High Glucose Inhibits Insulin-stimulated Nitric Oxide Production without Reducing Endothelial Nitric-oxide Synthase Ser 1177 Phosphorylation in Human Aortic Endothelial Cells*

Recent studies have indicated that insulin activates endothelial nitric-oxide synthase (eNOS) by protein kinase B (PKB)-mediated phosphorylation at Ser 1177 in endothelial cells. Because hyperglycemia contributes to endothelial dysfunction and decreased NO availability in types 1 and 2 diabetes mellitus, we have studied the effects of high glucose (25 m M , 48 h) on insulin signaling pathways that regulate NO production in human aortic endothelial cells. High glucose inhibited insulin-stimu-lated NO synthesis but was without effect on NO synthesis stimulated by increasing intracellular Ca 2 (cid:1) concentration. This was accompanied by reduced expression of IRS-2 and attenuated insulin-stimulated recruitment of PI3K to IRS-1 and IRS-2, yet insulin-stimulated PKB activity and phosphorylation of eNOS at Ser 1177 were unaffected. Inhibition of insulin-stimulated NO synthesis by high glucose was unaffected by an inhibitor of PKC. Furthermore, high glucose down-regulated the expression of CAP and Cbl, and insulin-stimulated Cbl phosphorylation, components of an insulin signaling cascade previously characterized in adipocytes. These data suggest that high glucose specifically inhibits insu-lin-stimulated NO synthesis and down-regulates some aspects of insulin

2 diabetes (4,5) and also in cardiovascular disorders such as hypertension and atherosclerosis associated with insulin resistance (6,7).
Insulin is a direct-acting vasodilator in intact vessels (8,9) and has been demonstrated to stimulate wortmannin-sensitive NO production in cultured endothelial cells (10). Recently, several laboratories have demonstrated activation of eNOS by phosphorylation at Ser 1177 (human sequence) in response to shear stress, vascular endothelial growth factor, insulin-like growth factor-I, and insulin (11)(12)(13)(14)(15). Phosphorylation at this site is accompanied by a decrease in the dependence of eNOS for Ca 2ϩ /calmodulin (12,16). Incubation of bovine aortic endothelial cells (BAECs) with insulin has been demonstrated to result in the rapid stimulation of insulin receptor substrate-1 (IRS-1) tyrosine phosphorylation, IRS-1-associated phosphatidylinositol 3-kinase (PI3K) activity, phosphorylation and activation of protein kinase B (PKB), eNOS phosphorylation at the equivalent bovine site to Ser 1177 , and NO production (15). It is proposed that insulin stimulates PKB-mediated phosphorylation and activation of eNOS via PI3K recruitment to IRS-1 (15,16).
Hyperglycemia contributes to endothelial dysfunction and decreased NO availability in types 1 and 2 diabetes mellitus, but the mechanisms remain uncharacterized. It has been proposed that high glucose inhibits basal eNOS expression, activity, and NO production in cultured endothelial cells (17,18). However, other studies have demonstrated the opposite effects of experimental hyperglycemia on NO production in human aortic endothelial cells (HAECs) (19). Additionally, a recent report has suggested that acute (15-30 min) stimulation with high glucose inhibits insulin-stimulated NO production as a result of reduced eNOS serine phosphorylation in human umbilical vein endothelial cells (HUVECs) (20).
The proposed pathway of insulin-stimulated NO synthesis exhibits striking similarities with the characterized mechanism of insulin-stimulated glucose transport in adipocytes. PI3K stimulation is necessary but not sufficient for insulinstimulated glucose transport in adipocytes (21). Similarly, it has been proposed that PKB stimulation is necessary but not sufficient for insulin-stimulated NO synthesis in endothelial cells (16). Recent studies in adipocytes have demonstrated a novel insulin signaling pathway required in parallel with PI3K activation to stimulate glucose transport (22,23). In this pathway, Cbl is recruited to the insulin receptor by interaction with the adapter protein CAP. Upon insulin stimulation Cbl becomes tyrosine-phosphorylated, and the CAP-Cbl complex dissociates from the insulin receptor and moves to a caveolin-and flotillinrich lipid raft subdomain of the plasma membrane (22). Phosphorylated Cbl recruits the CrkII-C3G complex to lipid rafts where C3G specifically activates the small G-protein TC10 (23).
