Associations of the UCP2 Gene Locus With Asymptomatic Carotid Atherosclerosis in Middle-Aged Women

From the Paracelsus Private Medical University and Salzburger Landeskliniken (H.O., B.I., K.K., J.U., B.P., W.P.), Salzburg; Konventhospital Barmherzige Brueder (M.H.), Linz; and Krankenhaus Hallein (F.K.), Hallein, Austria.

Correspondence to Wolfgang Patsch, MD, Paracelsus Private Medical University, Muellner Hauptstrasse 48, A-5020 Salzburg, Austria. E-mail w.patsch@salk.at

    Abstract

Objective— Reactive oxygen species (ROS) contribute to atherogenesis. Uncoupling protein 2 (UCP2) reduces mitochondrial ROS generation and protects against the disease in animal models. A common –866G/A promoter polymorphism that has been associated with obesity and ?-cell function may also affect UCP2 gene expression in cells of the arterial wall.

Methods and Results— Genotype distributions of the –866G/A and of a 45nt-del/ins polymorphism in the 3'-untranslated region of the UCP2 gene were determined in 1334 participants of the Salzburg Atherosclerosis Prevention Program in Subjects at High Individual Risk (SAPHIR). We observed a modest association of the –866G/A promoter polymorphism and 2-loci haplotypes with asymptomatic carotid atherosclerosis in female study participants. Functional studies revealed increased expression of the –866G wild-type allele in human umbilical vein endothelial cells and differentiated THP-1 cells. Electrophoretic mobility shift assay studies and antibody-interference assays performed with nuclear extracts of various cell lines showed binding of cell-type specific protein complexes to the region encompassing the –866 site and suggested involvement of hypoxia inducible factor 1 in the regulation of UCP2 gene expression in endothelial cells and macrophages.

Conclusions— Our results suggest a role of UCP2 in atherogenesis as originally proposed from studies in animal and cell culture models.

Uncoupling protein 2 (UCP2) reduces mitochondrial reactive oxygen species production and protects against atherosclerosis in animal models. We report that the UCP2 gene locus is associated with asymptomatic carotid atherosclerosis in females of a cross-sectional study and that a common promoter polymorphism affects UCP2 gene expression in endothelial cells and macrophages.

Key Words: reactive oxygen species ? uncoupling protein ? single nucleotide polymorphism ? gene expression ? haplotype

    Introduction

Virtually all cell types of the vessel wall produce reactive oxygen species (ROS). At low levels, intracellular ROS act as signaling molecules that modulate a diverse array of biochemical functions. However, increased ROS production is associated with significant damage to endothelial and smooth muscle cells1,2 and has been implicated in the pathogenesis of vascular disorders, such as atherosclerosis.3,4 Among cellular sources of ROS generation, the electron transport chain may play an important role, because up to 10% of the reducing equivalents from NADH leak to form superoxide anions and H2O2.5 When the electrochemical potential difference generated by the proton gradient along the inner mitochondrial membrane is high, electron transport becomes inhibited at complex III. As a result, the lifespan of ROS-generating intermediates is prolonged, and the flux of ROS from mitochondria is increased.6 In cultured bovine aortic endothelial cells, increases in ROS are prevented by chemical uncouplers of oxidative phosphorylation or by ectopic expression of uncoupling protein 1 (UCP1).7 Whereas UCP1 is solely expressed in brown adipose tissue,8 the related UCP2 is expressed in many tissues and cells, including macrophages.9 Indeed, disruption of the UCP2 gene enhances ROS generation in transgenic mice.10 Moreover, bone marrow transplants from UCP2-deficient mice increased atherosclerotic lesion size and markers of oxidative stress in low-density lipoprotein receptor-deficient mice.11

We previously described a common functional G/A polymorphism in the UCP2 gene at nucleotide position –866 relative to the transcription start site. The variant A allele was associated with enhanced adipose tissue mRNA expression, increased transcription of a reporter gene in the human adipocyte cell line PAZ6, and decreased risk of obesity in middle-aged humans.12 In a ?-cell environment, the variant allele was more effectively activated by the transcription factor PAX6 and was associated with a reduced disposition index (the product of the acute insulin response to glucose and insulin sensitivity) in obese subjects.13 The latter results are supported by studies in animal models showing an inverse association of UCP2 gene expression and ?-cell function.14 Among other polymorphic sites in the UCP2 gene that showed associations with various metabolic parameters in diverse populations,15,16 a 45nt deletion/insertion polymorphism located in the 3'-untranslated region appeared to be functional, because mRNA transcribed from the insertion allele displayed a shorter half-life in a fetal myoblast cell line than mRNA transcribed from the deletion allele.12

