Vascular Lipotoxicity: Endothelial Dysfunction via Fatty-Acid-Induced Reactive Oxygen Species Overproduction in Obese Zucker Diabetic Fatty Rats

Vascular endothelial dysfunction has been demonstrated in obesity, but the molecular basis for this link has not been clarified. We examined the role of free fatty acids (FFA) on vascular reactivity in the obese fa/fa Zucker diabetic fatty (ZDF) rat. Addition of acetylcholine produced a dose-dependent relaxation in aortic rings of ZDF and lean +/+ rats, but the ED₅₀ value was higher in ZDF (−6.80 ± 0.05 vs. −7.11 ± 0.05 log₁₀ mol/liter, P = 0.033). A 2-wk treatment with a 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor, pitavastatin (3 mg/kg/d) or a reduced nicotinamide adenine dinucleotide phosphate (NADPH) oxidase inhibitor, apocynin (5 mmol/liter in drinking water), improved the response in ZDF (ED₅₀, −7.16 ± 0.03 and −7.14 ± 0.05 log₁₀ mol/liter, P = 0.008 and P = 0.015 vs. vehicle, respectively). Vasodilator response to sodium nitroprusside was identical between ZDF and +/+ rats. Vascular reactive oxygen species (ROS) levels and NADPH oxidase activity in aorta were increased in ZDF rats but were decreased by pitavastatin. In in vitro cell culture, intracellular ROS signal and NADPH oxidase subunit mRNA were increased by palmitate, but this palmitate-induced ROS production was inhibited by NADPH oxidase inhibitor or pitavastatin. In conclusion, FFA-induced NADPH oxidase subunit overexpression and ROS production could be involved in the endothelial dysfunction seen in obese ZDF rats, and this could be protected by pitavastatin or NADPH oxidase inhibitors.

THE PRESENCE OF vascular endothelial dysfunction has been demonstrated in subjects with insulin resistance/visceral fat obesity (1–3). Namely, blood flow response to an endothelium-dependent vasodilator such as methacholine chloride but not to an endothelium-independent vasodilator such as sodium nitroprusside was impaired in obese insulin-resistant subjects (1). One common metabolic feature of insulin resistance/obesity is a deficit in insulin-mediated glucose disposal. Recent evidence raised the possibility that tissue accumulation of free fatty acids (FFA) mainly causes abnormalities of insulin secretion and actions and consecutive metabolic derangements, named lipotoxicity (4–6). Because FFA are supplied excessively and persistently to the bloodstream from visceral fat tissues, we and others assumed that an elevation of circulating FFA might be causally related to the onset and progression of endothelial dysfunction in patients with insulin resistance/visceral fat obesity (2, 3).

Although direct inhibitory effects of FFA on endothelial function have already been shown in humans (2, 3, 7), the mechanism by which FFA cause such inhibition has not been clear. Several in vitro studies reported that FFA can enhance production of reactive oxygen species (ROS) (8, 9), but a functional link of circulating FFA to endothelial function through ROS production has not been evaluated. It has been reported that 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statin) improve endothelial function and also reduce vascular superoxide production via inhibition of vascular NADPH oxidase activation (10, 11). Thus, statin might attenuate FFA-induced endothelial dysfunction via inhibition of vascular superoxide production.

In the present study, we examined 1) the role of FFA and the oxidases responsible for ROS production in vascular reactivity and 2) effects of statin on vascular ROS production in a rodent model of visceral fat obesity, Zucker diabetic fatty (ZDF) rat, which shows hyperphagia and obesity-related diabetes, dyslipidemia, and hypertension resulting from a loss-of-function mutation in the leptin receptor (12).

Materials and Methods

Animals

Studies were carried out in male obese homozygous (fa/fa) ZDF rats and lean wild-type (+/+) littermates (Charles River Laboratories, Wilmington, MA). All rats were fed standard laboratory chow and given tap water ad libitum. Their genotype was determined as described (12). ZDF fa/fa and +/+ rats received either pitavastatin (3 mg/kg/d) or vehicle from 7–9 wk of age via an orogastric tube. A group of fa/fa ZDF rats received apocynin (5 mmol/liter in drinking water). Rats were housed individually in metabolic cages for monitoring food intake, urine volume, and body weight. All procedures were performed in accordance with the guidelines of the University of the Ryukyus Committee on Animal Care and Handling.

