Nitric Oxide Production and Regulation of Endothelial Nitric-oxide Synthase Phosphorylation by Prolonged Treatment with Troglitazone

Recently, peroxisome proliferator-activated receptor γ (PPARγ) ligands have been reported to increase endothelial NO, but the signaling mechanisms involved are unknown. Using troglitazone, a PPARγ ligand known as an antidiabetic compound, we investigated the molecular mechanism of its effect on NO production in bovine aortic endothelial cells. Troglitazone increased endothelial NO production in a dose- and time-dependent manner with no alteration in endothelial nitric-oxide synthase (eNOS) expression. The maximal increase (∼3.1-fold) was achieved with 20 μm troglitazone treatment for 12 h, and this increase was accompanied by increases in the expression of vascular endothelial growth factor (VEGF) and its receptor, KDR/Flk-1, and in Akt phosphorylation. Analysis with antibodies specific for each phosphorylated site demonstrated that troglitazone (20 μm treatment for 12 h) significantly increased both the phosphorylation of Ser1179 of eNOS (eNOS-Ser1179) and the dephosphorylation of eNOS-Ser116 but did not alter eNOS-Thr497 phosphorylation. Treatment with anti-VEGF antibody to scavenge the increased VEGF induced by troglitazone partially inhibited troglitazone-stimulated NO production. This was accompanied by the attenuation of troglitazone-stimulated increases in the phosphorylation of Akt and eNOS-Ser1179 with no alteration in eNOS-Ser116 dephosphorylation. We also found that bisphenol A diglycidyl ether, a PPARγ antagonist, partially inhibited troglitazone-stimulated NO production with a concomitant reduction in VEGF-KDR/Flk-1-Akt-mediated eNOS-Ser1179 phosphorylation but with no alteration in eNOS-Ser116 dephosphorylation induced by troglitazone. Taken together, our results demonstrate that prolonged treatment with troglitazone increases endothelial NO production by at least two independent signaling pathways: PPARγ-dependent, VEGF-KDR/Flk-1-Akt-mediated eNOS-Ser1179 phosphorylation and PPARγ-independent, eNOS-Ser116 dephosphorylation.

Troglitazone, the first thiazolinedione (TZD) 1 compound synthesized, was used clinically as an oral antidiabetic drug to improve insulin resistance in patients with type 2 diabetes mellitus (DM) (1,2). However, it was later withdrawn from the market because of fatal hepatic injury (3). The effects of this drug are known to be mediated, in part, through its binding to peroxisome proliferator-activated receptor ␥ (PPAR␥) (4), a member of the nuclear hormone receptor superfamily. Because people with type 2 DM often have at least one cardiovascular disease risk factor (5), troglitazone is likely to be implicated in the pathophysiology of both cardiovascular disease and DM. Troglitazone has been reported to lower blood pressure (6), increase insulin sensitivity, and correct hyperinsulinemia (7,8), although the results are conflicting (7). Furthermore, another TZD drug, pioglitazone, attenuated vasoconstriction in vitro (9) and reduced L-type currents in rat vascular smooth muscle cells (VSMC) (10). Recently, troglitazone has also been reported to inhibit microvascular endothelial cell proliferation (11).
