Abstract
Context:
Obesity is a complex disease that involves both genetic and environmental perturbations to gene networks in adipose tissue and is proposed as a trigger for metabolic sequelae.
Objective:
We hypothesized that expression of adipose tissue transcripts in gene networks for adaptive response would correlate with the percent fat mass (PFAT) in healthy nondiabetic subjects to maintain metabolic equilibrium and would overlap with genes modulated in response to elevated fatty acid.
Design, Settings, and Patients:
Genome-wide transcript profiles were determined in sc adipose tissue of 136 nondiabetics and in palmitate-induced cells. Genotype information and gene expression data in nondiabetic subjects were integrated to characterize the function of 41 obesity-associated polymorphisms.
Results:
Genes involved in inflammation-immune response, endoplasmic reticulum stress, and cell-extracellular matrix interactions were significantly correlated with PFAT. The NRF2 (nuclear factor erythroid 2-related factor-2)-mediated oxidative stress response pathway was strongly enriched among genes correlated with PFAT in adipose and also emerged as the most enriched pathway among genes differentially expressed by palmitate in vitro. Thioredoxin reductase-1 (TXNRD1) was the most strongly correlated gene (ρ = 0.65). Genes coregulated with TXNRD1 expression indicated a significant interaction network of genes involved in thioredoxin-mediated oxidative stress defense mechanisms and angiogenesis. Pro- and antiangiogenic factors were negatively and positively correlated, respectively, with obesity. Eight obesity genome-wide association study single-nucleotide polymorphisms (SNP) were associated with expression of 10 local transcripts. SNP rs6861681 was the strongest cis-eQTL (expression quantitative trait loci) for CPEB4 (P = 3.02 × 10−9).
Conclusions:
Our study suggests a novel interaction of up-regulated TXN-TXNRD1 system-mediated oxidative stress defense mechanisms and down-regulated angiogenesis pathways as an adaptive response in obese nondiabetic subjects. A subset of obesity-associated SNP regulated expression of transcripts as cis-eQTL.
Obesity is characterized by an increase in adipose tissue mass and has reached epidemic proportions worldwide (1). Heritability studies show evidence for a substantial genetic contribution to obesity risk; however, it is likely that recent lifestyle changes have driven its prevalence. Adipose tissue serves as an integrator of various physiological pathways, including glucose and fatty acid homeostasis. The failure of adipocytes to sequestrate excess fuel during nutritional abundance results in an alteration of gene expression networks within the tissue and is proposed as a trigger for the metabolic sequelae associated with obesity at the systems level (2, 3).
Genome-wide analysis using microarrays permits an unbiased interrogation of gene expression patterns to illuminate the physiological mechanisms that result in obesity. However, recent reviews indicate that most published genome-wide studies compared adipose tissue gene expression in a limited number of obese and lean subjects (4, 5). The microarray studies show the importance of inflammation and immune response pathways in obesity. However, studies in genetically and diet-induced animal models of obesity and pathway-specific validation studies in humans identified several other obesity-associated changes in adipose tissue (2, 6). Recent genome-wide association studies (GWAS) and metaanalyses also identified a large number of single-nucleotide polymorphisms (SNP) associated with obesity (7, 8); however, the role of these SNP in modulating the expression of transcripts associated with obesity is largely unexplored. Thus, identifying perturbations in gene expression networks and delineating the role of obesity-associated genetic polymorphisms in regulating gene expression in adipose tissue of metabolically well-characterized nondiabetic individuals may lead to improved understanding of the early pathophysiological basis of obesity.
Based on the above considerations, we hypothesized that percent fat mass (PFAT) would correlate with expression of transcripts in inflammation-immune response and other pathways and expression networks involved in adaptive response to maintain metabolic equilibrium in adipose tissue of healthy nondiabetic subjects. Furthermore, we hypothesized that genes correlated with PFAT would at least partially overlap with genes modulated in response to elevated fatty acid. To test our hypothesis, we used genome-wide transcript profiling in sc adipose tissue from 136 nondiabetic subjects and in in vitro cell culture systems, where cells were challenged by palmitate (C16:0). Finally, to develop causal models of obesity, we sought to define the role of obesity-associated polymorphisms as cis-regulatory elements in modulating the expression of adipose tissue transcripts. We integrated genotype information and gene expression data from nondiabetic subjects to characterize the function of these variants in the pathophysiology of obesity.
