Interactions Between Vascular Wall and Perivascular Adipose Tissue Reveal Novel Roles for Adiponectin in the Regulation of Endothelial Nitric Oxide Synthase Function in Human Vessels
Adiponectin is an adipokine with potentially important roles in human cardiovascular disease states. We studied the role of adiponectin in the cross-talk between adipose tissue and vascular redox state in patients with atherosclerosis.
Methods and Results—
The study included 677 patients undergoing coronary artery bypass graft surgery. Endothelial function was evaluated by flow-mediated dilation of the brachial artery in vivo and by vasomotor studies in saphenous vein segments ex vivo. Vascular superoxide (O2−) and endothelial nitric oxide synthase (eNOS) uncoupling were quantified in saphenous vein and internal mammary artery segments. Local adiponectin gene expression and ex vivo release were quantified in perivascular (saphenous vein and internal mammary artery) subcutaneous and mesothoracic adipose tissue from 248 patients. Circulating adiponectin was independently associated with nitric oxide bioavailability and O2− production/eNOS uncoupling in both arteries and veins. These findings were supported by a similar association between functional polymorphisms in the adiponectin gene and vascular redox state. In contrast, local adiponectin gene expression/release in perivascular adipose tissue was positively correlated with O2− and eNOS uncoupling in the underlying vessels. In ex vivo experiments with human saphenous veins and internal mammary arteries, adiponectin induced Akt-mediated eNOS phosphorylation and increased tetrahydrobiopterin bioavailability, improving eNOS coupling. In ex vivo experiments with human saphenous veins/internal mammary arteries and adipose tissue, we demonstrated that peroxidation products produced in the vascular wall (ie, 4-hydroxynonenal) upregulate adiponectin gene expression in perivascular adipose tissue via a peroxisome proliferator-activated receptor-γ–dependent mechanism.
We demonstrate for the first time that adiponectin improves the redox state in human vessels by restoring eNOS coupling, and we identify a novel role of vascular oxidative stress in the regulation of adiponectin expression in human perivascular adipose tissue.
Circulating adiponectin, an adipokine involved in diabetes mellitus and insulin resistance, appears to be a link between obesity and atherosclerosis.1,2 Reduced plasma adiponectin levels have been associated with increased cardiovascular risk,3 and genetic variants decreasing plasma adiponectin levels increase the risk for diabetes mellitus4 and coronary heart disease,5 whereas adiponectin produced in perivascular adipose tissue may exert paracrine effects on the vascular wall.6 Although the expression of the adiponectin gene in adipocytes is highly regulated by peroxisome proliferator-activated receptor-γ (PPAR-γ),7 the local mechanisms regulating adiponectin release and controlling its potential impact on vascular function in humans remain unclear.
Clinical Perspective on p 2221
Experimental studies suggest that adiponectin stimulates nitric oxide (NO) production in endothelial cell cultures.8 This is believed to be due in part to endothelial NO synthase (eNOS) activation through PI3 kinase/Akt-mediated phosphorylation.9,10 However, whether adiponectin exerts the same effects in the human vasculature remains to be established. Indeed, the biological role of adiponectin appears to be much more complex in human cardiovascular disease than in experimental models. In particular, the possible protective effect of high circulating adiponectin in healthy individuals is lost (or even reversed) in advanced cardiovascular disease states such as heart failure.11
Under conditions of increased vascular oxidative stress observed in human atherosclerosis, eNOS is uncoupled mostly as a result of oxidation of its cofactor tetrahydrobiopterin (BH4) and produces superoxide radicals (O2−) instead of NO.12–14 In this biological setting, activation of eNOS by adiponectin may increase O2− generation from uncoupled eNOS15 if there is no parallel increase in vascular BH4. Therefore, the biological role of adiponectin in the regulation of eNOS functional status and activity in human atherosclerosis remains unexplored.
In this study, we examine the impact of adiponectin and vascular NO bioavailability/redox state in patients with advanced atherosclerosis and investigate, for the first time in humans, the relationship between local adiponectin synthesis in perivascular adipose tissue (AT) and O2−generation in human vessels. We then explore the molecular mechanisms by which adiponectin regulates eNOS activity and coupling in the human vascular endothelium and describe a novel role of vascular oxidative stress in the regulation of adiponectin synthesis in human perivascular AT.
Population and Protocol of Study 1
The population of study 1 consisted of 677 patients (Table) undergoing elective coronary artery bypass graft surgery. The day before surgery, endothelium-dependent flow-mediated dilation (FMD) and endothelium-independent dilatation of the brachial artery were determined (see Methods in the online-only Data Supplement). Fasted blood samples were obtained on the morning of the surgery. During coronary artery bypass graft surgery, segments of saphenous vein (SV) and internal mammary artery (IMA) were obtained. Exclusion criteria were any inflammatory, infectious, liver, or renal disease or malignancy. Patients with recent unstable coronary syndrome (within the previous 8 weeks) or clinical heart failure syndrome and those receiving nonsteroidal anti-inflammatory drugs, dietary supplements, or antioxidant vitamins were also excluded.
|Clinical Studies (Study 1)||Ex Vivo Studies (Study 2)|
|Male sex, n (%)||563 (83)||35 (78)|
|Hypertension, n (%)||465 (68)||36 (80)|
|Hyperlipidemia, n (%)||433 (64)||22 (49)|
|Diabetes mellitus, n (%)||221 (33)||12 (27)|
|Smoking (active/ex), n (%)||161/325 (24/48)||8/20 (18/44)|
|Triglycerides, mg/dL*||124 (94–166)||107 (56–128)|
In a subgroup of 248 patients, samples of perivascular AT surrounding the IMA (peri–IMA-AT) and SV (peri–SV-AT), subcutaneous AT (Sc-AT; from the site of the chest incision), and mesothoracic AT (Ms-AT; attached to the pericardium) were also obtained. AT samples from all sites were snap-frozen and stored at −80°C for gene expression studies or were used for tissue culture experiments (Sc-AT, peri–SV-AT, and Ms-AT) as described below.
