Download pdf - Genome-wide association study identifies African-ancestry specific variants for metabolic syndrome

Molecular Genetics and Metabolism xxx (2015) xxx–xxx

YMGME-05966; No. of pages: 9; 4C:

Contents lists available at ScienceDirect

Molecular Genetics and Metabolism

j ourna l homepage: www.e lsev ie r .com/ locate /ymgme

Genome-wide association study identifies African-ancestry specificvariants for metabolic syndrome

Fasil Tekola-Ayele a,⁎, Ayo P. Doumatey a, Daniel Shriner a, Amy R. Bentley a, Guanjie Chen a, Jie Zhou a,Olufemi Fasanmade b, Thomas Johnson b, Johnnie Oli c, Godfrey Okafor c, Benjami A. Eghan Jr d,Kofi Agyenim-Boateng d, Clement Adebamowo e, Albert Amoah f, Joseph Acheampong d,Adebowale Adeyemo a, Charles N. Rotimi a,⁎a Center for Research on Genomics and Global Health, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD, USAb University of Lagos, Lagos, Nigeriac University of Nigeria Teaching Hospital, Enugu, Nigeriad University of Science and Technology, Department of Medicine, Kumasi, Ghanae Department of Epidemiology and Public Health, School of Medicine, University of Maryland, Baltimore, MD, USAf University of Ghana Medical School, Department of Medicine, Accra, Ghana

Abbreviations: cMetS, continuous metabolic syndromassociation study; MetS, the metabolic syndrome as definEducation Program; NCEP, the National Cholesterol Eddiabetes.⁎ Corresponding authors at: Center for Research on

National Human Genome Research Institute, National InsRoom 4047, 12 South Drive, Bethesda, MD 20892-5635, U

E-mail addresses: [email protected] (F. Tekola-Aye(C.N. Rotimi).

http://dx.doi.org/10.1016/j.ymgme.2015.10.0081096-7192/Published by Elsevier Inc.

Please cite this article as: F. Tekola-Ayele, esyndrome, Mol. Genet. Metab. (2015), http:/

a b s t r a c t
a r t i c l e i n f o
Article history:Received 12 June 2015Received in revised form 21 October 2015Accepted 21 October 2015Available online xxxx

Keywords:Metabolic syndromePleiotropyGenome-wide association studyAfrican ancestry

Themetabolic syndrome (MetS) is a constellation ofmetabolic disorders that increase the risk of developing sev-eral diseases including type 2 diabetes and cardiovascular diseases. Although genome-wide association studies(GWAS) have successfully identified variants associated with individual traits comprising MetS, the geneticbasis and pathophysiologicalmechanisms underlying the clustering of these traits remain unclear.We conductedGWAS ofMetS in 1427 Africans from Ghana and Nigeria followed by replication testing andmeta-analysis in an-other continental African sample from Kenya. Further replication testing was performed in an African Americansample from the Atherosclerosis Risk in Communities (ARIC) study. We found two African-ancestry specific var-iants that were significantly associated with MetS: SNP rs73989312[A] near CA10 that conferred increased risk(P = 3.86 × 10−8, OR = 6.80) and SNP rs77244975[C] in CTNNA3 that conferred protection against MetS(P=1.63 × 10−8, OR= 0.15). Given the exclusive expression of CA10 in the brain, our CA10 finding strengthenspreviously reported link between brain function and MetS. We also identified two variants that are not Africanspecific: rs76822696[A] near RALYL associated with increased MetS risk (P = 7.37 × 10−9, OR = 1.59) andrs7964157[T] near KSR2 associated with reduced MetS risk (P = 4.52 × 10−8, Pmeta = 7.82 × 10−9, OR =0.53). The KSR2 locus displayed pleiotropic associations with triglyceride and measures of blood pressure. RareKSR2mutations have been reported to be associated with early onset obesity and insulin resistance. Finally, wereplicated the LPL and CETP loci previously found to be associatedwithMetS in Europeans. These findings providenovel insights into the genetics ofMetS inAfricans and demonstrate the utility of conducting trans-ethnic diseasegene mapping studies for testing the cosmopolitan significance of GWAS signals of cardio-metabolic traits.

Published by Elsevier Inc.

1. Introduction

The metabolic syndrome (MetS) manifests as a clustering of riskfactors including abdominal obesity, atherogenic dyslipidemia, ele-vated blood pressure, insulin resistance, and prothrombotic and

e score; GWAS, genome-wideed by the National Cholesterolucation Program; T2D, type 2

Genomics and Global Health,titutes of Health, Building 12A,SA.le), [email protected]

t al., Genome-wide associat/dx.doi.org/10.1016/j.ymgme

proinflammatory states [1,2]. Individuals withMetS have at least a five-fold increased risk of developing type 2 diabetes (T2D), a twofold in-creased risk of cardiovascular diseases [3], and increased susceptibilityto several other disorders including fatty liver disease [4], sleep apnea[5], and some forms of cancer [6].

Recent studies suggest that complex networks of metabolic path-ways modulated by interacting genetic and environmental factors un-derlie MetS [7,8]. Early evidence for the potential contribution ofgenetics to MetS was provided in a seminal study of twin pairs with31.6% of monozygotic twin pairs compared to 6.3% of dizygotic pairsdisplaying concordance for clustering of three MetS components (hy-pertension, diabetes, and obesity) [9]. Heritability estimates for MetSvary, with estimates as low as 13% in the Dutch [10], 24% inCaribbean-Hispanic families [10,11], and as high as 48% in Omani

ion study identifies African-ancestry specific variants for metabolic.2015.10.008

http://dx.doi.org/10.1016/j.ymgme.2015.10.008

mailto:[email protected]

mailto:[email protected]

Journal logo


http://www.sciencedirect.com/science/journal/10967192

www.elsevier.com/locate/ymgme


2 F. Tekola-Ayele et al. / Molecular Genetics and Metabolism xxx (2015) xxx–xxx

families [12]. Earlier genome-wide linkage studies reported links be-tween MetS and several chromosomal regions including 1p34.1,10p11.2 and 19q13.4 [13], 1q21–q25 [14], and 1q23–q31 [15]. Candi-date gene and whole exome sequencing studies have also identifiedmutations in patients with rare forms of MetS [16,17]. Meta-analysesof genome-wide association studies (GWAS) in Europeans and Asianshave reported associations betweenMetS and variants in or near severalloci including ZPR1, BUD13, APOA5, LPL, and CETP genes [18–20].

In contrast to the growing success in the identification of variants as-sociated with the individual components of MetS, little progress hasbeen made in the identification of variants underlying the syndromicclustering of the component traits of MetS and variants with pleiotropiceffect that may shed light on dysregulated pathways in MetS [21,22].Furthermore, the prevalence of MetS shows ethnic disparity in individ-uals of African descent. For example, analysis of the US National Healthand Nutrition Survey (NHANES) serial data from the 1999–2000 and2009–2010 surveys revealed modest decline in prevalence of MetS inCaucasians (25.6% to 21.8%) but a slight increase in African-Americans(22.0% to 22.7%) [23,24]. Paradoxically, the high prevalence of hyper-tension and diabetes in African-Americans contrasts with the observedlow prevalence of high triglyceride levels [25]. Low prevalence of hightriglyceride levels is also observed in west Africans, the ancestral popu-lations of African Americans despite dietary and other differencesbetween the two groups. These observed ethnic disparities in theburden of MetS [25] and other cardiometabolic traits [26] persistseven after adjustment for modifiable risk factors, implying the role ofbackground genetic predisposition.