Subjects and Methods
Experimental subjects
We conducted our genome-wide gene expression study in sc adipose tissue from 136 European-American or African-American individuals between 19 and 60 yr of age who had a body mass index (BMI) between 19 and 45 kg/m2 and with nondiabetic oral glucose tolerance tests. A summary of our study cohort is provided in Supplemental Table 1 (published on The Endocrine Society's Journals Online web site at http://jcem.endojournals.org). Methods for subject recruitment were described previously (6, 9) and are briefly described in Supplemental Methods. All study participants provided written informed consent under protocols originally approved by the University of Arkansas for Medical Sciences.
Microarray studies
Genome-wide transcriptome analysis was performed by using HumanHT-12 v4 Expression BeadChip (Illumina, San Diego, CA) whole-genome gene expression array according to the vendor-recommended standard protocol (see Supplemental Methods).
SNP selection and genotyping
We selected novel obesity-associated SNP from two large metaanalyses of GWAS on BMI and waist to hip ratio (WHR) of Caucasian subjects published by the Genetic Investigation of Anthropometric Traits consortium (7, 8). In all, we genotyped 41 SNP using the Sequenom MassARRAY system or pyrosequencing (see Supplemental Methods and Supplemental Table 2 for details of primer sequences) (9). Details of other methods are described in Supplemental Methods.
Data analysis
Microarray data analyses
Raw expression intensity was background subtracted and normalized by the average normalization algorithm, as implemented in GenomeStudio Gene Expression Module version 1.0 application software (Illumina). Normalized data were used for further analysis. Gene expression data of all expressed probes (see Supplemental Methods) in adipose tissue were first inverse normal transformed; then a standard (Pearson) or a nonparametric (Spearman) correlation analysis was performed with log-transformed values for PFAT, BMI, and WHR after adjusting for age, gender, and ethnicity.
Bioinformatic analyses
Pathway and interaction network analyses were performed by using Ingenuity Pathway Analysis (IPA version 8.7; https://analysis.ingenuity.com) as described elsewhere (Supplemental Methods) (10). Gene set enrichment analysis was performed by GeneTrail using a list of Entrez ID of genes ranked by Spearman's correlation coefficient between expression and obesity for all expressed probes (Supplemental Methods) (10, 11). Physical locations of genes within ±500 kb of 48 SNP associated with obesity traits (BMI and/or WHR) in large metaanalyses were extracted by using the UCSC table browser query (http://www.genome.ucsc.edu/cgi-bin/hgTables).
cis-expression quantitative trait loci (eQTL) analyses
The association between the genotypes of obesity-associated SNP and genes expressed within ±500 kb in adipose tissue was further tested using linear regression models implemented in PLINK version 1.07 software (12) and included age, gender, and PFAT as covariates (Supplemental Methods).
Results
Adipose gene expression correlates with PFAT in nondiabetic individuals
In our study, transcript level expression of 1,595 of 13,637 expressed genes showed significant correlations (Spearman's ρ ≥ 0.3, and P value <0.0001) with PFAT after adjusting for age, gender, and ethnicity (Supplemental Table 3 and Supplemental Fig. 1). Genes showing strong positive correlations included thioredoxin reductase-1 (ρ = 0.65) and integrin β-5 (ρ = 0.65), whereas genes showing highly significant negative correlations included 1-acylglycerol-3-phosphate O-acyltransferase-9 (ρ = −0.62) and angiogenin (ρ = −0.57). Consistent with our previous study, 31 of 64 genes involved in endoplasmic reticulum stress and unfolded protein response were significantly correlated with PFAT (Supplemental Table 4) (6). A secondary analysis in our European-American subset (n = 99) showed significant correlation (Spearman's ρ ≥ 0.3, and P value <0.0001) of 1112 genes (Supplemental Table 5), whereas in the African-American subset (n = 37), we found correlations of 1948 genes (Spearman's ρ ≥ 0.3, and P value <0.05) (Supplemental Table 6). Among the 1595 genes showing correlation with PFAT in our total cohort, most also showed correlations in European-American and African-American subset analyses (1053 and 1065 genes, respectively; Supplemental Fig. 2).