Biochemical measurements

Plasma glucose levels were measured by the glucose oxidase method with the Glucose Analyzer II (Beckman Coulter Inc., Fullerton, CA). Plasma insulin levels were assessed using an insulin ELISA kit. Serum levels of cholesterol and triglyceride, and FFA were measured using routine enzymatic assays. Plasma levels of lipid peroxidation were measured as thiobarbituric acid reactive substance (TBARS) using the LPO test (Wako Pure Chemical Industries, Osaka, Japan) (13). Plasma and urinary 8-epi-prostaglandin-F₂α (8-epi-PGF₂α) was extracted on C-18 SPE cartridges (Waters Corp., Milford, MA) and assayed by competitive immunoassay using a Cayman Chemical 8-epi-PGF₂α EIA kit (Ann Arbor, MI) (13). Plasma levels of adiponectin (Otsuka Pharmaceutical Co., Ltd, Tokyo, Japan) and TNF-α (JIMRO Co., Ltd., Takasaki, Gunma, Japan) were measured by sandwich ELISA as previously described (3).

Vascular reactivity

After a midlaparotomy under pentobarbital sodium anesthesia, the aorta was rapidly excised for vascular reactivity measurements (14), and nonfasting blood samples were obtained from the inferior vena cava. A portion of aorta was frozen in liquid nitrogen and stored at −70 C. Fresh aorta were cleared of periadventitial tissue and cut transversely into rings 1.5–2.0 mm in diameter. Vascular rings, handled carefully to avoid damage to the inner surface, were mounted on wires in the chambers of a multivessel myograph (J.P. Trading, Tokyo, Japan) and bathed in Krebs’ buffer. The medium was gassed with 95% O₂ and 5% CO₂ and maintained at 37 C (pH 7.4). After equilibration (30 min), the rings were set to an isometric force-displacement transducer (TB-611T; Nihon Kohden, Tokyo, Japan) for measurement of changes in tension and allowed to stabilize for another 30 min. The rings were then depolarized with potassium chloride (60 mmol/liter) to evaluate maximal contraction. After washing with a Krebs’ buffer, the vascular preparations were contracted with phenylephrine (10^–6 mol/liter), and when the contractile response was stabilized (steady-state phase, 15 min), vasorelaxing responses to cumulative increments in the concentration of acetylcholine or sodium nitroprusside were examined. The resting tension of the rings was adjusted to 1.0 g. Changes in vascular tension were recorded on a pen-writing recorder (WT-645G; Nihon Kohden).

Human umbilical vein endothelial cells (HUVEC) study

HUVEC are plated in a 100-mm culture dish at the density of 2.0 × 10⁶ cells per dish. After 16–24 h, the cells were incubated with 0.1–1 mm palmitate, a major fraction of saturated FFA in plasma, with a 1-h prior incubation each of vehicle, 1 mmol/liter pitavastatin, 10 μmol/liter diphenyleneiodium (DPI), 20 mmol/liter N-acetyl-l-cysteine (NAC). After treatment of indicated conditions, cells were harvested from the dish with 0.5× Trypsin-EDTA and then immediately subjected to cytoplasmic mRNA extraction by RNA-Easy kit (QIAGEN GmbH, Hilden, Germany).

ROS signals

The intracellular ROS formation in HUVEC was detected using the fluorescent probe 5-(and 6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate, acetyl ester (Molecular Probes, Inc., Eugene, OR) according to the manufacturer’s protocol. Preliminarily, we confirmed the satisfactory efficacy of this probe to detect intracellular ROS signal induced by H₂O₂ in several cell lines including HUVEC.