In this study, we performed a GWAS of MetS in continental Africansenrolled from Ghana and Nigeria (AF1), and replication testing andmeta-analysis with another continental African sample from Kenya(AF2) using ~15million directly genotyped and imputed single nucleo-tide polymorphisms (SNPs). Further replicationwas tested in an AfricanAmerican sample from the Atherosclerosis Risk in Communities (ARIC)study. We also performed a GWAS of MetS in a subset of the samples inthe tails of the continuous metabolic syndrome risk scores (cMetS) de-rived from the sum of standardized residuals of MetS component traits.

2. Materials and methods

2.1. Study samples

Individuals included in this study were participants enrolled in theAfrica America Diabetes Mellitus (AADM) study with centers in Ghanaand Nigeria (AF1) and Kenya (AF2) [27]. The AADM study has been on-going for over a decade, andmost of participants included in the presentanalysis were recruited in the year 2008. The study populations, datacollection procedures, and ethical processes have been described in de-tail elsewhere [27,28].

For developing continuous metabolic syndrome scores and testingtheir predictive accuracy, we analyzed all 4820 individuals in the co-horts with non-missing phenotype values (4023 AF1 and 797 AF2).The discovery genome-wide association analysis was done in 1427AF1. Independent replication of genome-wide significant loci wastested in 174 AF2 and 2475 African Americans enrolled in the ARICstudy.

2.2. MetS phenotypes

Based on the definition of the National Cholesterol Education Pro-gram (NCEP) improved threshold, an individual was considered tohave MetS if they have the following measures for three or more ofthe five component traits [1]: waist circumference ≥102 cm for menor ≥88 cm for women; fasting plasma glucose ≥100 mg/dL; plasmatriglyceride levels ≥150 mg/dL; HDL cholesterol b40 mg/dL for menor b50 mg/dL for women; systolic BP ≥ 130 mm Hg or diastolicBP ≥ 85 mm Hg. In our analysis, cases were individuals with MetS and

Please cite this article as: F. Tekola-Ayele, et al., Genome-wide associatsyndrome, Mol. Genet. Metab. (2015), http://dx.doi.org/10.1016/j.ymgme

controls were individuals without MetS. We also developed a continu-ous metabolic syndrome risk scores (cMetS). Previous studies used dif-ferent approaches including Z-scores, principal components, andpercentile rankings to derive cMetS; the scores obtained by usingthese methods displayed strong correlation with one another [29]. Wedeveloped cMetS using the sum of the standardized scores from thecomponents of MetS. Prior to deriving cMetS, normality of each traitwas assessed using descriptive statistics and distribution plots (histo-grams and scatter plots). Plasma glucose and triglyceride values werelog10-transformed because their distributions did not conform to nor-mality. Next, standardized residuals were obtained by regressing eachtrait on age, sex, and broad ethnicity. The sign of the standardized resid-uals for HDL was reversed so that higher values indicate greater risk ofmetabolic syndrome. Finally, all standardized residuals of MetS compo-nent traits were summed to form cMetS. The higher the cMetS, thehigher the tendency for the components to cluster indicating a higherMetS risk. To evaluate the accuracy of cMetS, we developed anothercontinuous metabolic syndrome score based on principal componentsanalysis of MetS component traits. Previous studies have shown thatthe power of detecting pleiotropic effects of genetic variants on corre-lated traits can be enhanced by reducing the dimension of the originaltraits into a single trait using principal components analysis and subse-quent analysis of the top principal component in place of the originaltraits [30,31]. We used the loadings of the first principal component(PC1) that explain most of the total trait variance as PC-based continu-ous metabolic syndrome score (pc1MetS).

2.3. Genotyping, quality control, and imputation

The discovery GWAS was done using 1637 unrelated AF1 in theAADM study genotyped on the Affymetrix Axiom® PANAFR SNP array.The array was chosen because it offers ≥90% genetic coverage of vari-ants with minor allele frequency (MAF) N 2% of the Yoruba (WestAfrican) genome and N85% coverage of the Luhya and Maasai (EastAfrican) genome. DNA sample preparation and quality control weredone in our laboratory at theNational Institutes of Health (NIH).We ex-cluded 210 samples (one duplicated, three sex discordant, and 206withone or more missing MetS component trait values), leaving 1427 AF1for the discovery GWAS. Replication of top signals and meta-analysiswere done using 185 unrelated AF2 genotyped on the AffymetrixAxiom® PANAFR SNP array. We excluded 11 samples (10 sex discor-dant, and one with one or more missing MetS component trait values),leaving 174 AF2 for independent replication and meta-analysis. Addi-tional replication was done using 2612 African Americans in the ARICstudy with ~24 million imputed SNP dosage data accessed throughthe Database of Genotypes and Phenotypes. We excluded 137 sampleswith one or more missing MetS component trait values, leaving 2475ARIC study individuals for replication.

Of the 2,217,748 SNPs in the raw genotype data of the AADMcohort,we excluded 46,562 SNPs that had a minor allele frequency (MAF) ofb1%, 22,509 that were missing in more than 10% of individuals, 10,282that had a Hardy-Weinberg p-value b1 × 10−6, and 61,087 non-autosomal SNPs. The remaining 2,077,552 SNPs that passed qualityfilers were used as the basis for imputation. Imputation was performedusing the 1000 Genomes Consortium phase 1, version 3 cosmopolitanreference using the programs MaCH [32]/MaCH-ADMIX [33]. Theresulting imputed dosage data were filtered for imputed allelic dosagefrequency b 0.01 and r2 b 0.3, yielding ~15 million SNPs for analysis.

We generated principal components (PCs) from multi-dimensionalscaling analysis of a pruned set of uncorrelated genome-wide SNPs(i.e., r2 threshold of 0.2, and a sliding window of 50 SNPs by skipping5 SNPs between consecutive windows) as implemented in the programPLINK. A plot of the first three PCs is shown in Fig. S1. To minimize theeffect of population structure in the GWAS, all analyses were adjustedfor the first three PCs in AF1 and AF2 and the first two PCs in AA.Quantile-quantile plot of the distribution of the test statistic was



Fig. 1. Receiver operating characteristics (ROC) curves of continuousmetabolic risk scorescorresponding to the metabolic syndrome.

3F. Tekola-Ayele et al. / Molecular Genetics and Metabolism xxx (2015) xxx–xxx

consistent with the expected distribution under the null hypothesis(Fig. S2), and the genomic control inflation was 1.014 (AF1) and 1.017(AF2), indicating a lack of population structure. The power of thestudy at different scenarios is shown in Fig. S3.

2.4. Statistical analyses

Statistical significance of differences between groups was testedusing student's t-test for continuous variables and the chi-square testfor discrete variables. The level of significance was set at 0.05.

The ability of the cMetS variable to be discriminative between indi-vidualswith andwithoutMetSwas evaluated using a receiver operatingcharacteristic (ROC) curve. The area under the curve (AUC) of the ROCcurvewas estimated using a nonparametric approach [34] to determinethe probability that cMetS was higher in individuals with MetS thanthose without MetS.

The imputed SNP dosage data were analyzed using a logistic regres-sion model with adjustment for age, sex, and the first three PCs as im-plemented in mach2dat. For association analyses of MetS componenttraits, in addition to the above covariates, adjustment was done forBMI and the analysis was performed using linear regression modelsunder the additive genetic model as implemented in mach2qtl.Genome-wide significance was set as P-value b5 × 10−8.