Enrichment of novel biological pathways among genes correlated with obesity
Consistent with published results, we observed significant enrichment of pathways related to inflammation among genes correlated with obesity (13). Specifically, we identified Fcγ receptor-mediated phagocytosis in macrophages and monocytes as the most enriched canonical pathway [Benjamini and Hochberg (B-H) corrected P value = 0.000076; Supplemental Table 7]. Interestingly, we also identified enrichment of several pathways indicating their novel role in obesity. Genes up-regulated in obesity were highly enriched for pathways related to cell-cell and cell-extracellular matrix interaction (integrin signaling, B-H P value = 0.000076), oxidative stress [nuclear factor erythroid 2-related factor-2 (NRF2)-mediated oxidative stress response, B-H P value = 0.00066], and cell proliferation-oncogenesis. Similarly, genes down-regulated in obesity were enriched for pathways related to branched-chain amino acid metabolism (valine, leucine, and isoleucine degradation, B-H P value = 0.00077), propanoate metabolism (B-H P value = 0.00087), and fatty acid metabolism and biosynthesis. Intriguingly, expression of a set of 178 genes that showed correlation with PFAT was also modulated by excess saturated fatty acid (0.2 mm, palmitate) in our in vitro experiment (Supplemental Results and Supplemental Table 8).
Interaction network analysis indicates a link between thioredoxin and angiogenesis
The oxidative stress response pathway was strongly enriched among genes correlated with PFAT in adipose of nondiabetic subjects and also emerged as the most enriched pathway among genes differentially expressed by palmitate in vitro (Supplemental Table 9). For further analyses, we thus prioritized this pathway among 80 different biological pathways that were enriched among genes correlated with obesity. The strongest correlation was observed for thioredoxin reductase-1 (TXNRD1) of the 36 genes in the oxidative stress response pathway that showed correlation with PFAT. To understand the genes coregulated with TXNRD1, we performed a correlation of TXNRD1 transcript level with all other expressed genes in adipose and found significant correlations (r2 ≥ 0.25) of 409 genes (Supplemental Table 10). Genes positively correlated with TXNRD1 expression included thioredoxin (TXN) and thioredoxin domain-containing proteins (TXNDC9 and TXNDC17), whereas the thioredoxin-interacting protein (TXNIP), an endogenous inhibitor of thioredoxin (14), was negatively correlated. Several other genes with putative roles in the NRF2-mediated antioxidant defense system, including sulfiredoxin 1 (SRXN1) and sulfotransferases SULT1A1 and SULT1A2, were also positively correlated with TXNRD1. The protein level expression of TXNRD1 also corresponded to its transcripts and was significantly higher in individuals with high PFAT (Supplemental Fig. 3). Biological interaction network analysis within genes correlated with TXNRD1 expression revealed a highly significant interaction network that included TXNRD1, TXN, and TXNIP. Interestingly, this network also included angiogenin (ANG) and ribonuclease/angiogenin inhibitor-1 (RNH1) (Supplemental Fig. 4). We found a significant negative correlation of proangiogenic factors like VEGFA (ρ = −0.44) and ANG (ρ = −0.60), and positive correlations of antiangiogenic factors including TNMD (ρ = 0.51) and RNH1 (ρ = 0.42) with PFAT (Table 1). Real-time PCR results validated microarray signals (Supplemental Table 11, A–D, and Supplemental Results).
Table 1.