Immunoprecipitations and Western blotting

Western blot analysis was performed as described previously (15). Protein samples (50 μg) were prepared from thoracic aortas of four groups of mice and denatured and run on polyacrylamide gels. After transfer onto polyvinylidene difluoride transfer membranes, the membranes were blocked for 90 min in 5% nonfat milk solution. For immunoprecipitation, the primary antibodies [endothelial nitric oxide synthase (eNOS) or p47phox] were used at a 1:1000 dilution in 5% nonfat milk solution for 12 h at 4 C (16, 17). Bound antibodies were detected with horseradish-peroxidase-conjugated antimouse IgG and visualized with an enhanced chemiluminescence detection system (SuperSignal West Pico Chemiluminescent Substrate; Pierce, Rockford, IL). For anti-phosphotyrosine or anti-phosphoserine blots, nitrocellulose membranes were blocked by incubation in Tris-buffered saline/Tween 20 containing 1% BSA for 2 h, followed by a 20-min incubation in anti-phosphotyrosine (eNOS) or anti-phosphoserine (p47phox) antibody diluted in blocking buffer. The membranes were washed extensively in Tris-buffered saline/Tween 20 and developed by using the above system. Band intensity was quantified by NIH ImageJ 1.32j, and the ratio of anti-phosphotyrosine or anti-phosphoserine blot intensity to those of eNOS or p47phox blot intensity was used to represent the enzyme catalytic activities. Lipid peroxidation level of aorta was determined using the TBARS assay kit (ZeptoMetrix Corp., Buffalo, NY).

Real-time RT-PCR

RT-PCR was done with SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen Japan K.K., Tokyo, Japan) and SYBR green on an ABI PRISM 7000 real-time PCR system (Applied Biosystems Japan Ltd., Tokyo, Japan) (13). Primers used were as follows: p22phox (GenBank NM_000101) forward, ATTACTATGTTCGGGCCGTCCT, and reverse, GGTAGATGCCGCTCGCAAT; p40phox (NM_000631) forward, ATGCGGATACCTGCCCTCAA, and reverse, CTCTGAGTCATAGGGCGACTGGTAA; p47phox (NM_000265) forward, GATGCCCAAAGATGGCAAGAGTA, and reverse, GCTTTCATCTGACAGAACCACCAA; p67phox (NM_000433) forward, AGCTCCGGCTGGAACACACTA, and reverse, GGCACCAGCTCATTGCTGTC; gp91phox (NM_000397) forward, AAATGGATCGCATCTGTGTGAC, and reverse, TGGCCACACTAACAGTGATTTAGAG; and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (NM_002046) forward, GGCCTCCAAGGAGTAAGACC, and reverse, AGGGGTCTACATGGCAACTG. Results are expressed as fold change in gene expression by determining the ratio of copy number of the gene of interest corrected for expression of GAPDH in the samples.

Statistical analysis

Values are expressed as the mean ± se. Two-tailed unpaired Student’s t test or one-way factorial ANOVA, followed by Bonferroni’s post hoc comparisons, was used to compare group means. Comparisons of dose-response curves were made by two-factor repeated-measures ANOVA. A P value < 0.05 was considered statistically significant. Analyses were processed using StatView J-5.0 software package (SAS Institute Inc., Cary, NC) or InStat 3 for Macintosh version 3.0b (GraphPad Software, Inc., San Diego, CA).

Results

Metabolic features in obese ZDF (fa/fa) rats

The mean body weight, plasma glucose, insulin, FFA, and triglyceride of obese male ZDF (fa/fa) rats at 9 wk of age were all higher than age-matched (+/+) controls (Table 1). Pitavastatin treatment did not significantly change plasma levels of total and high-density lipoprotein (HDL)-cholesterol, triglyceride, and FFA in ZDF rats.

TABLE 1.