The study was approved by the Institutional Review Committee. All subjects gave written informed consent.
Blood Sampling and Adiponectin Measurements
Venous blood samples were obtained after 8 hours of fasting on the morning of the surgery, and serum adiponectin was measured by ELISA (see Methods in the online-only Data Supplement).
DNA Extraction and Genotyping
Genomic DNA extraction from whole blood and genotyping were performed by standard methodology (see Methods in the online-only Data Supplement).
Vessel Harvesting and Vasomotor Studies
Human SV and IMA samples were obtained at the time of coronary artery bypass graft surgery, and the vasorelaxations in response to acetylcholine and sodium nitroprusside were studied in an organ-bath setting as described previously (see Methods in the online-only Data Supplement).16
Vascular Superoxide Measurements
Vascular O2− production was measured in fresh, intact IMA and SV segments by using lucigenin (5 μmol/L)-enhanced chemiluminescence, as previously described16,17 (see Methods in the online-only Data Supplement).
Adipose Tissue Culture
Samples of Sc-AT, Ms-AT, and peri–SV-AT obtained from patients in study 1 were used to estimate the biosynthetic rate of adiponectin in an ex vivo bioassay (see Methods in the online-only Data Supplement). AT from these depots was routinely cultured for 4 hours, and culture supernatants were collected to estimate the release of adiponectin.
RNA Isolation and Quantitative Real-Time Polymerase Chain Reaction
Samples of Sc-AT, Ms-AT, peri–SV-AT, and peri–IMA-AT were used for gene expression studies (see Methods in the online-only Data Supplement).
Measurement of Vascular Biopterins
Vascular BH4, dihydrobiopterin, and biopterin levels were each determined separately from the same sample through the use of high-performance liquid chromatography followed by serial electrochemical and fluorescent detection, as we have previously described12 (see Methods in the online-only Data Supplement).
Population in Study 2 and Experimental Procedures
To examine the direct effects of adiponectin on the mechanisms regulating NO bioavailability and O2− production in human vessels, we used ex vivo models of human SVs and IMAs, as previously described.16 For these experiments we recruited 46 patients undergoing coronary artery bypass graft surgery (Table) using the same exclusion criteria as for study 1. Serial SV/IMA segments were incubated ex vivo in the presence or absence of recombinant full-length adiponectin 10 μg/mL for 6 hours. The effect of adiponectin on vascular O2− (basal and N-nitro-l-arginine methyl ester [LNAME]–inhibitable O2−) was quantified by lucigenin-enhanced chemiluminescence and visualized with oxidative fluorescent dye, dihydroethidium, staining (see below). The changes in Akt and eNOS phosphorylation status were determined by Western blotting; vascular biopterins were quantified by high-performance liquid chromatography, as described above. In additional studies, SV and IMA segments were incubated with adiponectin 10 μg/mL in the presence and absence of wortmannin (100 nmol/L) to inhibit PI3 kinase/Akt signaling or 2,4-diamino hydroxyl pyrimidine (DAHP) to inhibit GTP-cyclohydrolase (the rate-limiting enzyme in the biosynthetic pathway of biopterins; see Methods in the online-only Data Supplement).
To examine the impact of vascular oxidative stress on local adiponectin synthesis in perivascular AT, we first incubated peri–SV-AT and peri–IMA-AT with 4-hydroxynonenal (4-HNE; 30 µmol/L) for 16 hours in the presence and absence of the PPAR-γ inhibitor T0070907 (10 μmol/L for peri–SV-AT) and examined the changes in ADIPOQ, PPAR-γ, and CD36 (which is downstream of PPAR-γ) gene expression. To examine whether 4-HNE is produced by human SVs and IMAs in the presence of oxidative stress, we quantified 4-HNE protein adducts content in 11 SVs and 18 IMAs and related these measurements to vascular O2− production from these vessels (see Methods in the online-only Data Supplement).
Oxidative Fluorescent Microtopography
In situ O2− production was determined in vessel cryosections by oxidative fluorescent dye, dihydroethidium, staining, as previously described (see Methods in the online-only Data Supplement).
Western blots in human vessels for Akt/phosphorylated Akt (Ser473), eNOS/phosphorylated eNOS (Ser1177), and 4-HNE adducts were performed as described in Methods in the online-only Data Supplement.
Continuous variables were tested for normal distribution with the Kolmogorov-Smirnov test. Nonnormally distributed variables were log-transformed for analysis.
Sample size calculations were based on previous data from our laboratory. For the clinical studies, we estimated that a total number of 500 subjects would allow us to detect a 2.38 absolute difference in FMD between the highest and lowest tertiles of circulating adiponectin with an α=0.05, a power of 0.9, and an assumed standard deviation of 2.3. For the AT experiments, we estimated that with 150 patients we would be able to detect a 0.13 difference in log(O2−) in SV samples between the extreme tertiles of adiponectin released from peri–SV-AT with an α=0.05, a power of 0.9, and an assumed standard deviation of 0.2. For the ex vivo experiments, sample size calculations were performed on the basis of our previous experience on this model,16 and we estimated that with a minimum of 5 pairs of samples (serial rings from the same vessel) we would be able to identify a change in log(O2−) of 0.48 with an α=0.05, a power of 0.9, and a standard deviation for a difference in the response of the pairs of 0.25.