Endothelial nitric-oxide synthase (eNOS) is an enzyme essential to the maintenance of cardiovascular integrity by producing NO in vivo, a key molecule with multiple functions, including vasodilation, and many antiatherogenic properties (12). Therefore, the dysregulation of eNOS is thought to contribute to the pathogenesis of certain vascular diseases, such as atherosclerosis and hypertension (13). eNOS is regulated not only at the level of expression (14,15) but also nongenomically by subcellular targeting (16), protein-protein interactions (17), fatty acylation (18), and phosphorylation (19). Recently, specific sites of phosphorylation have been identified; among these sites, Ser 1179 (eNOS-Ser 1179 ; bovine sequence) and eNOS-Thr 497 are the most studied. The phosphorylation of eNOS-Ser 1179 reduces the Ca 2ϩ -calmodulin dependence of the enzyme (20) and increases the rate of electron flux from the reductase domain to the oxygenase domain (21), thereby increasing NO production (22)(23)(24). This phosphorylation is mediated by several specific protein kinases, including protein kinase B (Akt), AMP-activated protein kinase, calmodulin-dependent kinase II, and protein kinase A (20,(25)(26)(27)(28). In contrast, the phosphorylation of eNOS-Thr 497 decreases eNOS activity by increasing Ca 2ϩ -calmodulin dependence (20,27,29). Phosphatases such as protein phosphatase 1 and protein phosphatase 2B increase the dephosphorylation of eNOS-Thr 497 , resulting in an increase in NO production (19,27). Another phosphorylation site, eNOS-Ser 116 , has also been reported, and its dephosphorylation by protein phosphatase 2B increases eNOS activity (30). Recently, two other sites, eNOS-Ser 635 and eNOS-Ser 617 , were also identified as phosphorylation targets of protein kinase A and Akt, respectively (31). However, the roles of these protein kinases and phosphatases as signaling molecules for eNOS phosphorylation and dephosphorylation at several potential sites are dependent on the experimental conditions used, such as the presence of agonists, the endothelial cell type, and treatment time. For example, vascular endothelial growth factor (VEGF) (32) and fluid shear stress (25) stimulated Akt-dependent eNOS-Ser 1179 phosphorylation at an earlier time, but they also stimulated protein kinase A-dependent eNOS-Ser 635 phosphorylation at a later time (33). In contrast to VEGF and shear stress, which cause no dephosphorylation of eNOS-Thr 497 , 8-bromo-cAMP rapidly dephosphorylates eNOS-Thr 497 . All of these findings suggest that the activity of eNOS in cells may be controlled through a coordinated regulation of, and interaction between, several protein kinases and phosphatases.
Most studies of troglitazone have focused on the inhibition of cytokine-induced NO production via a decrease in the expression of inducible nitric-oxide synthase (iNOS) in adipocytes (34) and VSMC (35). However, troglitazone is also reported to increase in vivo forearm blood flow (36). Furthermore, the PPAR␥ ligands 15-deoxy-⌬ 12,14 -prostaglandin J 2 (15d-PGJ 2 ) and ciglitazone were recently shown to increase endothelial NO production (37). However, no detailed mechanism underlying this increase has yet been reported. Together with the previous observation that TZDs in vivo increased the expression of VEGF, a well known agonist for endothelial NO production (38), these results prompted us to characterize the molecular mechanism underlying the troglitazone-stimulated increase in endothelial NO production. Our current data demonstrate, for the first time, that troglitazone increases NO production by at least two independent signaling pathways: PPAR␥-dependent, VEGF-KDR/Flk-1-Akt-mediated eNOS-Ser 1179 phosphorylation and PPAR␥-independent, eNOS-Ser 116 dephosphorylation.

EXPERIMENTAL PROCEDURES
Materials-Troglitazone was obtained as a gift from Sankyo Co. (Tokyo, Japan), and bisphenol A diglycidyl ether (BADGE) was obtained from Sigma. Antibodies against eNOS and Akt were purchased from Transduction Laboratories (Lexington, KY) and New England Biolabs (Beverly, MA), respectively. Antibodies against VEGF and KDR/Flk-1 were purchased from Sigma and Santa Cruz Biotechnology (La Jolla, CA), respectively. Antibodies against Akt phosphorylated at Ser 473 (p-Akt-Ser 473 ) and p-eNOS-Ser 1179 were obtained from Cell Signaling Technology (Beverly, MA), and those against p-eNOS-Thr 497 and p-eNOS-Ser 116 were from Upstate Biotechnology, Inc. (Lake Placid, NY). Trizol reagent for RNA extraction and SuperScript TM II RNase H Ϫ reverse transcriptase were obtained from Invitrogen. Recombinant Taq (rTaq) DNA polymerase was purchased from TaKaRa Biomedicals (Shiga, Japan), and collagenase (type 2) was purchased from Worthington Biochemical Corporation (Freehold, NJ). Minimal essential medium, Dulbecco's phosphate-buffered saline, newborn calf serum, penicillin and streptomycin antibiotics, L-glutamine, trypsin-EDTA solution, and plasticware for cell culture were purchased from Invitrogen. All other chemicals were of the purest analytical grade.
