Hexosamine-Induced Fibronectin Protein Synthesis in Mesangial Cells Is Associated With Increases in cAMP Responsive Element Binding (CREB) Phosphorylation and Nuclear CREB
The Involvement of Protein Kinases A and C
The Involvement of Protein Kinases A and C
Hyperglycemia-induced alterations in mesangial (MES) cell function and extracellular matrix protein accumulation are seen in diabetic glomerulopathy. Recent studies have demonstrated that some of the effects of high glucose (HG) on cellular metabolism are mediated by the hexosamine biosynthesis pathway (HBP), in which fructose-6-phosphate is converted to glucosamine 6-phosphate by the rate-liming enzyme glutamine:fructose-6-phosphate amidotransferase (GFA). In this study, we investigated the role of HBP on HG-stimulated fibronectin protein synthesis, a matrix component, in SV-40–transformed rat kidney MES cells. Treatment of MES cells with 25 mmol/l glucose (HG) for 48 h increases cellular fibronectin levels by two- to threefold on Western blots when compared with low glucose (5 mmol/l). Glucosamine (GlcN; 1.5 mmol/l), which enters the hexosamine pathway distal to GFA action, also increases fibronectin synthesis. Azaserine (AZA; 0.5 μmol/l), an inhibitor of GFA, blocks the HG- but not the GlcN-induced fibronectin synthesis. Fibronectin contains cAMP responsive element (CRE) consensus sequences in its promoter and the phosphorylation of CRE-binding protein (CREB) may regulate its expression. On Western blots, HG and GlcN stimulate two- to threefold the phosphorylation of CREB at Ser 133, whereas CREB protein content was unaltered by either HG or GlcN. In addition, nuclear CREB activity was increased by HG and GlcN on gel-shift assays using 32P-CRE oligonucleotides. AZA impeded the HG-enhanced CREB phosphorylation and CRE binding but had no effect on GlcN-mediated CREB phosphorylation and CRE binding. Pharmacologic inhibition of protein kinase C (PKC) and protein kinase A (PKA), which are involved in hexosamine-mediated matrix production, blocked the CREB phosphorylation and fibronectin synthesis seen in HG and GlcN conditions. We conclude that the effects of HG on fibronectin synthesis in the mesangium are mediated by the HBP possibly via hexosamine regulation of CREB and PKC/PKA signaling pathways. These results support the hypothesis that the HBP is a sensor and regulator of the actions of glucose in the kidney.
Diabetic nephropathy is characterized by the accumulation of extracellular matrix (ECM) proteins in the glomerulus and is represented morphologically by thickening and expansion of the glomerular basement membrane and the mesangium (1,2). Hyperglycemia is an important contributor to the development of diabetic nephropathy in both type 1 and type 2 diabetes (3). However, the mechanisms underlying the effects of chronic hyperglycemia on the kidney are not fully understood. Transforming growth factor β (TGF-β), protein kinase C (PKC), and protein kinase C (PKC) have been implicated in ECM production and the development of diabetic nephropathy (4,5,6,7). The addition of neutralizing antibodies against TGF-β in mesangial (MES) cells cultured in high-glucose (HG) medium blocks the synthesis of matrix proteins (5). Recent studies indicate that PKC is activated in diabetic glomeruli (5,6) and glomerular MES cells cultured under HG conditions probably via de novo synthesis of diacylglycerol (7,8,9,10,11). In addition, the stimulation of the cAMP-activated PKA system leads to transcriptional activation of type IV collagen in MES cells (12) and suggests a role for PKA in ECM production.
We previously demonstrated that some of the effects of HG on cellular metabolism are mediated by the hexosamine biosynthesis pathway (HBP), in which fructose-6-phosphate is converted to glucosamine-6-phosphate by the rate-liming enzyme glutamine:fructose-6-phosphate amidotransferase (GFA) (13). For example, chronic exposure of rat-1 fibroblasts to HG decreases both basal and insulin-stimulated glycogen synthase activity, and glucosamine (GlcN) or overexpression of GFA in these cells mimics the effects of HG (13,14,15). GlcN has also been shown to be more potent than glucose in the transcriptional regulation of TGF-α (16) and TGF-β (17). Moreover, transgenic mice that overexpress GFA in skeletal muscle and fat are insulin-resistant (18).