In the clinical studies, continuous variables between 3 groups were compared by use of 1-way ANOVA followed by the Bonferroni post hoc test for individual comparisons, whereas comparisons between 2 groups were performed by unpaired t tests. Categorical variables were compared by use of the χ2 test as appropriate. Correlations between continuous variables were assessed by bivariate analysis, and the Pearson coefficient was estimated. For the organ-bath experiments, the effect of circulating adiponectin tertile on vasorelaxations in response to acetylcholine was evaluated by use of 2-way ANOVA for repeated measures (examining the effect of acetylcholine or sodium nitroprusside concentration by circulating adiponectin tertile or genotype interaction on vasorelaxations) in a full factorial model. For the ex vivo experiments (in which serial rings from the same vessel were incubated with multiple interventions, we performed repeated measures ANOVA and paired t tests for individual comparisons, followed by the Bonferroni post hoc correction for multiple testing as appropriate.
In the clinical studies, correlations between continuous variables were tested by calculating the Pearson correlation coefficient. Linear regression was performed by using as dependent variables FMD or log(vascular O2-) and as independent variables log(circulating adiponectin) and the clinical demographic characteristics (age, sex, diabetes mellitus, smoking, dyslipidemia, hypertension) that showed a simple association with the dependent variable at the level of 15%. A backward elimination procedure was then used by having P=0.1 as the threshold to remove a variable from the respective model. All statistical tests were performed with SPSS version 20.0, and values of P<0.05 were considered statistically significant.
Adiponectin and NO-Mediated Vasorelaxations in Human Vessels
We first examined the association between circulating adiponectin and endothelial function (as evaluated by FMD and vasorelaxations of SV rings ex vivo). Circulating adiponectin was positively correlated with FMD (Figure 1). In multivariable analysis, the independent predictors of FMD were log(circulating adiponectin) (β, 1.75; SE, 0.69; P=0.012), hypertension (β, −2.16; SE, 0.47; P=0.0001), and smoking (β, −1.51; SE, 0.31; P=0.0001). This finding was confirmed in organ-bath studies in which the vasorelaxations of SV segments in response to acetylcholine were significantly greater in vessels from patients in the highest tertile of circulating adiponectin compared with patients in the lowest tertile. There were no associations between circulating adiponectin and either endothelium-independent dilatation of the brachial artery in vivo or vasorelaxations of SVs in response to sodium nitroprusside ex vivo (Figure 1).
Circulating Versus Local Adiponectin in Perivascular AT and Vascular Superoxide
We next examined whether circulating adiponectin was related to vascular O2− generation in human arteries (IMA) and veins (SV). We observed that circulating adiponectin was closely related to basal O2− in both vessel types (Figure 2). In multivariable analysis, the independent predictors of log(O2−) in SV were log(circulating adiponectin) (β, −0.2984; SE, 0.06; P=0.0001), smoking (β, 0.068; SE, 0.024; P=0.005), and treatment with statins (β, −0.219; SE, 0.044; P=0.0001). Similarly, the independent predictors of log(O2−) in IMAs were log(circulating adiponectin) (β, −0.265; SE, 0.077; P=0.001), statin treatment (β, −0.199; SE, 0.061; P=0.001), diabetes mellitus (β, 0.112; SE, 0.047; P=0.019), and smoking (β, 0.053; SE, 0.032; P=0.097).
To test for possible paracrine effects of local adiponectin secreted by perivascular AT, we quantified both the expression of ADIPOQ gene and local adiponectin protein secretion by peri–SV-AT after 4 hours of culture ex vivo and tested for their associations with vascular redox state. In contrast to what was observed with circulating adiponectin, increased vascular O2− in human SVs and IMAs was related to increased expression of the ADIPOQ gene and local adiponectin secretion from the perivascular AT (Figure 2). These discordant relationships between vascular O2− production and circulating versus local adiponectin production in perivascular AT (confirmed in both human arteries and veins) suggest that local production of adiponectin in perivascular AT and circulating adiponectin are differentially regulated and imply that vascular O2− has the potential to stimulate local adiponectin production in the neighboring perivascular AT.
Circulating Versus Local Adiponectin in Perivascular AT and eNOS Coupling
To assess whether there are interactions between circulating/local adiponectin and eNOS coupling, we examined the association between both circulating adiponectin and its local expression (in peri–SV-AT and peri–IMA-AT) and LNAME-inhibitable vascular O2− (which provides an estimate of eNOS uncoupling). We observed that low circulating adiponectin levels were related to greater LNAME-inhibitable O2− in both SV and IMA segments, which is indicative of more eNOS uncoupling (Figure 3). In multivariable analysis, we observed that the independent predictors of LNAME-Δ(O2−) in IMAs were log(circulating adiponectin) (β, 4.37; SE, 0.73; P=0.0001) and treatment with statins (β, 2.73; SE, 0.57; P=0.0001), whereas the predictors of LNAME-Δ(O2−) in SVs were also log(circulating adiponectin) (β, 2.12; SE, 0.54; P=0.0001) and statin treatment (β, 2.25; SE, 0.39; P=0.0001). These findings suggest that increased circulating adiponectin and regular statin treatment are independently associated with improved eNOS coupling in these human arteries and veins. In contrast, we observed that higher LNAME-inhibitable O2− in either SVs or IMAs (indicative of more eNOS uncoupling) was related to higher ADIPOQ gene expression in the perivascular AT (peri–SV-AT and peri–IMA-AT) of these vessels, a finding compatible with the increased local adiponectin synthesis observed in peri–SV-AT from these patients (Figure 3).
Linking Circulating and Local Adiponectin Production in AT: The Role of PPAR-γ
To further explore the possible contribution of the different AT depots to the circulating adiponectin pool, we examined the association between circulating adiponectin and its local release/ADIPOQ gene expression in Ms-AT, Sc-AT, peri–SV-AT, and peri–IMA-AT. We observed a significant, albeit weak, correlation between local adiponectin release and ADIPOQ gene expression in Ms-AT and Ac-AT, suggesting that the circulating pool may be driven partly by adiponectin produced by these remote (nonperivascular) depots (Figure 4). In contrast, circulating adiponectin was not related to either local adiponectin release/ADIPOQ gene expression in peri–SV-AT (Figure 4) or ADIPOQ gene expression in peri–IMA-AT (r=0.072, P=0.602), implying that different, probably local, mechanisms control the release of adiponectin in perivascular AT.
