oxide synthesis. Biochem J 353:245–252
Partridge EA, Le Roy C, Di Guglielmo GM, Pawling J, Cheung P, Granovsky M, Nabi IR, Wrana JL, Dennis JW 2004 Regulation of cytokine receptors by Golgi N-glycan processing and endocytosis. Science 306:120–124
Marshall S, Nadeau O, Yamasaki K 2005 Glucosamine-induced activation of glycogen biosynthesis in isolated adipocytes. Evidence for a rapid allosteric control mechanism within the hexosamine biosynthesis pathway. J Biol Chem 280:11018–11024
Masson E, Wiernsperger N, Lagarde M, El Bawab S 2005 Glucosamine induces cell-cycle arrest and hypertrophy of mesangial cells: implication of gangliosides. Biochem J 388:537–544
Eddy AA 2002 Plasminogen activator inhibitor-1 and the kidney. Am J Physiol 283:F209–F220
Paueksakon P, Revelo MP, Ma LJ, Marcantoni C, Fogo AB 2002 Microangiopathic injury and augmented PAI-1 in human diabetic nephropathy. Kidney Int 61:2142–2148
Tan J, Yang HS, Patel MS 1998 Regulation of mammalian pyruvate dehydrogenase subunit gene expression by glucose in HepG2 cells. Biochem J 336:49–56
Love DC, Kochran J, Cathey RL, Shin, SH, Hanover JA 2002 Mitochondrial and nucleocytoplasmic targeting of O-linked GlcNAc transferase. J Cell Sci 116:647–654
Tseng TS, Swain SM, Olszewski NE 2001 Ectopic expression of the tetratricopeptide repeat domain of SPINDLY causes defects in gibberellin response. Plant Physiol 126:1250–1258
Semizarov, D, Frost L, Sarthy A, Kroeger P, Halbert DN, Fesik SW 2003 Specificity of short interfering RNA determined through gene expression signatures. Proc Natl Acad Sci USA 100:6347–6352.
Majumdar G, Harmon A, Candelaria R, Martinez-Hernandez A, Raghow R, Solomon SS 2003 O-glycosylation of Sp1 and transcriptional regulation of the calmodulin gene by insulin and glucagon. Am J Physiol 285:E584–E591
Brasse-Lagnel C, Fairand A, Lavoinne A, Husson A 2003 Glutamine stimulates argininosuccinate synthetase gene expression through cytosolic O-glycosylation of Sp1 in Caco-2 cells. J Biol Chem 278:52504–52510
Weigert C, Klopfer K, Kausch C, Brodbeck K, Stumvoll M, Haring HU, Schleicher ED 2003 Palmitate-induced activation of the hexosamine pathway in human myotubes: increased expression of glutamine:fructose-6-phosphate aminotransferase. Diabetes 52:650–656
Haltiwanger RS, Grove K, Philipsberg GA 1998 Modulation of O-linked N-acetylglucosamine levels on nuclear and cytoplasmic proteins in vivo using the peptide O-GlcNAc--N-acetylglucosaminidase inhibitor O-(2-Acetoamido-2-deoxy-D-glucopyranosylidene)amino-N-phenylcarbamate. J Biol Chem 273:3611–3617
Yang X, Su K, Roos MD, Chang Q, Paterson AJ, Kudlow JE 2001 O-linkage of N-acetylglucosamine to Sp1 activation domain inhibits its transcriptional capability. Proc Natl Acad Sci USA 98:6611–6616
Yang X, Zhang F, Kudlow JE 2002 Recruitment of O-GlcNAc transferase to promoters by corepressor mSin3A: coupling protein O-GlcNAcylation to transcriptional repression. Cell 110:69–80
Singh LP, Green K, Alexander M, Bassly S, Crook ED 2004 Hexosamines and TGF-1 use similar signaling pathways to mediate matrix protein synthesis in mesangial cells. Am J Physiol 286:F409–F416
Gewinner C, Hart G, Zachara N, Cole R, Beisenherz-Huss C, Groner B 2003 The coactivator of transcription CREB binding protein interacts preferentially with the glycosylated form of Stat5. J Biol Chem 279:3563–3572
Hsieh TJ, Fustier P, Zhang SL, Filep JG, Tang SS, Ingelfinger JR, Fantus IG, Hamet P, Chan JS 2003 High glucose stimulates angiotensinogen gene expression and cell hypertrophy via activation of the hexosamine biosynthesis pathway in rat kidney proximal tubular cells. Endocrinology 144:4338–4349
Kneass ZT, Marchase RB 2005 Protein O-GlcNAc modulates motility-associated signaling intermediates in neutrophils. J Biol Chem 280:14579–14585(Howard J. Goldberg, Catharine I. Whitesi)
Banting and Best Diabetes Centre (H.J.G., C.I.W., I.G.F.), Toronto, Ontario, Canada M5G 2C4
Department of Physiology (I.G.F.), University of Toronto, Toronto, Ontario, Canada M5S 1A8
Department of Biological Chemistry (G.W.H.), Johns Hopkins University School of Medicine, Baltimore, Maryland 21205-2185
Abstract
Metabolic flux through the hexosamine biosynthetic pathway (HBP) is increased in the presence of high glucose (HG) and potentially stimulates the expression of genes associated with the development of diabetic nephropathy. A number of synthetic processes are coupled to the HBP, including enzymatic intracellular O-glycosylation (O-GlcNAcylation), the addition of single O-linked N-acetylglucosamine monosaccharides to serine or threonine residues. Despite much data linking flow through the HBP and gene expression, the exact contribution of O-GlcNAcylation to HG-stimulated gene expression remains unclear. In glomerular mesangial cells, HG-stimulated plasminogen activator inhibitor-1 (PAI-1) gene expression requires the HBP and the transcription factor, Sp1. In this study, the specific role of O-GlcNAcylation in HG-induced PAI-1 expression was tested by limiting this modification with a dominant-negative O-linked N-acetylglucosamine transferase, by overexpression of neutral -N-acetylglucosaminidase, and by knockdown of O-linked -N-acetylglucosamine transferase expression by RNA interference. Decreasing O-GlcNAcylation by these means inhibited the ability of HG to increase endogenous PAI-1 mRNA and protein levels, the activity of a PAI-1 promoter-luciferase reporter gene, and Sp1 transcriptional activation. Conversely, treatment with the -N-acetylglucosaminidase inhibitor, O-(2-acetamido-2-deoxy-D-glucopyranosylidene)amino-N-phenylcarbamate, in the presence of normal glucose increased Sp1 O-GlcNAcylation and PAI-1 mRNA and protein levels. These findings demonstrate for the first time that among the pathways served by the HBP, O-GlcNAcylation, is obligatory for HG-induced PAI-1 gene expression and Sp1 transcriptional activation in mesangial cells.