Cell Culture and Drug Treatments-Bovine aortic endothelial cells (BAEC) were isolated exactly as described previously (39) and maintained in minimal essential medium supplemented with 5% newborn calf serum at 37°C under 5% CO 2 in air. The endothelial cells were confirmed by their typical cobblestone configuration when viewed by light microscopy and by a positive indirect immunofluorescence test for von Willebrand factor VIII. The cells between passages 5 and 9 were used for all experiments. When BAEC were grown to confluence, the cells were further maintained for the indicated times in minimal essen-tial medium supplemented with 0.5% newborn calf serum containing various concentrations of troglitazone. In some experiments, the cells were co-treated with either anti-VEGF antibody (200 g/ml) or BADGE (5 M).
RNA Extraction and Semi-quantitative Reverse Transcription-PCR-After treatment with troglitazone for the indicated times, the culture medium was removed, and total cellular RNA was isolated with Trizol reagent (Invitrogen) according to the manufacturer's instructions. A reverse transcription reaction was performed with 2 g of total RNA in a final volume of 40 l using 20 pmol of oligo(dT) 15 in the presence of 200 units of SuperScript TM II RNase H Ϫ reverse transcriptase (Invitrogen). Subsequent PCR amplification of cDNA encoding VEGF was carried out in a total volume of 20 l containing 0.5 units of rTaqDNA polymerase and 10 pmol of each primer. Primer pairs for PCR were as follows (40): for VEGF (475 bp), forward 5Ј-ACGACAGAAGGGGAGCA-GAAAG-3Ј, reverse 5Ј-GGAACGTTGCGCTCAGACACA-3Ј. Amplification of cDNA encoding glyceraldehyde-3-phosphate dehydrogenase (494 bp) was performed for semi-quantitative normalization using the following primers: forward, 5Ј-ACCACAGTCCATGCCATCAC-3Ј, and reverse, 5Ј-TCCACCACCCTGTTGCTGTA-3Ј. The amplified fragments were separated on a 2% agarose gel containing ethidium bromide and visualized with an image analyzing device (Vilber Lourmat, France) under UV illumination. The bands on the images were quantitated with the image analyzing software, ImageJ (National Institutes of Health, Bethesda, Washington, D. C.).
Measurement of NO Release-NO production by BAEC was measured as nitrite (a stable metabolite of NO) concentration in cell culture supernatants, as described in many previous studies (41), with minor modifications. Briefly, after cells were treated with troglitazone for the indicated times in the absence or presence of various chemicals, the culture medium was changed to Kreb's solution (pH 7.4; 1.5 ml/60-mm dish), which contained 118 mM NaCl, 4.6 mM KCl, 27.2 mM NaHCO 3 , 1.2 mM MgSO 4 , 2.5 mM CaCl 2 , 1.2 mM KH 2 PO 4 , and 11.1 mM glucose, and was equilibrated for 1 h at 37°C. At the end of the incubation, 200 l of each supernatant (in Kreb's solution) was carefully transferred into a 96-well plate, with the subsequent addition of 100 l of Griess reagent (50 l of 1% sulfanilamide containing 5% phosphoric acid and 50 l of 0.1% N-(1-naphthyl)ethylenediamine). After color development at room temperature for 10 min, the absorbance was measured on a microplate reader at a wavelength of 520 nm. Each sample was assayed in triplicate wells. A calibration curve was plotted using known amounts of sodium nitrate solution. With this protocol, the measured values represent the amounts of NO produced by the cells during the 1-h incubation in Kreb's solution, following troglitazone treatment of a specified duration in the absence or presence of various chemicals. Therefore, subsequent NO production was solely dependent on eNOS activity at the end of these treatments.