To test the hypothesis that the UCP2 gene locus contributes to the pathogenesis of human atherosclerosis, we studied associations of both functional nucleotide polymorphisms (SNPs) with asymptomatic carotid artery atherosclerosis in a cross-sectional Austrian sample and characterized effects of the –866G/A polymorphism on transcription and trans-factor binding in a human macrophage cell line and endothelial cells.

    Materials and Methods

Study Population

We studied 40- to 65-year-old male and 45- to 70-year-old female participants of the Salzburg Atherosclerosis Prevention Program in Subjects at High Individual Risk (SAPHIR). The study objectives, its recruitment procedures, and population characteristics have been detailed.12 All study subjects provided informed consent, and the study was approved by the local ethics committee. Diabetes was diagnosed by fasting plasma glucose concentrations of 7.0 mmol/L or use of hypoglycemic medications. Between 7AM and 10PM, ambulatory blood pressure was measured as described.17 Means of systolic and diastolic blood pressure measurements (SBP-day and DBP-day) were used for analyses. Hypertension was defined by SBP-day >140 mm Hg or DBP-day >90 mm Hg or antihypertensive drug therapy.

Intima-media thickness (IMT) and plaque extent of the near and far walls of the common and internal carotid arteries and the bifurcations on both sides were measured by high resolution B-mode ultrasound using the HDI 3000 CV System (ATL) according to the Asymptomatic Carotid Artery Plaque Study (ACAPS) protocol.18 All measurements were conducted and read by a single experienced ultrasound operator who was blinded to all clinical and laboratory measurements. Because of technical difficulties in some subjects, complete IMT measurements were available in 810 men and 463 women. B-scores were obtained by grading on a 5-point scale ranging from 0 (normal) to 5 (complete luminal obstruction). Adding the B-score of all segments resulted in a sum B-score. For risk calculation, a sum B-score >1 or a score of 1 together with a maximum IMT above the 90th sex-specific percentile was considered as evidence for atherosclerotic disease. To reduce possible confounding resulting from therapeutic interventions, 32 male and 61 female subjects with symptomatic atherosclerotic diseases (myocardial infarction, angina pectoris, stroke, transitory ischemic attack, or peripheral arterial disease) were excluded. Tissue samples were obtained from the musculus rectus abdominis as described.19

Laboratory Determinations

Laboratory parameters in fasting plasma and the –866G/A and 45nt-del/ins polymorphisms were determined as described.12 High-sensitivity C-reactive protein (CRP) was measured using a Hitachi 917 analyzer and the respective kit (Roche Diagnostics).

Cell Culture, Transfections, and mRNA Quantification

INS1-E and PAZ6 cells were cultured as described.13,20 THP-1 cells were grown in RPMI medium 1640 supplemented with 0.05 mmol/L ?-mercaptoethanol/10% FBS and differentiated by the addition of 12-O-tetradecanoylphorbol-13 acetate for 48 hours. Human umbilical vein endothelial cells (HUVECs) were cultured in endothelial cell growth medium (Technoclone, Vienna, Austria) supplemented with 0.1 ng/mL epidermal growth factor and 1.0 ng/mL basic fibroblast growth factor. Cells were transfected with 0.5 μg of UCP2 –866G and –866A reporter plasmids,12 respectively, using LipofectAMINE (Invitrogen, Carlsbad, Calif). pRL-TK (20 ng per well; Promega, Madison, Wis) was cotransfected as internal control. Cells were harvested 24 hours later and luciferase activities were determined.20 For details on RNA isolation and quantitative real-time polymerase chain reaction analyses, see http://atvb.ahajournals.org.