To further explore the subcellular mechanisms controlling ADIPOQ gene expression in the various AT depots, we quantified the expression of PPAR-γ, which is known to regulate ADIPOQ gene expression in cell culture models.18 We observed a strong correlation between PPAR-γ and ADIPOQ gene expression in peri–SV-AT (r=0.576, P<0.0001), peri–IMA-AT (r=0.751, P<0.0001), Ms-AT (r=0.514, P<0.0001), and Sc-AT (r=0.344, P<0.0001). To confirm that PPAR-γ gene expression provides a good estimate of PPAR-γ activity, we then quantified the expression of its downstream molecule, CD36 (known to be highly regulated by PPAR-γ activity),19 and confirmed collinearity between the expression of PPAR-γ and CD36 genes in peri–SV-AT (r=0.976, P<0.0001), peri–IMA-AT (r=0.956, P<0.0001), Ms-AT (r=0.933, P<0.0001), and Sc-AT (r=0.953, P<0.0001). These results confirm that ADIPOQ gene expression remains under the direct control of PPAR-γ in all types of human AT studied.
Effects of Adiponectin on Endothelial Function and Vascular Redox State by Using the Genetic Variability of ADIPOQ
To explore the discordant associations between vascular O2− and circulating levels and local (perivascular) adiponectin biosynthesis, we searched for genetic single nucleotide polymorphisms with a known impact on adiponectin circulating levels. The genetically determined variability of adiponectin levels could then be used as a model system to test indirectly the effect of adiponectin on vascular NO bioavailability and vascular O2− generation in human vessels. We genotyped the entire population in study 1 for 2 common genetic polymorphisms: rs17366568 in ADIPOQ and rs266717 in ADIPOQ gene promoter, both known to affect circulating adiponectin levels in genome-wide association studies.20,21 These 2 polymorphisms were not in linkage disequilibrium (data not shown), so their effect on ADIPOQ gene expression could be independent from each other and therefore additive. Indeed, we observed that the number of rs266717T and rs17366568G alleles had an additive effect on local adiponectin biosynthesis in Ms-AT that was also reflected in circulating adiponectin levels (Figure 5). However, there was no effect of these single nucleotide polymorphisms on local adiponectin synthesis/ADIPOQ expression in perivascular AT (peri–SV-AT or peri–IMA-AT; Figure 5). These findings suggest that local factors may overwhelm the influence of genetic variability that is observed in the nonperivascular AT tested.
The number of rs266717T plus rs17366568G alleles was also positively associated with FMD, the ex vivo vasorelaxations in response to acetylcholine, and total/LNAME-inhibitable O2− in SV and IMA segments (indicative of eNOS uncoupling when the genetic background leads to lower circulating adiponectin). These findings document for the first time that genetically determined hypoadiponectinemia may actually lead to endothelial dysfunction and eNOS uncoupling in human vessels.
Direct Effects of Adiponectin on Redox State and eNOS Coupling by Using Ex Vivo Models of Human Vessels
To further explore the discordant associations between vascular endothelial function/redox state and circulating/local adiponectin production in perivascular AT (in study 1), we performed a number of mechanistic experiments in human IMAs and SVs ex vivo (study 2). We first incubated segments of these vessels with or without adiponectin 10 μg/mL for 6 hours and examined its effect on basal and LNAME-inhibitable O2−. Adiponectin induced a striking reduction in basal O2− by restoring eNOS coupling in both human SVs and IMAs (Figure 6). This effect was also confirmed by dihydroethidium staining of these vessels, in which adiponectin rapidly reversed LNAME-inhibitable O2− in the endothelium (Figure 6).
To further explore the mechanisms by which adiponectin affects eNOS physiology in the human endothelium, we explored its impact on eNOS phosphorylation in these vessels. Ex vivo incubation of human SVs and IMAs with adiponectin increased Akt phosphorylation at Ser473 and eNOS phosphorylation at Ser1177 (Figure 7). The effect of adiponectin on eNOS phosphorylation was blocked by wortmannin, an inhibitor of PI3-mediated Akt phosphorylation, suggesting that adiponectin induces a PI3-Akt–mediated phosphorylation of eNOS, resulting in its activation. The ability of adiponectin to induce eNOS phosphorylation remained unchanged in the presence of DAHP (an inhibitor of GTP-cyclohydrolase blocking the biosynthetic pathway of biopterins), suggesting that a reduction in vascular BH4 levels does not modify the effect of adiponectin on eNOS phosphorylation.
However, activation of eNOS by Ser1177 phosphorylation does not necessarily increase NO synthesis because phosphorylated eNOS in the presence of BH4 deficiency may even lead to increased O2− generation by the uncoupled enzyme.15 Therefore, we examined whether circulating adiponectin was related to vascular BH4 content. In a subgroup of 176 patients from study 1, circulating adiponectin was positively correlated with BH4 and the ratio of BH4 to total biopterins in both SV and IMA segments (Figure 7). Previous reports suggest that BH4 administration increases circulating adiponectin in an animal model,22 but it is unclear whether circulating adiponectin has any direct effect on vascular BH4 bioavailability in humans. We observed a significant but weak association between circulating adiponectin and vascular biopterins in study 1 (Figure 7). To further explore this association, we incubated human SVs and IMAs with adiponectin (10 μg/mL) for 6 hours. Adiponectin increased vascular BH4 and the ratio of BH4 to total biopterins (Figure 7), whereas inhibition of GTP-cyclohydrolase with DAHP resulted in the expected reduction of vascular BH4 levels, even in the presence of adiponectin. Taken together, these findings suggest that adiponectin improves NO and reduces O2− bioavailability in human vessels through a combined effect on eNOS activation (via PI3/Akt phosphorylation) and coupling (by increasing BH4 bioavailability).