Genes in the thioredoxin-mediated oxidative stress defense and angiogenesis pathway correlated with obesity
| Symbol | Spearman's correlation (ρ) |
P value |
Gene name | ||
|---|---|---|---|---|---|
| Min | Max | Min | Max | ||
| TXNRD1 | 0.646 | 0.655 | <0.0001 | <0.0001 | Thioredoxin reductase 1 |
| TXN | 0.473 | 0.482 | <0.0001 | <0.0001 | Thioredoxin |
| TXNDC9 | 0.368 | <0.0001 | Thioredoxin domain containing 9 | ||
| TXNDC17 | 0.357 | <0.0001 | Thioredoxin domain containing 17 | ||
| TXNIP | −0.328 | <0.0001 | Thioredoxin interacting protein | ||
| SRXN1 | 0.560 | <0.0001 | Sulfiredoxin 1 | ||
| SULT1A2 | 0.458 | 0.525 | <0.0001 | <0.0001 | Sulfotransferase family, cytosolic, 1A, phenol-preferring, member 2 |
| SULT1A4 | 0.317 | 0.496 | <0.0001 | 0.0002 | Sulfotransferase family, cytosolic, 1A, phenol-preferring, member 4 |
| ANG | −0.603 | −0.573 | <0.0001 | <0.0001 | Angiogenin |
| VEGFA | −0.439 | −0.421 | <0.0001 | <0.0001 | Vascular endothelial growth factor A |
| TNMD | 0.532 | <0.0001 | Tenomodulin | ||
| ANGPT1 | 0.440 | 0.494 | <0.0001 | <0.0001 | Angiopoietin 1 |
| ANGPT2 | 0.312 | 0.362 | <0.0001 | 0.0002 | Angiopoietin 2 |
| RNH1 | 0.340 | 0.416 | <0.0001 | <0.0001 | Ribonuclease/angiogenin inhibitor 1 |
| AMOT | 0.359 | <0.0001 | Angiomotin | ||
| MMP9 | 0.573 | <0.0001 | Matrix metallopeptidase 9 | ||
| MMP28 | 0.357 | 0.463 | <0.0001 | <0.0001 | Matrix metallopeptidase 28 |
| NLRP3 | 0.440 | 0.456 | <0.0001 | <0.0001 | NLR family, pyrin domain containing 3 |
Genes significantly correlated (Spearman's ρ ≥ 0.3, and P value <0.0001) with PFAT after adjusting for age, gender, and ethnicity in all nondiabetic subjects (n = 136) are shown. Correlation and significance of genes represented by multiple probes in the array are shown as minimum (min) and maximum (max) from Spearman's correlation analysis.
SNP associated with obesity acts as cis-eQTL in adipose
Metaanalyses of GWAS confirmed the association of several SNP with obesity (7, 8). We found significant correlations of 55 of 269 expressed genes within ±500 kb of these obesity-associated SNP with PFAT (Supplemental Table 12). We examined the genotypic association of 41 obesity GWAS SNP with the expression of adjacent (±500 kb) transcripts. In European-Americans (n = 99), eight of the 41 SNP were associated with expression of 10 local transcripts (Table 2). The strongest association was observed for SNP rs6861681 in regulating the transcription of CPEB4 (P = 3.02 × 10−9) in adipose tissue under the dominant model of inheritance. The genotype of SNP rs7359397 on chromosome 16 was associated with expression of SULT1A1 (P-dom = 0.001) and SULT1A2 (P-dom = 0.004). None of these associations was validated in our smaller African-American (n = 37) cohort.
Table 2.
Genotypic association of obesity GWAS SNP with the expression of adjacent transcripts in adipose tissue of nondiabetic European-American subjects
| SNP | Chromosome | SNP location (bp)a | Allele (freq.)b | Gene | Gene location (bp)c | Effectd |
P valuee |
|||
|---|---|---|---|---|---|---|---|---|---|---|
| Dom. | Rec. | Add. | General 2Df | |||||||
| rs6861681 | 5 | 173362458 | A (0.30) | CPEB4 | 173351322 | + | 3.02 × 10−9 | 0.001 | 2.89 × 10−7 | 1.80 × 10−11 |
| rs7359397 | 16 | 28885659 | T (0.37) | SULT1A1 | 28620346 | + | 0.001 | 0.26 | 0.02 | 0.004 |
| rs2241423 | 15 | 68086838 | A (0.26) | MAP2K5 | 67967236 | − | 0.004 | 0.18 | 0.06 | 0.011 |
| rs7359397 | 16 | 28885659 | T (0.37) | SULT1A2 | 28605680 | + | 0.004 | 0.07 | 0.008 | 0.008 |
| rs7138803 | 12 | 50247468 | A (0.34) | BCDIN3D | 50233369 | − | 0.007 | 0.08 | 0.02 | 0.014 |
| rs29941 | 19 | 34309532 | A (0.35) | PEPD | 33945327 | − | 0.02 | 0.002 | 0.001 | 0.004 |
| rs925946 | 11 | 27667202 | A (0.29) | LGR4 | 27440921 | − | 0.07 | 0.003 | 0.002 | 0.005 |
| rs7359397 | 16 | 28885659 | T (0.37) | SPNS1 | 28990982 | + | 0.98 | 0.006 | 0.03 | 0.013 |
| rs1443512 | 12 | 54342684 | A (0.24) | TARBP2 | 53897572 | − | 0.66 | 0.006 | 0.011 | 0.009 |
| rs9491696 | 6 | 127452639 | G (0.45) | RSPO3 | 127479116 | + | 0.04 | 0.012 | 0.004 | 0.013 |
Only SNP associated (P ≤ 0.01) with expression of a transcript are shown. General 2Df is a two degrees of freedom joint test of additive and dominance implemented in PLINK (corresponding to the general genotypic model).