Electrophoretic Mobility Shift Assays

Nuclear extracts and electrophoretic mobility shift assays were performed as described.13 Double-stranded competitor probes designed for aryl hydrocarbon receptor/aryl hydrocarbon receptor nuclear translocator (AhR/ARNT) and hypoxia inducible factor (HIF)-1/ARNT binding matrix consensus sequences were 5'-AGCTTGCTGGGGGCATTGCGTGACA-3'21 and 5'-AGCTTGC-CGTAGGTACTGTCTCAG -3',22 respectively. For antibody-interference assays, nuclear extracts were preincubated with 1 μg of monoclonal anti-AhR (Alexis Biochemicals, Montreal, Canada) or anti-HIF-1 (BD Transduction Laboratories, San Jose, Calif) antibodies as indicated. Intensities of bands were quantified by scanning autoradiographs as described.12

Statistics

Differences of continuous variables between study subjects with and without carotid atherosclerosis as well as effects of genotypes on clinical parameters were ascertained by 2-way ANOVA. Logarithmic transformations were made if the equal variance and normality assumptions of ANOVA were rejected. Measurements were adjusted for effects of age, sex, and body mass index (BMI) as indicated. For testing the significance of multiple comparisons, the Bonferroni correction was used. Allele frequencies were estimated by gene counting. Agreement with Hardy-Weinberg expectations was tested using a 2 goodness-of-fit test. The standardized pair-wise linkage disequilibrium statistics (D') and haplotype frequencies were estimated according to Terwilliger and Ott.23 Differences in genotype frequencies between subjects with and without carotid atherosclerosis were determined using a 2 distribution with 2 degrees of freedom. For details on the estimation of odds ratios (ORs) for each genotype and testing of associations of haplotypes with quantitative and binary traits, refer to http://atvb.ahajournals.org.

    Results

Average values for age and BMI as well as the prevalence of hypertension were higher in subjects with asymptomatic carotid artery atherosclerosis. Among laboratory parameters, average values for cholesterol were higher in subjects with asymptomatic disease, whereas average values for high-density lipoprotein (HDL) cholesterol were lower (Table 1). Sex-specific differences were observed for BMI and the proportion of smokers. Women displayed lower average values for glucose, triglycerides, and apolipoprotein B but higher average values for HDL cholesterol, apolipoprotein A-I, and CRP.

   TABLE 1. Characteristics of the Study Population

The common UCP2 –866G/A promoter and a 45nt-del/ins polymorphism were typed in 1334 SAPHIR participants. Both polymorphisms fulfilled Hardy-Weinberg expectations and were in positive standardized linkage disequilibrium (D/Dmax>0.733). Asymptomatic carotid atherosclerosis was ascertained by measures of maximum IMT and a B-score describing morphological alterations of the carotid arteries. Subjects with and without asymptomatic carotid atherosclerosis showed a borderline significant difference in the distribution of –866G/A genotypes, but a significant difference of 2-loci haplotypes, whereas 45nt-del/ins genotype frequencies did not differ (Table I, available online at http://atvb.ahajournals.org). Strong statistical evidence for an interaction of sex with the –866G/G genotype (P<0.0001) was observed in multivariate logistic analyses. Sex-specific comparison revealed different distributions of genotypes associated with the –866 SNP in women but not in men (Table 2). Global haplotype score statistics revealed a disease association in women. The lowest and highest haplotype-specific scores were observed for the –866G/45nt-del and –866A/45nt-ins haplotypes, respectively. No associations were observed in men (Table 3). To verify these results and to calculate ORs, we used noninteger haplotype predictions in multivariate logistic models that considered established risk factors. Again, the –866G/45nt-del haplotype was associated with a reduced OR in women but not in men (Table 4). Consideration of all possible haplotype interactions in the full model showed no significant interactions in either male or female subjects (not shown). As the –866A/45nt-ins haplotype was dropped in the logistic model because of collinearity, we orthogonalized haplotypes and again observed a reduced OR only for the –866G/45nt-del haplotype. In women, ORs (95% CI) for the –866G/45nt-del haplotype relative to the –866G/45nt-ins, –866A/45nt-del, and –866A/45nt-ins haplotypes were 0.69 (0.53 to 0.92, P=0.011), 0.739 (0.56 to 0.97, P=0.032), and 0.64 (0.49 to 0.85, P=0.002), respectively.