Statistical Analysis-All results are expressed as the means Ϯ standard deviation (S.D.), with n indicating the number of experiments. Statistical significance was determined by Student's t test for two points. All differences were considered significant at a p value of Ͻ 0.05.

Troglitazone Increases NO Production in BAEC with
No Alteration in eNOS Expression-Troglitazone increased NO production by BAEC in a dose-and time-dependent manner, as shown in Fig. 1. The maximal increase in NO levels (3.1 Ϯ 0.42-fold of the control) was observed with 20 M troglitazone treatment for 12 h. Longer incubation (24 h) of cells with 20 M troglitazone caused no further increase. Therefore, all of the subsequent experiments were performed using these conditions. Western blot analysis revealed that the troglitazonestimulated increase in NO production did not result from an increase in eNOS protein expression (Fig. 1B), suggesting that classical intracellular genomic activity is not responsible for the observed effect.
Troglitazone Increases NO Production by Up-regulating the Expression of VEGF and Its Receptor, KDR/Flk-1-Because troglitazone has been previously reported to increase the in vivo and in vitro expression of VEGF (42), a well known agonist of NO production in endothelial cells, we next tested whether this troglitazone-stimulated NO increase is mediated by the up-regulation of VEGF. Troglitazone increased the expression of VEGF mRNA in a dose-and time-dependent manner, as shown in Fig. 2 (A and B). The maximal increase (6.23 Ϯ 1.33-fold of the control) was observed with 20 M troglitazone treatment for 24 h, although a significant increase (4.12 Ϯ 0.53-fold of the control) was also found with the 12-h treatment. In concert with the increase in VEGF mRNA levels, troglitazone also increased VEGF protein expression in a dose-dependent manner (Fig. 2C). The maximal increase (3.73 Ϯ 1.09-fold of the control) was observed with 20 M troglitazone treatment for 12 h. A similar dose response was also apparent in the troglitazone-stimulated increase in the expression of the VEGF receptor, KDR/Flk-1 (Fig. 2C), peaking at 20 M. To determine whether these increases in VEGF and KDR/Flk-1 expression mediate troglitazone-stimulated NO production in BAEC, we used anti-VEGF antibody to scavenge the increased VEGF induced by troglitazone. Anti-VEGF antibody partially (ϳ40%), but significantly, reduced troglitazone-stimulated NO production, as shown in Fig. 2D, suggesting that troglitazone-stimulated NO production in BAEC may be mediated, at least in part, by the VEGFϪKDR/Flk-1 signaling pathway.
Troglitazone Increases NO Production in BAEC by Increasing Either eNOS-Ser 1179 Phosphorylation or eNOS-Ser 116 Dephosphorylation-Because phosphorylation of eNOS-Ser 1179 is a major mechanism for VEGF-mediated increase in NO production (43,44), we examined whether troglitazone increases NO production by the stimulation of the phosphorylation of this residue. Troglitazone (20 M) increased eNOS-Ser 1179 phosphorylation in a time-dependent manner, as shown in Fig. 3, suggesting a potential role for troglitazone in eNOS-Ser 1179 phosphorylation. The maximal increase (2.4 Ϯ 0.53-fold of the control) was observed with the 12-h treatment. With longer incubation (24 h) of the cells with troglitazone, however, a slight but significant attenuation was observed (1.8 Ϯ 0.17-fold of the control). Furthermore, using antibodies specific to p-eNOS-Thr 497 and p-eNOS-Ser 116 , we also observed that troglitazone significantly stimulated the dephosphorylation of eNOS-Ser 116 in a time-dependent manner (Fig. 3), whereas the phosphorylation levels of eNOS-Thr 497 were not altered by troglitazone treatment for up to 12 h. It should be noted that eNOS-Ser 116 dephosphorylation by troglitazone occurred at an earlier time point (6 h) than eNOS-Ser 1179 phosphorylation and that a slight but significant increase in the phosphorylation of eNOS-Thr 497 was also observed with longer exposure (24 h). Our current results, together with previous findings of a positive role for both eNOS-Ser 1179 phosphorylation and eNOS-Ser 116 dephosphorylation in increasing NO production (30,32), indicate that troglitazone-dependent coordinated changes in the phosphorylation levels of eNOS at these two different residues are primarily involved in the troglitazone-stimulated NO increase in BAEC.