In the present study, we investigated the effects of prolonged (48 h) experimental hyperglycemia on insulin-stimulated NO production in HAECs. Our results suggest that insulin-stimulated NO production is specifically abolished by culture in high glucose. Importantly, we found that there was no reduction in PKB activity, total phosphorylation of eNOS, or specific phosphorylation of eNOS at Ser 1177 in response to insulin that could account for this inhibition. Therefore, insulin-stimulated phosphorylation of eNOS at Ser 1177 is not sufficient to stimulate NO production in endothelial cells cultured in 25 mM glucose. We also demonstrate that HAECs express CAP, Cbl, C3G, and flotillin. Culture of HAECs in high glucose concentrations resulted in reduced expression of CAP and Cbl and inhibited insulin-stimulated Cbl phosphorylation. We propose that the CAP-Cbl pathway may represent a candidate parallel pathway required in addition to PKB activation for insulin-stimulated NO synthesis.
HAEC Cell Culture-HAECs were grown in large vessel endothelial cell medium at 37°C in 5% CO 2 and passaged when at 80% confluence. Prior to use the cells were maintained in medium containing 5 or 25 mM glucose for 48 h. During this period, culture medium containing 5 mM glucose was supplemented with 20 mM mannitol to maintain osmolarity. The cells were used for experiments between passages 3 and 6.
Nitric Oxide Release-The cells cultured in 6-well plates in medium containing either glucose concentration were preincubated for 1 h at 37°C in 2 ml/well Krebs-Ringer-HEPES (KRH) buffer (119 mM NaCl, 20 mM NaHEPES, pH 7.4, 5 mM NaHCO 3 , 4.7 mM KCl, 1.3 mM CaCl 2 , 1.2 mM MgSO 4 , 1 mM KH 2 PO 4 , 100 M L-arginine, 5 or 25 mM glucose) at 37°C. The medium was removed and replaced with 1 ml of fresh KRH buffer in the presence or absence of insulin or A23187. Aliquots (100 l) of the medium were removed after various incubation times and refluxed in glacial acetic acid containing NaI. Under these conditions, NO 2 Ϫ , a stable breakdown product of NO, is quantitatively reduced to NO. NO-specific chemiluminescence was then analyzed using a Sievers 280A NO meter. The appropriate control experiments were performed in the presence of the eNOS inhibitor, L-NMMA. NO production (adjusted for L-NMMA-insensitive production) is expressed per mg of protein.
Preparation of HAEC Lysates-The cells cultured in 10-cm-diameter dishes in medium containing either glucose concentration were preincubated for 1 h at 37°C in 5 ml of KRH buffer containing 5 or 25 mM glucose. The medium was replaced with 5 ml of KRH buffer containing test substances and incubated for various durations at 37°C. The medium was removed, and 0.5 ml of buffer A (50 mM Tris-HCl, pH 7.4 at 4°C, 150 mM NaCl, 50 mM NaF, 5 mM sodium pyrophosphate, 2 mM sodium orthovanadate, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 0.1 mM benzamidine, 0.1 mM phenylmethylsulfonyl fluoride, 5 g/ml soybean trypsin inhibitor, 1% (v/v) Triton X-100, 1% (v/v) glycerol) was added. The cell extract was scraped off and transferred to a microcentrifuge tube. The extracts were vortex-mixed and centrifuged (14,000 ϫ g, 3 min, 4°C). The supernatants were snap frozen in liquid N 2 and stored at Ϫ80°C prior to use. 32 P Labeling of HAECs-The cells cultured in 10-cm-diameter dishes in medium containing either glucose concentration were preincubated for 1 h at 37°C in 5 ml of phosphate-free KRH buffer containing 0.2 mCi/ml 32 PO 4 3Ϫ and 5 or 25 mM glucose. After incubation in the presence or absence of insulin (100 nM) for 5 min, the medium was removed, and the cell lysates were prepared.