Physical location on GRCh37/hg19 assembly.
Frequency (freq.) of minor allele in 99 nondiabetic European-Americans.
Average location of center of all transcripts.
Effect direction of minor allele.
Statistical significance of genotypic association from linear regression analysis under dominant (Dom.), recessive (Rec.), additive (Add.), or general model inheritance of minor allele after adjusting for age, gender, and PFAT in PLINK.
Discussion
Previous genome-wide studies to understand obesity- and adiposity-associated gene expression changes in adipose tissue compared lean and extremely obese subjects or performed correlation analyses in population-based samples. Those studies largely identified the enrichment of genes in inflammation and immune response pathways or macrophage-enriched metabolic networks among obese subjects (4, 5, 13). Our study of metabolically well-characterized nondiabetic subjects confirms the role of inflammation-immune response pathways and also identified the enrichment of several other pathways, including cell-cell/cell-extracellular matrix interaction, oxidative stress response, endoplasmic reticulum stress, and cell proliferation-oncogenesis among genes strongly correlated with PFAT or BMI measured by dual-energy x-ray absorptiometry. Pathway-specific studies published by our laboratory and others have already pointed out the importance of some of these pathways in human obesity (6, 15, 16). In this study, we also observed significant overlap of gene members among those pathways (Supplemental Table 7). Several positively regulated oncogenes (HRAS, MRAS, NRAS, RRAS2, GRB2, and RAC2), MAPK (MAPK3 and MAP2K1), and phosphatidylinositol-3-kinase subunit (PIK3CG) were identified as members of multiple (>20) pathways enriched in obesity. Thus, our study indicates that these pathways are interconnected and influence the function of each other as a part of a large gene expression network.
Intriguingly, a subset of genes correlated with PFAT was also significantly modulated in vitro by palmitate and indicated a role for the NRF2-mediated oxidative stress response pathway in adipose of obese nondiabetic subjects. It supports our hypothesis that alteration in the adipose tissue transcriptome of obese subjects is caused at least partially by elevated fatty acid. In this pathway, expression of TXNRD1 was most strongly correlated with PFAT and showed higher protein level expression in obese subjects. The TXNRD1 is an oxidoreductase and plays a major role in intracellular redox balance by reactive oxygen species scavenging (17). Thioredoxin (the substrate of TXNRD1), thioredoxin domain-containing proteins, sulfiredoxin, and sulfotransferases with a putative role in the reduction of oxidized thiol-containing protein were positively correlated, whereas TXNIP (the endogenous inhibitor of thioredoxin) was negatively correlated with obesity. Integration of genotype data with gene expression additionally showed that the BMI-associated SNP rs7359397 may function as a cis-eQTL for phenol sulfotransferase genes (SULT1A1 and SULT1A2, Table 2) in adipose tissue. Published studies indicate the role of sulfotransferases in cytoprotection against oxidative damage (18). Interestingly, interaction network analysis within genes correlated with TXNRD1 expression indicated a significant network that includes genes involved in thioredoxin-mediated oxidative stress defense mechanisms and angiogenesis. Recent studies also support a role for the thioredoxin system in angiogenesis through an uncharacterized mechanism (19). In concordance with published studies (20, 21), proangiogenic factors were negatively correlated with obesity, and additionally, we showed positive correlation of antiangiogenic (RNH1 and TNMD) transcripts.