General characteristics of animals after a 2-wk treatment with vehicle or pitavastatin

	Lean +/+		ZDF fa/fa
	Vehicle	Pitavastatin	Vehicle	Pitavastatin	Apocynin
Body weight (g)	153 ± 28	150 ± 6	220 ± 19b	203 ± 21b	227 ± 12
Food intake (g/d)	18 ± 1	17 ± 1	24 ± 2a	23 ± 1a	28 ± 2a
Glucose (mmol/liter)	7.9 ± 0.5	7.8 ± 0.4	10.3 ± 0.8	7.9 ± 0.8	10.6 ± 1.3
Insulin (pmol/liter)	9.6 ± 1.8	9.7 ± 1.6	83.3 ± 13.6a	79.2 ± 8.0a	34.5 ± 8.6a
Total cholesterol (mmol/liter)	2.85 ± 0.11	2.72 ± 0.07	3.02 ± 0.35	3.43 ± 0.67	2.59 ± 0.09
Triglyceride (mmol/liter)	0.41 ± 0.09	0.32 ± 0.08	1.44 ± 0.40b	1.13 ± 0.59b	—
HDL-cholesterol (mmol/liter)	0.91 ± 0.03	0.82 ± 0.09	1.08 ± 0.14	1.02 ± 0.11	1.28 ± 0.08
FFA (mmol/liter)	0.33 ± 0.03	0.30 ± 0.03	0.92 ± 0.16a	0.50 ± 0.18a	0.53 ± 0.11a
Plasma TBARS (nmol MDA/ml)	1.0 ± 0.1	1.3 ± 0.3	8.08 ± 1.533b	7.29 ± 1.64b	7.00 ± 0.35b
Urinary 8-epi-PGF2α (pg/mg creatinine)	533 ± 132	545 ± 147	890 ± 75b	795 ± 157b	892 ± 24b

Values are the mean ± sem of three to six rats. MDA, Malondialdehyde.

^a

p < 0.017 vs. lean +/+ vehicle (post hoc Bonferroni test).

^b

p < 0.05 vs. lean +/+ vehicle (t test).

Effects of pitavastatin or apocynin on vascular reactivity

In aortic rings, vasoconstrictive response to phenylephrine was almost identical between +/+ and ZDF rats (Fig. 1, left). Addition of acetylcholine produced a dose-dependent relaxation (10⁻⁹ to 10⁻⁴ mol/liter) in both +/+ and ZDF rats (Fig. 1, center), but the ED₅₀ value was significantly higher in ZDF rats (Table 2). A 2-wk treatment of pitavastatin or apocynin improved the dose-relaxation response in ZDF rats (Table 2). The vasodilator response to sodium nitroprusside was almost identical between +/+ and ZDF rats and was not affected by pitavastatin or apocynin treatment (Fig. 1, right).

Fig. 1.

Vascular reactivity in aorta isolated from +/+ and fa/fa ZDF rats. Rats were treated with vehicle (○), pitavastatin (•), or apocynin (▵) from 7–9 wk of age. Vascular reactivity to phenylephrine (Phe), acetylcholine (Ach), or sodium nitroprusside (SNP) was determined in +/+ (upper panels) and ZDF rats (lower panels). Data represent the mean ± sem (n = 3–6). The P values for curve difference by two-factor repeated-measures ANOVA are shown.

TABLE 2.

ED₅₀ value of vascular reactivity

	Lean +/+		ZDF fa/fa
	Vehicle	Pitavastatin	Vehicle	Pitavastatin	Apocynin
Phenylephrine (log10 mol/liter)	−8.247 ± 0.126	−8.255 ± 0.122	−7.431 ± 0.074	−7.391 ± 0.080	−7.194 ± 0.128
95% confidential interval	−9.842 to −6.651	−9.806 to −6.705	−8.374 to −6.489	−8.411 to −6.371	−8.825 to −5.563
Acetylcholine (log10 mol/liter)	−7.111 ± 0.053	−7.021 ± 0.017	−6.803 ± 0.050a	−7.156 ± 0.030b	−7.143 ± 0.048b
95% confidential interval	−7.278 to −6.943	−7.075 to −6.967	−7.016 to −6.590	−7.286 to −7.027	−7.348 to −6.938
Sodium nitroprusside (log10 mol/liter)	−7.942 ± 0.025	−7.870 ± 0.038	−8.155 ± 0.035	−8.153 ± 0.016	−7.904 ± 0.074
95% confidential interval	−8.022 to −7.863	−7.989 to −7.750	−8.268 to −8.042	−8.205 to −8.102	−8.138 to −7.670

Values are the mean ± sem of three to six rats.