Troglitazone-induced Increases in eNOS-Ser 1179 Phosphorylation and eNOS-Ser 116 Dephosphorylation Are Mediated by at Least Two Separate Signaling Pathways-We examined whether either the phosphorylation or dephosphorylation of eNOS at the Ser 1179 or Ser 116 residues, respectively, is the consequence of a troglitazone-stimulated VEGFϪKDR/Flk-1mediated signaling pathway. Co-treatment with anti-VEGF antibody and troglitazone significantly reversed the eNOS-Ser 1179 phosphorylation induced by troglitazone, whereas no alteration in the phosphorylation levels of either eNOS-Thr 497 or eNOS-Ser 116 was observed, as shown in Fig. 4 (A-C). Furthermore, anti-VEGF antibody also completely reversed the troglitazone-stimulated increase in the phosphorylation of Akt, an intermediate signaling molecule that mediates VEGF-eNOS-Ser 1179 phosphorylation (Fig. 4D). Troglitazone also stimulated Akt phosphorylation in a time-dependent manner, showing a very similar pattern to the phosphorylation of eNOS-Ser 1179 induced by troglitazone (data not shown). These results, together with the finding that anti-VEGF antibody partially attenuates troglitazone-stimulated NO increase (Fig.  2D), suggest that troglitazone increased NO production in BAEC in part by increasing VEGF-KDR/Flk-1-Akt-mediated eNOS-Ser 1179 phosphorylation. In contrast, because anti-VEGF antibody did not alter the level of phosphorylation of  B). The cell proteins were separated on a sodium dodecyl sulfatepolyacrylamide gel and electrophoretically transferred onto nitrocellulose membranes. eNOS protein was detected by Western blot analysis using antibody specific for eNOS and quantitated by densitometry (B). Each bar represents the mean NO production (after normalization to total cellular protein) as fold increases above control Ϯ S.D. The values were considered statistically significant at p Ͻ 0.05 (*) and p Ͻ 0.01 (**) (n ϭ 5-7). In BAEC, troglitazone increased NO production in a doseand time-dependent manner, peaking at 20 M treatment for 12 h, without increasing eNOS protein.
eNOS-Ser 116 (Fig. 4C) and eNOS-Ser 116 dephosphorylation has a potential role in the troglitazone-stimulated increase in NO production, we hypothesize that the VEGF-KDR/Flk-1-Akt-mediated pathway is not an upstream signaling pathway for eNOS-Ser 116 dephosphorylation.