Meta-analysis of the discovery and replication datasets was doneusing the METAL software program [35]. We combined the evidence ofassociation in two sided p-values using a fixed-effects model. Samplesize of each SNP was used as a weight, and the sign of the beta valuewas used as the direction of association of the effect allele. Evidencefor the between-study heterogeneity was tested using Cochran's Qstatistic and I2, and the P-value of the chi-square was used to declaresignificance [36].

3. Results

3.1. Characteristics of the study participants

Characteristics of the AF1 subjects included in the discovery GWASand the AF2 subjects included in the replication testing is presented inTables S1 & S2. Of the 1427 AF1, 602 were cases (i.e., had MetS) and825 were controls (i.e., did not have MetS). Of the 174 AF2, 118 werecases and 56 were controls. On average, subjects with MetS were~3 years older than those without MetS (Tables S1 & S2); as a result,age was adjusted for in all analyses models. The prevalence of MetSwas 35.6% in the overall sample, and was greater in women than men(43.2% vs. 25.0%; P b 0.001) (Tables 1 & S3).

Table 1Characteristics of subjects included in developing and testing accuracy of the continuous meta

All (n = 4820) Sample AF1 (Ghana and

Mean ± SDAge, year 47.33 ± 15.64 45.73 ± 16.02Waist circumference, cm 88.86 ± 13.07 88.53 ± 12.94Systolic BP, mmHg 134.13 ± 23.65 133.56 ± 23.76Diastolic BP, mmHg 80.05 ± 13.98 79.39 ± 14.27Plasma glucose, mg/dL 127.22 ± 78.62 125.69 ± 75.35HDL cholesterol, mg/dL 46.50 ± 16.92 46.24 ± 16.89Triglyceride, mg/dL 107.43 ± 73.13 94.24 ± 51.18cMetSb −0.028 ± 3.20 −0.06 ± 3.14pcMetSc −0.016 ± 1.00 −0.04 ± 1.00

n (%)MetSd 1716 (35.6) 1318 (32.8)Women 2811 (58.3) 2389 (59.4)

a Based on t-test for differences of means and Pearson's chi-square test statistic for differencKenya).

b cMetS: continuous metabolic syndrome score based on sum of standardized componentsc pcMetS: continuous metabolic syndrome score based on the first principal component load MetS: the metabolic syndrome (present/absent) based on definition of the National Chole


Characteristics of the study participants included in the develop-ment of cMetS and testing of its predictive accuracy are presented inTables 1, S3 & S4. Our assessment of cMetS showed that it is a valid pre-dictor ofMetS as follows: (i) cMetS for individualswithMetSwas signif-icantly higher than for those without MetS (mean difference 3.86, P b

0.001) (Table S5); (ii) cMetS was strongly correlated with another con-tinuous metabolic syndrome score metric derived from the first princi-pal component loadings of MetS component traits (Pearson correlation(r) = 0.92, Fig. S4); (iii) the AUC of cMetS corresponding to MetS was0.85 (95% CI 0.84–0.87) (Figs. 1 & S5); and (iv) cMetS was moderatelycorrelated with individual component traits of MetS ranging from acorrelation coefficient (r) of 0.40 for fasting plasma glucose to 0.62 fordiastolic blood pressure (Table S6).

3.2. Genome-wide significant associations

In the discovery sample, two strongly correlated low-frequency var-iants [rs73989312 (1.64%) and rs73989319 (1.63%)] near CA10 con-ferred a more than six-fold increased risk of developing MetS (P =3.86 × 10−8, OR= 6.80; and P=3.97 × 10−8; OR= 6.85 respectively)(Figs. 2 & S6, Tables 2 & S7). These variants had frequencies of 1.70% and1.64% respectively in the 1000 Genomes west African ancestry samples(YRI— Yoruba in Ibadan, Nigeria and ASW— African ancestry in South-west USA), but were absent in all non-African and east African (LWK—

Luhya inWebuye, Kenya) samples. In general, we observed that variantfrequencies were similar between our samples and those of west

bolic syndrome score.

Nigeria) (n = 4023) Sample AF2 (Kenya) (n = 797) P-valuea

55.43 ± 10.33 b0.00190.55 ± 13.61 b0.001137.00 ± 22.87 b0.00183.37 ± 11.88 b0.001134.97 ± 93.05 0.00847.83 ± 17.04 0.016173.97 ± 117.59 b0.0010.14 ± 3.46 0.0950.09 ± 1.02 0.001

398 (49.9) b0.001422 (52.9) b0.001

es of proportions between AF1 (sample from Ghana and Nigeria) and AF2 (sample from

of the metabolic syndrome.dings of the metabolic syndrome components in principal components analysis.sterol Education Program (NCEP).



Fig. 2. Plots of genome-wide association of single nucleotide polymorphisms with the metabolic syndrome. Panels A and B represent plots from analyses of the full data set from the dis-covery GWAS andmeta-analysis, respectively. Panels C and D represent plots from analyses of the two 33% tails of the data set from the discovery GWAS andmeta-analysis, respectively.Panels C and D represent plots from analyses of the two 25% tails of the data set from the discovery GWAS andmeta-analysis, respectively. Each single nucleotide polymorphism (SNP) isrepresented by a point, with higher points (higher negative log10 P values), showingmore significant association. Points above the bluehorizontal line represent SNPswith a P value of lessthan 10−5 and points above the red horizontal line indicate SNPs with a P value of less than 5 × 10−8. Blue-lettered genes represent loci with genome-wide significant association(P b 5 × 10−8) and those in purple represent loci with suggestive association (5 × 10−8 b P b 9 × 10−8).


African ancestry samples in the 1000 Genomes dataset, validating thegenotype accuracy in our datasets (Table S8).

The meta-analysis of AF1 and AF2 identified six variants withgenome-wide significant associations with MetS: rs77244975 in the in-tron of CTNNA3 (P = 1.63 × 10−8, OR = 0.15, C allele frequency =1.71%) and five strongly correlated (r2 ≥ 0.8) common variants (fre-quencies from 53.61% to 58.54%) near RALYL (lead SNP = rs76822696,P = 7.37 × 10−9, OR = 1.59) (Tables 1 & S9). The CTNNA3rs77244975[C] variant was present at low frequency (b3%) in the1000 Genomes African ancestry samples (YRI, ASW, and LWK) but


was absent in all non-African samples (Table S8). The rs77244975[C]in CTNNA3 (allele frequency 1.82%) was successfully replicated inAfrican Americans (ARIC study) with directionally consistent significantassociation with MetS (P = 0.011, OR = 0.52).

Dichotomous analyses of the extreme tails (⅓— 33.3% and¼— 25%)of the empirical distribution of cMetS identified two genome-wide sig-nificant associations; the first is an intronic SNP (rs7964157) in KSR2(Pdiscovery = 4.52 × 10−8, Pmeta =7.82 × 10−9, OR= 0.53, frequency=29.73%) and the second is a SNP (rs146816516) near MBNL1 (Pmeta =3.51 × 10−8, OR = 0.22, G allele frequency = 3.95%) (Tables 2, S7 &



Table 2SNPs showing genome-wide significant association with the metabolic syndrome.

