Rosiglitazone Reduces Glucose-Induced Oxidative Stress Mediated by NAD(P)H Oxidase via AMPK-Dependent Mechanism

【摘要】  Objective— Hyperglycemia is the main determinant of long-term diabetic complications, mainly through induction of oxidative stress. NAD(P)H oxidase is a major source of glucose-induced oxidative stress. In this study, we tested the hypothesis that rosiglitazone (RSG) is able to quench oxidative stress initiated by high glucose through prevention of NAD(P)H oxidase activation.

Methods and Results— Intracellular ROS were measured using the fluoroprobe TEMPO-9-AC in HUVECs exposed to control (5 mmol/L) and moderately high (10 mmol/L) glucose concentrations. NAD(P)H oxidase and AMPK activities were determined by Western blot. We found that 10 mmol/L glucose increased significantly ROS production in comparison with 5 mmol/L glucose, and that this effect was completely abolished by RSG. Interestingly, inhibition of AMPK, but not PPAR, prevented this effect of RSG. AMPK phosphorylation by RSG was necessary for its ability to hamper NAD(P)H oxidase activation, which was indispensable for glucose-induced oxidative stress. Downstream of AMPK activation, RSG exerts antioxidative effects by inhibiting PKC.

Conclusions— This study demonstrates that RSG activates AMPK which, in turn, prevents hyperactivity of NAD(P)H oxidase induced by high glucose, possibly through PKC inhibition. Therefore, RSG protects endothelial cells against glucose-induced oxidative stress with an AMPK-dependent and a PPAR -independent mechanism.

The present study was designed to characterize the molecular mechanisms underlying the effects of rosiglitazone on hyperglycemia-induced ROS production in HUVECs. We demonstrate that rosiglitazone reduces glucose-induced oxidative stress through inhibition of NAD(P)H oxidase. This effect is not mediated by PPAR but is dependent on AMPK activation and downstream PKC inhibition.

【关键词】  diabetes AMPK NAD(P)H oxidase oxidative stress rosiglitazone

Introduction

Chronic hyperglycemia induces protein glycation, systemic low grade inflammation, and endothelial dysfunction. 1 As a consequence, diabetes is one of the main risk factor for cardiovascular disease. Hyperglycemia-induced endothelial dysfunction is characterized by an enhanced production of reactive oxygen species (ROS), which are important actors in the development of vascular damage. Consistently, antioxidant agents are able to rescue hyperglycemia-induced vascular dysfunction. 1,2

Recently, we have shown that, in human umbilical vein endothelial cells (HUVECs), high glucose (10 mmol/L) increases ROS generation through a NAD(P)H oxidase–dependent mechanism. 3 Furthermore, vascular NAD(P)H oxidase activity is increased in diabetic patients in vivo, and endothelial NAD(P)H oxidase activity is markedly increased by high glucose levels in vitro. 4,5 Therefore, NAD(P)H oxidase appears as one of the major sources of ROS production after exposure to hyperglycemia. 6

Thiazolidinediones (TDZs) are used clinically in type 2 diabetic patients by virtue of their insulin-sensitizing action, conveyed by the activation of the nuclear transcription factor peroxisome proliferator-activated receptor- (PPAR ). 7 In addition, these agents have remarkable pleiotropic activities: by improving endothelial function and systemic inflammation, they are expected to exert direct beneficial effects on cardiovascular risk, which are not mediated by the improvement in glucose metabolism. In this regard, pioglitazone was shown to abolish ROS production in 3T3-L1 adipocytes, 8 whereas rosiglitazone (RSG) reduced NAD(P)H-stimulated superoxide production in aortas from diabetic mice, 9 and troglitazone diminished ROS generation in leukocytes from obese subjects. 10 However, the molecular mechanism by which TZDs attenuate oxidative stress is not clear.

TDZs are specific ligands for the PPAR family of nuclear receptors, which are intimately involved in the regulation of energy homeostasis. 7 However, several evidences suggest that TDZs affect nitric oxide (NO), tumor necrosis factor (TNF)-alpha production and endothelial cell proliferation by PPAR -independent mechanisms. 11–14 Recently, it has been demonstrated that TDZs also activate 5'-AMP-activated protein kinase (AMPK), which represents the major regulator of cellular and systemic energy homeostasis in liver and muscle. Moreover, AMPK plays an important role in protecting endothelial cells against the adverse effects of sustained hyperglycemia. 15–18 Therefore, the aim of the present study was to dissect the molecular mechanisms underlying the effects of RSG on hyperglycemia-induced ROS production.