^a

P < 0.05 vs. lean +/+ vehicle.

^b

P < 0.05 vs. ZDF fa/fa vehicle.

Circulating ROS and vascular activities of NADPH oxidase and eNOS

Levels of plasma TBARS and urinary 8-epi-PGF₂α were increased in ZDF rats (Table 1). A 2-wk treatment with pitavastatin did not change plasma TBARS and urinary 8-epi-PGF₂α in ZDF rats. The level of eNOS phosphorylation was not different between +/+ and ZDF rats (Fig. 2, left), but the level of p47phox serine phosphorylation, a marker of NADPH oxidase activity, was increased in aorta homogenates of ZDF rats, and it was decreased by pitavastatin treatment (Fig. 2, center). Vascular ROS production was also increased in ZDF rats, and the increase was inhibited by pitavastatin to the comparable level of +/+ (Fig. 2, right).

Fig. 2.

Activities of eNOS and p47phox and ROS signal in aorta isolated from +/+ and fa/fa ZDF rats. Rats were treated either with vehicle or pitavastatin from 7–9 wk of age. RDS signal was expressed in pmol malondialdehyde/μg protein. Data represent the mean ± sem(n = 3–6). *, P < 0.05 vs. 9-wk +/+ vehicle; †, P < 0.05 vs. 9-wk fa/fa ZDF rats.

HUVEC study

Using in vitro cell culture, palmitate increased ROS signals dose dependently (0.1–2 mmol/liter) and time dependently (up to 24 h) (data not shown). The palmitate-induced ROS formation was inhibited completely by pitavastatin and a general antioxidant, NAC, and partially by DPI, a NADPH oxidase inhibitor (Fig. 3). Palmitate increased expression levels of p22phox, p40phox, p47phox, p67phox, and gp91phox subunit gene (Fig. 3). A prior treatment with pitavastatin inhibited palmitate-induced up-regulation in p22phox, p40phox, and p47phox mRNA, but did not change p67phox and gp91phox.

Fig. 3.

Effects of palmitate on ROS production and levels of vascular NADPH oxidase subunit gene in HUVEC. HUVEC were incubated for 0, 15, and 120 min with vehicle or 1 mm palmitate (C16:0) during a 1-h previous incubation each of 1 mmol/liter pitavastatin (P), 20 mmol/liter N-acetyl-l-cysteine (NAC), or 10 μmol/liter diphenyleneiodium (DPI). The intracellular ROS formation was detected using the fluorescent probe, 5-(and 6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate, acetyl ester. Expression levels of p22phox, p40phox, p47phox, p67phox, and gp91phox subunit gene were quantified by real-time PCR and corrected by GAPDH level. Data represent the mean ± sem (n = 3–6). *, P < 0.0125 vs. vehicle.

Discussion

The major findings of the present study were that 1) the vasodilator response to acetylcholine, but not to sodium nitroprusside, was impaired in prediabetic obese ZDF rats, 2) elevations in circulating FFA and ROS and an enhancement of NADPH oxidase activation and vascular ROS production were also observed in ZDF rats, 3) pitavastatin recovered vasodilator responses in ZDF rats with a reduction of vascular NADPH oxidase activity and ROS production, and 4) FFA enhanced production of ROS and expression of NADPH oxidase subunit mRNA, and those were inhibited by pitavastatin.

Vascular reactivity and vascular ROS

Using a rodent model of visceral fat obesity, the ZDF (fa/fa) rat (4, 5), we measured the vascular response to vasodilatory and vasoconstrictive agents. The vasodilator response to acetylcholine, but not to sodium nitroprusside, was impaired in ZDF rats. This indicates that endothelium-dependent vasodilatation, frequently represented by the response to acetylcholine, was impaired, but endothelium-independent vasodilatation, represented usually by the response to sodium nitroprusside, was preserved in ZDF rats. Levels of plasma TBARS and urinary 8-epi-PGF₂α were increased in ZDF rats, indicating an increase in circulating ROS. Because accumulated fat is possibly a principal source of circulating ROS in obesity (13), the circulating ROS might be coming mainly from accumulated fat. However, the level of vascular ROS production was also increased in ZDF rats. Serine phosphorylation of p47phox, which is a critical step for cytoplasmic complex formation of NADPH oxidase and serves as NADPH oxidase activation (17), was enhanced in ZDF aorta, indicating that ROS production was also locally amplified. Increased ROS, regardless of whether it was locally produced or fat-derived remote ROS (13), may be associated with endothelial dysfunction (18).