Immunoprecipitation of IRS-1 and IRS-2-HAEC lysate (50 g) was added to 20 l of 25% (v/v) protein A-Sepharose prebound to 2.5 g of rabbit anti-IRS-1 or anti-IRS-2 antibodies, and the volume was made up to 300 l with buffer A and mixed for 2 h at 4°C on a rotating mixer. The mixture was then centrifuged (14,000 ϫ g, 30 s, 4°C), and the pellet was washed three times with 1 ml of buffer A at 4°C. Immunoprecipitated protein was separated by SDS-PAGE and transferred to nitrocellulose.
Immunoprecipitation and Assay of PKB-HAEC lysate (50 g) was added to 20 l of 25% (v/v) protein G-Sepharose prebound to 5 g of sheep anti-PKB antibody, and the volume was made up to 300 l with buffer A and mixed for 2 h at 4°C on a rotating mixer. The mixture was then centrifuged (14,000 ϫ g, 30 s, 4°C), and the pellet was washed and assayed as described previously by incorporation of 32 P into the RPRAATF substrate peptide (24).
Immunoprecipitation of 32 P-Labeled Cbl-HAEC lysate (300 g) from 32 P-labeled cells was added to 20 l of 25% (v/v) protein A-Sepharose prebound to 2 g of rabbit anti-Cbl, and the volume was made up to 300 l with buffer A and mixed for 2 h at 4°C on a rotating mixer. The mixture was then centrifuged (14,000 ϫ g, 30 s, 4°C), and the pellet was washed three times with 1 ml of buffer A supplemented with 0.5 M NaCl and three times with 1 ml of buffer A at 4°C. Immunoprecipitated protein was separated by SDS-PAGE, the resultant was gel-dried, and 32 P-labeled proteins were visualized by autoradiography.
Immunoprecipitation of 32 P-Labeled eNOS-HAEC lysate (200 g) from 32 P-labeled cells was added to 20 l of 25% (v/v) protein G-Sepharose prebound to 5 g of mouse anti-eNOS, and the volume was made up to 300 l with buffer A and mixed for 2 h at 4°C on a rotating mixer. The mixture was then centrifuged (14,000 ϫ g, 30 s, 4°C), and the pellet was washed three times with 1 ml of buffer A supplemented with 0.5 M NaCl and three times with 1 ml of buffer A at 4°C. Immunoprecipitated protein was separated by SDS-PAGE, the resultant was gel-dried, and 32 P-labeled proteins were visualized by PhosphorImager analysis using a Fuji FLA-5000 analyzer.
Statistics-Unless stated otherwise, the results are expressed as the means Ϯ S.D. Statistically significant differences were determined using a two-tailed independent samples Student's t test, using p Ͻ 0.05 as significant using Statview software.

RESULTS
To study the effects of experimental hyperglycemia on insulin-stimulated NO production in HAECs, we examined the ability of insulin to stimulate NO synthesis in HAECs cultured for 48 h in 5 or 25 mM glucose using a Sievers 280A NO meter (Fig. 1A). As shown, insulin (100 nM) rapidly stimulated NO synthesis in HAECs cultured in 5 mM glucose. Insulin-stimulated NO production could be demonstrated at concentrations as low as 5 nM and was maximal at 100 nM (not shown), a concentration at which all further experiments were performed. In contrast, insulin had no significant effect on NO synthesis in HAECs cultured in 25 mM glucose. There was no significant difference in basal NO synthesis at either glucose concentration. To investigate whether this inhibitory effect of hyperglycemia was restricted to insulin-stimulated NO synthesis, we next examined the effect of culture glucose concentration on NO synthesis stimulated by the Ca 2ϩ ionophore, A23187 (Fig. 1B). As shown, A23187 rapidly stimulated NO synthesis in HAECs cultured in either 5 or 25 mM glucose. There was no significant difference between A23187-stimulated NO production at either glucose culture concentration.
In an attempt to identify differences between HAECs cultured in either glucose concentration that could account for the observed inhibition of insulin-stimulated NO production, we characterized the expression of proteins involved in insulin signaling in HAECs cultured in either glucose concentration. Quantitative analysis of Western blots illustrated that there was no significant difference in the expression of IRS-1 (Fig. 2), PI3K (p85 and p110␤ subunits), PTEN, PDK-1, PKB␣, or eNOS in HAECs cultured in either 5 or 25 mM glucose (data not shown). However, expression of IRS-2 in cells cultured in 25 mM glucose was reduced to 29 Ϯ 19% (n ϭ 6, p Ͻ 0.01) of that in HAECs cultured in 5 mM glucose (Fig. 2). No expression of the p110␣/␥ subunit isoforms of PI3K or PKB␤ protein could be demonstrated in HAECs.