Materials and Methods

For detailed methods, please see the supplemental Materials and Methods, available at online at http://atvb.ahajournals.org. Human umbilical vein endothelial cells (HUVECs) were incubated for 48 hours in the presence of either normal (5 mmol/L) or elevated (10 mmol/L) glucose concentration, with or without RSG (20 µmol/L). Oxidative stress was measured by fluorescence microscopy and by electron spin resonance (ESR) using the free radical probe TEMPO-9AC. For siRNA experiments, cells were transfected with 15 nmol/L siRNA using HiPerFect transfection reagent (Qiagen) according to the manufacturer?s protocol. Quantitative RT-PCR was performed using an iCycler iQ system (BIO-RAD) and SYBR Green detection. Proteins were analyzed by Western immunoblot using standard procedures.

Results

Effects of RSG on ROS Production Induced by High Glucose

We have previously demonstrated that high glucose increases ROS production. 3 Here we first investigated the effect of RSG on this phenomenon, by incubating HUVECs for 48 hours with RSG (20 µmol/L) plus high glucose (10 mmol/L). Figure 1 shows a representative fluorescence image measured by inverted microscope at the beginning (1 minute) and after 10 minutes of the exposure to the TEMPO-9-AC probe. As shown, fluorescence markedly increased in cells exposed to 10 mmol/L glucose ( Figure 1 B) in comparison to 5 mmol/L glucose ( Figure 1 A). After 10 minutes incubation, fluorescence increases in proportion to the probe influx ( Figure 1 B). RSG significantly prevented the increase in ROS production induced by 10 mmol/L glucose ( Figure 1 C). Figure 1 D summarizes these effects: 10 mmol/L glucose-stimulated ROS production in comparison to cells exposed to 5 mmol/L glucose (from 14.8±1.5 to 45±2 fluorescence rate, A.U., P<0.005) and RSG completely blunted this effect (from 45±2 to 17.01±2.5 fluorescence rate, A.U., P<0.01). Supplemental Figure IA and IB shows the dose-response and the time-course of RSG action on ROS production induced by 10 mmol/L glucose: ROS generation was significantly inhibited after at least 24-hour incubation with RSG concentrations between 20 and 50 µmol/L.

Figure 1. Measurement of ROS in HUVECs. Fluorescence images for ROS in cells grown at 5 mmol/L glucose (A); 10 mmol/L glucose (B); 10 mmol/L glucose plus 20 µmol/L RSG (C), at 1 (left) and 10 (right) minutes. D, Bar graph representations. Mean±SE of 10 experiments. * P <0.01. RSG indicates rosiglitazone.

To dissect the mechanisms involved in the inhibition of glucose-induced ROS production by RSG, we explored the 2 molecular pathways activated by TDZs: AMPK and PPAR. Therefore, we determined the effects of a specific AMPK activator, 5-aminoimidazole-4-carboxamide-1-beta- D -ribofuranoside (AICAR) and GW9662, a specific PPAR inhibitor, on ROS production induced by 10 mmol/L of glucose. As shown in supplemental Figure II, AICAR (500 µmol/L, 48 hours) decreased ROS production induced by high glucose (10 mmol/L, 48 hour), similarly to what observed with RSG alone. On the contrary, GW9662 (2 µmol/L, 48 hour) did not alter the effects of RSG on ROS production, suggesting that PPAR receptors are not directly involved in this phenomenon.

To validate these results, we also studied the effects of GW1929, a PPAR tyrosine derivate agonist by ESR technique. As shown in supplemental Figure IIB, the PPAR -agonist GW1929 (20 µmol/L, 48 hours) did not reduce glucose-induced ROS production, indicating that PPAR is not the mediator of the inhibition of glucose-dependent ROS production by RSG.