Variant Chr Position Frequency (%) Nearest gene P-value ORa SE

DiscoveryFull datasetb

rs73989312[A]c 17 50464800 1.64 CA10 3.86E−08 6.80 0.398rs73989319[C]c 17 50477498 1.63 CA10 3.97E−08 6.85 0.401Extreme 33.3% tails of cMetSd

rs7964157[T] 12 117982220 29.73 KSR2 4.52E−08 0.53 0.121

Meta-analysisFull datasetrs77244975[C] 10 69207975 1.71 CTNNA3 1.63E−08 0.15 0.430rs76822696[A] 8 84929136 53.61 RALYL 7.37E−09 1.59 0.080rs16912410[T] 8 84926391 55.42 RALYL 8.00E−09 1.56 0.077rs62526240[G] 8 84927369 56.22 RALYL 8.94E−09 1.58 0.079rs55752635[A] 8 84926274 56.46 RALYL 1.16E−08 1.55 0.077rs188622356[T] 8 84931781 58.54 RALYL 3.28E−08 1.56 0.080Extreme 33.3% tails of cMetSrs7964157[T] 12 117982220 29.73 KSR2 7.82E−09 0.52 0.116Extreme 25% tails of cMetSe

rs146816516[G] 3 152262078 3.95 MBNL1 3.51E−08 0.22 0.363

Chr: chromosome; position: base-pair position in NCBI Build 37 coordinates; OR: odds ratio; SE: standard error.a In the meta-analysis, odds ratio was calculated using the SCHEME STDERR command that weights effect size estimates using the inverse of the corresponding standard errors as

implemented inMETAL.b Full dataset (602 cases from Ghana and Nigeria (AF1 sample), 825 AF1 controls).c This has not replicated because the variant allele is absent in east African populations.d Extreme 33.3% tails of cMetS (402 AF1 cases, 550 AF1 controls).e Extreme 25% tails of cMetS (310 AF1 and 54 cases from Kenya (AF2), 405 AF1 and 32 AF2 controls).


S9). The rs146816516[G] variant was absent in the 1000 Genome's non-African samples (Table S8). We also identified variants with suggestivegenome-wide significant associations (5 × 10−8 b P b 9 × 10−8) includ-ing rs9354671 (P=5.56 × 10−8) and rs2061117 (P=6.74 × 1−8) nearBAI3 and rs149307971 (P=6.66× 10−8) near EDEM1-GRM7 (Table S7).

3.3. MetS-associated variants displaying pleiotropic effects

The GWAS significant MetS variants and neighboring SNPs within±100 kb were tested for association with the individual traits used todefine the MetS. We found several SNPs near or within two genesdisplaying Bonferroni corrected significant pleiotropic associations:1) KSR2 SNPs were associated with both triglycerides and measures ofblood pressure; 2)MBNL1 SNPswere associatedwith BMI andmeasuresof blood pressure.

3.4. Replication of European MetS meta-analyses GWAS loci in African an-cestry individuals

We used the local replication strategy to test whether three loci re-ported to be associated with MetS in Europeans were also associatedwith MetS in our African ancestry samples. The index European associ-ated SNPs along with neighboring (± 100 kb) variants were evaluatedfor replication [18,19]. Correlated variants (r2 ≥ 0.4 in the 1000 Ge-nomes' CEU samples) in two of the three loci were replicated by fiveSNPs: LPL (rs294, P = 8.1 × 10−3; rs4523270, P = 2.2 × 10−2;rs2165558, P = 2.3 × 10−2; rs35237252, P = 3.3 × 10−2) and CETP(rs4783961, P=1.3× 10−2). The BUD13-ZPR1-APOA5 locus did not rep-licate, mainly because the variants showing significant associations inEuropeans were very rare in our African-ancestry population dataset(e.g., rs10790162 and rs2266788 have minor allele frequencies of0.0045 and 0.0028 in our dataset). Directionally consistent and strongerreplication P-values were achieved with meta-analysis of the westAfricans (AF1) and African Americans (ARIC) (LPL: rs294, P =3.5 × 10−3; CETP: rs4783961, P = 3.0 × 10−4). Interestingly, wefound neighboring SNPs uncorrelated with the index SNPs that hadlower P-values: rs306 (LPL, P = 8.16 × 10−4), rs734104 (APOC3, P =1.52 × 10−4), and rs149959798 (NLRC5, P = 1.75 × 10−3)(Table S10). Notably, the rs306[A] allele associated with reduced risk


of MetS was present in African ancestry populations at a frequency ofabout 10% but was absent in non-African 1000 Genomes' samples. Inthe ENCODE and Roadmap projects, these variants have been shownto be enhancer histone marks: rs306 in left ventricle and adipose cells,and rs734104 in liver, adipose and pancreatic cells.

4. Discussion

This first GWAS of MetS in continental Africans identified novelcommon and low-frequency variants that conferred increased risk aswell as protection against MetS. Two of these variants (CA10 andCTNNA3) are African ancestry specific. The identified variants are func-tionally plausible given their regulatory functional predictions andtheir locations in or near genes involved in glucose homeostasis(KSR2), adipocyte differentiation (RALYL), brain function (CA10,EDEM1-GRM7 locus in 3p26–25& BAI3), and cardiac and skeletalmusclecontractions (CTNNA3, MBNL1). KSR2 and MBNL1 variants displayedpleiotropic properties given their association with two or more cardio-metabolic traits. In addition to these novel findings, we replicated theLPL and CETP loci previously found to be associated with MetS inEuropean ancestry individuals providing evidence for the trans-ethnicproperties of these variants.

The rs76822696 A variant (RALYL) associated with increased risk ofMetS is predicted to alter binding motifs of C/EBPβ and the glucocorti-coid receptor that modulate adipocyte differentiation by regulatingthe transcriptional induction of PPARG [37,38]. PPARG is a master regu-lator of adipogenesis and rare variants in PPARG that reduce adipocytedifferentiation have been associated with increased risk of T2D, amajor consequence of MetS [39]. Our finding parallels the hypothesisthat reduced adipocyte differentiation leads to diabetes [40], and furthersuggests that reduced adipocyte differentiation attributed to a novelcommon variant in RALYL increases the risk of the metabolic syndrome.

We found a common variant (rs7964157) in KSR2 associated with a47% reduced risk of developing MetS. KSR2 (Kinase suppressor of Ras2) is a molecular scaffold of the RAS-MEK-ERK pathway that regulatesglucose uptake and glycolysis [41,42], and promotes activation of theAMPK enzyme that regulates cellular energy homeostasis [43]. RareKSR2 loss-of-functionmutations have been linkedwith early onset obe-sity and insulin resistance in humans [44]. These KSR2mutations impair



Table 3Metabolic syndrome-associated loci showing Bonferroni-corrected significant association with individual components of the MetS and related cardiometabolic traitsa.

Gene MetSb GWAS SNP (chr: position) Replication trait Replicated SNP Position Alleles A1c freq P-value Data set

MBNL1 rs146816516 (chr3:152262078) Systolic BP rs76204242 152291018 A,C 0.98 3.74E−05 ABody Mass Index rs116815255 152192393 T,A 0.98 1.29E−05 BBody Mass Index rs145202083 152193084 C,A 0.98 1.32E−05 BBody Mass Index rs150947994 152195940 T,A 0.98 1.25E−05 BBody Mass Index rs114991118 152199365 T,A 0.98 1.16E−05 BBody Mass Index rs184693776 152201330 C,T 0.98 1.07E−05 BBody Mass Index rs114284899 152203823 G,T 0.98 1.32E−05 BBody Mass Index rs192244039 152209252 C,T 0.98 1.17E−05 BBody Mass Index rs183566325 152221527 G,T 0.98 2.66E−05 BBody Mass Index rs6804557 152334938 C,A 0.98 1.44E−05 BBody Mass Index rs6773118 152339917 C,T 0.98 1.05E−05 B