To determine whether there were any differences in the activity of the previously characterized insulin signaling pathway in HAECs cultured in 5 or 25 mM glucose, we characterized the recruitment of PI3K to IRSs in HAECs cultured in either glucose concentration under basal and insulin-stimulated conditions. Insulin-stimulated recruitment of PI3K to IRS-1 is reduced by 24 Ϯ 7% (n ϭ 6, p Ͻ 0.01) in HAECs cultured in 25 mM glucose, compared with those cultured in 5 mM glucose (Fig.  3). Insulin also stimulated the recruitment of PI3K to IRS-2 in cells cultured in 5 mM glucose but could not be demonstrated in cells cultured in 25 mM glucose (Fig. 3A). We next characterized the effects of insulin on PKB activity in HAECs cultured in either glucose concentration. Stimulation of HAECs with insulin resulted in the time-dependent activation of PKB␣, maximal by 10 min (not shown). Insulin-stimulated PKB activity was maximal at 100 nM and could be demonstrated at concentrations as low as 1 nM (data not shown). PKB␤ activity could not be detected in HAECs by immunoprecipitation kinase assays with a PKB␤-specific antibody. Culture of HAECs at 25 mM glucose had no significant effect on total insulin-stimulated PKB␣ activity, compared with HAECs cultured in 5 mM glucose, although basal PKB activity was significantly increased (2.4-fold, p Ͻ 0.01) (Fig. 4).
To determine whether phosphorylation of eNOS was altered by culture glucose concentration, we determined the total phosphorylation of immunoprecipitates from 32 P-labeled HAECs. Insulin stimulated eNOS phosphorylation in cells cultured in either 5 or 25 mM glucose (1.6-fold, p Ͻ 0.05 and 1.4-fold, p Ͻ 0.05, respectively, n ϭ 3) (Fig. 5). There was no significant difference in basal or insulin-stimulated eNOS phosphorylation at either glucose concentration. To determine whether insulin stimulated tyrosine phosphorylation of eNOS, we used a phosphotyrosine antibody on Western blots of eNOS immunoprecipitates from HAEC lysates. No anti-phosphotyrosinereactive species of the correct molecular mass was detected in lysates prepared from cells cultured in either glucose concentration in the presence or absence of insulin (not shown). To determine whether phosphorylation of eNOS at Ser 1177 was altered by culture glucose concentration we used a phospho-Ser 1177 -specific antibody. Insulin stimulated phosphorylation of eNOS at Ser 1177 in HAECs cultured in either 5 or 25 mM glucose (1.6-fold, p Ͻ 0.05 and 2.4-fold, p Ͻ 0.01, respectively, n ϭ 6) (Fig. 6). There was no significant difference in basal or insulin-stimulated Ser 1177 phosphorylation at either glucose concentration.
Activation of protein kinase C (PKC) has been demonstrated in vascular cells exposed to elevated glucose (25). Furthermore, it has been proposed that activation of PKB␤ 1 inhibits insulinstimulated eNOS expression in BAECs (26) and that phosphorylation of eNOS at Thr 495 by PKC inhibits activity (27). We therefore characterized the expression of PKC isoforms and phosphorylation of eNOS at Thr 495 in HAECs cultured in either glucose concentration. Phosphorylation of eNOS at Thr 495 was unaltered by medium glucose concentration in HAECs as judged by quantitative analysis of Western blots using a phospho-Thr 495 -specific anti-eNOS antibody (not shown). We next determined whether inhibition of PKC with bisindolylmaleimide I would reverse the inhibition of insulin-stimulated NO production in HAECs cultured in 25 mM glucose. Preincubation of cells for 1 h with bisindolylmaleimide I had no significant effect on basal or insulin-stimulated NO production in cells cultured in 5 mM glucose (Fig. 7). Additionally, bisindolylmaleimide I had no effect on the inhibition of insulin-stimulated NO production by culture in 25 mM glucose (Fig. 7).