This hypothesis was further examined using BADGE, a PPAR␥ antagonist (45,46). BADGE (5 M) partially (ϳ40%) attenuated the troglitazone-stimulated increase in NO production, as shown in Fig. 5A. Furthermore, BADGE completely blocked KDR/Flk-1 expression stimulated by troglitazone (Fig.  5B). Like anti-VEGF antibody, BADGE significantly attenuated the phosphorylation of Akt and eNOS-Ser 1179 induced by troglitazone but not the phosphorylation of eNOS-Thr 497 and eNOS-Ser 116 (Fig. 5, C-F). These data further support the premise that an increase in eNOS-Ser 1179 phosphorylation, but not in eNOS-Ser 116 dephosphorylation, is the consequence of an effect of troglitazone as a PPAR␥ agonist. Taken together, our data clearly show that the troglitazone-stimulated increase The nitrocellulose membranes were reprobed with antibody detecting total eNOS to monitor equal loading of the samples. Densitometry was performed to quantitate the phosphorylated bands, and graphs show the mean fold increases above control Ϯ S.D., as described in legend to Fig. 1 (n ϭ 5-8). In BAEC, troglitazone increased dephosphorylation of eNOS-Ser 116 at the earliest time point (6 h), followed by phosphorylation of eNOS-Ser 1179 at 12 h and phosphorylation of eNOS-Thr 497 at 24 h. Whereas troglitazone continuously reduced the phosphorylation of eNOS-Ser 116 over the entire experimental period (for 24 h), the pattern of increasing eNOS-Ser 1179 phosphorylation by troglitazone was reversed after 12 h. genase mRNA was used for semi-quantitative normalization (A and B). The expression of VEGF and its receptor, KDR/Flk-1 (C), was measured by Western blot analysis using specific antibodies, as described in the legend to Fig. 1. NO released by troglitazone was measured in the absence or presence of anti-VEGF antibody (200 g/ml) (D). The data are presented as fold increases above control Ϯ S.D. Statistically significant at p Ͻ 0.05 (* and #) and p Ͻ 0.01 (** and ##) (n ϭ 7-12). In BAEC, troglitazone increased VEGF mRNA in a dose-(A) and time-dependent manner (B). It also increased VEGF protein and its receptor in a dose-dependent manner (C). Anti-VEGF antibody significantly attenuated the troglitazone-stimulated increase in NO production (D), suggesting that troglitazone increases NO production in part by a VEGF-KDR/Flk-1 signaling pathway. DMSO, dimethyl sulfoxide. in NO production is mediated by two independent signaling pathways: PPAR␥-dependent, VEGF-KDR/Flk-1-Akt-mediated eNOS-Ser 1179 phosphorylation and PPAR␥-independent, eNOS-Ser 116 dephosphorylation. DISCUSSION Recently, a single study showed that the PPAR␥ ligands 15d-PGJ 2 and ciglitazone increase NO release in endothelial cells (36). The mechanisms underlying this effect of PPAR␥ ligands remain undefined. However, the authors of that report suggested that PPAR␥ ligands regulate the expression of another target gene rather than eNOS itself, which promotes NO release, particularly because 15d-PGJ 2 increased NO production but decreased eNOS protein expression. In this study, we have further characterized this molecular mechanism and shown for the first time that troglitazone, another PPAR␥ ligand, increases NO production in endothelial cells by the molecular activation of eNOS through alterations in its phosphorylation status.
Two lines of evidence in this study suggest that there are two independent signaling pathways that may account for troglitazone-stimulated NO production: PPAR␥-dependent, VEGF-KDR/Flk-1-Akt-mediated eNOS-Ser 1179 phosphorylation and PPAR␥-independent, eNOS-Ser 116 dephosphorylation. Firstly, we found that only VEGF-KDR/Flk-1-Akt-mediated eNOS-Ser 1179 phosphorylation was inhibited by anti-VEGF antibody, resulting in a partial attenuation of troglitazone-stimulated NO production. However, anti-VEGF antibody did not alter eNOS-Ser 116 dephosphorylation. Secondly, BADGE, a PPAR␥ antagonist, also partially attenuated troglitazone-stimulated NO increase by suppressing the signaling pathway responsible for VEGF-KDR/Flk-1-Akt-mediated eNOS-Ser 1179 phosphorylation but did not alter eNOS-Ser 116 dephosphorylation.