KSR2 rs7964157 (chr12:117982220) Systolic BP rs78683574 117980191 T,A 0.89 1.39E−05 ASystolic BP rs7964157 117982220 C,T 0.70 3.56E−05 ADiastolic BP rs7964157 117982220 C,T 0.70 3.56E−05 ASystolic BP rs1501635 117973298 T,C 0.60 2.70E−05 BSystolic BP rs78683574 117980191 T,A 0.89 1.35E−05 BTriglycerides rs11610896 117973915 A,G 0.64 1.70E−05 BTriglycerides rs7964157 117982220 C,T 0.70 9.92E−06 BTriglycerides rs11610896 117973915 A,G 0.64 1.14E−05 CTriglycerides rs7964157 117982220 C,T 0.70 2.06E−05 C

EDEM1-GRM7 rs149307971 (chr3:5854876) HDL cholesterol rs1377212 5767686 G,C 0.27 7.20E−06 AHDL cholesterol rs1377212 5767686 G,C 0.27 1.54E−05 BHDL cholesterol rs1377212 5767686 G,C 0.27 1.71E−05 CHDL cholesterol rs116357511 5769812 G,T 0.87 2.16E−05 BHDL cholesterol rs116357511 5769812 G,T 0.87 2.11E−05 C

CTNNA3 rs77244975 (chr10:69207975) Waist circumference rs188455459 69273893 C,T 0.98 1.40E−05 B

Position: base-pair position in NCBI Build 37 coordinates.A: Full dataset (602 cases from Ghana and Nigeria (AF1), 825 AF1 controls).B: 33.3% tails of cMetS (402 AF1 cases, 550 AF1 controls).C: 25% tails of cMetS (310 AF1 cases, 405 AF1 controls).cMetS: continuous metabolic syndrome score based on the sum of standardized residuals of the metabolic syndrome component traits.

a SNPswithin 100 kb distance from the variants showing genome-wide significant associationwithMetSwere tested for associationwith components ofMetS in the discovery datasets.Here we reported those showing Bonferroni-corrected (0.05/number of SNPs tested in the respective locus) significant association.

b MetS: the metabolic syndrome based on definition of the National Cholesterol Education Program (NCEP).c A1 freq: frequency of the first of the two alleles listed under the ‘Alleles’ column in the West African discovery dataset.


the RAS-MEK-ERK pathway similar to gain-of-function mutations inDYRK1B that was recently associated with a form of metabolic syn-drome characterized by early onset coronary artery disease, central obe-sity, hypertension, and diabetes [17]. Individuals carrying the KSR2 loss-of-function mutations exhibited childhood hyperphagia, reduced basalmetabolic rate, and severe insulin resistance [44]. Recapitulating thesemetabolic phenotypes observed in humans, mice that are deficient inKsr2 develop profound obesity [43,45], glucose intolerance, and insulinresistance in liver, muscle and adipose tissues [46,47], possibly due toreduced AMPK activation leading to impairment of fatty acid oxidationand enhancement of triglyceride storage [43]. Identification of a novelvariant associated with MetS in our study showed that, in addition tothe previously identified rare KSR2 mutations [44], common variantsin KSR2 contribute to metabolic disorders. Our KSR2 GWAS locusrs7964157 is predicted to alter binding motifs of Maf, a family of tran-scription factors involved in regulating pancreatic β cell function andβ-cell-specific insulin gene transcription (ENCODE/RoadMap project).MafA knockout mice displayed pancreatic islet abnormalities andglucose intolerance, and developed T2D [48].

An important pleiotropic role of the KSR2 locus is shown by ourfind-ings of statistically significant associations of KSR2 variants with MetScomponent traits (systolic, diastolic blood pressure, and triglycerides)(Table 3) as well as published associations with LDL cholesterol andwaist circumference (Tables S13& S14). Previously,we reported linkagesignals at chromosome 12q24, the locus harboring KSR2, in a genome-wide linkage study of T2D in a cohort of 343 affected sibling pairsfrom west Africa [27]. Other family-based studies have also linked the12q23–24 locus to obesity [49–51], T2D [52–54], percent body fat[55], and total cholesterol [56]. Genome-wide studies have found sug-gestive associations (that fell short of genome-wide significance thresh-old) of KSR2 SNPs with visceral adipose tissue [57], LDL cholesterol [58],and total cholesterol [58] (Table S11). A recentmeta-analysis reported a


pleiotropic effect of a nearby locus in 12q24.31 (rs12310367 in ZNF664)for lipid and inflammatory traits (triglycerides, HDL cholesterol, andadiponectin) [22]. Taken together, our finding and those of others pro-vide strong evidence in support of the pleiotropic properties of theKSR2-dependent signaling.

Two Africa-specific GWAS variants we found in the CTNNA3 andMBNL1 genes that modulate cardiac muscle contraction are likely rele-vant to the understanding of the patho-biology of heart diseases. Thealpha-T-catenin protein encoded by CTNNA3 is involved in muscle con-traction and is predominantly expressed in the heart [59]. PublishedGWAS reported suggestive associations between CTNNA3 variants andserum levels of the cardiac function markers alanine aminotransfer-ase/aspartate aminotransferase [60] (Tables S11 & S12). Furthermore,the 10q21 locus harboring CTNNA3 is a common fragile site linked to di-lated cardiomyopathy [61]. Similarly,MBNL1 is highly expressed in car-diac muscle [62]. MBNL1 inhibits pre-mRNA expression of cardiactroponin T (cTNT) that regulates contractile function of cardiac muscle[63,64]. Overexpression of MBNL1 in muscle cells reduces inhibition ofinsulin receptor (IR) protein [63,65] leading to enhanced glucose uptakein skeletal muscle [66]. According to the ENCODE project, our GWAS as-sociated variant rs146816516 [G] inMBNL1 is predicted to alter bindingmotifs of Foxa and STAT transcription factors that regulate glucose ho-meostasis [67] and adipocyte metabolism [68]. SNP rs116264158which is in strong LD (r2 = 0.87) with rs146816516 has beenshown to be a promoter/enhancer histone mark in adipose and skel-etal muscle cells, in a DNAse hypersensitivity site in pancreatic islets,and is predicted to alter bindingmotifs of the glucocorticoid receptorprotein. SNP rs150505636, which is strongly correlated withrs146816516 (r2 = 0.87), is a weak enhancer in pancreatic islets(ENCODE and RoadMap).

Our finding that variants near CA10 (Carbonic anhydraseX) substantially increased the risk (OR = 6.80) of developing MetS




provides evidence in support of the reported pathophysiological linkbetween brain function and metabolic syndrome. CA10 is exclusivelyexpressed in the brain and is considered to have an essential, yet un-known, functional role because it is highly conserved across animal spe-cies [69]. The encoded CA10 protein is catalytically inactive due toabsence of zinc-binding histidine residues.When themissing histidinesare restored in the sequence, the mutated CA10 exhibits high catalyticactivity [70]. Most carbonic anhydrases (CAs) are involved in importantbiosynthetic processes including lipogenesis and gluconeogenesis. Inhi-bition of mitochondrial CA isozymes using sulfonamide analogs such asthe antiepileptic drugs topiramate (TPM) and zonisamide (ZNS) in-duces weight loss [71–75]. FDA approved chronic weight managementdrug Qsymia (phentermine and TPM extended-release) leads to weightloss, and improvements in weight-related comorbidities such as im-proved glycemia, decreased blood pressure, and improved cholesterollevels [76]. The rs73989312 [A] variant that we found to be associatedwith a substantially increased risk (OR = 6.80) of MetS is predicted toalter binding motifs of transcription factor activator protein-2 (AP-2)(ENCODE). The AP-2 transcription factors are involved in the pathogen-esis of metabolic diseases through dysregulation of adipocyte function[77] and regulates nervous system and other developmental processes[78]. The obesity-linked Drosophila homolog TfAP-2 is expressed inoctopaminergic neurons that regulate satiation and themouse homologAP-2β is located in parts of the brain involved in modulating feedingbehavior [79].