Recently, we showed that HG and GlcN increased laminin synthesis (19) and that PKC and PKA signaling pathways may participate in the hexosamine-induced synthesis of this ECM component. Hexosamine regulation of the ECM is not restricted to laminin, as we showed that fibronectin protein also is increased by HG and GlcN. The mechanisms whereby PKC and PKA mediate hexosamine-induced ECM synthesis are not known, but it is possible that transcriptional regulation of some gene(s) is involved. The transactivation of genes through the cAMP-regulated enhancer or cAMP responsive element (CRE) is thought to occur by the phosphorylation of the transcription factor CRE-binding protein (CREB) and its binding to CRE consensus sequences in the promoter region. In the present study, we investigated the role of HBP in HG- and GlcN-mediated fibronectin synthesis in SV-40–transformed rat kidney MES cells. The fibronectin gene contains CRE consensus sequences (20), and we therefore examined the role of the HBP on phosphorylation of CREB and nuclear CRE binding. The results demonstrate that both HG and GlcN increase fibronectin protein synthesis, and this increase is associated with parallel changes in the level of CREB phosphorylation in MES cells. The effects of hexosamines on fibronectin and CREB seem to be mediated by PKC and PKA.
RESEARCH DESIGN AND METHODS
Goat anti-rat fibronectin antibodies, Bis I, H-89, Calphostin C, H-6 hydrochloride, and Protein A agarose were purchased from Calbiochem-Novabiochem (San Diego, CA). Antiphosphorylated CREB antibodies and anti-CREB antibodies, which recognize both phosphorylated and nonphosphorylated CREB, were from Upstate Biotechnology (Lake Placid, NY). An enhanced chemiluminescence (ECL) detection system for Western Blotting was obtained from Amersham. The gel-shift assay system for determining the nuclear CRE-binding activity was purchased from Promega (Madison, WI). γ-32P[ATP] was purchased from ICN, and 3H-leucine was from Amersham Pharmacia Biotech. All other reagents and chemicals were of reagent or analytical grade.
SV-40–transformed rat kidney MES cells (SV-40 MES obtained from ATCC, Rockville, MD) were cultured in media containing Dulbecco’s modified Eagle’s medium and F-12 Nutrient Mixture (HAM) (3:1 ratio) with 10% fetal calf serum and 0.5 mg/ml gentamicin. Cells were routinely passaged at confluence every 4 days using 10-cm culture dishes. Approximately 50% confluent monolayers were incubated in the above medium supplemented with 2.25% fetal calf serum and the desired concentrations of glucose and GlcN for 48 h. To examine the effect of inhibition of GFA on HG- or GlcN-induced fibronectin synthesis, 0.5 μmol/l azaserine (AZA) was added along with sugars and was present throughout the time of incubation. AZA is a glutamine analog of glutamine and an inhibitor of GFA and other glutamine-sensitive enzymes and usually has no effect on GlcN action as it enters the HBP distal to GFA action (19). For the blockade of PKC or PKA activities, Bis-I (1 μmol/l) or H-89 (2 μmol/l) was added to the culture. At the end of the incubation, the dishes were rinsed twice with extraction buffer A (50 mmol/l β-glycerophosphate, pH 7.3, 1.5 mmol/l EGTA, 1 mmol/l dithiothreitol, 0.2 mmol/l Na orthovanadate, 1 mmol/l benzamidine, 10 μg/ml aprotonin, 20 μg/ml leupeptin, 1 mmol/l NaF, 0.5 μg/ml microcystine, and 2 μg/ml pepstatin A) and then harvested in 1 ml of the same buffer using a rubber policeman. The cells were centrifuged at 16,000g for 5 s, resuspended in 200 μl of extraction buffer A, immediately frozen in liquid nitrogen, and stored at −80°C until use. Cells were subsequently thawed, sonicated for 20 s, and centrifuged at 4°C for 10 min. The supernatant was removed as the cytosolic fraction. For obtaining nuclear fractions, the pellet was washed and resuspended in buffer A plus 1% (vol/vol) Triton X-100 and 400 mmol/l KCl, sonicated and centrifuged as above. The supernatant was collected as the nuclear fraction. Protein concentrations in cell extracts were determined by the method of Bradford using bovine serum albumin as the standard.