Effects of RSG on AMPK Activation, In Vitro and In Vivo

In vitro, in cells grown at 5 mmol/L of glucose, the time-course of AMPK activation induced by RSG is shown in supplemental Figure III, which reports a representative immunoblot of AMPK phosphorylation (p-AMPK). AMPK phosphorylation increased after 20 minutes and reached a significant activation after 24 hours RSG (20 µmol/L; supplemental Figure IIIA). In addition, RSG-induced AMPK phosphorylation was inhibited by Compound C (100 µmol/L, 48 hours), a specific AMPK inhibitor, although it was not affected by GW9662 (2 µmol/L, 48 hours) (supplemental Figure IIIB).

In vivo, in adult male Sprague-Dawley rats, ischemia/reperfusion significantly increased AMPK by 228±26% (P<0.01) as compared with basal state. Seven-day RSG treatment, at the dose of 15 mg/kg body weight/d, activated AMPK in rat skeletal muscles in basal conditions (+147±15%, P<0.05) and during ischemia/reperfusion injury (+258±18%, P<0.01 versus basal) with a net +30% increase, although not significantly different, when compared with control study (supplemental Figure IV).

Next, to investigate the functional role of AMPK activation by RSG on the ability of RSG to inhibit glucose-induced ROS production, we silenced AMPK gene expression using siRNA. As shown in Figure 2, transfection with AMPK 1 siRNA for 72 hours efficiently reduced AMPK 1 mRNA level by 77% and AMPK 1 protein expression by 70%. Furthermore, in AMPK 1 siRNA cells, AMPK phosphorylation by RSG was markedly decreased in comparison with control siRNA cells ( Figure 2C and 2 D).

Figure 2. AMPK expression and phosphorylation. HUVECs were transfected with siAMPK-1 or control scrambled siRNA (siCnt). A, AMPK-1 expression in siAMPK1 transfected cells at 24 hours, 48 hours, and 72 hours. B, Immunoblots of transfected cells for AMPK and GADPH. C, RSG-induced AMPK phosphorylation in siAMPK1 and control siRNA cells at 10 mmol/L glucose, with RSG or with AICAR. (n=6) * P <0.01.

Then, we measured glucose-induced ROS production in AMPK-silenced cells by fluorescence microscope and ESR techniques. In AMPK1 siRNA-transfected cells, reduction of AMPK 1 expression abolished the inhibitory effects of 20 µmol/L RSG on ROS production induced by 10 mmol/L glucose (supplemental Figure VC) in comparison to cells transfected with control siRNA in presence of RSG (supplemental Figure VC). A similar effect was observed when the cells were incubated for 48 hours with 100 µmol/L Compound C, an AMPK inhibitor (supplemental Figure VC). The fluorescence results were always confirmed by ESR measurements (supplemental Figure VE and VF).

Effect of RSG on p22 phox, p47 phox, gp91 phox, p67 phox, rac-1 Protein Expression, and Translocation

The NAD(P)H oxidase is made up of 2 membrane subunits (p22 phox and gp91 phox ) and 3 cytosolic subunits (p47 phox, p67 phox, and Rac-1). When the NAD(P)H oxidase is stimulated, the cytosolic subunits translocate from the cytosol to the membrane and induce ROS production. 20 To further describe the effects of RSG on NAD(P)H oxidase, we studied expression and translocation of NAD(P)H oxidase subunits in the presence of glucose 10 mmol/L with or without RSG. p22 phox protein expression was increased 2-fold (P<0.01) by 10 mmol/L glucose in comparison with 5 mmol/L glucose, and this effect was inhibited by RSG. Remarkably, this action of RSG was antagonized by Compound C (100 µmol/L, 48 hours). Total p47 phox, gp91 phox, p67 phox, and rac-1 expression were unaffected by high glucose (supplemental Figure VI).

Then, we found that high glucose induced the translocation of p47 phox and Rac-1 from the cytosol to the membrane ( Figure 3A through 3 C). The activation of these NAD(P)H isoforms was prevented by the presence of the PKC inhibitor GF109203X, but not by the presence of PP2, a Src inhibitor, and by the presence of L-NAME, an NO inhibitor ( Figure 3A through 3 C). RSG prevented glucose-induced p47 phox and rac-1 translocation to the plasma membrane, but not in AMPK1 siRNA- transfected cells or in the presence of Compound C (100 µmol/L, 48 hour; Figure 3B through 3 D).