Introduction
PROLONGED HYPERGLYCEMIA plays an essential role in the development of diabetic nephropathy and other microvascular complications of diabetes (1, 2, 3). The adverse effects of high glucose (HG) have been proposed to be due to a number of biochemical alterations, including increased reactive oxygen species, increased polyol pathway activity, activation of protein kinase C, formation of advanced glycation end-products, and, potentially, increased flux through the hexosamine biosynthetic pathway (HBP) (2, 3). Approximately 2–5% of glucose normally enters the HBP, of which the first and rate-limiting enzyme, glutamine:fructose-6-phosphate amidotransferase, catalyzes the formation of glucosamine-6-phosphate from fructose-6-phosphate and glutamine (4, 5). Further modifications generate the end-product, UDP-N-acetylglucosamine, a substrate for a number of processes, including O-linked glycosylation of intracellular proteins, complex N- and O-linked protein glycosylation of membrane and secreted proteins, and the synthesis of glycolipids, proteoglycans, and glycosylphosphatidylinositol anchors (6). HG increases flux through the HBP in vitro and in vivo in both insulin-responsive cells and in noninsulin-responsive cells, such as mesangial, endothelial, and smooth muscle cells (7, 8, 9, 10, 11, 12). The HBP was initially proposed to be a mediator of insulin resistance and was found to contribute to the HG-induced expression of leptin, plasminogen activator inhibitor-1 (PAI-1), and other genes in muscle and adipose tissue (5, 13, 14, 15, 16, 17, 18). More recently, HG-stimulated HBP flux has been postulated to be involved in the development of diabetic nephropathy (2, 7, 9, 10, 19, 20, 21, 22, 23). This hypothesis is based on the ability of HBP flux to stimulate the expression of prosclerotic genes in mesangial cells in vitro, such as transforming growth factor-1, PAI-1, fibronectin, and laminin (7, 9, 10, 19, 20, 21, 22, 23). Such changes would serve to accelerate glomerular extracellular matrix accumulation and mesangial expansion in vivo, which are characteristic of diabetic nephropathy (24).
Enzymatic intracellular O-glycosylation (O-GlcNAcylation), the addition of single O-linked -N-acetylglucosamine (O-GlcNAc) monosaccharides to serine or threonine residues on cytosolic and nuclear proteins, is one of the processes coupled to the HBP (4). HG increases protein O-GlcNAcylation in a variety of cell types, including mesangial cells (7, 8, 10, 11, 12). O-GlcNAcylation is concentrated in the nucleus, where it has been identified on transcription factors, such as Sp1, c-myc, YY1, and CRE-binding protein, as well as other nuclear proteins (4). Certain cytosolic proteins have also been reported to be subject to O-GlcNAcylation, including endothelial nitric oxide synthase (8), insulin receptor substrate-1 (16), glycogen synthase (16), the Rpt2 ATPase subunit of the 26S proteasome (25), and heat shock protein 70 (4). Unlike traditional N-linked glycosylation of extracellular proteins, O-GlcNAcylation is dynamic, reversible, and responsive to extracellular stimuli (4). O-GlcNAc transferase (OGT), which catalyzes O-GlcNAcylation, has been cloned and determined to consist of 11.5 amino-terminal tetratricopeptide repeats (TPRs) and a carboxyl-terminal catalytic domain (26, 27). Conversely, the O-GlcNAc-selective deglycosylating enzyme, neutral cytosolic -N acetylglucosaminidase (O-GlcNAcase), removes O-GlcNAc from proteins (28).
HBP-regulated gene expression has been documented by treating cells with glucosamine or glutamine and by overexpression of glutamine:fructose-6-phosphate amidotransferase, which increase cellular UDP-N-acetylglucosamine pools in the absence of HG (9, 10, 11, 14, 17, 18, 19, 20, 21, 22, 23). Furthermore, reversal of HG-induced gene expression with antisense RNA or antisense oligonucleotides specific for glutamine:fructose-6-phosphate amidotransferase or through the use of pharmacological inhibitors of glutamine:fructose-6-phosphate amidotransferase, such as 6-diazo-5-oxonorleucine and azaserine, support the link between the HBP and increased gene expression (7, 9, 11, 19, 20, 21, 22, 23, 29). Yet, the molecular mechanisms governing HBP-induced gene expression are poorly understood. To elucidate these mechanisms, it is essential to first establish which aspect of the HBP regulates gene expression in response to altered HBP flux. O-GlcNAcylation has been correlated with HBP-stimulated transcription and is an attractive candidate to mediate this effect of the HBP (4, 7, 11, 23, 29). However, the reagents used in the studies of HBP-induced gene expression affect overall HBP flux and are not entirely specific for O-GlcNAcylation. Therefore, these would have modulated other consequences of increased HBP flux, such as the development of oxidative stress (30, 31), N-glycosylation (32), glucosamine-6-phosphate accumulation (31, 33), or ganglioside synthesis (34), which have the potential to alter gene expression through changes in intracellular signaling. For example, an enzyme participating in N-glycosylation, 1,6 N-acetylglucosaminyltransferase V, was found to be rate limiting for cytokine signaling (32). Synthesis of the glycolipids, GM1 and GM2 ganglioside, in response to glucosamine has been linked to mesangial cell hypertrophy and cell cycle arrest (34). High levels of glucosamine-6-phosphate generated in response to glucosamine inhibit glucose-6-phosphate dehydrogenase, thereby depleting NADPH and enhancing oxidative stress (30, 31, 33). Thus, a causal relationship between O-GlcNAcylation and increased gene expression has not been established.
Elevated PAI-1 gene expression is involved in the development of diabetic nephropathy and atherosclerosis associated with diabetes (35, 36). Using the PAI-1 gene as a model for HBP-mediated gene expression, we previously found that Sp1 binding sites in the PAI-1 promoter were targets of HG- and glucosamine-induced HBP flux in mesangial cells (19). Increased HBP flux was accompanied by elevated Sp1 transcriptional activation, measured with an Sp1-GAL4 fusion protein (22). To address the role of O-GlcNAcylation in these effects of HG, several approaches were used in this study to specifically target this posttranslational modification. Thus, we demonstrate that O-GlcNAcylation is indispensable for HG-induced PAI-1 gene expression, PAI-1 promoter activation, and Sp1 transcriptional activation.
Materials and Methods
Cell culture
Primary rat mesangial cells were isolated from Sprague-Dawley rat kidney glomeruli, characterized, and grown as described previously (19, 22). For HG experiments, mesangial cells were grown in DMEM Base (Sigma, St. Louis, MO) with 2 mM L-glutamine, and either 1 (control) or 15 (HG) mM glucose for 96 h. Mesangial cells were generally used between passages 15–20, with additional studies performed in passage 8–12 cells comparing 5 with 25 mM glucose. O-(2-acetamido-2-deoxy-D-glucopyranosylidene)amino-N-phenylcarbamate (PUGNAc) was purchased from Toronto Research Chemicals (Toronto, Canada).