   TABLE 2. UCP2 Polymorphisms and Associated Risk of Asymptomatic Atherosclerosis in Male and Female Subjects

   TABLE 3. UCP2 Haplotype Scores for Asymptomatic Atherosclerosis in Male and Female Subjects

   TABLE 4. Multivariate Logistic Regression Model for Asymptomatic Atherosclerosis, Traditional Risk Factors, and UCP2 Haplotypes

IMTs of the carotid arteries have been shown to be surrogate markers for subclinical atherosclerosis.24 In our study population, common and internal carotid arteries IMTs were strongly associated with case-control status defined by B-score and maximum IMT (Table II, available online at http://atvb.ahajournals.org). Unexpectedly, no associations of the 2 polymporphisms with average IMTs of the common or internal carotid arteries or maximum IMT were observed in our male and female study participants (Table III, available online at http://atvb.ahajournals.org), and only trends were observed for haplotypes in multivariate regression models (Table IV, available online at http://atvb.ahajournals.org). Because the determinants for carotid phenotypes, such as IMT, carotid plaque area, and carotid stenosis, may differ,25 we considered the possibility that the UCP2 haplotypes were mainly associated with plaques rather than IMT. We adjusted for average IMTs of the common or internal carotid arteries in the model and observed an OR of 0.506 (0.335 to 0.766, P<0.001) for the –866G/45nt-del haplotype in women. Moreover, haplotype scores were similar as in the model not adjusted for IMT (Table 3, Models I and II). In addition, we reclassified the case-control status in that at least 1 score 2-lesion (1.5 to 2 mm) had to be detected in cases. Using this definition, the lowest haplotype score of and a reduced OR for the –866G/45nt-del haplotype was observed in females (Tables V and VI, available online at http://atvb.ahajournals.org).

Quantitative real-time RT-PCR analyses showed that UCP2 mRNA expression levels were 6-fold higher in differentiated THP-1 cells and 4-fold lower in HUVECs than in human skeletal muscle tissue. Transient transfection studies with –866G/A reporter constructs revealed that transcriptional activities of variant and wild-type promoters differed in both THP-1 cells and HUVECs in that transactivation of the wild-type promoter was more effective compared with the variant promoter (Figure 1). This observation is in contrast to previous results in PAZ6 cells.12

   Figure 1. UCP2 promoter studies in HUVECs and THP-1 cells. HUVECs (A) or THP-1 cells (B) were transiently transfected with UCP2 promoter constructs –866G or –866A. P<0.02, –866G vs –866A in both cell types; the relative luciferase activity of the –866G construct was set to 100. Results are representative for 1 of 3 independent experiments, each performed in quadruplicate. Black and white bars represent the –866G and the –866A reporter construct, respectively. Results are means±SD.

Electrophoretic mobility shift assays with nuclear extracts from THP-1 cells and HUVECs revealed similar DNA-protein complexes, whereas the pattern of complexes with nuclear extracts from INS1-E and PAZ6 cells clearly differed, indicating differences in transcription factor binding (Figure 2). Unlabeled bandshift probe was a highly effective competitor in all cell lines, arguing for the specificity of these complexes. Our previous studies identified PAX6 as the transcription factor that accounted for the different activities of the 2 promoter alleles in INS1-E cells, but computational analyses suggested putative binding sites for additional transcription factors, including the HIF-1/ARNT- and AhR/ARNT-heterodimers.12 Therefore, consensus probes for HIF-1/ARNT and AhR/ARNT were used in competition experiments in a 5- to 100-fold molar excess (Figure 2). The HIF-1/ARNT consensus oligonucleotide competed as effectively as the unlabeled bandshift probe, whereas competition with the AhR/ARNT probe was two times less effective. In antibody-interference assays, formation of DNA-protein complexes in HUVECs was inhibited by antibodies against HIF-1, but antibodies against Ahr showed little or no effect. Similar findings were obtained in THP-1 cells (data not shown). These results strongly suggest that HIF-1/ARNT binds to the multifunctional site around –866 in HUVECs and THP-1 macrophages.