It has previously been reported that TZDs up-regulate VEGF mRNA in VSMC (47) and blood VEGF in patients with type 2 DM (42). These results are reproduced reasonably closely in our current study using BAEC. Furthermore, the up-regulation of VEGF protein by troglitazone coincides with the up-regulation of its mRNA, suggesting that troglitazone increases VEGF expression, perhaps by regulating its transcription. Nonetheless, it seems likely that the effect of troglitazone on VEGF mRNA expression may be mediated by unknown transcription factor(s) rather than by a direct binding of this drug and its receptor complex to the promoter region of the VEGF gene, because VEGF mRNA is devoid of a PPAR␥ response element in its promoter (48 -51). In this regard, several transcription factors and co-activators such as activator protein-1, cAMP response element binding protein/p300, hypoxia inducible factor-1, and stimulatory protein-1 were reported to be involved in the PPAR␥-activated increase in VEGF mRNA expression (49). Up-regulation of the VEGF receptor, KDR/Flk-1, was also observed in response to chronic troglitazone treatment, to our zone for 12 h, as described in the legend to Fig. 1, in the absence or presence of anti-VEGF antibody (200 g/ml). After treatment, the cell lysates were prepared, and eNOS and p-eNOS were analyzed as described in the legend to Fig. 3. The expression of Akt and p-Akt-Ser 473 was measured by Western blot analysis using antibodies specific for those proteins. The data are presented as described in the legend to Fig.  1 (n ϭ 5-8). In BAEC, co-treatment with anti-VEGF antibody significantly blocked the troglitazone-stimulated increase in eNOS-Ser 1179 phosphorylation (A) and completely blocked the increase in Akt-Ser 473 phosphorylation (D), but it did not alter the phosphorylation status of eNOS-Thr 497 (B) or eNOS-Ser 116 (C). These results, together with the finding that anti-VEGF antibody partially inhibited the troglitazonestimulated increase in NO production (Fig. 2D), indicate that troglitazone increases NO production in part by VEGF-Akt-mediated eNOS-Ser 1179 phosphorylation, which is independent of eNOS-Ser 116 dephosphorylation. DMSO, dimethyl sulfoxide. The mechanism underlying this up-regulation remains unclear, but it is speculated that the increase in VEGF induced by troglitazone may in turn induce the expression of its receptor (52,53). At present, it is reasonable to infer that the activation of Akt, as a downstream signaling molecule of VEGF, follows the VEGF up-regulation elicited by troglitazone treatment. However, it should be noted that chronic exposure to troglitazone may also utilize the same Akt-dependent mechanism to exert its cellular effect, such as an alteration in Akt phosphorylation but not in Akt expression, a phenomenon observed during acute exposure to various agonists (25). Further studies using other signaling molecules are required to generalize this concept. The activation of the VEGF signaling pathway by troglitazone, together with the partial inhibition of troglitazone-stimulated NO production by anti-VEGF antibody, clearly suggests that the increase in VEGF After treatment, the cell lysates were further processed as described in the legend to Fig. 4. The data are presented as described in the legend to Fig. 1 (n ϭ 5-8). In BAEC, co-treatment with BADGE partially blocked the troglitazone-stimulated increase in NO production (A) and the increase in eNOS-Ser 1179 (C) and Akt-Ser 473 phosphorylation (F). Similarly, BADGE also completely blocked the troglitazone-stimulated increase in KDR/Flk-1 expression (B). In contrast, BADGE did not alter the phosphorylation status of eNOS-Thr 497 (D) or eNOS-Ser 116 (E), suggesting that troglitazone increases NO production, in part, by PPAR␥-dependent, VEGF-KDR/ Flk-1-Akt-mediated eNOS-Ser 1179 phosphorylation and PPAR␥-independent, eNOS-Ser 116 dephosphorylation. DMSO, dimethyl sulfoxide. expression induced by troglitazone may account, at least in part, for troglitazone-stimulated NO production. However, unlike previous studies that showed a significant up-regulation of eNOS protein by chronic VEGF treatment in human umbilical vein endothelial cells (44) and bovine adrenal cortex endothelial cells (54), we were unable to detect any up-regulation of eNOS expression (Fig. 2). The reasons for these apparently conflicting results are unclear but may be related to the cell type studied, i.e. BAEC as opposed to either human umbilical vein endothelial cells or bovine adrenal cortex endothelial cells. Alternatively, it may also be speculated that the amount of VEGF increase elicited by troglitazone under our experimental conditions was insufficient to cause an increase in eNOS expression.