Plasmids
Plasmid expression vectors for human O-GlcNAcase and Sp1-GAL4 (amino acids 83–778 of Sp1 fused to Gal4 amino acids 1–147), and the plasmids, pCMV-Luc and pPAI-Luc (which contain nucleotides –740 to +44 of the human PAI-1 promoter fused to a luciferase reporter gene), were described previously (19, 22, 28). To produce a dominant-negative OGT (DNOGT) mutant, nucleotides 170-2212 of human nucleocytoplasmic OGT (GenBank accession no. BC014434), consisting of the TPR domain without the catalytic domain, were amplified from a cDNA template (Invitrogen, Carlsbad, CA) by PCR with pfu turbo DNA polymerase (Stratagene, La Jolla, CA), ligated between the BamHI and HindIII sites of pcDNA3MycHis (Invitrogen), and verified by sequencing. A luciferase reporter gene, pFR-Luc, driven by five GAL4 binding sites was obtained from Stratagene. The expression vector for green fluorescent protein (GFP), pEGFP-N1, was from BD Biosciences Clontech (Palo Alto, CA).
Transient expression of DNOGT and O-GlcNAcase
Mesangial cells were grown in 10-cm dishes, and after 24 h, the cells were transfected at approximately 40% confluence with 1.1 μg plasmid DNA expression vector for DNOGT or O-GlcNAcase, or the empty vector pcDNA3, using Effectene (Qiagen, Valencia, CA) as specified by the manufacturer.
Immunofluorescence
Mesangial cells were seeded on coverslips in six-well plates. For the PAI-1 studies, the cells were transiently transfected after 24 h with an expression vector for GFP, pEGFP-N1, and either O-GlcNAcase or the empty vector, pcDNA3, as described above. The cells were then incubated for an additional 72 h in either 5 or 25 mM glucose and transferred to 0.1% serum during the final 48 h. In additional experiments, the cells were stimulated with 10% serum for 24 h. For the Sp1 studies, the mesangial cells were incubated in 1 or 15 mM glucose for 96 h and transferred to 0.1% serum during the final 48 h. In both cases, the cells were washed with PBS, fixed with 4% formaldehyde for 10 min, permeabilized in 100% methanol at –20 C for 10 min, and blocked in PBS containing 1% goat serum and 0.1% BSA for Sp1 or in 2.5% BSA for PAI-1 for 45 min. The cells were incubated for 1.5 h with either polyclonal Sp1 antibodies, 1/100 (sc-59, Santa Cruz Biochemicals, Santa Cruz, CA), or for the PAI-1 studies, monoclonal GFP antibodies, 1/100 (sc-9996, Santa Cruz Biochemicals) and rat PAI-1 polyclonal antibodies, 1/100 (no. 1062, American Diagnostica, Greenwich, CT). After washing, the cells were incubated with secondary antibodies for 1 h, 1/300 goat antirabbit IgG-Texas Red (Jackson ImmunoResearch, West Grove, PA) for the Sp1 studies or for the PAI-1 studies, and 1/300 goat antirabbit IgG-Texas Red and 1/300 goat antimouse IgG-fluorescein isothiocyanate (Santa Cruz Biochemicals). The coverslips were then mounted onto glass slides with fluorescence mounting media (DakoCytomation, Mississauga, Ontario, Canada). In some experiments, the nuclei were stained with 1 μg/ml 4',6'-diamidino-2-phenylindole for 15 min. Images were captured with a Nikon Eclipse microscope (Nikon, Melville, NY) equipped for epifluorescence and with a monochrome charge-coupled device camera, using SimplePCI software (Compix, Cranberry Township, PA). The images were colorized with SimplePCI and quantified with Scion Image software (Scion Corp., Frederick, MD), which is a version of NIH Image for the Windows operating system. PAI-1 and Sp1 staining were assessed in the cytosol and nucleus, respectively. No staining was observed when boiled Sp1 or PAI antibodies were used (data not shown).
Sp1 immunoprecipitation and Western blotting
To analyze Sp1 O-GlcNAcylation, mesangial cells were lysed in a buffer consisting of 50 mM HEPES (pH 7.9), 150 mM NaCl, 2 mM EDTA, 10% glycerol, 0.1% SDS, 1% NP-40, 0.5% sodium deoxycholate, and protease inhibitor mixture (Roche Applied Science, Indianapolis, IN). Sp1 was immunoprecipitated by rotating cell extracts containing equal amounts of protein (300 μg) with 2.4 μg polyclonal Sp1 antibodies and 90 μl protein G PLUS-agarose beads (Santa Cruz Biochemicals). Western blotting was carried out as previously described (22) using RL-2 O-GlcNAc antibodies, 1/1000 (Affinity BioReagents, Golden, CO), or Sp1 antibodies, 1/1000 (sc-59, Santa Cruz Biochemicals). The resulting x-ray films were scanned and quantified using Scion Image software.
Semiquantitative RT-PCR
Total RNA was isolated with TRIzol Reagent (Invitrogen), and 1.5 μg RNA was reverse transcribed into cDNA with a First Strand cDNA Synthesis kit (Fermentas, Hanover, MD) using random hexamers according to the manufacturer’s protocol. Subsequently, a 2-μl aliquot of cDNA was amplified by PCR in a total volume of 50 μl containing 10 mM Tris (pH 8), 50 mM KCl, 0.08% NP-40, 0.2 mM dNTP, 1.5 mM MgCl2, 8 ng of each primer, and 5 U recombinant Taq DNA polymerase (Fermentas). For PAI-1 and OGT, parallel control PCRs were carried out with primers for 18S rRNA. PCR was performed with cycles consisting of: denaturation at 94 C for 0.5 min, annealing at the temperature indicated below for 1 min, and primer extension at 72 C for 2 min. All cycle numbers were in the linear range. The primers used were: rat PAI-1, sense, 5'-ACCCTCAGCATGTTCATTGC-3' and antisense, 5'-CTCGTTCACCTCGATCTTGAC-3'; 18S rRNA, sense, 5'-CGGCTACCACATCCAAGGAA-3' and antisense, 5'-GCTGGAATTACCGCGGCT-3'; and rat OGT 3' region, sense, 5'-GCCTGTTGAAGTCACTGAGTCAGCCT-3' and antisense, 5'-ATCACCCATCTGCAACACAGTACTATAC-3'. DNA oligonucleotides were synthesized by The Centre for Applied Genomics (The Hospital for Sick Children, Toronto, Canada). The PCR parameters (number of cycles, annealing temperature, and product size) were: rat PAI-1, 29 cycles, 61 C, and 297 bp; 18S rRNA, 21 cycles, 60 C, and 187 bp; and rat OGT 3' region, 30 cycles, 60 C, and 145 bp. Both the PAI-1 and OGT primers crossed exon boundaries. The PCR products were fractionated on a 1.3% agarose gel, stained with ethidium bromide, and visualized with UV light and a charge-coupled device camera using a Gel Doc 2000 instrument (Bio-Rad, Hercules, CA). The resulting TIFF images were quantified with Scion Image software, and the data were normalized to 18S rRNA levels.