Cell extracts (30 μg protein) were applied on a 10% SDS-PAGE and blotted to a polyvinylidene difluoride filter membrane. The blot was blocked with 5% nonfat dry milk in 10 mmol/l Tris-HCl (pH 7.6) containing 150 mmol/l NaCl and 0.05% Tween 20 (buffer B) for 20 min. The filter was washed in Buffer B and incubated with antifibronectin (1:5,000 dilution) or antiphosphorylated CREB (1:2,000 dilution) antibodies at 4°C overnight with continuous shaking in buffer B containing 5% nonfat dried milk. The membrane was then washed with buffer B (5 min × four times) and incubated with appropriate horseradish peroxidase–conjugated secondary antibodies (1:3,000 dilution) at room temperature for 1.5 h. Immunoreactive bands were detected with the ECL system, and intensity of the bands was measured by a densitometer for quantitation.
Immunoprecipitation of 3H-leucine–labeled fibronectin.
Approximately 30–40% confluent MES monolayers were incubated as described above in medium supplemented with 2.25% fetal calf serum and the desired concentrations of glucose and GlcN for 48 h. Six hours before the harvest, 1 μCi/ml 3H-leucine was added to each culture. The dishes were washed with ice-cold phosphate-buffered saline, and cells were detached with pancreatin. After centrifugation, the cells were washed twice with cold phosphate-buffered saline to remove residual radiolabeled leucine. Finally, cells were resuspended in 250 μl of buffer A with 0.2% Tween 20, vortexed, and centrifuged. Protein contents were determined in the supernatant, and concentrations were adjusted to 0.5 mg/ml protein and incubated with preabsorbed Protein A agarose and antifibronectin complexes at 4°C overnight. The beads were collected by centrifugation, washed twice with Tris-buffered saline, resuspended in 100 μl of SDS sample buffer without the dye, and heated over boiling water for 2 min. Fifty microliters of the supernatant were counted in an LS 6500 Model Beckman Scintillation Counter.
The gel-shift, or electrophoretic mobility-shift, assay provides a simple and rapid method for detecting DNA-binding activity of proteins (21). We investigated the CREB activity of MES nuclear proteins by using a commercially available gel-shift assay kit from Promega. An 18-mer oligonucleotide containing CRE consensus sequence 5′-TGACGTCA-3′ was 5′-labeled with T4 polynucleotide kinase and [γ-32P]ATP, and the radiolabeled CRE were separated on a G-25 Sephadex Spin column (Boehringer Mannheim). After incubation at 30°C for 10 min with 32P-CRE, nuclear extracts (7.5 μg protein) were subjected to 6% nondenaturing polyacrylamide gels prepared with TBE gel formulation (supplied by the manufacturer). The gels were run at 100 V for ∼1 h until the Bromophenol blue dye front is three quarters down the gel, using 0.5× TBE running buffer. The gel was covered with Saran Wrap and exposed to X-ray films. Alternately, the gel was dried in Whatman 3 MM papers before exposure. For competitive and noncompetitive assays, cold CRE and SP-1, an unrelated oligonucleotide sequence, were used, respectively. In some experiments, antiphosphorylated CREB antibodies were included in the reaction to establish the specificity of reaction before subjecting to gel electrophoresis and exposed as above. We also synthesized a CRE-containing sequence from the rat fibronectin promoter −171 ACCTGACCCCGTGACGTCACCCGGACTCCGG −141. Single-strand oligonucleotides were annealed, gel-purified, and 5′-labeled with T4 polynucleotide kinase as described above. Electrophoretic mobility-shift assay was performed with nuclear extracts (6 μg protein) and ∼40,000 cpm of radiolabeled probe.
Results are expressed as means ± SE of the indicated number of experiments. Student’s t test was used to compare differences between cultures. A value of P < 0.05 was considered statistically significant.
HG-induced increases in fibronectin synthesis are mediated by the HBP.