Figure 3. Effect of RSG and AMPK inhibitor on p47 phox and rac-1 translocation. A through C, Cells exposed to 5 and 10 mmol/L glucose ±20 µmol/L RSG, GF109203X 5 µmol/L, PP2 5 µmol/L, and L-NAME 100 µmol/L for 48 hours. B through D, Cells exposed to 5 (Glu 5) and 10 mmol/L of glucose (Glu 10) for 48 hours, to RSG (20 µmol/L) alone, in the presence of Compond C (100 µmol/L), and in siAMPK1 cells. Data expressed as mean±SE of 3 experiments. * P <0.01. RSG indicates rosiglitazone.

These observations were confirmed by the determination of glucose-induced ROS production in the presence of GF109203X, and of L-NAME by fluorescence microscopy. GF109203X, a PKC inhibitor, significantly attenuated ROS production induced by glucose, while L-NAME did not have any effect (supplemental Figure VII).

Further, we measured cellular DAG levels and PKC activity to verify the interactions between RSG and PKC. As expected, cells exposed for 48 hours to 10 mmol/L of glucose resulted in significant increase in total DAG levels and PKC activity in comparison with the cells exposed to 5 mmol/L of glucose (supplemental Figure VIII). Treatment with RSG significantly reduced both the DAG levels and PKC activity. This effect was not inhibited by the incubation of cells with GW9662, but it was abolished by the presence of Compound C, suggesting that PKC is a target of RSG downstream of AMPK activation.

Discussion

In this study, we have demonstrated that RSG prevents glucose-induced oxidative stress in endothelial cells, an effect independent from PPAR, but distinctively dependent on AMPK activation. We also showed that the ability of RSG to quench oxidative stress is conveyed through the inhibition of NAD(P)H oxidase. Furthermore, we demonstrated that, downstream of AMPK activation, the effect of RSG on glucose-induced NAD(P)H oxidase–derived ROS production is mediated by the inhibition of the DAG-PKC pathway ( Figure 4 depicts this model schematically).

Figure 4. Putative mechanisms by which RSG prevents glucose-induced oxidative stress. RSG activates PPAR and AMPK through distinct intracellular pathways. Distinctively, RSG blunts PKC activation by high glucose via AMPK, thus preventing NAD(P)H oxidase to release excessive ROS.

Many studies have reported that TZDs act through PPAR -dependent mechanisms, and this is also true in endothelial cells. For instance, RSG increased NO production in HUVECs through a transcriptional mechanism unrelated to eNOS expression but dependent on PPAR activation. 12,20 Interestingly, this effect has been attributed to the inhibition of NO quenching by NAD(P)H oxidase–derived ROS. 21 The PPAR agonists pioglitazone and RSG also exert direct antiinflammatory effects by interfering with monocyte chemoattractant protein-1 and its receptor, CCR2. 22,23 TZDs may have cardiovascular pleiotropic effects that are independent of their actions on glucose and lipid metabolism. In facts, clinical trials show that RSG ameliorated vascular function beyond its anti-hyperglycemic effects. 24,25 The clinical potential benefits of TZDs have been underscored by several studies: in type 2 diabetic patients, RSG reduced serum levels of matrix metallo-proteinase-9 and C-reactive protein, 26 whereas pioglitazone reduced carotid intima-media thickness, 27 further suggesting a possible role in slowing atherosclerosis. A substantial part of these positive effects of TZDs is mediated by distinctive antioxidative properties, 28,29 especially in the setting of glucose-induced oxidative stress. 30,31

In this study, we demonstrate that RSG significantly decreases glucose-induced oxidative stress and that this effect is independent of its ability to activate PPAR. In support of this, we show that a specific PPAR antagonist (GW9662), which is able to fully prevent PPAR transactivation by RSG, 12 did not abolish the antioxidant action of RSG. This observation is also sustained by previous works. Davies and colleagues have shown that, in isolated hepatocytes, troglitazone inhibits the expression of the PEPCK gene by a PPAR -independent, antioxidant-related mechanism. 32 Similarly, Lennon and colleagues demonstrated that another PPAR agonist, ciglitazone, activates p38 MAP kinase through a PPAR -independent mechanism. 33 Our present data suggest that most of the antioxidative activity of RSG is determined by its ability to activate AMPK. This is authenticated by the decreased production of ROS when cells were incubated with AICAR, an artificial activator of AMPK, and by the neutral effect of RSG on oxidative stress when AMPK was inhibited by Compound C or by knockdown by siRNA. Consistent with the notion that different TZDs have distinct pleiotropic effects, 7 our data indicate that the antioxidative property of RSG is not a "class effect" because the PPAR -agonist GW1929 was unable to prevent ROS induction by high glucose.