Under normal conditions, NO released by eNOS stimulates soluble guanylyl cyclase, increasing cGMP, activating cGMP-dependent protein kinase 1, and finally eliciting vasodilation (18). When vascular ROS is in excess, it can react with NO, thereby generating peroxynitrite, the most stable and potent oxidant. Peroxynitrite uncouples eNOS, switching the NO-producing process to a ROS reproduction process (18). In ZDF rats, excess vascular ROS can come from increased activity of NADPH oxidase and normal activity of eNOS. A 2-wk treatment with a NADPH oxidase inhibitor, apocynin (13), almost completely recovered the vascular response to acetylcholine in ZDF rats, supporting the notion that vascular ROS is the major cause of endothelial dysfunction.

Hyperglycemia is the other possible mechanism of vascular endothelial dysfunction in visceral obesity. Our ZDF rats at 9 wk of age were at the phase of glucose intolerance, showing mild hyperglycemia (nonfasting was 7.9 mmol/liter vs. age-matched control was 10.3 mmol/liter), hyperinsulinemia (9.6 vs. 83.3 pmol/liter) and hyperlipacidemia (0.33 vs. 0.92 mmol/liter) (4–6). We confirmed that this level of mild hyperglycemia did not impair vascular reactivity to acetylcholine and did not increase ROS production (data not shown). Although we cannot completely exclude the role of hyperglycemia in impairing vascular reactivity, it is likely that mild hyperglycemia is not the primary cause of ROS-associated endothelial dysfunction in prediabetic ZDF rats.

It had been shown that 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statin) reduce ROS production (10, 11). In our study, pitavastatin did not change levels of plasma TBARS and urinary 8-epi-PGF₂α in ZDF rats but improved the vasodilator response to acetylcholine. The level of vascular ROS was decreased by pitavastatin simultaneously with a decrease of vascular p47phox serine phosphorylation, indicating that pitavastatin somehow inhibited the activation process of NADPH oxidase (10, 11). Pitavastatin did not change the plasma levels of adipocyte-derived cytokines such as TNF-α (vehicle 4.00 ± 0.75 vs. pitavastatin 4.77 ± 1.69 pg/ml) and adiponectin (8.75 ± 0.95 vs. 7.27 ± 0.48 μg/ml) in ZDF rats (19, 20) (see also IDF Worldwide Definition of the Metabolic Syndrome at http://www.idf.org/home/). Collectively, pitavastatin did not affect circulating ROS level but decreased vascular ROS production, suggesting direct vascular effects.

FFA and vascular NADPH oxidase activity

A common feature of visceral fat obesity is an oversupply of FFA to the bloodstream from adipose tissues, and FFA can enhance vascular production of ROS (8, 9). We thus tested whether FFA do directly activate vascular ROS production and, if so, whether it can be through NADPH oxidase activation.

First we confirmed that palmitate, a major fraction of saturated FFA in plasma, directly increased intracellular ROS signals dose dependently and time dependently (data not shown) (8). The palmitate-induced increases in vascular ROS signals were inhibited completely by pitavastatin and a general antioxidant, NAC, and partially by DPI, a NADPH oxidase inhibitor.

The major source of superoxide anion in the vasculature is the NADPH oxidase family of enzymes (21, 22). Vascular NADPH oxidase is a multisubunit enzyme complex that includes the membrane-bound flavocytochrome b558 formed by gp91phox and p22phox and the cytosolic proteins p47phox, p67phox, and Rac. We thus determined the effects of palmitate on the expression levels of the vascular NADPH oxidase subunit gene. Palmitate increased expression levels of p22phox, p40phox, p47phox, p67phox, and gp91phox. A prior treatment with pitavastatin inhibited the palmitate-induced increases in p22phox, p40phox, and p47phox mRNA but did not change p67phox and gp91phox.