Recent studies in adipocytes have illustrated a novel insulinsignaling pathway that localizes to caveolin and flotillin-rich lipid raft domains and is required for insulin-stimulated glucose transport, in parallel with PI3K activation (22,23). We therefore determined whether proteins involved in this path-way were expressed in HAECs cultured at 5 or 25 mM glucose. There was no significant difference in the expression of C3G, caveolin, or flotillin in HAECs cultured in either 5 or 25 mM glucose (Fig. 8). However, expression of CAP and Cbl in cells cultured in 25 mM glucose was reduced to 54 Ϯ 12% (n ϭ 9, p Ͻ 0.05) and 60 Ϯ 9% (n ϭ 9, p Ͻ 0.01), respectively, of that in HAECs cultured in 5 mM glucose (Fig. 8). Next we determined the phosphorylation of Cbl in response to insulin in HAECs cultured in either glucose concentration. Insulin stimulated Cbl phosphorylation in cells cultured in 5 mM glucose (1.5-fold, n ϭ 3, p Ͻ 0.01) but was ineffective in cells cultured in 25 mM glucose (Fig. 9). DISCUSSION Abnormal vascular endothelial function, as defined by measures of NO bioavailability, is well described in patients with types 1 and 2 diabetes and in cardiovascular disorders that are associated with insulin resistance including hypertension and atherosclerosis (6, 7, 28 -30). Although endothelial dysfunction may occur in association with insulin resistance even in normoglycemic subjects (31), hyperglycemia is an obvious candidate contributor. This prompted us to determine the effects of experimental hyperglycemia on insulin-stimulated NO production in HAECs.
Culture of HAECs in 22.2 mM glucose for 5 days was previously reported to stimulate basal eNOS expression and NO production (19). In contrast, other groups have demonstrated reduced basal NO production in response to high glucose, with a concomitant reduction in eNOS expression in human coronary artery endothelial cells and BAECs (17,18). In the pres- 3Ϫ and subsequent incubation with insulin (100 nM) for 5 min. eNOS was immunoprecipitated from HAEC lysates, and the immunoprecipitated proteins were resolved by SDS-PAGE and subjected to Western blotting with anti-eNOS antibodies or PhosphorImager analysis. A, a representative Western blot and autoradiograph is shown. B, quantification of eNOS phosphorylation. Insulin stimulated eNOS phosphorylation in cells cultured in 5 or 25 mM glucose (1.6-fold, p Ͻ 0.05 and 1.4-fold, p Ͻ 0.05 respectively, n ϭ 3). There were no significant differences between basal eNOS phosphorylation at either glucose concentration. Similarly there were no significant differences between insulin-stimulated eNOS phosphorylation at either glucose concentration. ent study we could not demonstrate any effect of culture glucose concentration on eNOS expression in HAECs. These differences may be due to the different endothelial cell types used or the varied duration of experimental hyperglycemia applied.
In this study, we have demonstrated that insulin-stimulated NO production was abolished in HAECs cultured in 25 mmol/ liter glucose, yet stimulation of eNOS by the Ca 2ϩ ionophore A23187 was unaffected by culture glucose concentration, which suggests that the inhibitory effect of experimental hyperglycemia is specific to insulin stimulation and that eNOS is still capable of being activated by increased intracellular Ca 2ϩ . We therefore hypothesized that this inhibition of insulin-stimulated NO production was a consequence of reduced expression or activity of insulin signaling proteins. It is proposed that insulin stimulates PKB-mediated phosphorylation and activa-FIG. 6. Effects of culture glucose concentration on insulinstimulated eNOS phosphorylation at Ser 1177 . HAEC lysates were prepared from cells cultured for 48 h in medium containing 5 or 25 mM glucose and incubated with insulin (100 nM) for 5 min. 10 g of protein were resolved by SDS-PAGE, transferred to nitrocellulose, and probed with sheep anti-phospho-Ser 1177 antibodies. A, a representative Western blot is shown of three independent samples. B, quantification of Ser 1177 phosphorylation. Insulin stimulated Ser 1177 phosphorylation in cells cultured in 5 or 25 mM glucose (1.6-fold, p Ͻ 0.05 and 2.4-fold, p Ͻ 0.01, respectively, n ϭ 6). There were no significant differences between basal Ser 1177 phosphorylation at either glucose concentration. Similarly there were no significant differences between insulin-stimulated Ser 1177 phosphorylation at either glucose concentration.  8. Expression of CAP/Cbl signaling proteins in HAECs cultured in 5 or 25 mM glucose. HAEC lysates were prepared from cells cultured for 48 h in medium containing 5 or 25 mM glucose. The lysate proteins were resolved by SDS-PAGE, transferred to nitrocellulose, and probed with the antibodies indicated. A Western blot is shown of three independent samples. There was no significant difference in the expression of the proteins examined, except for CAP expression (*), reduced to 54 Ϯ 12% (n ϭ 6, p Ͻ 0.05) of that in HAEC lysates cultured in 5 mM glucose and Cbl expression ( †), reduced to 60 Ϯ 9% (n ϭ 6, p Ͻ 0.01) of that in HAEC lysates cultured in 5 mmol/liter glucose.