Reporter gene analysis
Luciferase reporter gene assays were performed as described previously (19, 22). To take into account transfection efficiency differences between plasmids (such as between O-GlcNAcase and pcDNA3), relative luciferase activity was adjusted based on the luciferase activity obtained when these plasmids were cotransfected in parallel with a luciferase reporter gene driven by a cytomegalovirus promoter, pCMV-Luc. All transfections were repeated at least three times and with different DNA preparations.
EMSA
Mesangial cells were exposed to 5 or 25 mM glucose for 4 d. Nuclear extracts were prepared and EMSAs performed as previously described (19) using the human PAI-1 promoter –85 to –63 Sp1 binding site (sense, 5'-CAGTGAGTGGGTGGGGCTGGAAC-3'; antisense, 5'-GTTCCAGCCCCACCCACTCACTG-3'). All buffers included 50 mM N-acetylglucosamine and 100 nM okadaic acid to prevent deglycosylation and dephosphorylation, respectively. After drying, the gels were exposed to x-ray film, scanned, and quantified with Scion Image software.
PAI-1 ELISA
For determination of PAI-1 protein content in conditioned media, mesangial cells were seeded in 60-mm dishes in DMEM base, 14% fetal bovine serum, 2 mM L-glutamine, and the indicated concentration of glucose for 24 h. This was replaced with the same medium supplemented with 5% fetal bovine serum for 24 h, 0.1% fetal bovine serum for 24 h, and serum-free medium supplemented with 20 mM HEPES (pH 7.4), 2 mM L-glutamine, and 24 mM NaHCO3 for the final 24 h. After centrifugation at 1000 rpm to remove debris, PAI-1 protein was detected in the conditioned medium supernatant with a rat PAI-1 sandwich ELISA kit (American Diagnostica). Results were normalized to the total cellular protein content of each dish.
RNA interference (RNAi)
To reduce OGT expression, a small interfering RNA (siRNA) was targeted to the sequence, 5'-AAAGGGAACTAGATAACATGC-3', corresponding to nucleotides 3479–3499 of the 3' region of rat OGT (GenBank accession no. U76557). This siRNA was prepared by in vitro transcription with a Silencer siRNA synthesis kit (Ambion, Austin, TX) according to the manufacturer’s instructions. The sequence of the DNA oligonucleotide used against OGT was sense, 5'-AAAGGGAACTAGATAACATGCCCTGTCTC-3'. For the control scrambled siRNA, the sequence was sense, 5'-AATCAAGATAGAGCGAACATGCCTGTCTC-3'. A BLAST search of GenBank was carried out to avoid exact matches with other known sequences. The siRNAs were transfected at a final concentration of 20 nM with 48 μl GeneEraser siRNA transfection reagent (Stratagene) according to the manufacturer’s instructions. After 72 h, the cells were used to prepare RNA or cell lysates. OGT antibody, used at a 1/4000 dilution in Western blots, was previously described (26).
Statistics
Results are presented as mean ± SE of at least three independent experiments. Differences between groups were assessed by Student’s t test, with P < 0.05 considered to be statistically significant. Where indicated, results were analyzed by ANOVA with Newman-Keuls post hoc comparisons of means. Statistics were determined with Statistica software (Statsoft, Tulsa, OK).
Results
Dominant-negative OGT interferes with HG-induced Sp1 O-GlcNAcylation and PAI-1 gene expression in mesangial cells
The role of O-GlcNAcylation in HG-stimulated PAI-1 gene expression was examined in rat glomerular mesangial cells. We compared the effect of exposure to 15 mM glucose (HG) for 96 h to 1 mM glucose (control) because the glucose dose-response curve is shifted to lower concentrations in the mesangial cells studied, as previously demonstrated (19). This phenomenon, which has been described in other cultured cell lines (12, 37), is due to high-level glucose transporter 1 expression (data not shown) and/or altered hexokinase activity that results in high basal glucose uptake. Data were also obtained in earlier passage mesangial cells using 5 mM glucose in the control group (see below).
To inhibit O-GlcNAcylation, a nucleocytoplasmic OGT (38) mutant that included the amino-terminal TPRs, but lacked most of the carboxy-terminal catalytic domain, was transiently transfected. This construct was designed to impair OGT function in a dominant-negative manner by competing with wild-type OGT for trimerization, targeting, and substrate recognition (4). An analogous construct containing the TPRs from the Arabidopsis OGT homologue, SPINDLY, exerts dominant-negative effects in Arabidopsis (39). Preliminary experiments with GFP revealed a transient transfection efficiency of approximately 65% in the mesangial cells using Effectene (data not shown). Expression of DNOGT was documented by immunoblotting with an antibody directed against the OGT myc-epitope tag. A band with an approximate molecular mass of 68 kDa was identified in DNOGT-transfected cells but not in cells transfected with pcDNA3 (Fig. 1A). Sp1 O-GlcNAcylation, detected by immunoprecipitating Sp1 and immunoblotting with monoclonal RL-2 O-GlcNAc antibodies, was used as a marker for cellular O-GlcNAcylation. This and other Western blots presented below were quantified by scanning the x-ray films and using Scion Image software, giving a semiquantitative assessment of protein expression. As depicted in Fig. 1B, mesangial cells incubated in HG displayed increased Sp1 O-GlcNAcylation, 2.9 ± 0.8-fold above control (P < 0.05, n = 3), and this was diminished to 83 ± 42% of basal levels (P < 0.05 compared with HG, n = 3) by transfected DNOGT.
To determine whether endogenous PAI-1 mRNA levels are regulated by O-GlcNAcylation, PAI-1 mRNA levels were measured by semiquantitative RT-PCR using 18S rRNA amplified in parallel as an internal control. In pcDNA3-transfected cells, PAI-1 mRNA levels were elevated 2.1-fold by a 96-h treatment with HG (Fig. 1C, P < 0.001). Transient expression of DNOGT did not reduce baseline PAI-1 mRNA levels significantly but inhibited the increase in response to HG (Fig. 1C).
Overexpression of O-GlcNAcase blocks HG-induced increases in PAI-1 gene expression, promoter activity, and protein levels, and Sp1 transcriptional activation in mesangial cells
The inhibitory effects of DNOGT suggested that O-GlcNAcylation was required for stimulation of PAI-1 expression by HG. However, another possibility could have been that OGT interfered with gene expression by protein-protein interactions via the overexpressed TPR domain (4). Thus, O-GlcNAcase, which has previously been demonstrated to diminish O-GlcNAcylation when overexpressed, was also transiently transfected (28). As shown in Fig. 2A, in the presence of HG Sp1 O-GlcNAcylation was reduced by O-GlcNAcase to 40 ± 13% of the levels in pcDNA3-transfected cells (P < 0.01, n = 3). Transfected O-GlcNAcase did not significantly reduce baseline PAI-1 mRNA levels (Fig. 2B). At the same time, HG failed to increase PAI-1 mRNA levels when O-GlcNAcase was overexpressed (Fig. 2B).