Our EDEM1-GRM7 and BAI3 findings provide additional evidence forthe link between neural connectivity in the brain andMetS. GRM7 is themajor excitatory neurotransmitter in the central nervous system, andBAI3 regulates excitatory synapses and is involved in brain angiogene-sis. EDEM1 is directly involved in endoplasmic reticulum-associateddegradation (ERAD) of misfolded proteins. Dysfunctional ERAD medi-ates the link between chronic ER stress and impairment of cellular func-tions that lead to metabolic dysregulations and neurodegenerativediseases [80–82]. The 3p26–25 locus that harbors the EDEM1-GRM7 re-gion has been shown to have pleiotropic effects on obesity, lipids, andblood pressure traits [83]. These SNPs are located 3 kb–17 kb from theless frequent Africa-specific SNP (rs149307971) that showed associa-tion with MetS in our study. In all, our findings open novel researchareas for understanding the genetic basis of the roles of the brain in reg-ulating energy metabolism, body fat content, and glucose metabolismand the link between defects in brain function and metabolic disorders[84].

We acknowledge that our replication sample fromAfrica is small be-cause of the paucity of well phenotyped genome-wide data from conti-nental Africa. Despite this limitation, our study identified novel lowfrequency and common variants in genes with plausible pathophysio-logical role in metabolic and cardiovascular diseases. The functional di-versity of the proteins encoded by the identified genes includingglucose homeostasis, adipocyte differentiation, cardiac andmuscle con-tractions, neuro-endocrine signaling, brain and pancreatic β cell func-tion indicates that the underlying pathology in MetS is a complexnetwork of dysregulated pathways. The identification of African ances-try specific variants influencing MetS is noteworthy for several reasonsincluding providing strong supportive evidence for including multi-ethnic global populations in disease mapping strategies to facilitatediscovery of novel variants influencing complex traits. Finally, weshowed that a continuous measure of metabolic syndrome (cMetS)can be used to identify individuals at the extremes of metabolic dis-ease risk profile in GWAS, and that cMetS is a valid predictor of MetSin continental Africans as previously observed in other ancestralpopulations [85–87].

Funding

This work was supported in part by the Intramural Research Pro-gram of the National Institutes of Health in the Center for Research on


Genomics and Global Health (CRGGH). The CRGGH is supported bythe National Human Genome Research Institute, the National Instituteof Diabetes and Digestive and Kidney Diseases, the Center for Informa-tion Technology, and the Office of the Director at the National Institutesof Health (1ZIAHG200362). Support for participant recruitment and ini-tial genetic studies of the AADM study was provided by NIH grant No.3T37TW00041-03S2 from the Office of Research on Minority Health.The project was also supported in part by the National Center for Re-search Resources. The Atherosclerosis Risk in Communities (ARIC)Study is carried out as a collaborative study supported by NationalHeart, Lung, and Blood Institute contracts (HHSN268201100005C,HHSN268201100006C, HHSN268201100007C, HHSN268201100008C,HHSN268201100009C, HHSN268201100010C, HHSN268201100011C,and HHSN268201100012C). The authors thank the staff and partici-pants of the ARIC study for their important contributions. Funding forARIC Gene Environment Association Studies (GENEVA) was providedby National Human Genome Research Institute grant U01HG004402(E. Boerwinkle).

Disclosures

The authors declare that they have no conflict of interest.

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.ymgme.2015.10.008.

References

[1] S.M. Grundy, H.B. Brewer Jr., J.I. Cleeman, S.C. Smith Jr., C. Lenfant, Definition of met-abolic syndrome: report of the national Heart, Lung, and Blood Institute/AmericanHeart Association conference on scientific issues related to definition, Arterioscler.Thromb. Vasc. Biol. 24 (2004) e13–e18.

[2] National Cholesterol Education Program (NCEP) Expert Panel, Third report of theNational Cholesterol Education Program (NCEP) expert panel on detection, evalua-tion, and treatment of high blood cholesterol in adults (Adult Treatment Panel III)final report, Circulation 106 (2002) 3143–3421.

[3] R.H. Eckel, S.M. Grundy, P.Z. Zimmet, The metabolic syndrome, Lancet 365 (2005)1415–1428.

[4] A. Kotronen, H. Yki-Jarvinen, Fatty liver: a novel component of the metabolicsyndrome, Arterioscler. Thromb. Vasc. Biol. 28 (2008) 27–38.

[5] A.N. Vgontzas, E.O. Bixler, G.P. Chrousos, Sleep apnea is a manifestation of themetabolic syndrome, Sleep Med. Rev. 9 (2005) 211–224.

[6] K. Esposito, P. Chiodini, A. Colao, A. Lenzi, D. Giugliano, Metabolic syndrome and riskof cancer: a systematic review and meta-analysis, Diabetes Care 35 (2012)2402–2411.

[7] T. Joy, R.A. Hegele, Genetics of metabolic syndrome: is there a role for phenomics?Curr. Atheroscler. Rep. 10 (2008) 201–208.

[8] A.J. Lusis, A.D. Attie, K. Reue, Metabolic syndrome: from epidemiology to systemsbiology, Nat. Rev. Genet. 9 (2008) 819–830.

[9] D. Carmelli, L.R. Cardon, R. Fabsitz, Clustering of hypertension, diabetes, and obesityin adult male twins: same genes or same environments? Am. J. Hum. Genet. 55(1994) 566–573.

[10] P. Henneman, Y.S. Aulchenko, R.R. Frants, K.W. van Dijk, B.A. Oostra, C.M. van Duijn,Prevalence and heritability of the metabolic syndrome and its individual compo-nents in a Dutch isolate: the Erasmus Rucphen Family Study, J. Med. Genet. 45(2008) 572–577.

[11] H.F. Lin, B. Boden-Albala, S.H. Juo, N. Park, T. Rundek, R.L. Sacco, Heritabilities of themetabolic syndrome and its components in the Northern Manhattan Family Study,Diabetologia 48 (2005) 2006–2012.

[12] J.C. Lopez-Alvarenga, C. Solis-Herrera, J.W. Kent, D. Jaju, S. Albarwani, S. Al Yahyahee,M.O. Hassan, R. Bayoumi, A.G. Comuzzie, Prevalence and heritability of clusters fordiagnostic components of metabolic syndrome: the Oman family study, Metab.Syndr. Relat. Disord. 6 (2008) 129–135.

[13] R.J. Loos, P.T. Katzmarzyk, D.C. Rao, T. Rice, A.S. Leon, J.S. Skinner, J.H. Wilmore, T.Rankinen, C. Bouchard, Genome-wide linkage scan for the metabolic syndrome inthe HERITAGE family study, J. Clin. Endocrinol. Metab. 88 (2003) 5935–5943.

[14] M.C. Ng, W.Y. So, V.K. Lam, C.S. Cockram, G.I. Bell, N.J. Cox, J.C. Chan, Genome-widescan for metabolic syndrome and related quantitative traits in Hong Kong Chineseand confirmation of a susceptibility locus on chromosome 1q21–q25, Diabetes 53(2004) 2676–2683.