FIG. 9. Effects of culture glucose concentration on insulinstimulated Cbl phosphorylation. HAEC lysates were prepared from cells cultured for 48 h in medium containing 5 or 25 mM glucose prior to incubation in media containing 0.2 mCi/ml 32 PO 4 3Ϫ and subsequent incubation with insulin (100 nM) for 5 min. Cbl was immunoprecipitated from HAEC lysates, the immunoprecipitated proteins were resolved by SDS-PAGE, and the gel was dried and subjected to autoradiography. A, a representative autoradiograph is shown. B, quantification of Cbl phosphorylation. Insulin stimulated Cbl phosphorylation in cells cultured in 5 mM glucose (1.5-fold, n ϭ 3, p Ͻ 0.01). There was no significant alteration in Cbl phosphorylation in response to insulin in those cells cultured in 25 mM glucose. tion of eNOS via PI3K recruitment to IRS-1 (15,16). In this study, we find that in cells cultured in 5 mM glucose, the time course and dose dependence of insulin-stimulated NO production in HAECs is comparable with that previously reported in HUVECs and BAECs (10,15). We have also shown that insulin stimulates PI3K recruitment to IRS-1, PKB activity, and NO phosphorylation on Ser 1177 as previously reported in BAECs (15). Our data also represent, to our knowledge, the first demonstration of insulin-stimulated PI3K recruitment to IRS-2 in endothelial cells. The importance of IRS-2 in endothelial cell insulin signaling has yet to be determined, although studies in knockout mice have demonstrated that IRS-2 appears to mediate whole body insulin action, whereas IRS-1 is of importance primarily in muscle (32).
The reduction in insulin-stimulated PI3K recruitment to IRS-1 mediated by experimental hyperglycemia is consistent with recent data in HUVECs (21) and not a result of decreased IRS-1 or PI3K expression. The total inhibition of PI3K recruitment to IRS-2 observed in this study is likely to be a result of the markedly reduced expression of IRS-2. This reduction of PI3K recruitment did not, however, result in reduced insulinstimulated PKB activity in HAECs subjected to high glucose. Although this may seem surprising, it has recently been demonstrated that PKB activation is normal in insulin-resistant mice with impaired insulin-stimulated IRS-1 phosphorylation and PI3K activity (33). There are several potential explanations for the seemingly divergent effects of elevated culture glucose concentration on insulin-stimulated PI3K recruitment to IRSs and PKB activity. First, only partial activation of PI3K may be necessary for full stimulation of PKB. Second, changes in total IRS-associated PI3K activity and PKB activity may not reflect changes within specific membrane compartments. Third, a PI3K-independent mechanism may be able to offset the decrease in recruitment of PI3K to IRSs.
In contrast to our study, a recent report has demonstrated that basal eNOS activity and phosphorylation at Ser 1177 were reduced in BAECs cultured in 30 mM glucose for 48 h (18). Du et al. (18) interpreted their findings as demonstrating increased O-linked N-acetylglucosamine modification of eNOS. Additionally, inhibition of insulin-stimulated eNOS serine phosphorylation by acute exposure of HUVECs to high glucose concentration has recently been reported (20). In the present study, chronic exposure of HAECs to 25 mM glucose had no effect on either insulin-stimulated total eNOS phosphorylation or specific Ser 1177 phosphorylation. These contrasting effects are likely to reflect the very different durations of experimental hyperglycemia applied or the different sources of endothelial cells used. It should also be noted that Schnyder et al. (20) used an anti-phosphoserine antibody rather than a phospho-Ser 1177specific antibody in their study; thus, the hyperglycemia-induced reduction in insulin-stimulated eNOS serine phosphorylation that they reported may represent phosphorylation at sites other than Ser 1177 .