Effects of Vascular Oxidative Stress on Adiponectin Expression in Perivascular AT
To explore the positive association observed between O2− production in human SVs/IMAs and adiponectin release/ADIPOQ expression in the respective perivascular AT, we then examined whether products of peroxidation released from the vascular wall may regulate ADIPOQ expression in perivascular AT. Recent data suggested that 4-HNE (a product of lipid peroxidation) may upregulate ADIPOQ gene expression in both primary human adipocytes18 and skeletal muscle cells23; therefore, we hypothesized that 4-HNE may be involved in the cross-talk between vascular wall and perivascular AT in humans. To examine whether increased vascular O2− leads to increased 4-HNE production in human vessels, we first quantified the levels of 4-HNE protein adducts in SV (n=11) and IMA (n=18) segments from our cohort in study 1 and demonstrated a good correlation between vascular O2− and 4-HNE protein adducts in both SVs (r=0.720, P=0.042) and IMAs (r=0.489, P=0.039; representative examples are shown in Figure 8), suggesting that there is increased production of 4-HNE in human arteries and veins in the presence of increased vascular oxidative stress. We then incubated peri–SV-AT (n=6) and peri–IMA-AT (n=6) with 4-HNE 30 µmol/L for 16 hours and observed that ADIPOQ gene expression was upregulated by ≈2 fold in both peri–SV-AT (Figure 8) and peri–IMA-AT (by 1.9±0.3 fold; P<0.05 versus control). This was accompanied by a parallel upregulation of PPAR-γ in peri–SV-AT (Figure 8) and peri–IMA-AT (by 1.7±0.4 fold; P<0.05 versus control), whereas in the presence of T0070908 (an inhibitor of PPAR-γ activity) at 10 µmol/L, the effect of 4-HNE on ADIPOQ gene expression in peri–SV-AT was abolished (Figure 8). As a positive control in these experiments, we used CD36 (a downstream molecule with an expression that is under the direct control of PPAR-γ), which showed responses similar to ADIPOQ in both peri–SV-AT (Figure 8) and peri–IMA-AT (upregulated by 1.8±0.4-fold after incubation with 4-HNE). These experiments suggest that 4-HNE produced in human SVs and IMAs in the presence of increased vascular oxidative stress may upregulate the expression of ADIPOQ gene in perivascular AT through a PPAR-γ–dependent mechanism.
In the present study, we examine the role of both circulating and locally produced adiponectin in the regulation of vascular redox state in patients with atherosclerosis. We demonstrate for the first time in humans that, in addition to its association with NO bioavailability, circulating adiponectin is inversely related to vascular O2− (derived from uncoupled eNOS) in human arteries and veins, independently of atherosclerosis risk factors. These findings are also confirmed by linking the genetic variability of ADIPOQ with vascular redox state and NO bioavailability and by using ex vivo experiments with human arteries and veins. We further demonstrate that the effect of adiponectin on eNOS coupling is mediated by its combined impact on PI3/Akt-mediated phosphorylation of eNOS and vascular BH4 bioavailability. In contrast, increased vascular O2− is associated with increased local adiponectin release/ADIPOQ gene expression in perivascular AT, implying that local mechanisms related to the vascular redox state may control adiponectin synthesis in perivascular AT in patients with atherosclerosis. In additional ex vivo experiments in human vessels and perivascular AT, we demonstrated that vascular oxidative stress induces the release of products of lipid peroxidation (ie, 4-HNE) that upregulate ADIPOQ gene in the perivascular AT via a PPAR-γ–dependent mechanism. This cycle of cross-talk between the vessel and perivascular AT may represent a novel defense mechanism of the human vascular wall against oxidative stress (Figure 8).
Adiponectin and Endothelial Dysfunction
Adiponectin is an adipokine with antiatherogenic properties in experimental models but has a controversial role in the clinical setting.1 Although circulating adiponectin is reduced in obesity24 and is related to reduced cardiovascular risk in the general population,3 increased circulating adiponectin is inversely correlated with mortality and the overall clinical outcome in advanced cardiovascular disease states such as heart failure.11 Moreover, pharmacological treatments that improve endothelial function and reduce cardiovascular risk (such as lipophilic statins) appear to reduce circulating adiponectin in patients with hypercholesterolemia.25,26 This discordance between experimental data and clinical observations has introduced the concept that adiponectin may behave as a rescue hormone in advanced disease states,1 although the mechanisms behind this hypothesis have not been explored in the clinical setting.
Experimental evidence suggests that adiponectin stimulates eNOS-derived NO in endothelial cells through stimulation of PI3/Akt-mediated eNOS phosphorylation.9,27 In keeping with these first observations from cell culture models, adiponectin knockout mice exhibit reduced eNOS phosphorylation status (at Ser1177) and impaired endothelial function.10 However, the relevance of these observations to humans is totally unknown. In clinical studies, circulating adiponectin is directly related to endothelial function in the general population,28 but this association is reversed in patients with type 2 diabetes mellitus and diabetic nephropathy,29 introducing further controversy into the role of adiponectin in different clinical disease states.