We recently demonstrated that treatment of MES cells with HG (25 mmol/l) increases laminin synthesis, and the effect of HG is mimicked by low concentrations of GlcN (1.5 mmol/l) (19). To determine whether the hexosamine-mediated matrix protein synthesis in MES cells is specific for laminin, we also examined the effect of HG and GlcN on fibronectin, another ECM glycoprotein component in glomerular mesangium. On Western blots, 25 mmol/l glucose (HG) and 5 mmol/l glucose plus 1.5 mmol/l GlcN increased fibronectin synthesis in MES cells (Fig. 1A). A densitometric analysis of the Western blots revealed that the level of fibronectin in cellular extracts from cells that were treated with HG or GlcN for 48 h increased ∼2.45 ± 0.19 (P < 0.02 vs. low glucose [LG]) and 1.95 ± 0.29-fold (P < 0.03 vs. LG), respectively, when compared with cells that were cultured in LG (Fig. 1B). The effects of glucose are not due to osmolar stress, as 25 mmol/l l-glucose had no effect on fibronectin synthesis (Fig. 1C). Neither HG nor GlcN altered the level of α-actinin, a component of the cytoskeleton (Fig. 1D). Although fibronectin levels are increased in these conditions, it does not necessarily mean that synthesis of fibronectin is increased. Increased fibronectin synthesis was verified by the observation that incorporation of 3H-leucine into fibronectin was increased in HG- and GlcN-treated cells (Table 1).
To further support the role of the HBP in glucose regulation of fibronectin, we used an inhibitor of GFA. Cells were cultured in AZA with LG, HG, or GlcN for 48 h, and fibronectin levels were determined. In the presence of AZA, the effect of HG on fibronectin was decreased by 68%, whereas the effect of GlcN on fibronectin was not altered (Fig. 2). This suggests that the HBP is involved in HG-mediated fibronectin synthesis in MES cells. The physiologic effect of AZA on hexosamine flux was verified by high-performance liquid chromatography (15). Uridine-diphospho-N-acetyl-GlcN, a downstream product of the HBP, was increased more than twofold and approximately eightfold by HG and GlcN, respectively, when compared with LG. AZA completely abolished the HG- but not the GlcN-mediated increase in uridine-diphospho-N-acetyl-hexosamine levels (data not shown).
HG and GlcN increase CREB phosphorylation and nuclear CREB activity.
The fibronectin gene contains CRE elements in its promoter. Phosphorylation of CREB may increase the interaction and assembly of transcription factors at the CRE site and enhance fibronectin gene expression. Therefore, we determined the level of CREB and its phosphorylation state at Ser 133, the site responsible for the expression of CRE-regulated genes (20). Specific antibodies for nonphosphorylated and phosphorylated CREB were used in Western blots. As shown in Fig. 3A, exposure of MES cells to HG (25 mmol/l for 48 h) increases the extent of CREB phosphorylation approximately two- to threefold. GlcN (1.5 mmol/l) mimics the effect of glucose on CREB phosphorylation (Fig. 3A). The effect of HG but not of GlcN on CREB phosphorylation is abolished by the addition of AZA in the culture (Fig. 3), indicating a role of HBP in the regulation of CREB phosphorylation.
We further investigated the 32P-CREB activity of nuclear extracts from cells that were treated with HG or GlcN on gel-shift assays. After incubation at 30°C for 10 min with 32P-CRE, nuclear extracts were subjected to nondenaturing PAGE and exposed to X-ray films. Figure 4 shows 32P-CRE mobility shift after binding to nuclear proteins. In nuclear extracts of MES cells that were treated with HG or GlcN, the amount of 32P-CRE associated with proteins was increased 1.4 ± 0.17 (P < 0.03) and 1.9 ± 0.3 (P < 0.2), respectively, compared with controls at LG (Fig. 4A, lanes 1–3). The addition of a 10-fold excess of unlabeled CRE for competition abolished radioactivity associated with the nuclear proteins (Fig. 4A, lane 4) but not by an excess of cold SP-1 oligonucleotides, a noncompetitive DNA sequence specific for SP-1 transcription factor (Fig. 4A, lane 5). Again, AZA inhibited the effects of glucose (1.2 ± 0.4 vs. LG; NS) but not those of GlcN (2.3 ± 0.4 vs. LG; P < 0.01; Fig. 4A, lanes 6–8). The specificity of the binding of CREB to 32P-CRE in nuclear extracts was confirmed by a supershift with the addition of anti-CREB(p) antibodies (Fig. 4B, lanes 5 and 6). Similar results also were obtained when gel-shift assays were performed with a 32P-CRE oligonucleotide developed from the rat fibronectin promoter sequence (nucleotides −171 to −141; Fig. 4C).