   Figure 2. Electrophoretic mobility shift and super-shift assays with INS1-E, PAZ6, THP-1, and HUVEC nuclear extracts. Nuclear extracts were incubated with a 32P-labeled double-stranded UCP2 –866G probe, both in the absence or presence of specific competitor oligonucleotides or monoclonal antibodies. Specificity of binding of nuclear proteins from INS1-E cells and differentiated PAZ6 and THP-1 cells was demonstrated by competition with a 100-fold molar excess of cold target DNA. Competition studies in HUVECs included double-stranded competitor probes for either UCP2 –866G, AhR/ARNT, or HIF-1/ARNT in increasing amounts (5- and 100-fold molar excess) as indicated. For antibody-interference assays HUVEC nuclear extracts were preincubated at 4°C for 1 hour with 1 μg of anti-HIF-1 or anti-AhR monoclonal antibodies, respectively, as indicated.

    Discussion

UCP2 belongs to the mitochondrial anion carrier family and has been shown to modulate the production of ROS by decreasing the mitochondrial membrane potential.26 UCP2 overexpression in macrophages limits steady-state levels of intracellular ROS,27 whereas UCP2 antisense oligonucleotides increased intracellular ROS generation in murine endothelial cells.28 Earlier studies in macrophages of UCP2 knock-out mice (UCP2–/–) also demonstrated an excess production of ROS in response to Toxoplasma gondii infection.10 In ob/ob mice, UCP2 expression in macrophages was lower, and mitochondrial ROS production was higher than in control mice.29 UCP2 overexpression in THP-1 macrophages inhibited chemoattractant protein-1–induced transendothelial migration, thereby reducing atherosclerotic plaque formation.27 Taken together, these and other findings11 suggest a protective role for UCP2 in atherosclerosis by reducing ROS generation.

We previously reported associations of the –866G/A polymorphism with obesity12 and ?-cell dysfunction.13 Several studies have since reported associations of the –866G/A SNP with various metabolic prameters.30–32 While this work was in progress, Dhamrait et al reported a doubling in the risk of coronary heart disease for –866A/A homozygotes in a prospective study of 2695 healthy men.33 Our study shows that the –866G/A site alone or in a haplotype context with the 45nt-del/ins polymorphism is associated with asymptomatic carotid artery atherosclerosis in female subjects of the SAPHIR population. These associations remained significant after consideration of possible UCP2 genotype or haplotype effects on obesity and type 2 diabetes. Our functional analyses show that the –866G allele displays stronger transcriptional activity in both THP-1 cells and HUVECs. Such increased UCP2 mRNA expression would be expected to reduce ROS generation in endothelial cells and macrophages, thereby protecting against oxidative damage and atherosclerosis. Based on our previous studies, the 45nt-del/ins polymorphism would have a weaker effect than the –866 SNP, but the 45nt-del allele would be predicted to enhance UCP2 expression in comparison to the 45nt-ins allele. The lower frequency of the –866G/45nt-del haplotype in women with asymptomatic atherosclerosis compared with controls and the lack of an association of the 45nt-del/ins site with disease is consistent with this concept. We have no clear explanation why the associations were only observed in women. In our population, the effects of established risk factors such as hypertension, plasma cholesterol, and CRP appeared to be much stronger in men than in women. It is therefore possible that effects of UCP2 haplotypes on atherogenesis in men, as reported by Dhamrait, were concealed in our male population.

The lack of associations between UCP2 haplotypes and IMTs was unexpected, because modest to moderate associations of IMTs with various disease end points have been documented.24,25,34 However, it is possible that the phenotypes of carotid atherosclerosis established by ultrasound determinations and probably representing different stages of disease have common and distinct determinants. It has been shown that total carotid plaque area and carotid stenosis displayed different associations with specific risk factors.25,35 Moreover, IMT was a stronger predictor of stroke than of myocardial infarction,24 whereas total plaque area was a stronger predictor of myocardial infarction than of stroke.25,36 Our initial definition of cases rested primarily on the presence of subclinical plaques. The sex-specific association of UCP2 haplotypes with disease was maintained after adjustment for IMT and after case-control reclassification that was solely based on the presence of slightly more advanced plaques. Hence, our results suggest but do not prove that enhanced UCP2 gene expression may reduce plaque development. Such a localized effect of UCP2 could be expected because of its intracellular localization. Nevertheless, the cross-sectional study design warrants cautious interpretation of results. Clearly, additional studies are necessary to substantiate these findings.