In summary, using an integrative biology approach, we present a more comprehensive characterization of obesity-associated changes of adipose tissue transcriptome in nondiabetic subjects. Our study indicates a novel interaction of up-regulated TXN-TXNRD1 system-mediated oxidative stress defense mechanisms and down-regulated angiogenesis pathways in nondiabetic obese subjects, possibly as an adaptive response to surplus free fatty acids. Functional analysis of this interaction network is required to delineate its precise role in the pathophysiology of obesity.
Acknowledgments
We thank the Clinical Research Center staff of University of Arkansas for Medical Sciences for their outstanding support in the physiological studies and assistance with data management. We thank Prof. Siqun Zheng, Director, Genotyping Laboratory, and the technical staff of the Center for Human Genomics, Wake Forest School of Medicine, especially Ms. Shelly Smith and Dr. Ge Li for their extensive support in genotyping and gene expression analysis. We also thank Prof. Thomas DuBose (Chair, Department of Internal Medicine, Wake Forest School of Medicine) for infrastructural and administrative support and Amanda Goode and Karen Klein for critical reading and editing of our manuscript.
This work was supported by Grant R01 DK039311 from the National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases. Clinical studies were supported by General Clinical Research Center Grant M01RR14288 from the National Center for Research Resources, National Institutes of Health, to the University of Arkansas for Medical Sciences.
Steven C. Elbein, M.D., passed away unexpectedly on June 6, 2010.
Disclosure Summary: The authors have nothing to disclose.
Footnotes
- B-H
- Benjamini and Hochberg
- BMI
- body mass index
- eQTL
- expression quantitative trait loci
- GWAS
- genome-wide association studies
- NRF2
- nuclear factor erythroid 2-related factor-2
- PFAT
- percent fat mass
- SNP
- single-nucleotide polymorphism
- TXN
- thioredoxin
- TXNRD1
- thioredoxin reductase-1
- WHR
- waist to hip ratio.
References
- 1. Kelly T, Yang W, Chen CS, Reynolds K, He J. 2008. Global burden of obesity in 2005 and projections to 2030. Int J Obes (Lond) 32:1431–1437 [DOI] [PubMed] [Google Scholar]
- 2. Hummasti S, Hotamisligil GS. 2010. Endoplasmic reticulum stress and inflammation in obesity and diabetes. Circ Res 107:579–591 [DOI] [PubMed] [Google Scholar]
- 3. Guilherme A, Virbasius JV, Puri V, Czech MP. 2008. Adipocyte dysfunctions linking obesity to insulin resistance and type 2 diabetes. Nat Rev Mol Cell Biol 9:367–377 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Keller MP, Attie AD. 2010. Physiological insights gained from gene expression analysis in obesity and diabetes. Annu Rev Nutr 30:341–364 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Kim Y, Park T. 2010. DNA microarrays to define and search for genes associated with obesity. Biotechnol J 5:99–112 [DOI] [PubMed] [Google Scholar]
- 6. Sharma NK, Das SK, Mondal AK, Hackney OG, Chu WS, Kern PA, Rasouli N, Spencer HJ, Yao-Borengasser A, Elbein SC. 2008. Endoplasmic reticulum stress markers are associated with obesity in nondiabetic subjects. J Clin Endocrinol Metab 93:4532–4541 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Speliotes EK, Willer CJ, Berndt SI, Monda KL, Thorleifsson G, Jackson AU, Allen HL, Lindgren CM, Luan J, Mägi R, Randall JC, Vedantam S, Winkler TW, Qi L, Workalemahu T, Heid IM, Steinthorsdottir V, Stringham HM, Weedon MN, Wheeler E, Wood AR, Ferreira T, Weyant RJ, Segrè AV, Estrada K, et al. 2010. Association analyses of 249,796 individuals reveal 18 new loci associated with body mass index. Nat Genet 42:937–948 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Heid IM, Jackson AU, Randall JC, Winkler TW, Qi L, Steinthorsdottir V, Thorleifsson G, Zillikens MC, Speliotes EK, Mägi R, Workalemahu T, White CC, Bouatia-Naji N, Harris TB, Berndt SI, Ingelsson E, Willer