In the present study, we demonstrate that circulating adiponectin is an independent predictor of endothelial function in a well-phenotyped cohort of patients with coronary artery disease without renal or heart failure. In a first attempt to prove a causal link between adiponectin and endothelial function, we examined whether genetic polymorphisms known from genome-wide association studies to regulate circulating adiponectin levels (rs17366568 in ADIPOQ and rs266717 in ADIPOQ promoter) had an impact on endothelial function. We observed that genetically determined lower circulating adiponectin levels were related to worse endothelial function. To further investigate the molecular mechanisms underlying this observation, we exposed human SVs and IMAs to adiponectin ex vivo and observed a rapid PI3/Akt-mediated increase in eNOS phosphorylation at Ser1177, which is known to activate the enzyme.15 Although eNOS is the source of NO in the vascular endothelium, deficiency of its cofactor BH4 leads to its uncoupling, turning it into a source of O2− instead of NO.30 We demonstrate for the first time that adiponectin increases vascular BH4, a critical cofactor necessary for eNOS enzymatic coupling in human vessels.12 Although the biological significance of circulating adiponectin in the regulation of vascular BH4 bioavailability in humans in vivo needs further validation, our finding is compatible with the strong association we observe between circulating adiponectin and eNOS coupling in human arteries and veins. Therefore, it is likely that even this rather weak effect of adiponectin on vascular biopterins may have a significant impact on eNOS coupling in vivo. The combined phosphorylation of eNOS and increase in its enzymatic coupling also explain the impact of circulating adiponectin on vascular NO bioavailability in humans.
Adiponectin and Vascular Superoxide Generation
Evidence from cell culture models suggests that activation of eNOS via Ser1177 phosphorylation leads to either increased O2− or NO, depending on the underlying coupling status of the enzyme.15 Given that vascular BH4 is oxidized by free radicals and the balance between its synthesis and oxidation defines its net bioavailability,12 any change in vascular redox state could drive eNOS uncoupling and trigger a proatherogenic vicious cycle.
We demonstrate for the first time in humans that circulating adiponectin is an independent predictor of vascular O2− and eNOS coupling in arteries and veins from patients with coronary atherosclerosis. By using the genetic variability of ADIPOQ and ex vivo models of human arteries and veins, we also demonstrate that adiponectin reduces vascular O2− by its combined impact on BH4-mediated eNOS coupling and PI3/Akt-mediated activation of the enzyme.
Effects of Vascular Oxidative Stress on Adiponectin Expression in Perivascular AT
Evidence suggests that ischemia/reperfusion injury leads to increased adiponectin expression in brain vessels,31 a vascular bed where uncoupled eNOS is a key contributor to vascular O2− generation.32 Moreover, it has recently been shown that 4-HNE, a product of peroxidation released from tissues with high oxidative burden, upregulates ADIPOQ in primary human and 3T3 adipocytes,18 as well as in skeletal muscle.23 These findings support the notion that adiponectin may behave as a rescue hormone, being upregulated in conditions of increased oxidative stress.
In our study, vascular O2− production (in both arteries and veins) was paradoxically positively correlated with adiponectin release and ADIPOQ gene expression in perivascular AT. We have shown that peroxidation products (ie, 4-HNE, produced in the vascular wall in the presence of increased vascular oxidative stress) upregulate ADIPOQ gene in perivascular AT in a PPAR-γ–dependent mechanism. These findings introduce the concept of a novel, local defense mechanism of the vascular wall against oxidative stress that relies on the continuous cross-talk between the vessel and its perivascular fat.
Given that both vessel types used in our study (IMA and SV) were free of atherosclerosis, it is unclear whether these findings are applicable to the human coronary arteries, especially in the presence of coronary atherosclerosis. Although our ex vivo models use human vessels obtained from patients with advanced atherosclerosis (therefore chronically exposed to a proatherogenic environment), the absence of atherosclerotic lesions in these vessels is a potential limitation of the study. Moreover, the extent to which adiponectin produced in perivascular AT is sufficient to effectively control redox state in the human vascular wall in vivo is hard to estimate, given the complexity of the mechanisms implicated in redox state regulation in the human vascular wall. Finally, it is unclear whether our findings are applicable to human arteries and veins from healthy individuals because these vessels are harvested only during coronary artery bypass graft surgery, and these patients are also under multiple pharmacological treatments.
This is the first study demonstrating that adiponectin has a direct impact on redox state in human arteries and veins through its combined effect on BH4-mediated improvement of eNOS coupling and PI3/Akt-mediated phosphorylation of eNOS. It also introduces the novel concept that increased oxidative stress in the vessel wall leads to the release of peroxidation products (ie, 4-HNE) that upregulate adiponectin gene expression in perivascular AT via a PPAR-γ–dependent mechanism. These findings propose for the first time a bidirectional cross-talk between the human vascular wall and perivascular AT, with potentially important implications in vascular biology.