AhR and HIF-1 are both ligand-activated members of the basic helix-loop-helix/Per-ARNT-Sim transcription factor family.37 AhR mediates the biological and toxic effects of environmental pollutants, such as dioxin, whereas HIF-1 is a central regulator of cellular responses to hypoxia.37,38 Our studies with HIF-1/ARNT and AhR/ARNT consensus binding sequences and antibodies directed against the respective heterodimer partners strongly suggest a role of HIF-1 in the regulation of UCP2 gene expression in human endothelial cells and macrophages.

An imbalance between demand and supply of oxygen in the arterial wall is critical to the development of atherosclerotic lesions.39 Under normoxic conditions, HIF-1 is rapidly degraded by the proteasome.40 Because hypoxia stabilizes HIF-1, its target genes, including vascular endothelial growth factor, matrix metalloproteinases, and plasminogen activator inhibitor-1 are induced in response to oxygen deprivation.41 Hypoxia also enhances the intracellular ROS production, and HIF-1 activation is correlated with changes in ROS.42 In transgenic mice, vascular ROS production was induced by smooth muscle cell-specific overexpression of p22phox, a critical component of NAD(P)H oxidase.43 Tgp22vsmc mice displayed elevated levels of hydrogen peroxide in smooth muscle cells, and expression of vascular endothelial growth factor and matrix metalloproteinase-9 was induced by increased levels of HIF-1 and accompanied by enhanced arterial lesions.43 Transactivation of the UCP2 gene by HIF-1, as suggested by our studies, may therefore represent a feedback loop that reduces cellular ROS levels originating from mitochondria.

    Acknowledgments

This study was supported by Jubilaeumsfondsprojekt No. 10932 and No. 10678 from the Oesterreichische Nationalbank and by grants from the Medizinische Forschungsgesellschaft Salzburg and the Verein für Medizinische Fortbildung, Salzburg. We thank Prof D.O. Stram for his help and guidance in statistical analyses of haplotype effects.

References

Lum H, Roebuck KA. Oxidant stress and endothelial cell dysfunction. Am J Physiol Cell Physiol. 2001; 280: C719–C741.

Lander HM. An essential role for free radicals and derived species in signal transduction. FASEB J. 1997; 11: 118–124.

Griendling KK, FitzGerald GA. Oxidative stress and cardiovascular injury: part II: animal and human studies. Circulation. 2003; 108: 2034–2040.

Cai H, Harrison DG. Endothelial dysfunction in cardiovascular diseases: the role of oxidant stress. Circ Res. 2000; 87: 840–844.

Marnett LJ, Riggins JN, West JD. Endogenous generation of reactive oxidants and electrophiles and their reactions with DNA and protein. J Clin Invest. 2003; 111: 583–593.

Brownlee M. A radical explanation for glucose-induced beta cell dysfunction. J Clin Invest. 2003; 112: 1788–1790.

Nishikawa T, Edelstein D, Du XL, Yamagishi S, Matsumura T, Kaneda Y, Yorek MA, Beebe D, Oates PJ, Hammes HP, Giardino I, Brownlee M. Normalizing mitochondrial superoxide production blocks three pathways of hyperglycaemic damage. Nature. 2000; 404: 787–790.

Bouillaud F, Ricquier D, Thibault J, Weissenbach J. Molecular approach to thermogenesis in brown adipose tissue: cDNA cloning of the mitochondrial uncoupling protein. Proc Natl Acad Sci U S A. 1985; 82: 445–448.

Fleury C, Neverova M, Collins S, Raimbault S, Champigny O, Levi-Meyrueis C, Bouillaud F, Seldin MF, Surwit RS, Ricquier D, Warden CH. Uncoupling protein-2: a novel gene linked to obesity and hyperinsulinemia. Nat Genet. 1997; 15: 269–272.

Arsenijevic D, Onuma H, Pecqueur C, Raimbault S, Manning BS, Miroux B, Couplan E, Alves-Guerra MC, Goubern M, Surwit R, Bouillaud F, Richard D, Collins S, Ricquier D. Disruption of the uncoupling protein-2 gene in mice reveals a role in immunity and reactive oxygen species production. Nat Genet. 2000; 26: 435–439.

Blanc J, Alves-Guerra MC, Esposito B, Rousset S, Gourdy P, Ricquier D, Tedgui A, Miroux B, Mallat Z. Protective role of uncoupling protein 2 in atherosclerosis. Circulation. 2003; 107: 388–390.