To our knowledge, this is also the first study showing that a member of the TZD family activates AMPK in endothelial cells. Several works suggest that PPAR agonists activate AMPK and have antioxidative properties, but the molecular connections between these 2 phenomena were previously unknown. We found that RSG activates AMPK after 24 hours of incubation and, in parallel, decreases glucose-induced ROS formation. LeBrasseur and colleagues found that troglitazone and pioglitazone activate AMPK in mammalian tissues within minutes. 34 It is possible that the PPAR -independent activation of AMPK requires a longer lag phase in endothelial cells. Our results emphasize the protective role of AMPK in endothelial cells, as already shown by an extensive literature. 35,36 Recently, we proved that ROS induction by high glucose is dependent on NAD(P)H oxidase. 3 The present results extend those observations by showing that the incubation of endothelial cells with RSG abolished glucose-induced activation of NAD(P)H oxidase and ROS production. A previous study reported that both PPAR and PPAR agonists decrease p22 phox gene expression in HUVECs. 37 Another study showed that ciglitazone downregulates phagocyte p47 phox oxidase. 38 Here, we add significant pieces of information to this picture, by showing that RSG inhibits not only p22 phox protein expression but also the translocation of both p47 phox and Rac-1, crucial components of NAD(P)H oxidase. Moreover, we show that the ability of RSG to inhibit NAD(P)H oxidase requires the activation of AMPK, because this effect is completely abolished by compound C. In fact, our data underscore that AMPK is an important inhibitor of NAD(P)H oxidase. This concept has been supported by previous works showing, for example, that -lipoic acid suppresses NAD(P)H oxidase activity by activating AMPK in human aortic endothelial cells. 39 AMPK is also able to modulate NAD(P)H in human neutrophils and in smooth muscle cell. 40,41

In compliance with our previous data showing that PKC is stimulated by hyperglycemia, 3,19 we noted that the hyperactivity of NAD(P)H oxidase induced by high glucose was prevented by PKC inhibition. In the attempt for a further dissection of the antioxidant mechanism of RSG, we show that RSG inhibits the DAG-PKC pathway and that, again, this effect requires AMPK activation. Therefore, downstream of AMPK phosphorylation, PKC is a critical target of the antioxidant effect of RSG. 42

Finally, to determine whether this effect of RSG occurs in vivo and whether there is evidence of AMPK activation in intact animals, we treated adult male Sprague-Dawley rats with oral RSG for 7 days. We show that RSG activates AMPK in vivo in basal conditions and during ischemia/reperfusion injury, a model in which oxidative stress plays a major role in determining tissue damage. AMPK was activated by ischemia per se, an event that is supposed to protect the target tissue from ischemic damage, and to stimulate neoangiogenesis. 43 Interestingly, it has been suggested that AMPK hyperactivation may also have negative impacts on some adaptive responses in the cardiovascular system. 44 Very recent clinical data warns that, despite its metabolic actions and its strong theoretically favorable effects of TZDs on the cardiovascular system, RSG may not decrease the incidence of myocardial infarction, or may increase it. 45 Complex effects of RSG on cell types other than the endothelium may be responsible for this outcome. Unfortunately, we could not monitor selectively AMPK activation in myocytes and endothelial cells. In any case, potentiation of AMPK signaling by RSG may have important clinical implications in the setting of ischemic diseases.

In conclusion, we report that RSG has a potent antioxidant effect which is not mediated by PPAR but is strictly dependent on its ability to activate AMPK. We also show that RSG reduces ROS mainly by inhibiting NAD(P)H oxidase, with a mechanism that is dependent on AMPK activation and related to PKC inhibition.

Acknowledgments

Sources of Funding

This work was funded by the Italian Ministry of University and Research (MIUR; to A.A., G.C., A.S.), Finanziamento di Ateneo 2005.

Disclosures

A.A. has received an unrestricted grant ( 15 000) from GSK, which was not used to carry out the present project.