CJ, Weedon MN, Luan J, Vedantam S, Esko T, Kilpeläinen TO, Kutalik Z, Li S, Monda KL, et al. 2010. Meta-analysis identifies 13 new loci associated with waist-hip ratio and reveals sexual dimorphism in the genetic basis of fat distribution. Nat Genet 42:949–960 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Sharma NK, Langberg KA, Mondal AK, Elbein SC, Das SK. 2011. Type 2 diabetes (T2D) associated polymorphisms regulate expression of adjacent transcripts in transformed lymphocytes, adipose, and muscle from Caucasian and African-American subjects. J Clin Endocrinol Metab 96:E394–E403 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Das SK, Mondal AK, Elbein SC. 2010. Distinct gene expression profiles characterize cellular responses to palmitate and oleate. J Lipid Res 51:2121–2131 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Keller A, Backes C, Al-Awadhi M, Gerasch A, Küntzer J, Kohlbacher O, Kaufmann M, Lenhof HP. 2008. GeneTrailExpress: a web-based pipeline for the statistical evaluation of microarray experiments. BMC Bioinformatics 9:552. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, Maller J, Sklar P, de Bakker PI, Daly MJ, Sham PC. 2007. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81:559–575 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Emilsson V, Thorleifsson G, Zhang B, Leonardson AS, Zink F, Zhu J, Carlson S, Helgason A, Walters GB, Gunnarsdottir S, Mouy M, Steinthorsdottir V, Eiriksdottir GH, Bjornsdottir G, Reynisdottir I, Gudbjartsson D, Helgadottir A, Jonasdottir A, Jonasdottir A, Styrkarsdottir U, Gretarsdottir S, Magnusson KP, Stefansson H, Fossdal R, Kristjansson K, et al. 2008. Genetics of gene expression and its effect on disease. Nature 452:423–428 [DOI] [PubMed] [Google Scholar]
- 14. Nishiyama A, Matsui M, Iwata S, Hirota K, Masutani H, Nakamura H, Takagi Y, Sono H, Gon Y, Yodoi J. 1999. Identification of thioredoxin-binding protein-2/vitamin D3 up-regulated protein 1 as a negative regulator of thioredoxin function and expression. J Biol Chem 274:21645–21650 [DOI] [PubMed] [Google Scholar]
- 15. Henegar C, Tordjman J, Achard V, Lacasa D, Cremer I, Guerre-Millo M, Poitou C, Basdevant A, Stich V, Viguerie N, Langin D, Bedossa P, Zucker JD, Clement K. 2008. Adipose tissue transcriptomic signature highlights the pathological relevance of extracellular matrix in human obesity. Genome Biol 9:R14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Tinahones FJ, Murri-Pierri M, Garrido-Sanchez L, Garcia-Almeida JM, Garcia-Serrano S, Garcia-Arnes J, Garcia-Fuentes E. 2009. Oxidative stress in severely obese persons is greater in those with insulin resistance. Obesity (Silver Spring) 17:240–246 [DOI] [PubMed] [Google Scholar]
- 17. Arnér ES. 2009. Focus on mammalian thioredoxin reductases–important selenoproteins with versatile functions. Biochim Biophys Acta 1790:495–526 [DOI] [PubMed] [Google Scholar]
- 18. Yeh CT, Yen GC. 2006. Involvement of p38 MAPK and Nrf2 in phenolic acid-induced P-form phenol sulfotransferase expression in human hepatoma HepG2 cells. Carcinogenesis 27:1008–1017 [DOI] [PubMed] [Google Scholar]
- 19. Dunn LL, Buckle AM, Cooke JP, Ng MK. 2010. The emerging role of the thioredoxin system in angiogenesis. Arterioscler Thromb Vasc Biol 30:2089–2098 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Miranda M, Escote X, Ceperuelo-Mallafre V, Megia A, Caubet E, Naf S, Gomez JM, Gonzalez-Clemente JM, Vicente V, Vendrell J. 2010. Relation between human LPIN1, hypoxia and endoplasmic reticulum stress genes in subcutaneous and visceral adipose tissue. Int J Obes (Lond) 34:679–686 [DOI] [PubMed] [Google Scholar]
- 21. Pasarica M, Sereda OR, Redman LM, Albarado DC, Hymel DT, Roan LE, Rood JC, Burk DH, Smith SR. 2009. Reduced adipose tissue oxygenation in human obesity: evidence for rarefaction, macrophage chemotaxis, and inflammation without an angiogenic response. Diabetes 58:718–725 [DOI] [PMC free article] [PubMed] [Google Scholar]