Next, the influence of O-GlcNAcylation on PAI-1 gene promoter activity was assessed by transiently transfecting a luciferase reporter gene driven by a (–740 to +44) PAI-1 promoter (19, 22). O-GlcNAcase had no effect on baseline PAI-1 promoter activity. A 1.8-fold increase in PAI-1 promoter activity was observed in HG-treated cells cotransfected with pcDNA3 (P < 0.001, Fig. 2C), whereas PAI-1 promoter activity was refractory to stimulation by HG in cells cotransfected with O-GlcNAcase. PAI-1 promoter activation by HG was specific because luciferase activity driven by a cytomegalovirus promoter was not affected by HG (1 ± 0.1-fold of control, data not shown). Sp1 transcriptional activity was evaluated by cotransfecting Sp1-GAL4, an expression vector for Sp1 (amino acids 83–778) fused to the DNA-binding domain of the yeast transcription factor, GAL4, and a luciferase reporter gene driven by multiple GAL4 binding sites, pFR-Luc (22). HG-stimulated Sp1-GAL4 activity was substantially greater (2.5-fold, P < 0.01) than control in pcDNA3-transfected cells (Fig. 2D). Significantly, cotransfected O-GlcNAcase diminished baseline Sp1-GAL4 activity and rendered Sp1-GAL4 activation unresponsive to HG (Fig. 2D). Similar results were obtained with DNOGT (data not shown).
To confirm that HG-stimulated PAI-1 gene expression is dependent on O-GlcNAcylation in mesangial cells cultured in a physiologic baseline glucose concentration, the effect of 25 mM glucose was compared with that of 5 mM glucose in passage 8–12 rat mesangial cells. These cells were transiently transfected with GFP to mark the transfected cells and with either O-GlcNAcase or pcDNA3. PAI-1 and GFP were detected by immunofluorescence staining of mesangial cells with specific antibodies, and PAI-1 levels were quantified in the cytosol of GFP-positive cells. The percentage of GFP-positive cells, 3%, indicated a lower transfection efficiency in these slower growing, lower passage mesangial cells. Accordingly, PAI-1 protein expression was examined in individual transfected cells. Figure 3 shows that 25 mM glucose triggered a 2.03- ± 0.09-fold (P < 0.001) increase in PAI-1 protein staining compared with 5 mM glucose in cells transfected with pcDNA3, whereas O-GlcNAcase suppressed this increase (1.04- ± 0.05-fold, n = 4) without any effect on baseline PAI-1 protein levels. In contrast, PAI-1 protein staining in mesangial cells exposed to 10% serum was not reduced by transiently transfected O-GlcNAcase (Fig. 4).
Silencing OGT expression with RNAi reduces PAI-1 mRNA levels in mesangial cells
Additional evidence that O-GlcNAcylation regulates PAI-1 gene expression in mesangial cells was obtained using RNAi to diminish endogenous OGT expression. A siRNA targeted to OGT was transfected into mesangial cells at a low concentration, 20 nM, to minimize off-target gene regulation (40). After 72 h, OGT siRNA significantly reduced OGT mRNA levels to 37 ± 0.07% of control (P < 0.001, n = 3) (Fig. 5A), decreased OGT protein expression (Fig. 5B), and concomitantly decreased Sp1 O-GlcNAcylation (Fig. 5C) compared with cells transfected with a scrambled siRNA. Consistent with a positive role for O-GlcNAc in PAI-1 gene expression, reduction of OGT levels by RNAi diminished PAI-1 mRNA abundance in mesangial cells grown under HG conditions to 41 ± 9% (P < 0.01, n = 3) of the levels in cells transfected with a scrambled siRNA (Fig. 5D).
HG does not increase nuclear Sp1 levels or DNA binding in mesangial cells
Previous studies of insulin-stimulated calmodulin gene expression and glutamine-stimulated argininosuccinate synthetase gene expression revealed that Sp1 O-GlcNAcylation correlated with and may have induced Sp1 nuclear translocation (41, 42). To ascertain whether HG affects Sp1 nuclear levels in mesangial cells, Sp1 immunofluorescence staining was performed. As illustrated in Fig. 6A and confirmed by 4',6'-diamidino-2-phenylindole staining (data not shown), Sp1 was primarily localized to the nucleus. The intensity of Sp1 nuclear or cytosolic staining was not elevated in HG-treated mesangial cells compared with control cells in three experiments (Fig. 6B). Similar results were obtained by immunoblotting for Sp1 in nuclear extracts (data not shown).
Based on reported in vitro data, O-GlcNAcylation may up-regulate Sp1 DNA binding activity (43). Therefore, EMSAs were carried out to assess Sp1 binding to an oligonucleotide containing one of the PAI-1 promoter Sp1 sites (–85 to –63). A previous study confirmed Sp1 binding to this PAI-1 promoter site in mesangial cells by supershifting of the Sp1 band with Sp1-specific antibodies (19). Binding to this PAI-1 oligonucleotide was specific because it was eliminated by a 50-fold excess of unlabeled oligonucleotide (Fig. 7B). However, the amount of Sp1-DNA complexes formed were similar in nuclear extracts prepared from mesangial cells incubated with either 25 or 5 mM glucose (Fig. 7, A and C).
Inhibition of O-GlcNAcase with PUGNAc up-regulates PAI-1 mRNA and protein levels
To determine whether elevated O-GlcNAc levels increase PAI-1 gene expression in the absence of HG, mesangial cells were treated with an inhibitor of O-GlcNAcase, PUGNAc. This inhibitor causes a generalized increase in steady-state O-GlcNAc-modified protein levels, reflecting the dynamic nature of O-GlcNAcylation (15, 44). Exposing mesangial cells to 100 μM PUGNAc for 24 h resulted in elevated Sp1 O-GlcNAcylation (Fig. 8A) and a 2.5-fold increase in PAI-1 mRNA levels (P < 0.01, Fig. 8B). There was no further increase in PAI-1 mRNA levels with the combination of PUGNAc and HG. To establish the impact of PUGNAc on PAI-1 protein levels, PAI-1 ELISAs were performed in conditioned media. HG and phorbol 12-myristate 13-acetate, used as positive controls, both increased PAI-1 protein concentration in conditioned media 2.1-fold above control, whereas 100 μM PUGNAc (for 24 h) elicited a similar, 2.3-fold increase (P < 0.05, Fig. 8C). Absolute PUGNAc-stimulated PAI-1 levels were 13.8 ± 1.7 ng/ml.