Pharmacologic inhibition of PKC and PKA blocks HG- and GlcN-induced CREB phosphorylation and fibronectin synthesis.
Agents that increase PKC and PKA activities in MES cells are known to have an effect on matrix protein synthesis (12,22). Protein kinases that target CREB include PKC, PKA, and Ca2+/Calmodulin-dependent protein kinases (23,24,25). We previously showed that HG and GlcN stimulate PKC and PKA activities in MES cells (19). Therefore, we investigated the roles of PKC and PKA on HG- and GlcN-induced CREB phosphorylation and fibronectin synthesis. Cells were incubated with HG or GlcN in the absence or presence of the PKC inhibitor Bis-I or the PKA inhibitor H-89 for 48 h and harvested. Cell extracts were prepared and analyzed by Western blotting for phosphorylated and nonphosphorylated CREB in the nuclear fraction and for fibronectin content in the cytosolic fraction. As shown in Fig. 5A, HG and GlcN stimulate CREB phosphorylation approximately threefold, and the addition of 1 μmol/l Bis-I impedes the HG- and GlcN-induced CREB phosphorylation by 60% (P < 0.03, HG vs. HG + Bis-I) and 50% (P < 0.03, GlcN vs. GlcN + Bis-I), respectively. Similarly, the addition of 2 μmol/l H-89, an inhibitor of PKA, blocks the HG- and GlcN-induced CREB phosphorylation by 69% (P < 0.04, HG vs. HG + H-89) and 42% (P < 0.05, GlcN vs. GlcN + H-89), respectively. These alterations in CREB phosphorylation occurred without any significant change in the total CREB protein content (Fig. 5B).
Because many PKC and PKA inhibitors have other effects independent of their activity on PKC/A, we used other kinase inhibitors. Calphostin C (0.1 μmol/l) and H-8 (2 μmol/l) were used to inhibit PKC and PKA, respectively, and results similar to those with Bis-I and H-89 were obtained for the inhibition of CREB phosphorylation (data not shown). Neither HG- nor GlcN-mediated CREB phosphorylation was completely inhibited with any of the PKC or PKA inhibitors used individually. To investigate the possibly that some other kinase(s) may be involved in this process, we cultured cells with PKC and PKA inhibitors in combination. As shown in Fig. 5C, the combination of PKC and PKA inhibitors does not completely inhibit HG/GlcN-induced CREB phosphorylation, indicating the involvement of some other kinase(s) in this process.
We hypothesized that inhibition of hexosamine-induced CREB phosphorylation by downregulation of PKC and PKA would result in inhibition of fibronectin synthesis. As shown in Fig. 6, HG and GlcN increase fibronectin level approximately two- to threefold. In the presence of Bis-I or H-89, the effects of HG and GlcN on fibronectin were blocked. Neither Bis-I nor H-89 had any significant effect on fibronectin content under LG conditions.
Numerous studies have demonstrated that HG levels cause an increase in the synthesis and accumulation of ECM proteins in cultured kidney MES and tubular cells (3,12,26,27). Some of the effects of HG in the mesangium are mediated via synthesis of growth factors such as TGF-β, which acts in an autocrine/paracrine manner (28,29). It is not completely understood how HG regulates the ECM and growth factors. The HBP, which converts fructose-6-phosphate to GlcN-6-phosphate with glutamine as the amino donor, has been hypothesized to be a sensor for glucose and therefore a mediator of glucose regulation in a variety of cell types (14,15,17,30). In kidney MES cells, GlcN was more potent than glucose in stimulating TGF-β mRNA transcription and bioactivity, and the inhibition of GFA activity by the glutamine analogue azaserine or antisense oligonucleotide against GFA blocked the HG-induced expression of TGF-β and matrix protein synthesis (17,31). Likewise, hexosamines regulate TGF-β1 transcription in rat MES and proximal tubule cells and vascular cells via regulation of the TGF-β promoter (17).