These data suggest that there is an as yet uncharacterized pathway that regulates eNOS, at least in high glucose concentrations. This may be an alternative inhibitory phosphorylation site on eNOS that is activated under hyperglycemic conditions and specifically interferes with activation mediated by Ser 1177 phosphorylation yet does not inhibit eNOS activation by Ca 2ϩ / calmodulin. In the current study, culture glucose concentration had no effect on total eNOS phosphorylation, arguing against the mediation of inhibition of insulin-stimulated NO production by an alternative phosphorylation site. However, these data do not rule out changes in the extent of eNOS phosphorylation at sites other than Ser 1177 that result in no alteration in the total extent of eNOS phosphorylation. It has been dem-onstrated that hyperglycemia stimulates PKC activity (25) and proposed that phosphorylation of eNOS at Thr 495 by PKC inhibits activity (27). However, the observed inhibition of insulinstimulated NO synthesis was not affected by the PKC inhibitor, bisindolylmaleimide I, and was not due to increased phosphorylation of Thr 495 , which is unaltered by medium glucose concentration in this study. These data indicate that neither PKC activation nor inhibitory phosphorylation of eNOS at Thr 495 mediate the inhibition of insulin-stimulated NO production by high culture glucose concentration.
Alternatively, an alternative insulin signaling pathway, which is impaired in hyperglycemic conditions, may be required for eNOS activation in parallel with PKB activation. The proposed pathway of insulin-stimulated NO synthesis exhibits striking similarities with the mechanism of insulin-stimulated glucose transport in adipocytes. PI3K stimulation is necessary but not sufficient for insulin-stimulated glucose transport in adipocytes (21). Similarly, it has been proposed that PKB stimulation is necessary but not sufficient for insulin-stimulated NO synthesis in endothelial cells (16). Studies in adipocytes have demonstrated a novel insulin-signaling pathway that localizes to flotillin-rich lipid raft domains in the plasma membrane, required in parallel with PI3K stimulation for insulin-stimulated glucose transport (22,23). We therefore reasoned that this novel insulin-signaling pathway characterized in adipocytes might be activated by insulin in parallel with PKB activation in endothelial cells and impinge on insulinstimulated eNOS activity. We demonstrate that flotillin, CAP, Cbl, and C3G, components of this pathway are present in endothelial cells. Intriguingly, culture in elevated glucose concentrations markedly reduced the expression of CAP and Cbl and insulin-stimulated Cbl phosphorylation required for the recruitment of the CrkII-C3G complex to lipid rafts in adipocytes (23). The regulation of CAP and Cbl expression is largely uncharacterized. Both CAP gene expression and Cbl tyrosine phosphorylation have been demonstrated to be stimulated by the thiazolidinedione insulin-sensitizing drugs in adipocytes (34). The regulation of these genes by culture glucose concentration is a novel finding that requires further study in other tissues such as adipocytes.
In summary, we have shown that culture in high glucose quantitatively inhibits insulin-stimulated NO production in HAECs. This is not due to a general reduced responsiveness of eNOS, because A23187-stimulated NO production is unaffected. Importantly, however, there is no effect of high glucose on insulin-stimulated PKB activation or eNOS phosphorylation at Ser 1177 . We therefore propose that PKB-stimulated phosphorylation of eNOS at Ser 1177 is not sufficient to stimulate NO synthesis in cells exposed to high glucose for 48 h. Moreover, the CAP-Cbl insulin signaling pathway is downregulated by high glucose in these cells. The role of the CAP-Cbl pathway in insulin signaling in endothelial cells is unclear but provides a candidate pathway required for insulin-stimulated NO production in parallel with PKB activation, as is the case in insulin-stimulated glucose transport in adipocytes. Future experiments are required to characterize the role of the CAP-Cbl pathway in insulin signaling and insulin-stimulated NO production in HAECs to help understand the mechanism of hyperglycemia-induced endothelial dysfunction.