Two general mechanisms underlying activation of NADPH oxidase are either an acute increase in oxidase complex formation secondary to posttranslational modification of regulatory subunits (p47phox and Rac) or a chronic increase in the expression and abundance of component subunits (18, 21, 22). As we and others reported previously (5, 8), palmitate directly increases diacylglycerol levels and protein kinase C activation, which is the well-known signal for activation of NADPH oxidase. A key mechanism of acute activation by palmitate can be that protein kinase C-dependent phosphorylation of the p47phox regulatory subunit and its translocation to the Nox2/p22phox heterodimer to form fully assembled complexes. Increased expression of NADPH oxidase subunits might be the mechanism of chronic NADPH activation by palmitate.

As recently reported, mitochondrial uncoupling could be another source of ROS production in FFA-treated endothelial cells (23). Adenoviral overexpression of uncoupling protein 1 (UCP-1) or inhibition of mitochondrial FFA oxidation by carnitine palmitoyltransferase I (CPT-I) inhibitor (etomoxir) could inhibit such FFA-induced ROS production. NADPH oxidase and mitochondrial uncoupling could independently contribute to FFA-induced ROS production in vascular system.

Activated Rac in its GTP-bound state binds to the cytosolic p67phox subunit and activates the oxidase. Pitavastatin may inhibit palmitate-induced activation of NADPH oxidase through Rac inactivation, because Rac activation requires its posttranslational modification by isoprenylation, a process that is inhibited by statin (10, 11). Inhibition of palmitate-induced up-regulation of NADPH oxidase subunits may be another mechanism of NADPH inactivation by pitavastatin.

Conclusion

Endothelial dysfunction and NADPH oxidase activation were concomitantly observed in obese ZDF rats, but those were improved by pitavastatin and apocynin treatment. It is suggested that pitavastatin might inhibit FFA-induced NADPH oxidase subunit gene expression and ROS production in endothelial cells and then protect the endothelial dysfunction seen in obese ZDF rats. Visceral fat obesity is the essential component of the metabolic syndrome including hypertension, dyslipidemia, and glucose intolerance (Ref. 19 and http://www.idf.org/home/). Endothelial dysfunction, which is a systemic disorder and a key variable in the initiation and progression of atherosclerosis and its complications (20), occurs frequently in subjects with visceral fat obesity (1–3). As shown in Fig. 4, we suggest that FFA-induced ROS overproduction might be a possible underlying mechanism(s) for the impaired endothelial function in visceral fat obesity, vascular lipotoxicity (Fig. 4).

Fig. 4.

A working hypothesis of vascular lipotoxicity. Visceral fat is the major source of circulating FFA and ROS. Circulating FFA induces vascular ROS production via up-regulation of vascular NADPH oxidase. The locally produced ROS and fat-derived ROS concomitantly react with NO, generate peroxynitrite, and finally impair cGMP-dependent vasodilatation. Statin may block FFA-induced ROS production via inhibition of NADPH oxidase expression and Rac inactivation. There is controversy about effects of FFA on insulin-mediated eNOS activation and angiotensin type 1 receptor (AT1R) signaling. AII, Angiotensin II; DAG, diacylglycerol.

This work was supported by a grant from the Japanese Society for the Promotion of Science (14571103).

Disclosure Summary: The authors have nothing to declare.

Abbreviations:

 
• DPI,

  Diphenyleneiodonium;

 
• eNOS,

  endothelial nitric oxide synthase;

 
• 8-epi-PGF₂α,

  8-epi-prostaglandin-F₂α;

 
• FFA,

  free fatty acids;

 
• GAPDH,

  glyceraldehyde-3-phosphate dehydrogenase;

 
• HDL,

  high-density lipoprotein;

 
• HUVEC,

  human umbilical vein endothelial cells;

 
• NAC,

  N-acetyl-l-cysteine;

 
• ROS,

  reactive oxygen species;

 
• TBARS,

  thiobarbituric acid reactive substance;

 
• ZDF,

  Zucker diabetic fatty.