We sought to investigate further the mechanisms whereby HG induces ECM synthesis in the mesangium. We showed that the effects of glucose on fibronectin are mediated by the HBP. This was supported by the observation that GlcN mimics the effects of glucose on fibronectin and that inhibition of GFA blocks the effects of glucose. HG and GlcN increase fibronectin protein synthesis approximately two- to threefold. This is consistent with results of other investigators (31); however, in those studies, much higher concentrations GlcN (12 mmol/l) were used. We found that GlcN concentrations >2.0 mmol/l reduced cell growth and viability drastically during the 48-h incubation. Our observation that GlcN exerts its effects at low concentrations (1.5 mmol/l) is important, as GlcN has been shown to have certain untoward effects on certain cells (32). Further support of the role of the HBP in mediating the effects of glucose is that GlcN is much more potent than glucose in its effects (1.5 mmol/l GlcN vs. 25 mmol/l glucose). This indicates that the role of the HBP is not simply dependent on glucose flux but rather on other regulatory events.
The underlying mechanisms by which HG and GlcN enhance matrix protein synthesis are not fully understood. We previously showed that HG and GlcN increase both PKC and PKA activities in MES cells and that agents that block the activity of these kinases blunt HG- and GlcN-induced laminin synthesis (19). In this study, inhibitors of PKC and PKA blocked HG- and GlcN-induced fibronectin synthesis, suggesting that the effects of PKC and PKA on hexosamine-mediated ECM synthesis are not unique to laminin. In addition, the current work supports that hexosamine regulation of fibronectin involves CREB.
The many isoforms of PKC that are known to exist are categorized into three subclasses (conventional, novel, and atypical) according to their structure and function (12,33). These PKC isoforms have different enzymatic properties, and their distribution changes after cell activation. Some isoforms are translocated from cytosol to membrane, whereas others are translocated into the nucleus, where they play a major role in signaling (33,34). Although it is not completely understood how hexosamines regulate PKC in the mesangium, it seems that translocation within the cytosol and possibly within the nucleus (CREB regulation) is important. In GlcN-treated MES cells, membrane translocation of several PKC isoforms was observed within hours (35). Similarly, we have observed that the effects of HG and GlcN on PKC activity are more pronounced in membrane fractions (19). Future work will focus on identification of which isoforms are regulated by the HBP and how those isoforms may regulate CREB.
Hyperglycemia also increases the intracellular concentration of cAMP (12,19,36) and, therefore, may activate the PKA signaling pathway in MES cells. HG and GlcN increase cAMP levels in MES cells by two- to threefold (19). The ability of HG and GlcN to increase PKA activity and the ability of the PKA inhibitor H-89 to decrease their effects on fibronectin synthesis suggest that a PKA-mediated pathway also may be important in ECM regulation in the kidney. Because the fibronectin promoter contains CRE consensus sequences, phosphorylation of CREB may play a direct role in the transcriptional regulation of the fibronectin gene. Consistent with this hypothesis, both HG and GlcN increased CREB phosphorylation, and the HG effect is blunted by GFA inhibitors. Furthermore, there is an increase in nuclear CREB activity after treatment of MES cells with HG and GlcN. Again, AZA, an inhibitor of GFA and the HBP, blocks this increase. These results do not conclusively determine that hexosamine regulation of fibronectin occurs via effects on CREB. However, support of the role of CREB in mediating hexosamine-induced fibronectin synthesis is the observation that CREB phosphorylation is dependent on PKC and PKA pathways in a manner similar to fibronectin. In addition, there was enhanced binding of nuclear extracts from HG- and GlcN-treated cells to a CRE-containing DNA fragment derived from a portion of the fibronectin promoter sequence.