Esterbauer H, Schneitler C, Oberkofler H, Ebenbichler C, Paulweber B, Sandhofer F, Ladurner G, Hell E, Strosberg AD, Patsch JR, Krempler F, Patsch W. A common polymorphism in the promoter of UCP2 is associated with decreased risk of obesity in middle-aged humans. Nat Genet. 2001; 28: 178–183.

Krempler F, Esterbauer H, Weitgasser R, Ebenbichler C, Patsch JR, Miller K, Xie M, Linnemayr V, Oberkofler H, Patsch W. A functional polymorphism in the promoter of UCP2 enhances obesity risk but reduces type 2 diabetes risk in obese middle-aged humans. Diabetes. 2002; 51: 3331–3335.

Zhang CY, Baffy G, Perret P, Krauss S, Peroni O, Grujic D, Hagen T, Vidal-Puig AJ, Boss O, Kim YB, Zheng XX, Wheeler MB, Shulman GI, Chan CB, Lowell BB. Uncoupling protein-2 negatively regulates insulin secretion and is a major link between obesity, beta cell dysfunction, and type 2 diabetes. Cell. 2001; 105: 745–755.

Wang H, Chu WS, Lu T, Hasstedt SJ, Kern PA, Elbein SC. Uncoupling protein-2 polymorphisms in type 2 diabetes, obesity, and insulin secretion. Am J Physiol Endocrinol Metab. 2004; 286: E1–E7.

Walder K, Norman RA, Hanson RL, Schrauwen P, Neverova M, Jenkinson CP, Easlick J, Warden CH, Pecqueur C, Raimbault S, Ricquier D, Silver MH, Shuldiner AR, Solanes G, Lowell BB, Chung WK, Leibel RL, Pratley R, Ravussin E. Association between uncoupling protein polymorphisms (UCP2-UCP3) and energy metabolism/obesity in Pima indians. Hum Mol Genet. 1998; 7: 1431–1435.

Oberkofler H, Holzl B, Esterbauer H, Xie M, Iglseder B, Krempler F, Paulweber B, Patsch W. Peroxisome proliferator-activated receptor-gamma coactivator-1 gene locus: associations with hypertension in middle-aged men. Hypertension. 2003; 41: 368–372.

The ACAPS Group. Rationale and design for the Asymptomatic Carotid Artery Plaque Study (ACAPS). The ACAPS Group. Control Clin Trials. 1992; 13: 293–314.

Oberkofler H, Schraml E, Krempler F, Patsch W. Restoration of sterol-regulatory-element-binding protein-1c gene expression in HepG2 cells by peroxisome-proliferator-activated receptor-gamma co-activator-1alpha. Biochem J. 2004; 381: 357–363.

Oberkofler H, Esterbauer H, Linnemayr V, Strosberg AD, Krempler F, Patsch W. Peroxisome proliferator-activated receptor (PPAR) gamma coactivator-1 recruitment regulates PPAR subtype specificity. J Biol Chem. 2002; 277: 16750–16757.

Swanson HI, Yang J. Mapping the protein/DNA contact sites of the Ah receptor and Ah receptor nuclear translocator. J Biol Chem. 1996; 271: 31657–31665.

Yim S, Choi SM, Choi Y, Lee N, Chung J, Park H. Insulin and hypoxia share common target genes but not the hypoxia-inducible factor-1alpha. J Biol Chem. 2003; 278: 38260–38268.

Terwilliger JD, Ott J. Handbook of Human Genetic Linkage. Baltimore, Md: Johns Hopkins University Press; 1994.

O’Leary DH, Polak JF, Kronmal RA, Manolio TA, Burke GL, Wolfson SK Jr. Carotid-artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. Cardiovascular Health Study Collaborative Research Group. N Engl J Med. 1999; 340: 14–22.

Spence JD, Hegele RA. Noninvasive phenotypes of atherosclerosis: similar windows but different views. Stroke. 2004; 35: 649–653.

Negre-Salvayre A, Hirtz C, Carrera G, Cazenave R, Troly M, Salvayre R, Penicaud L, Casteilla L. A role for uncoupling protein-2 as a regulator of mitochondrial hydrogen peroxide generation. FASEB J. 1997; 11: 809–815.