G.C. and A.G. contributed equally to this study.

Original received November 17, 2006; final version accepted September 21, 2007.

【参考文献】
  Ceriello A, Motz E. Is oxidative stress the pathogenic mechanism underlying insulin resistance, diabetes, and cardiovascular disease? The common soil hypothesis revisited. Arterioscler Thromb Vasc Biol. 2004; 24: 816–823.

Aronson D, Rayfield EJ. How hyperglycemia promotes atherosclerosis: molecular mechanisms. Cardiovasc Diabetol. 2002; 1: 1–10.

Gallo A, Ceolotto G, Pinton P, Iori E, Murphy E, Rutter GA, Rizzuto R, Semplicini A, Avogaro A. Metformin prevents glucose-induced protein kinase C-beta2 activation in human umbilical vein endothelial cells through an antioxidant mechanism. Diabetes. 2005; 54: 1123–1131.

Martín-Gallán P, Carrascosa A, Gussinyé M, Domínguez C. Biomarkers of diabetes-associated oxidative stress and antioxidant status in young diabetic patients with or without subclinical complications. Free Radic Biol Med. 2003; 34: 1563–1574.

Schmidt AM, Yan SD, Stern DM. The dark side of glucose. Nat Med. 1995; 1: 1002–1004.

Wautier JL, Schmidt AM. Protein glycation: a firm link to endothelial cell dysfunction. Circ Res. 2004; 95: 233–238.

Yki-Jarvinen H. Thiazolidinedione. N Engl J Med. 2004; 351: 1106–1118.

Houstis N, Rosen ED, Lander ES. Reactive oxygen species have a causal role in multiple forms of insulin resistance. Nature. 2006; 440: 944–948.

Calkin AC, Forbes JM, Smith CM, Lassila M, Cooper ME, Jandeleit-Dahm KA, Allen TJ. Rosiglitazone attenuates atherosclerosis in a model of insulin insufficiency independent of its metabolic effects. Aterioscler Thromb Vasc Biol. 2005; 25: 1903–1309.

Miles PD, Higo K, Romeo OM, Lee MK, Rafaat K, Olefsky JM. Troglitazone prevents hyperglycemia-induced but not glucosamine-induced insulin resistance. Diabetes. 1998; 47: 395–400.

Sasaki M, Jordan P, Welbourne T, Minagar A, Joh T, Itoh M, Elrod JW, Alexander JS. Troglitazone, a PPAR-gamma activator prevents endothelial cell adhesion molecule expression and lymphocyte adhesion mediated by TNF-alpha. BMC Physiol. 2005; 5: 1–12.

Polikandriotis JA, Mazzella LJ, Rupnow HL, Hart CM. Peroxisome proliferator-activated receptor gamma ligands stimulate endothelial nitric oxide production through distinct peroxisome proliferator-activated receptor gamma-dependent mechanisms. Arterioscler Thromb Vasc Biol. 2005; 25: 1810–1816.

Liu HB, Hu YS, Medcalf RL, Simpson RW, Dear AE. Thiazolidinediones inhibit TNFalpha induction of PAI-1 independent of PPARgamma activation. Biochem Biophys Res Commun. 2005; 334: 30–37.

Artwohl M, Fürnsinn C, Waldhäusl W, Hölzenbein T, Rainer G, Freudenthaler A, Roden M, Baumgartner-Parzer SM. Thiazolidinediones inhibit proliferation of microvascular and macrovascular cells by a PPARgamma-independent mechanism. Diabetologia. 2005; 48: 586–594.

Lessard SJ, Chen ZP, Watt MJ, Hashem M, Reid JJ, Febbraio MA, Kemp BE, Hawley JA. Chronic rosiglitazone treatment restores AMPKalpha2 activity in insulin-resistant rat skeletal muscle. Am J Physiol Endocrinol Metab. 2006; 290: E251–E257.

Long YC, Zierath JR. AMP-activated protein kinase signaling in metabolic regulation. J Clin Invest. 2006; 116: 1776–1783.

Ouedraogo R, Wu X, Xu SQ, Fuchsel L, Motoshima H, Mahadev K, Hough K, Scalia R, Goldstein BJ. Adiponectin suppression of high-glucose-induced reactive oxygen species in vascular endothelial cells: evidence for involvement of a cAMP signalling pathway. Diabetes. 2006; 55: 1840–1846.