Discussion
In this study, we investigated the role of O-GlcNAcylation, one of the consequences of enhanced HBP flux, in HG-induced PAI-1 gene expression in mesangial cells. Suppression of O-GlcNAcylation by overexpression of DNOGT or O-GlcNAcase prevented HG-induced increases in PAI-1 mRNA levels and PAI-1 promoter activity. Reduced PAI-1 mRNA levels in the presence of HG was also demonstrated when OGT was depleted by RNAi. Sp1 transcriptional activation by HG, previously shown to be associated with PAI-1 promoter activation (22) and assessed with Sp1-GAL4 and pFR-Luc, was also attenuated by O-GlcNAcase. Furthermore, the ability of O-GlcNAcase to block HG-induced PAI-1 protein expression was demonstrated in early passage mesangial cells that were incubated in 5 and 25 mM glucose. Finally, treatment of mesangial cells with the O-GlcNAcase inhibitor, PUGNAc, led to increased Sp1 O-GlcNAcylation, PAI-1 mRNA levels, and PAI-1 protein levels in the presence of normal (basal) glucose concentrations. Taken together, these observations demonstrate for the first time that one aspect of the HBP, namely O-GlcNAcylation, is essential for HG-stimulated PAI-1 gene expression and Sp1 transcriptional activation in mesangial cells.
Overexpression of DNOGT decreased HG-stimulated Sp1 O-GlcNAcylation to a small extent below that detected in control cells not treated with HG (P = not significant) (Fig. 1). This probably reflects some reduction of basal O-GlcNAcylation, possibly in the context of greater transfection efficiency than determined in preliminary studies. It should also be noted that baseline O-GlcNAcylation is determined in large part by the rate of glucose uptake through glucose transporter 1 in each particular set of mesangial cells studied. Consequently, some decrease in basal Sp1 O-GlcNAcylation upon overexpression of DNOGT or O-GlcNAcase would not be unexpected in higher passage mesangial cells that tend to have high basal glucose uptakes. Nonetheless, basal PAI-1 mRNA levels did not decrease significantly as a consequence of DNOGT or O-GlcNAcase overexpression (Figs. 1 and 2). Similarly, no change in baseline PAI-1 protein expression was seen in the immunofluorescence study when O-GlcNAcase was overexpressed (Fig. 3). Furthermore, O-GlcNAcase did not impair serum-stimulated PAI-1 protein expression (Fig. 4). Thus, the ability of O-GlcNAcase to block HG-induced PAI-1 expression does not appear to be due to a nonspecific inhibition of basal cellular function or of PAI-1 response. A positive role of O-GlcNAcylation in the stimulation of PAI-1 expression by HG is also supported by the finding that PAI-1 protein and mRNA expression were increased in the presence of PUGNAc, which acts by raising O-GlcNAc levels above baseline. HG-stimulated Sp1-GAL4 activity was completely abolished by O-GlcNAcase overexpression, and even basal activity was somewhat suppressed (Fig. 2). These effects likely represent the high sensitivity of the artificial chimeric transcription factor, Sp1-GAL4, to Spl function, and support the notion that Sp1 is positively regulated by cellular O-GlcNAc levels in mesangial cells.
In addition to the positive effects of O-GlcNAcylation on gene expression found in the present experiments, O-GlcNAcylation can exert negative effects on Sp1-dependent transcription. Previous reports indicate that O-GlcNAcylation of Ser 484 in the glutamine-rich Sp1 B transcriptional activation domain disrupts its interaction with the transcription cofactor, Drosophila TATA-binding protein-associated factor II 110 (dTAFII110, human equivalent hTAFII130) (45). A GAL4 construct fused to this Sp1 domain was less active in an in vitro transcription assay when O-glycosylated (45). Overexpression of OGT inhibited the activity of a reporter gene driven by multiple Sp1 sites, in part due to the recruitment of the transcriptional repressor, mSin3A, by OGT (46). In contrast, we have found that O-GlcNAcylation is required for HG-induced PAI-1 gene expression and Sp1 transcriptional activation in mesangial cells. A number of differences between the present experiments and the previous reports may account for these different results. First, the present experiments argue that an overall increase in mesangial cell O-GlcNAcylation is necessary for HG-induced PAI-1 expression, but, it is important to note, do not demonstrate directly the role played by O-GlcNAcylation of Sp1 itself. Indeed, the O-GlcNAcylation status of Sp1 Ser 484 in the current experiments remains unknown. It is possible that in mesangial cells HG-induced O-GlcNAcylation of other Spl sites, or more likely other substrates, overcomes any negative effects of Sp1 Ser 484 O-GlcNAcylation. Second, we used various approaches to modify HG-induced O-GlcNAcylation but not overexpression of OGT as in the earlier reports. In preliminary experiments, we found that overexpression of full-length OGT did not increase Sp1 O-GlcNAcylation in mesangial cells (data not shown), perhaps because O-GlcNAcylation is limited primarily by substrate flux. The ability of full-length OGT to form complexes with other nuclear proteins, for example mSin3A (46), may cause OGT overexpression to affect transcription differently than HG and potentially in a manner independent of O-GlcNAcylation. Third, we assessed endogenous PAI-1 mRNA levels and the activity of the native PAI-1 promoter linked to a luciferase reporter rather than a reporter gene driven by a multimerized Sp1-binding site. Finally, O-GlcNAcylation or phosphorylation of Sp1 and its associated transcription cofactors may differ in mesangial cells compared with the previously used cell types, thereby modifying the overall impact of O-GlcNAcylation on transcription.
A number of mechanisms may explain the linkage of O-GlcNAcylation to PAI-1 gene transcription. Because HG did not increase Sp1 nuclear protein levels (Fig. 6), the present results are unlikely to be consequences of either O-GlcNAc-stimulated Sp1 nuclear translocation or O-GlcNAc-mediated inhibition of the proteasome (25), leading to Sp1 accumulation. Possibly, these latter effects only require basal levels of O-GlcNAcylation. The high basal levels of Sp1 observed in the nucleus in the present experiments are consistent with this concept and would have masked any effects of O-GlcNAc on Sp1 nuclear translocation. Similarly, the lack of effect of HG on Sp1 DNA binding in EMSAs (Fig. 7) argues against O-GlcNAc-induced Sp1 DNA binding to account for the observed effects of O-GlcNAcylation on transcription.
Because PAI-1 expression is known to be stimulated by transforming growth factor-, and the latter is up-regulated by increased HBP flux (10, 23, 47), it was possible that the O-GlcNAc effects could be indirect. However, our previous data argue against a role for transforming growth factor- in HBP-stimulated PAI-1 promoter activity under the current experimental conditions (19, 21).
The observation that HG-stimulated Sp1-GAL4 activity was blocked by cotransfection of O-GlcNAcase points to Sp1 and/or its coactivators as potential functional targets of O-GlcNAcylation. Theoretically, O-GlcNAcylation may augment Sp1 function by altering protein-protein interactions between Sp1 and its cofactors. A significant precedent is that O-GlcNAcylation of the transcription factor, Stat5a, increases its interaction with its cofactor, CRE-binding protein (48). Further work is necessary to investigate the modulation of Sp1 protein interactions by O-GlcNAcylation.