Ryu JW, Hong KH, Maeng JH, Kim JB, Ko J, Park JY, Lee KU, Hong MK, Park SW, Kim YH, Han KH. Overexpression of uncoupling protein 2 in THP1 monocytes inhibits beta2 integrin-mediated firm adhesion and transendothelial migration. Arterioscler Thromb Vasc Biol. 2004; 24: 864–870.

Duval C, Negre-Salvayre A, Dogilo A, Salvayre R, Penicaud L, Casteilla L. Increased reactive oxygen species production with antisense oligonucleotides directed against uncoupling protein 2 in murine endothelial cells. Biochem Cell Biol. 2002; 80: 757–764.

Lee FY, Li Y, Yang EK, Yang SQ, Lin HZ, Trush MA, Dannenberg AJ, Diehl AM. Phenotypic abnormalities in macrophages from leptin-deficient, obese mice. Am J Physiol. 1999; 276: C386–C394.

Sasahara M, Nishi M, Kawashima H, Ueda K, Sakagashira S, Furuta H, Matsumoto E, Hanabusa T, Sasaki H, Nanjo K. Uncoupling protein 2 promoter polymorphism -866G/A affects its expression in beta-cells and modulates clinical profiles of Japanese type 2 diabetic patients. Diabetes. 2004; 53: 482–485.

Le Fur S, Le Stunff C, Dos SC, Bougneres P. The common -866 G/A polymorphism in the promoter of uncoupling protein 2 is associated with increased carbohydrate and decreased lipid oxidation in juvenile obesity. Diabetes. 2004; 53: 235–239.

Sesti G, Cardellini M, Marini MA, Frontoni S, D’Adamo M, Del Guerra S, Lauro D, De Nicolais P, Sbraccia P, Del Prato S, Gambardella S, Federici M, Marchetti P, Lauro R. A common polymorphism in the promoter of UCP2 contributes to the variation in insulin secretion in glucose-tolerant subjects. Diabetes. 2003; 52: 1280–1283.

Dhamrait SS, Stephens JW, Cooper JA, Acharya J, Mani AR, Moore K, Miller GJ, Humphries SE, Hurel SJ, Montgomery HE. Cardiovascular risk in healthy men and markers of oxidative stress in diabetic men are associated with common variation in the gene for uncoupling protein 2. Eur Heart J. 2004; 25: 468–475.

Adams MR, Nakagomi A, Keech A, Robinson J, McCredie R, Bailey BP, Freedman SB, Celermajer DS. Carotid intima-media thickness is only weakly correlated with the extent and severity of coronary artery disease. Circulation. 1995; 92: 2127–2134.

Al Shali KZ, House AA, Hanley AJ, Khan HM, Harris SB, Zinman B, Mamakeesick M, Fenster A, Spence JD, Hegele RA. Genetic variation in PPARG encoding peroxisome proliferator-activated receptor gamma associated with carotid atherosclerosis. Stroke. 2004; 35: 2036–2040.

Spence JD, Eliasziw M, DiCicco M, Hackam DG, Galil R, Lohmann T. Carotid plaque area: a tool for targeting and evaluating vascular preventive therapy. Stroke. 2002; 33: 2916–2922.

Park H. Aromatic hydrocarbon nuclear translocator as a common component for the hypoxia- and dioxin-induced gene expression. Mol Cells. 1999; 9: 172–178.

Semenza GL. HIF-1 and mechanisms of hypoxia sensing. Curr Opin Cell Biol. 2001; 13: 167–171.

Bjornheden T, Levin M, Evaldsson M, Wiklund O. Evidence of hypoxic areas within the arterial wall in vivo. Arterioscler Thromb Vasc Biol. 1999; 19: 870–876.

Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, Salic A, Asara JM, Lane WS, Kaelin WG Jr. HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science. 2001; 292: 464–468.

Shweiki D, Itin A, Soffer D, Keshet E. Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature. 1992; 359: 843–845.

Paul SA, Simons JW, Mabjeesh NJ. HIF at the crossroads between ischemia and carcinogenesis. J Cell Physiol. 2004; 200: 20–30.

Khatri JJ, Johnson C, Magid R, Lessner SM, Laude KM, Dikalov SI, Harrison DG, Sung HJ, Rong Y, Galis ZS. Vascular oxidant stress enhances progression and angiogenesis of experimental atheroma. Circulation. 2004; 109: 520–525.