[15] C.D. Langefeld, L.E. Wagenknecht, J.I. Rotter, A.H. Williams, J.E. Hokanson, M.F. Saad,D.W. Bowden, S. Haffner, J.M. Norris, S.S. Rich, B.D.Mitchell, Linkage of themetabolicsyndrome to 1q23–q31 in Hispanic families: the Insulin Resistance AtherosclerosisStudy Family Study, Diabetes 53 (2004) 1170–1174.




http://refhub.elsevier.com/S1096-7192(15)30066-4/rf0005
















































[16] H. Cao, R.A. Hegele, Nuclear lamin A/C R482Q mutation in Canadian kindreds withDunnigan-type familial partial lipodystrophy, Hum. Mol. Genet. 9 (2000) 109–112.

[17] A.R. Keramati, M. Fathzadeh, G.W. Go, R. Singh, M. Choi, S. Faramarzi, S. Mane, M.Kasaei, K. Sarajzadeh-Fard, J. Hwa, K.K. Kidd, M.A. Babaee Bigi, R. Malekzadeh, A.Hosseinian, M. Babaei, R.P. Lifton, A. Mani, A form of the metabolic syndrome asso-ciated with mutations in DYRK1B, N. Engl. J. Med. 370 (2014) 1909–1919.

[18] K. Kristiansson, M. Perola, E. Tikkanen, J. Kettunen, I. Surakka, A.S. Havulinna, A.Stancakova, C. Barnes, E. Widen, E. Kajantie, J.G. Eriksson, J. Viikari, M. Kahonen, T.Lehtimaki, O.T. Raitakari, A.L. Hartikainen, A. Ruokonen, A. Pouta, A. Jula, A.J.Kangas, P. Soininen, M. Ala-Korpela, S. Mannisto, P. Jousilahti, L.L. Bonnycastle,M.R. Jarvelin, J. Kuusisto, F.S. Collins, M. Laakso, M.E. Hurles, A. Palotie, L. Peltonen,S. Ripatti, V. Salomaa, Genome-wide screen for metabolic syndrome susceptibilityLoci reveals strong lipid gene contribution but no evidence for common geneticbasis for clustering of metabolic syndrome traits, Circ. Cardiovasc. Genet. 5 (2012)242–249.

[19] A.T. Kraja, D. Vaidya, J.S. Pankow, M.O. Goodarzi, T.L. Assimes, I.J. Kullo, U. Sovio, R.A.Mathias, Y.V. Sun, N. Franceschini, D. Absher, G. Li, Q. Zhang, M.F. Feitosa, N.L. Glazer,T. Haritunians, A.L. Hartikainen, J.W. Knowles, K.E. North, C. Iribarren, B. Kral, L.Yanek, P.F. O'Reilly, M.I. McCarthy, C. Jaquish, D.J. Couper, A. Chakravarti, B.M.Psaty, L.C. Becker, M.A. Province, E. Boerwinkle, T. Quertermous, L. Palotie, M.R.Jarvelin, D.M. Becker, S.L. Kardia, J.I. Rotter, Y.D. Chen, I.B. Borecki, A bivariategenome-wide approach to metabolic syndrome: STAMPEED consortium, Diabetes60 (2011) 1329–1339.

[20] S.W. Jeong, M. Chung, S.J. Park, S.B. Cho, K.W. Hong, Genome-wide association studyof metabolic syndrome in koreans, Genomics Inform. 12 (2014) 187–194.

[21] D. Welter, J. MacArthur, J. Morales, T. Burdett, P. Hall, H. Junkins, A. Klemm, P. Flicek,T. Manolio, L. Hindorff, H. Parkinson, The NHGRI GWAS Catalog, a curated resourceof SNP-trait associations, Nucleic Acids Res. 42 (2014) D1001–D1006.

[22] A.T. Kraja, D.I. Chasman, K.E. North, A.P. Reiner, L.R. Yanek, T.O. Kilpelainen, J.A.Smith, A. Dehghan, J. Dupuis, A.D. Johnson, M.F. Feitosa, F. Tekola-Ayele, A.Y. Chu,I.M. Nolte, Z. Dastani, A. Morris, S.A. Pendergrass, Y.V. Sun, M.D. Ritchie, A. Vaez,H. Lin, S. Ligthart, L. Marullo, R. Rohde, Y. Shao, M.A. Ziegler, H.K. Im, R.B.Schnabel, T. Jorgensen, M.E. Jorgensen, T. Hansen, O. Pedersen, R.P. Stolk, H.Snieder, A. Hofman, A.G. Uitterlinden, O.H. Franco, M.A. Ikram, J.B. Richards, C.Rotimi, J.G. Wilson, L. Lange, S.K. Ganesh, M. Nalls, L.J. Rasmussen-Torvik, J.S.Pankow, J. Coresh, W. Tang, W.H. Linda Kao, E. Boerwinkle, A.C. Morrison, P.M.Ridker, D.M. Becker, J.I. Rotter, S.L. Kardia, R.J. Loos, M.G. Larson, Y.H. Hsu, M.A.Province, R. Tracy, B.F. Voight, D. Vaidya, C.J. O'Donnell, E.J. Benjamin, B.Z.Alizadeh, I. Prokopenko, J.B. Meigs, I.B. Borecki, Pleiotropic genes for metabolic syn-drome and inflammation, Mol. Genet. Metab. (2014).

[23] B. Falkner, N.D. Cossrow, Prevalence of metabolic syndrome and obesity-associatedhypertension in the racial ethnic minorities of the United States, Curr. Hypertens.Rep. 16 (2014) 449.

[24] H. Beltran-Sanchez, M.O. Harhay, M.M. Harhay, S. McElligott, Prevalence and trendsof metabolic syndrome in the adult U.S. population, J. Am. Coll. Cardiol. 62 (2013)(1999–2010) 697–703.

[25] Y.W. Park, S. Zhu, L. Palaniappan, S. Heshka, M.R. Carnethon, S.B. Heymsfield, Themetabolic syndrome: prevalence and associated risk factor findings in the US popu-lation from the Third National Health and Nutrition Examination Survey, 1988–1994, Arch. Intern. Med. 163 (2003) 427–436.

[26] A.J. Karter, A. Ferrara, J.Y. Liu, H.H. Moffet, L.M. Ackerson, J.V. Selby, Ethnic disparitiesin diabetic complications in an insured population, JAMA 287 (2002) 2519–2527.

[27] C.N. Rotimi, G. Chen, A.A. Adeyemo, P. Furbert-Harris, D. Parish-Gause, J. Zhou, K.Berg, O. Adegoke, A. Amoah, S. Owusu, J. Acheampong, K. Agyenim-Boateng, B.A.Eghan Jr., J. Oli, G. Okafor, E. Ofoegbu, B. Osotimehin, F. Abbiyesuku, T. Johnson, T.Rufus, O. Fasanmade, R. Kittles, H. Daniel, Y. Chen, G. Dunston, F.S. Collins, Agenome-wide search for type 2 diabetes susceptibility genes in West Africans: theAfrica America Diabetes Mellitus (AADM) Study, Diabetes 53 (2004) 838–841.

[28] C.N. Rotimi, G.M. Dunston, K. Berg, O. Akinsete, A. Amoah, S. Owusu, J. Acheampong,K. Boateng, J. Oli, G. Okafor, B. Onyenekwe, B. Osotimehin, F. Abbiyesuku, T. Johnson,O. Fasanmade, P. Furbert-Harris, R. Kittles, M. Vekich, O. Adegoke, G. Bonney, F.Collins, In search of susceptibility genes for type 2 diabetes inwest Africa: the designand results of the first phase of the AADM study, Ann. Epidemiol. 11 (2001) 51–58.