We found that BADGE partially blocked troglitazone-stimulated phosphorylations of Akt and eNOS at the Ser 1179 residue in troglitazone-stimulated cells, whereas it did not alter the phosphorylation of eNOS-Ser 116 . Furthermore, a higher dose (10 M) of BADGE even stimulated troglitazone-induced eNOS-Ser 116 dephosphorylation (data not shown). This result clearly suggests that an alteration in the phosphorylation status of eNOS-Ser 116 by troglitazone is not PPAR␥-dependent, whereas VEGF-KDR/Flk-1-Akt-mediated eNOS-Ser 1179 phosphorylation is PPAR␥-dependent. Although the effects of troglitazone on various cellular functions are predominantly mediated by PPAR␥, it has also been suggested that troglitazone can exert PPAR␥-independent effects. For example, 15d-PGJ 2 inhibited IB kinase, thus interrupting NF-B signaling by PPAR␥-independent signaling pathways (55,56). Furthermore, troglitazone inhibited protein kinase C (PKC) activity in a PPAR␥independent manner in VSMC (57). These data, together with an earlier finding that PKC inhibitor blocks the phosphorylation of eNOS-Ser 116 (30), suggest a potential role for PKC in troglitazone-induced eNOS-Ser 116 dephosphorylation in BAEC. Therefore, possible interactions between the regulation of PKC activity and troglitazone should be assessed in future to clarify this issue.
As shown in all of the figures in this study, a temporally distinct phosphorylation pattern of eNOS exists at any given time at these three residues. It should be noted that the observed NO levels are dependent on the total phosphorylation status of eNOS at these residues and at other known eNOS residues as well. For example, using this concept, the absence of any further increase in NO production induced by troglitazone with 24 h of treatment (compared with 12 h) (Fig. 1B) might be attributable to the combined effects of both decreased dephosphorylation at eNOS-Ser 1179 and eNOS-Ser 116 and increased phosphorylation at eNOS-Thr 497 . Although PKC (29) and AMP-activated protein kinase (20) regulate the phosphorylation status of eNOS-Ser 1179 and eNOS-Thr 497 , it is unclear at present whether these enzymes also play a role in troglitazone-induced eNOS regulation at a later time. One important outcome of the time course study of the troglitazone-stimulated increase in NO production is that the phosphorylation and dephosphorylation of eNOS at these three residues display different kinetics. The earliest response to troglitazone was eNOS-Ser 116 dephosphorylation, followed by eNOS-Ser 1179 phosphorylation and eNOS-Thr 497 phosphorylation.
Our current data show that NO production is stimulated by troglitazone via changes in eNOS phosphorylation. This offers a novel and potentially unifying mechanism for the reduction in blood pressure elicited by TZD drugs (9). In contrast to our findings, it has been also reported that TZD inhibits NO in VSMC by the suppression of iNOS expression (34,35). However, in this study, we were unable to detect iNOS expression in BAEC (data not shown), suggesting that the eNOS-dependent mechanism proposed in this study accounts for the ability of troglitazone to modulate NO production in endothelial cells. Our results, together with previous finding that troglitazone stimulates the suppression of iNOS, suggest that the overall vascular protective effects of troglitazone on lipid metabolism, inflammation, and vasodilation may be attributable to the coordinated actions of two vascular cells, endothelial cells and VSMC, through the regulation of eNOS and iNOS, respectively. Furthermore, our data also suggest that care should be exercised in the use of troglitazone in patients with type 2 DM carrying retinopathy or cancer, because NO in endothelial cells is implicated in stimulating angiogenesis (58).
In summary, the most important and original finding of this study is that chronic treatment with troglitazone increases NO production by at least two independent pathways: PPAR␥-dependent, VEGF-KDR/Flk-1-Akt-mediated eNOS-Ser 1179 phosphorylation and PPAR␥-independent, eNOS-Ser 116 dephosphorylation. A better understanding of the mechanisms involved in troglitazone-induced eNOS regulation may help to provide new strategies to modify the pathophysiology of cardiovascular diseases, such as hypertension and atherosclerosis, as well as type 2 DM.