Finally, several studies have reported that flux through the HBP stimulates a plethora of signal transduction pathways, including protein kinase C and reactive oxygen species formation (8, 10, 15, 20, 22, 30, 31, 49, 50), although the role of O-GlcNAc in most of these effects is not clear. In this regard, it was previously demonstrated that overall HBP flux increased protein kinase C- and - kinase activity; however, protein kinase C O-GlcNAcylation was not detected (22).
In summary, these results demonstrate that O-GlcNAcylation plays an essential role in HG-induced PAI-1 gene expression and Sp1 transcriptional activation in mesangial cells. These findings indicate that the search for the relevant mechanism by which HG-mediated increased HBP flux stimulates gene expression should focus on O-GlcNAc-modified proteins. Furthermore, these data motivate in vivo experiments designed to establish the role of O-GlcNAcylation as the major effector of HBP flux in the regulation of PAI-1 and the expression of other genes in the diabetic kidney. These data support the concept that HG-stimulated O-GlcNAcylation acts in concert with other pathways, such as reactive oxygen species, protein kinase C, transforming growth factor-, and advanced glycosylation end-products, to augment prosclerotic gene expression in mesangial cells, leading to diabetic nephropathy.
Footnotes
This work was supported by grants from the Juvenile Diabetes Foundation International and by the Canadian Institutes of Health Research (Grant 88811).
First Published Online September 22, 2005
Abbreviations: DMSO, Dimethyl sulfoxide; DNOGT, dominant-negative OGT; GFP, green fluorescent protein; HBP, hexosamine biosynthetic pathway; HG, high glucose; O-GlcNAc, O-linked -N-acetylglucosamine; O-GlcNAcase, neutral cytosolic -N acetylglucosaminidase; O-GlcNAcylation, O-glycosylation; OGT, O-GlcNAc transferase; PAI, plasminogen activator inhibitor; PUGNAc, O-(2-acetamido-2-deoxy-D-glucopyranosylidene)amino-N-phenylcarbamate; RNAi, RNA interference; siRNA, small interfering RNA; TPR, tetratricopeptide repeat.
Accepted for publication September 14, 2005.
References
The Diabetes Control and Complications Trial Research Group 1993 The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 329:977–986
Brownlee M 2001 Biochemistry and molecular cell biology of diabetic complications. Nature 414:813–820
Sheetz MJ, King GL 2002 Molecular understanding of hyperglycemia’s adverse effects for diabetic complications. JAMA 288:2579–2588
Zachara NE, Hart GW 2004 O-GlcNAc a sensor of cellular state: the role of nucleocytoplasmic glycosylation in modulating cellular function in response to nutrition and stress. Biochim Biophys Acta 1673:13–28
McClain DA 2002 Hexosamines as mediators of nutrient sensing and regulation in diabetes. J Diabetic Compl 16:72–80
Boehmelt G, Wakeham A, Elia A, Sasaki T, Plyte S, Potter J, Yang Y, Tsang E, Ruland J, Iscove NN, Dennis JW, Mak TW 2000 Decreased UDP-GlcNAc levels abrogate proliferation control in EMeg32-deficient cells. EMBO J 19:5092–5104
Du XL, Edelstein D, Rossetti L, Fantus IG, Goldberg H, Ziyadeh F, Wu J, Brownlee M 2000 Hyperglycemia-induced mitochondrial superoxide overproduction activates the hexosamine pathway and induces plasminogen activator inhibitor-1 expression by increasing Sp1 glycosylation. Proc Natl Acad Sci USA 97:12222–12226
Du XL, Edelstein D, Dimmeler S, Ju O, Sui C, Brownlee M 2001 Hyperglycemia inhibits endothelial nitric oxide synthase activity by posttranslational modification at the Akt site. J Clin Invest 108:1341–1348
Kolm-Litty V, Sauer U, Nerlich A, Lehmann R, Schleicher ED 1998 High glucose-induced transforming growth factor 1 production is mediated by the hexosamine pathway in porcine glomerular mesangial cells. J Clin Invest 101:160–169
Burt DJ, Gruden G, Thomas SM, Tutt P, Dell’Anna C, Viberti GC, Gnudi L 2003 P38 mitogen-activated protein kinase mediates hexosamine-induced TGF1 mRNA expression in human mesangial cells. Diabetologia 46:531–537
Sayeski PP, Kudlow JE 1996 Glucose metabolism to glucosamine is necessary for glucose stimulation of transforming growth factor- gene transcription. J Biol Chem 271:15237–15243
Han I, Oh ES, Kudlow JE 2000 Responsiveness of the state of O-linked N-acetylglucosamine modification of nuclear pore protein p62 to the extracellular glucose concentration. Biochem J 350:109–114
Marshall S, Bacote V, Traxinger RR 1991 Discovery of a metabolic pathway mediating glucose-induced densenitization of the glucose transport system. Role of hexosamine biosynthesis in the induction of insulin resistance. J Biol Chem 266:4706–4712
Wang J, Liu R, Hawkins M, Barzilai N, Rossetti L 1998 A nutrient-sensing pathway regulates leptin gene expression in muscle and fat. Nature 393:684–688
Vosseller K, Wells L, Lane MD, Hart, GW 2002 Elevated nucleocytoplasmic glycosylation by O-GlcNAc results in insulin resistance associated with defects in Akt activation in 3T3–L1 adipocytes. Proc Natl Acad Sci USA 99:5313–5318
McClain DA, Lubas WA, Cooksey RC, Hazel M, Parker GJ, Love DC, Hanover JA 2002 Altered glycan-dependent signaling induces insulin resistance and hyperleptinemia. Proc Natl Acad Sci USA 99:10695–10699
Obici S, Wang J, Chowdury R, Feng Z, Siddhanta U, Morgan K, Rossetti L 2002 Identification of a biochemical link between energy intake and energy expenditure. J Clin Invest 109:1599–1605
Einstein FH, Atzmon G, Yang XM, Ma XH, Rincon M, Rudin E, Muzumdar R, Barzilai N 2005 Differential responses of visceral and subcutaneous fat depots to nutrients. Diabetes 54:672–678
Goldberg HJ, Scholey J, Fantus IG 2000 Glucosamine activates the plasminogen activator-inhibitor 1 gene promoter through Sp1 DNA binding sites in glomerular mesangial cells. Diabetes 49:863–871
Singh LP, Crook ED 2000 Hexosamine regulation of glucose-mediated laminin synthesis in mesangial cells involves protein kinases A and C. Am J Physiol 279:F646–F654