Kukidome D, Nishikawa T, Sonoda K, Imoto K, Fujisawa K, Yano M, Motoshima H, Taguchi T, Matsumura T, Araki E. Activation of AMP-activated protein kinase reduces hyperglycemia-induced mitochondrial reactive oxygen species production and promotes mitochondrial biogenesis in human umbilical vein endothelial cells. Diabetes. 2006; 55: 120–127.

Ceolotto G, Bevilacqua M, Papparella I, Baritono E, Franco L, Corvaja C, Mazzoni M, Semplicini A, Avogaro A. Insulin generates free radicals by an NAD(P)H, phosphatidylinositol 3' kinase dependent mechanism in human skin fibroblasts ex vivo. Diabetes. 2004; 53: 1344–1351.

Calnek DS, Mazzella L, Roser S, Roman J, Hart CM. Peroxisome proliferator-activated receptor gamma ligands increase release of nitric oxide from endothelial cells. Arterioscler Thromb Vasc Biol. 2003; 23: 52–57.

Hwang J, Kleinhenz DJ, Lassègue B, Griendling KK, Dikalov S, Hart CM. Peroxisome proliferator-activated receptor-gamma ligands regulate endothelial membrane superoxide production. Am J Physiol Cell Physiol. 2005; 288: C899–C905.

Ishibashi M, Egashira K, Hiasa K, Inoue S, Ni W, Zhao Q, Usui M, Kitamoto S, Ichiki T, Takeshita A. Antiinflammatory and antiarteriosclerotic effects of pioglitazone. Hypertension. 2002; 40: 687–693.

Han KH, Chang MK, Boullier A, Green SR, Li A, Glass CK, Quehenberger O. Oxidized LDL reduces monocyte CCR2 expression through pathways involving peroxisome proliferator-activated receptor. J Clin Invest. 2000; 106: 793–802.

Pistrosch F, Passauer J, Fischer S, Fuecker K, Hanefeld M, Gross P. In type 2 diabetes, rosiglitazone therapy for insulin resistance ameliorates endothelial dysfunction independent of glucose control. Diabetes Care. 2004; 27: 484–490.

Marx N, Imhof A, Froehlich J, Siam L, Ittner J, Wierse G, Schmidt A, Maerz W, Hombach V, Koenig W. Effect of rosiglitazone treatment on soluble CD40L in patients with type 2 diabetes and coronary artery disease. Circulation. 2003; 107: 1954–1957.

Marx N, Imhof A, Froehlich J, Siam L, Ittner J, Wierse G, Schmidt A, Maerz W, Hombach V, Koenig W. Antidiabetic PPAR -activator rosiglitazone reduces MMP-9 serum levels in type 2 diabetic patients with coronary artery disease. Arterioscler Thromb Vasc Biol. 2003; 23: 283–288.

Koshiyama H, Shimono D, Kuwamura N, Minamikawa J, Nakamura Y. Inhibitory effect of pioglitazone on carotid arterial wall thickness in type 2 diabetes. J Clin Endocrinol Metab. 2001; 86: 3452–3456.

Zhao Y, Patzer A, Herdegen T, Gohlke P, Culman J. Activation of cerebral peroxisome proliferator-activated receptors gamma promotes neuroprotection by attenuation of neuronal cyclooxygenase-2 overexpression after focal cerebral ischemia in rats. FASEB J. 2006; 20: 1162–1175.

Mehta JL, Hu B, Chen J, Li D. Pioglitazone inhibits LOX-1 expression in human coronary artery endothelial cells by reducing intracellular superoxide radical generation. Arterioscler Thromb Vasc Biol. 2003; 23: 2203–2208.

El Midaoui A, Wu L, Wang R, de Champlain J. Modulation of cardiac and aortic peroxisome proliferator-activated receptor-gamma expression by oxidative stress in chronically glucose-fed rats. Am J Hypertens. 2006; 19: 407–412.

Majithiya JB, Paramar AN, Balaraman R. Pioglitazone, a PPARgamma agonist, restores endothelial function in aorta of streptozotocin-induced diabetic rats. Cardiovasc Res. 2005; 66: 150–161.