Sources of Funding
This study was funded by the
Antoniades C, Antonopoulos AS, Tousoulis D, Stefanadis C. Adiponectin: from obesity to cardiovascular disease.Obes Rev. 2009; 10:269–279.CrossrefMedlineGoogle Scholar
Han SH, Sakuma I, Shin EK, Koh KK. Antiatherosclerotic and anti-insulin resistance effects of adiponectin: basic and clinical studies.Prog Cardiovasc Dis. 2009; 52:126–140.CrossrefMedlineGoogle Scholar
Pischon T, Girman CJ, Hotamisligil GS, Rifai N, Hu FB, Rimm EB. Plasma adiponectin levels and risk of myocardial infarction in men.JAMA. 2004; 291:1730–1737.CrossrefMedlineGoogle Scholar
Gong M, Long J, Liu Q, Deng HC. Association of the ADIPOQ rs17360539 and rs266729 polymorphisms with type 2 diabetes: a meta-analysis.Mol Cell Endocrinol. 2010; 325:78–83.CrossrefMedlineGoogle Scholar
Chung CM, Lin TH, Chen JW, Leu HB, Yang HC, Ho HY, Ting CT, Sheu SH, Tsai WC, Chen JH, Lin SJ, Chen YT, Pan WH. A genome-wide association study reveals a quantitative trait locus of adiponectin on CDH13 that predicts cardiometabolic outcomes.Diabetes. 2011; 60:2417–2423.CrossrefMedlineGoogle Scholar
Greenstein AS, Khavandi K, Withers SB, Sonoyama K, Clancy O, Jeziorska M, Laing I, Yates AP, Pemberton PW, Malik RA, Heagerty AM. Local inflammation and hypoxia abolish the protective anticontractile properties of perivascular fat in obese patients.Circulation. 2009; 119:1661–1670.LinkGoogle Scholar
Guzik TJ, Marvar PJ, Czesnikiewicz-Guzik M, Korbut R. Perivascular adipose tissue as a messenger of the brain-vessel axis: role in vascular inflammation and dysfunction.J Physiol Pharmacol. 2007; 58:591–610.MedlineGoogle Scholar
Hattori Y, Suzuki M, Hattori S, Kasai K. Globular adiponectin upregulates nitric oxide production in vascular endothelial cells.Diabetologia. 2003; 46:1543–1549.CrossrefMedlineGoogle Scholar
Xi W, Satoh H, Kase H, Suzuki K, Hattori Y. Stimulated HSP90 binding to eNOS and activation of the PI3-Akt pathway contribute to globular adiponectin-induced NO production: vasorelaxation in response to globular adiponectin.Biochem Biophys Res Commun. 2005; 332:200–205.CrossrefMedlineGoogle Scholar
Cao Y, Tao L, Yuan Y, Jiao X, Lau WB, Wang Y, Christopher T, Lopez B, Chan L, Goldstein B, Ma XL. Endothelial dysfunction in adiponectin deficiency and its mechanisms involved.J Mol Cell Cardiol. 2009; 46:413–419.CrossrefMedlineGoogle Scholar
Beatty AL, Zhang MH, Ku IA, Na B, Schiller NB, Whooley MA. Adiponectin is associated with increased mortality and heart failure in patients with stable ischemic heart disease: data from the Heart and Soul Study.Atherosclerosis. 2012; 220:587–592.CrossrefMedlineGoogle Scholar
Antoniades C, Shirodaria C, Crabtree M, Rinze R, Alp N, Cunnington C, Diesch J, Tousoulis D, Stefanadis C, Leeson P, Ratnatunga C, Pillai R, Channon KM. Altered plasma versus vascular biopterins in human atherosclerosis reveal relationships between endothelial nitric oxide synthase coupling, endothelial function, and inflammation.Circulation. 2007; 116:2851–2859.LinkGoogle Scholar
Katusic ZS. Vascular endothelial dysfunction: does tetrahydrobiopterin play a role?Am J Physiol Heart Circ Physiol. 2001; 281:H981–H986.CrossrefMedlineGoogle Scholar
Milstien S, Katusic Z. Oxidation of tetrahydrobiopterin by peroxynitrite: implications for vascular endothelial function.Biochem Biophys Res Commun. 1999; 263:681–684.CrossrefMedlineGoogle Scholar
Chen CA, Druhan LJ, Varadharaj S, Chen YR, Zweier JL. Phosphorylation of endothelial nitric-oxide synthase regulates superoxide generation from the enzyme.J Biol Chem. 2008; 283:27038–27047.CrossrefMedlineGoogle Scholar
Antoniades C, Bakogiannis C, Leeson P, Guzik TJ, Zhang MH, Tousoulis D, Antonopoulos AS, Demosthenous M, Marinou K, Hale A, Paschalis A, Psarros C, Triantafyllou C, Bendall J, Casadei B, Stefanadis C, Channon KM. Rapid, direct effects of statin treatment on arterial redox state and nitric oxide bioavailability in human atherosclerosis via tetrahydrobiopterin-mediated endothelial nitric oxide synthase coupling.Circulation. 2011; 124:335–345.LinkGoogle Scholar
Skatchkov MP, Sperling D, Hink U, Mülsch A, Harrison DG, Sindermann I, Meinertz T, Münzel T. Validation of lucigenin as a chemiluminescent probe to monitor vascular superoxide as well as basal vascular nitric oxide production.Biochem Biophys Res Commun. 1999; 254:319–324.CrossrefMedlineGoogle Scholar
Wang Z, Dou X, Gu D, Shen C, Yao T, Nguyen V, Braunschweig C, Song Z. 4-Hydroxynonenal differentially regulates adiponectin gene expression and secretion via activating PPARγ and accelerating ubiquitin-proteasome degradation.Mol Cell Endocrinol. 2012; 349:222–231.CrossrefMedlineGoogle Scholar
Nagy L, Tontonoz P, Alvarez JG, Chen H, Evans RM. Oxidized LDL regulates macrophage gene expression through ligand activation of PPARgamma.Cell. 1998; 93:229–240.CrossrefMedlineGoogle Scholar
Heid IM, Henneman P, Hicks A, Coassin S, Winkler T, Aulchenko YS, Fuchsberger C, Song K, Hivert MF, Waterworth DM, Timpson NJ, Richards JB, Perry JR, Tanaka T, Amin N, Kollerits B, Pichler I, Oostra BA, Thorand B, Frants RR, Illig T, Dupuis J, Glaser B, Spector T, Guralnik J, Egan JM, Florez JC, Evans DM, Soranzo N, Bandinelli S, Carlson OD, Frayling TM, Burling K, Smith GD, Mooser V, Ferrucci L, Meigs JB, Vollenweider P, Dijk KW, Pramstaller P, Kronenberg F, van Duijn CM. Clear detection of ADIPOQ locus as the major gene for plasma adiponectin: results of genome-wide association analyses including 4659 European individuals.Atherosclerosis. 2010; 208:412–420.CrossrefMedlineGoogle Scholar