James LR, Fantus IG, Goldberg H, Ly H, Scholey JW 2000 Overexpression of glutamine:fructose-6-phosphate amidotransferase (GFAT) activates PAI promoter. Am J Physiol 279:F718–F727
Goldberg HJ, Whiteside CI, Fantus IG 2002 The hexosamine pathway regulates the plasminogen activator inhibitor-1 gene promoter and Sp1 transcriptional activation through protein kinase C- I and -. J Biol Chem 277:33833–33841
Weigert C, Brodbeck K, Sawadogo M, Haring HU, Schleicher ED 2004 Upstream stimulatory factor (USF) proteins induce human TGF-1 gene activation via the glucose-response element-1013/-1002 in mesangial cells: up-regulation of USF activity by the hexosamine biosynthetic pathway. J Biol Chem 279:15908–15915
Steffes MW, Bilous RW, Sutherland DE, Mauer SM 1992 Cell and matrix components of the glomerular mesangium in type I diabetes. Diabetes 41:679–684
Zhang F, Su K, Yang X, Bowe DB, Paterson AJ, Kudlow JE 2003 O-GlcNAc modification is an endogenous inhibitor of the proteasome. Cell 115:715–725
Kreppel LK, Blomberg MA, Hart GW 1997 Dynamic glycosylation of nuclear and cytosolic proteins. Cloning and characterization of a unique O-GlcNAc transferase with multiple tetratricopeptide repeats. J Biol Chem 272:9308–9315
Lubas WA, Frank DW, Krause M, Hanover JA 1997 O-linked GlcNac transferase is a conserved nucleocytoplasmic protein containing tetratricopeptide repeats. J Biol Chem 272:9316–9324
Wells L, Gao Y, Mahoney JA, Vosseller K, Chen C, Rosen A, Hart GW 2002 Dynamic O-glycosylation of nuclear and cytosolic proteins: further characterization of the nucleocytoplasmic -N-acetylglucosaminidase, O-GlcNAcase. J Biol Chem 277:1755–1761
James LR, Tang D, Ingram A, Ly H, Thai K, Cai L, Scholey JW 2002 Flux through the hexosamine pathway is a determinant of nuclear factor B-dependent promoter activation. Diabetes 51:1146–1155
Kaneto H, Xu G, Song KH, Suzuma K, Bonner-Weir S, Sharma A, Weir GC 2001 Activation of the hexosamine pathway leads to deterioration of pancreatic -cell function through the induction of oxidative stress. J Biol Chem 276:31099–31104
Wu G, Haynes TE, Li H, Yan, Meininger, CJ 2001 Glutamine metabolism to glucosamine is necessary for glutamine inhibition of endothelial nitric oxide synthesis. Biochem J 353:245–252
Partridge EA, Le Roy C, Di Guglielmo GM, Pawling J, Cheung P, Granovsky M, Nabi IR, Wrana JL, Dennis JW 2004 Regulation of cytokine receptors by Golgi N-glycan processing and endocytosis. Science 306:120–124
Marshall S, Nadeau O, Yamasaki K 2005 Glucosamine-induced activation of glycogen biosynthesis in isolated adipocytes. Evidence for a rapid allosteric control mechanism within the hexosamine biosynthesis pathway. J Biol Chem 280:11018–11024
Masson E, Wiernsperger N, Lagarde M, El Bawab S 2005 Glucosamine induces cell-cycle arrest and hypertrophy of mesangial cells: implication of gangliosides. Biochem J 388:537–544
Eddy AA 2002 Plasminogen activator inhibitor-1 and the kidney. Am J Physiol 283:F209–F220
Paueksakon P, Revelo MP, Ma LJ, Marcantoni C, Fogo AB 2002 Microangiopathic injury and augmented PAI-1 in human diabetic nephropathy. Kidney Int 61:2142–2148
Tan J, Yang HS, Patel MS 1998 Regulation of mammalian pyruvate dehydrogenase subunit gene expression by glucose in HepG2 cells. Biochem J 336:49–56
Love DC, Kochran J, Cathey RL, Shin, SH, Hanover JA 2002 Mitochondrial and nucleocytoplasmic targeting of O-linked GlcNAc transferase. J Cell Sci 116:647–654
Tseng TS, Swain SM, Olszewski NE 2001 Ectopic expression of the tetratricopeptide repeat domain of SPINDLY causes defects in gibberellin response. Plant Physiol 126:1250–1258
Semizarov, D, Frost L, Sarthy A, Kroeger P, Halbert DN, Fesik SW 2003 Specificity of short interfering RNA determined through gene expression signatures. Proc Natl Acad Sci USA 100:6347–6352.
Majumdar G, Harmon A, Candelaria R, Martinez-Hernandez A, Raghow R, Solomon SS 2003 O-glycosylation of Sp1 and transcriptional regulation of the calmodulin gene by insulin and glucagon. Am J Physiol 285:E584–E591
Brasse-Lagnel C, Fairand A, Lavoinne A, Husson A 2003 Glutamine stimulates argininosuccinate synthetase gene expression through cytosolic O-glycosylation of Sp1 in Caco-2 cells. J Biol Chem 278:52504–52510
Weigert C, Klopfer K, Kausch C, Brodbeck K, Stumvoll M, Haring HU, Schleicher ED 2003 Palmitate-induced activation of the hexosamine pathway in human myotubes: increased expression of glutamine:fructose-6-phosphate aminotransferase. Diabetes 52:650–656
Haltiwanger RS, Grove K, Philipsberg GA 1998 Modulation of O-linked N-acetylglucosamine levels on nuclear and cytoplasmic proteins in vivo using the peptide O-GlcNAc--N-acetylglucosaminidase inhibitor O-(2-Acetoamido-2-deoxy-D-glucopyranosylidene)amino-N-phenylcarbamate. J Biol Chem 273:3611–3617
Yang X, Su K, Roos MD, Chang Q, Paterson AJ, Kudlow JE 2001 O-linkage of N-acetylglucosamine to Sp1 activation domain inhibits its transcriptional capability. Proc Natl Acad Sci USA 98:6611–6616
Yang X, Zhang F, Kudlow JE 2002 Recruitment of O-GlcNAc transferase to promoters by corepressor mSin3A: coupling protein O-GlcNAcylation to transcriptional repression. Cell 110:69–80
Singh LP, Green K, Alexander M, Bassly S, Crook ED 2004 Hexosamines and TGF-1 use similar signaling pathways to mediate matrix protein synthesis in mesangial cells. Am J Physiol 286:F409–F416
Gewinner C, Hart G, Zachara N, Cole R, Beisenherz-Huss C, Groner B 2003 The coactivator of transcription CREB binding protein interacts preferentially with the glycosylated form of Stat5. J Biol Chem 279:3563–3572
Hsieh TJ, Fustier P, Zhang SL, Filep JG, Tang SS, Ingelfinger JR, Fantus IG, Hamet P, Chan JS 2003 High glucose stimulates angiotensinogen gene expression and cell hypertrophy via activation of the hexosamine biosynthesis pathway in rat kidney proximal tubular cells. Endocrinology 144:4338–4349
Kneass ZT, Marchase RB 2005 Protein O-GlcNAc modulates motility-associated signaling intermediates in neutrophils. J Biol Chem 280:14579–14585(Howard J. Goldberg, Catharine I. Whitesi)