Genes with CRE-containing promoters can be regulated differently by a wide variety of mechanisms depending on cell type and physiological response (23,24,25,37). Phosphorylation of CREB by PKC was reported to increase the binding of CREB to CRE in vitro and to stimulate CREB homodimerization or heterodimer complex formation with ATF-1, a transcription factor that shares many structural properties with CREB (37,38,39,40). Similarly, phosphorylation of CREB in vitro by PKA increased transcription efficiency of CRE-regulated genes 20-fold (38). Although it seems that PKC and PKA play major roles in hexosamine-mediated regulation of CREB in the mesangium, the involvement of other kinase(s) in the phosphorylation of CREB in MES cells may not be ruled out. CREB phosphorylation is not completely abolished in the presence of PKC and PKA inhibitors singly or in combination, suggesting that some other kinase(s) may be involved or that the kinases were not completely inhibited. Although the role of the phosphorylation of CREB in its interaction with other transcription factors and enhancement of gene expression is well documented, the effect of CREB phosphorylation on DNA binding itself is not entirely clear. Others have reported no alteration in nucleotide-binding activity of CRE after treatment of primary MES cells with HG or TGF-β (23). The discrepancy between the results may be due to differences in experimental designs and cell types used. Nevertheless, our results support a role for CREB phosphorylation in regulation of the effects of glucose in the mesangium.
It is hypothesized that the HBP acts as a cellular glucose sensor in target cells (13,41). Abnormalities in flux through or regulation of this pathway may lead to altered cellular responses to glucose. This is supported by the loss of glucose-induced increases in fibronectin levels when the rate-limiting enzyme in the HBP, GFA, is inhibited. Thus, downstream products of the HBP may upregulate second messenger proteins, growth factors, or transcription factors, resulting in enhanced ECM gene expression and ultimately in increased ECM levels. The role of the HBP in mediating ECM levels in the mesangium is complex. In addition to its effects on PKC, PKA, and TGF-β1 (17,19,31), the HBP regulates the gene expression of plasminogen activator inhibitor 1 in MES cells (42,43). The regulation of the plasminogen activator inhibitor 1 promoter by GlcN involved Sp-1, a transcription factor that has been shown to be regulated by the HBP (44). Furthermore, we also have observed that effects of hexosamines and TGF-β1 on fibronectin levels and CREB phosphorylation in MES cells are not additive (data not shown), suggesting that their effects occur via a common signaling pathway. Additional preliminary work has shown that blocking TGF-β signaling inhibits the effects of HG and GlcN on fibronectin synthesis. This indicates that TGF-β may be upstream of CREB with respect to hexosamine-induced fibronectin regulation. Additional evidence for the complexity of this regulatory pathway is that both glucose and TGF-β regulate GFA enzyme activity in MES cells (45). Elucidating how the HBP regulates these events will be important and is the focus of ongoing work.
In conclusion, the present study demonstrated that the HBP mediates HG-induced ECM synthesis in the MES cells. These effects of the HBP involve PKC and PKA pathways. The transcription factor CREB, a regulator of fibronectin gene, is also regulated by the HBP in a manner similar to fibronectin, supporting its role in mediating glucose- and hexosamine-induced ECM synthesis. Further delineation of the mechanisms by which the HBP has its effects may facilitate the development of novel therapeutic interventions for patients with diabetic nephropathy.
This work was supported by the Robert Wood Johnson Foundation (E.D.C.), the Kidney Care Foundation, Inc., and the Veterans Administration. E.D.C. is a recipient of a career development award from the Veterans Administration.
Address correspondence and reprint requests to Dr. Errol D. Crook, Department of Medicine, Division of Nephrology, 2500 N. State St., Jackson 39216-4505. E-mail:.
Received for publication 18 August 2000 and accepted in revised form 29 June 2001.
AZA, azaserine; CRE, cAMP responsive element; CREB, CRE-binding protein; ECL, enhanced chemiluminescence; ECM, extracellular matrix; GFA, glutamine:fructose-6-phosphate amidotransferase; GlcN, glucosamine; HBP, hexosamine biosynthesis pathway; HG, high glucose; LG, low glucose; MES, mesangial; PKA, protein kinase A; PKC, protein kinase C; TGF, transforming growth factor.