Richards JB, Waterworth D, O’Rahilly S, Hivert MF, Loos RJ, Perry JR, Tanaka T, Timpson NJ, Semple RK, Soranzo N, Song K, Rocha N, Grundberg E, Dupuis J, Florez JC, Langenberg C, Prokopenko I, Saxena R, Sladek R, Aulchenko Y, Evans D, Waeber G, Erdmann J, Burnett MS, Sattar N, Devaney J, Willenborg C, Hingorani A, Witteman JC, Vollenweider P, Glaser B, Hengstenberg C, Ferrucci L, Melzer D, Stark K, Deanfield J, Winogradow J, Grassl M, Hall AS, Egan JM, Thompson JR, Ricketts SL, König IR, Reinhard W, Grundy S, Wichmann HE, Barter P, Mahley R, Kesaniemi YA, Rader DJ, Reilly MP, Epstein SE, Stewart AF, Van Duijn CM, Schunkert H, Burling K, Deloukas P, Pastinen T, Samani NJ, McPherson R, Davey Smith G, Frayling TM, Wareham NJ, Meigs JB, Mooser V, Spector TD; GIANT Consortium. A genome-wide association study reveals variants in ARL15 that influence adiponectin levels.PLoS Genet. 2009; 5:e1000768.CrossrefMedlineGoogle Scholar
Wang X, Hattori Y, Satoh H, Iwata C, Banba N, Monden T, Uchida K, Kamikawa Y, Kasai K. Tetrahydrobiopterin prevents endothelial dysfunction and restores adiponectin levels in rats.Eur J Pharmacol. 2007; 555:48–53.CrossrefMedlineGoogle Scholar
Delaigle AM, Senou M, Guiot Y, Many MC, Brichard SM. Induction of adiponectin in skeletal muscle of type 2 diabetic mice: in vivo and in vitro studies.Diabetologia. 2006; 49:1311–1323.CrossrefMedlineGoogle Scholar
Weiss R, Dziura J, Burgert TS, Tamborlane WV, Taksali SE, Yeckel CW, Allen K, Lopes M, Savoye M, Morrison J, Sherwin RS, Caprio S. Obesity and the metabolic syndrome in children and adolescents.N Engl J Med. 2004; 350:2362–2374.CrossrefMedlineGoogle Scholar
Koh KK, Quon MJ, Han SH, Lee Y, Ahn JY, Kim SJ, Koh Y, Shin EK. Simvastatin improves flow-mediated dilation but reduces adiponectin levels and insulin sensitivity in hypercholesterolemic patients.Diabetes Care. 2008; 31:776–782.CrossrefMedlineGoogle Scholar
Koh KK, Quon MJ, Han SH, Lee Y, Kim SJ, Park JB, Shin EK. Differential metabolic effects of pravastatin and simvastatin in hypercholesterolemic patients.Atherosclerosis. 2009; 204:483–490.CrossrefMedlineGoogle Scholar
Cerqueira FM, Brandizzi LI, Cunha FM, Laurindo FR, Kowaltowski AJ. Serum from calorie-restricted rats activates vascular cell eNOS through enhanced insulin signaling mediated by adiponectin.PLoS One. 2012; 7:e31155.CrossrefMedlineGoogle Scholar
Saarikoski LA, Huupponen RK, Viikari JS, Marniemi J, Juonala M, Kähönen M, Raitakari OT. Adiponectin is related with carotid artery intima-media thickness and brachial flow-mediated dilatation in young adults: the Cardiovascular Risk in Young Finns Study.Ann Med. 2010; 42:603–611.CrossrefMedlineGoogle Scholar
Ran J, Xiong X, Liu W, Guo S, Li Q, Zhang R, Lao G. Increased plasma adiponectin closely associates with vascular endothelial dysfunction in type 2 diabetic patients with diabetic nephropathy.Diabetes Res Clin Pract. 2010; 88:177–183.CrossrefMedlineGoogle Scholar
Guzik TJ, Mussa S, Gastaldi D, Sadowski J, Ratnatunga C, Pillai R, Channon KM. Mechanisms of increased vascular superoxide production in human diabetes mellitus: role of NAD(P)H oxidase and endothelial nitric oxide synthase.Circulation. 2002; 105:1656–1662.LinkGoogle Scholar
Yatomi K, Miyamoto N, Komine-Kobayashi M, Liu M, Oishi H, Arai H, Hattori N, Urabe T. Pathophysiological dual action of adiponectin after transient focal ischemia in mouse brain.Brain Res. 2009; 1297:169–176.CrossrefMedlineGoogle Scholar
Santhanam AV, d’Uscio LV, Smith LA, Katusic ZS. Uncoupling of eNOS causes superoxide anion production and impairs NO signaling in the cerebral microvessels of hph-1 mice.J Neurochem. 2012; 122:1211–1218.CrossrefMedlineGoogle Scholar
Obesity is a risk factor for atherosclerosis, but the molecular mechanisms by which the various depots of adipose tissue contribute to atherogenesis are unclear. Adiponectin, an adipokine with a circulating level that is reduced in obesity, appears to have potentially important roles in human cardiovascular disease states. In this study, we explored the role of adiponectin in the cross-talk between adipose tissue and vascular redox state in patients with atherosclerosis. In a large clinical cohort undergoing coronary artery bypass graft surgery and ex vivo models of human arteries and veins obtained during coronary artery bypass graft surgery, we demonstrate a novel role of adiponectin as a key regulator of redox state in human vessels. Adiponectin stimulates the activity and improves enzymatic coupling of endothelial nitric oxide synthase in the human vascular endothelium, reducing vascular superoxide generation while increasing nitric oxide bioavailability. We also demonstrate that peroxidation products from the human vascular wall (ie, 4-hydroxynonenal) upregulate the expression of adiponectin in the local perivascular adipose tissue, which may act back to the vascular wall in a paracrine fashion to supress vascular oxidative stress. These findings describe novel, reciprocal interactions between the human vascular wall and perivascular adipose tissue with potentially important clinical implications.