Davies GF, Khandelwal RL, Wu L, Juurlink BH, Roesler WJ. Inhibition of phosphoenolpyruvate carboxykinase (PEPCK) gene expression by troglitazone: a peroxisome proliferator-activated receptor-gamma (PPARgamma)-independent, antioxidant-related mechanism. Biochem Pharmacol. 2001; 62: 1071–1079.

Lennon AM, Ramauge M, Dessouroux A, Pierre M. MAP kinase cascades are activated in astrocytes and preadipocytes by 15-deoxy-Delta(12–14)-prostaglandin J(2) and the thiazolidinedione ciglitazone through peroxisome proliferator activator receptor gamma-independent mechanisms involving reactive oxygenated species. J Biol Chem. 2002; 277: 29681–29685.

LeBrasseur NK, Kelly M, Tsao TS, Farmer SR, Saha AK, Ruderman NB, Tomas E. Thiazolidinediones can rapidly activate AMP-activated protein kinase in mammalian tissues. Am J Physiol Endocrinol Metab. 2006; 291: E175–E181.

Morrow VA, Foufelle F, Connell JM, Petrie JR, Gould GW, Salt IP. Direct activation of AMP-activated protein kinase stimulates nitric-oxide synthesis in human aortic endothelial cells. J Biol Chem. 2003; 278: 31629–31639.

Michell BJ, Chen ZP, Tiganis T, Stapleton D, Katsis F, Power DA, Sim AT, Kemp BE. Coordinated control of endothelial nitric-oxide synthase phosphorylation by protein kinase C and the cAMP-dependent protein kinase. J Biol Chem. 2001; 276: 17625–17628.

Inoue I, Goto S, Matsunaga T, Nakajima T, Awata T, Hokari S, Komoda T, Katayama S. The ligands/activators for peroxisome proliferator-activated receptor alpha (PPARalpha) and PPARgamma increase Cu2+,Zn2+-superoxide dismutase and decrease p22phox message expressions in primary endothelial cells. Metabolism. 2001; 50: 3–11.

Von Knethen A, Brune B. Activation of peroxisome proliferator-activated receptor gamma by nitric oxide in monocytes/macrophages down-regulates p47phox and attenuates the respiratory burst. J Immunol. 2002; 169: 2619–2626.

Lee WJ, Lee IK, Kim HS, Kim YM, Koh EH, Won JC, Han SM, Kim MS, Jo I, Oh GT, Park IS, Youn JH, Park SW, Lee KU, Park JY. -Lipoic Acid Prevents Endothelial Dysfunction in Obese Rats via Activation of AMP-Activated Protein Kinase. Arterioscler Thromb Vasc Biol. 2005; 25: 2488–2494.

Alba G, El Bekay R, Alvarez-Maqueda M, Chacón P, Vega A, Monteseirín J, Santa María C, Pintado E, Bedoya FJ, Bartrons R, Sobrino F. Stimulators of AMP-activated protein kinase inhibit the respiratory burst in human neutrophils. FEBS Letters. 2004; 573: 219–225.

Nagata D, Takeda R, Sata M, Satonaka H, Suzuki E, Nagano T, Hirata Y. AMP-Activated Protein Kinase Inhibits Angiotensin II–Stimulated Vascular Smooth Muscle Cell Proliferation. Circulation. 2004; 110: 444–451.

Verrier E, Wang L, Wadham C, Albanese N, Hahn C, Gamble JR, Chatterjee VK, Vadas MA, Xia P. PPARgamma agonists ameliorate endothelial cell activation via inhibition of diacylglycerol-protein kinase C signaling pathway: role of diacylglycerol kinase. Circ Res. 2004; 94: 1515–1522.

Ouchi N, Shibata R, Walsh K. AMP-Activated Protein Kinase Signaling Stimulates VEGF Expression and Angiogenesis in Skeletal Muscle. Circ Res. 2005; 96: 838–846.

Arad M, Seidman CE, Seidman JG. AMP-Activated Protein Kinase in the Heart: Role During Health and Disease. Circ Res. 2007; 100: 474–488.

Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med. 2007; 356: 2457–2471.

作者单位：Department of Clinical and Experimental Medicine (G.C., A.G., I.P., E.I., E.P., G.P.F., M.A., A.S., A.A.), Chemical Sciences (L.F.), University of Padova Medical School, Italy.