[29] I.S. Okosun, R. Lyn, M. Davis-Smith, M. Eriksen, P. Seale, Validity of a continuousmetabolic risk score as an index for modeling metabolic syndrome in adolescents,Ann. Epidemiol. 20 (2010) 843–851.

[30] L. Klei, D. Luca, B. Devlin, K. Roeder, Pleiotropy and principal components of herita-bility combine to increase power for association analysis, Genet. Epidemiol. 32(2008) 9–19.

[31] H. Aschard, B.J. Vilhjalmsson, N. Greliche, P.E. Morange, D.A. Tregouet, P. Kraft, Max-imizing the power of principal-component analysis of correlated phenotypes ingenome-wide association studies, Am. J. Hum. Genet. 94 (2014) 662–676.

[32] Y. Li, C.J. Willer, J. Ding, P. Scheet, G.R. Abecasis, MaCH: using sequence and genotypedata to estimate haplotypes and unobserved genotypes, Genet. Epidemiol. 34(2010) 816–834.

[33] E.Y. Liu, M. Li, W. Wang, Y. Li, MaCH-admix: genotype imputation for admixed pop-ulations, Genet. Epidemiol. 37 (2013) 25–37.

[34] J.A. Hanley, B.J. McNeil, The meaning and use of the area under a receiver operatingcharacteristic (ROC) curve, Radiology 143 (1982) 29–36.

[35] C.J. Willer, Y. Li, G.R. Abecasis, METAL: fast and efficient meta-analysis ofgenomewide association scans, Bioinformatics 26 (2010) 2190–2191.

[36] J.P. Higgins, S.G. Thompson, J.J. Deeks, D.G. Altman, Measuring inconsistency inmeta-analyses, BMJ 327 (2003) 557–560.

[37] J.K. Hamm, B.H. Park, S.R. Farmer, A role for C/EBPbeta in regulating peroxisomeproliferator-activated receptor gamma activity during adipogenesis in 3 T3-L1preadipocytes, J. Biol. Chem. 276 (2001) 18464–18471.


[38] S.H. Park, H.J. Choi, H. Yang, K.H. Do, J. Kim, D.W. Lee, Y. Moon, Endoplasmic reticu-lum stress-activated C/EBP homologous protein enhances nuclear factor-kappaBsignals via repression of peroxisome proliferator-activated receptor gamma, J BiolChem 285 (2010) 35330–35339.

[39] C.Y. Ou, T.C. Chen, J.V. Lee, J.C. Wang, M.R. Stallcup, Coregulator cell cycle and apo-ptosis regulator 1 (CCAR1) positively regulates adipocyte differentiation throughthe glucocorticoid signaling pathway, J Biol Chem 289 (2014) 17078–17086.

[40] E. Danforth Jr., Failure of adipocyte differentiation causes type II diabetes mellitus?Nat. Genet. 26 (2000) 13.

[41] M.K. Dougherty, D.A. Ritt, M. Zhou, S.I. Specht, D.M. Monson, T.D. Veenstra, D.K.Morrison, KSR2 is a calcineurin substrate that promotes ERK cascade activation inresponse to calcium signals, Mol. Cell 34 (2009) 652–662.

[42] J. Yun, C. Rago, I. Cheong, R. Pagliarini, P. Angenendt, H. Rajagopalan, K. Schmidt, J.K.Willson, S. Markowitz, S. Zhou, L.A. Diaz Jr., V.E. Velculescu, C. Lengauer, K.W.Kinzler, B. Vogelstein, N. Papadopoulos, Glucose deprivation contributes to the de-velopment of KRAS pathway mutations in tumor cells, Science 325 (2009)1555–1559.

[43] D.L. Costanzo-Garvey, P.T. Pfluger, M.K. Dougherty, J.L. Stock, M. Boehm, O. Chaika,M.R. Fernandez, K. Fisher, R.L. Kortum, E.G. Hong, J.Y. Jun, H.J. Ko, A. Schreiner, D.J.Volle, T. Treece, A.L. Swift, M. Winer, D. Chen, M. Wu, L.R. Leon, A.S. Shaw, J.McNeish, J.K. Kim, D.K. Morrison, M.H. Tschop, R.E. Lewis, KSR2 is an essential regu-lator of AMP kinase, energy expenditure, and insulin sensitivity, Cell Metab. 10(2009) 366–378.

[44] L.R. Pearce, N. Atanassova, M.C. Banton, B. Bottomley, A.A. van der Klaauw, J.P.Revelli, A. Hendricks, J.M. Keogh, E. Henning, D. Doree, S. Jeter-Jones, S. Garg, E.G.Bochukova, R. Bounds, S. Ashford, E. Gayton, P.C. Hindmarsh, J.P. Shield, E.Crowne, D. Barford, N.J. Wareham, S. O'Rahilly, M.P. Murphy, D.R. Powell, I.Barroso, I.S. Farooqi, KSR2 mutations are associated with obesity, insulin resistance,and impaired cellular fuel oxidation, Cell 155 (2013) 765–777.

[45] R. Brommage, U. Desai, J.P. Revelli, D.B. Donoviel, G.K. Fontenot, C.M. Dacosta, D.D.Smith, L.L. Kirkpatrick, K.J. Coker, M.S. Donoviel, D.E. Eberhart, K.H. Holt, M.R.Kelly, W.J. Paradee, A.V. Philips, K.A. Platt, A. Suwanichkul, G.M. Hansen, A.T.Sands, B.P. Zambrowicz, D.R. Powell, High-throughput screening of mouse knockoutlines identifies true lean and obese phenotypes, Obesity (Silver Spring) 16 (2008)2362–2367.

[46] M.D. Henry, D.L. Costanzo-Garvey, P.J. Klutho, R.E. Lewis, Obesity-dependent dysreg-ulation of glucose homeostasis in kinase suppressor of ras 2−/−mice, Physiol. Rep.2 (2014).

[47] J.P. Revelli, D. Smith, J. Allen, S. Jeter-Jones, M.K. Shadoan, U. Desai, M. Schneider, I.van Sligtenhorst, L. Kirkpatrick, K.A. Platt, A. Suwanichkul, K. Savelieva, B.Gerhardt, J. Mitchell, J. Syrewicz, B. Zambrowicz, B.D. Hamman, P. Vogel, D.R.Powell, Profound obesity secondary to hyperphagia in mice lacking kinase suppres-sor of ras 2, Obesity (Silver Spring) 19 (2011) 1010–1018.

[48] H. Kaneto, T.A. Matsuoka, S. Kawashima, K. Yamamoto, K. Kato, T. Miyatsuka, N.Katakami, M. Matsuhisa, Role of MafA in pancreatic beta-cells, Adv. Drug Deliv.Rev. 61 (2009) 489–496.

[49] W.D. Li, C. Dong, D. Li, H. Zhao, R.A. Price, An obesity-related locus in chromosomeregion 12q23–24, Diabetes 53 (2004) 812–820.

[50] Y.C. Chagnon, C.Merette, R.H. Bouchard, C. Emond,M.A. Roy,M.Maziade, A genomewidelinkage study of obesity as secondary effect of antipsychotics in multigenerational fami-lies of eastern Quebec affected by psychoses, Mol. Psychiatry 9 (2004) 1067–1074.

[51] S.G. Wilson, G. Adam, M. Langdown, R. Reneland, A. Braun, T. Andrew, G.L.Surdulescu, M. Norberg, F. Dudbridge, P.W. Reed, P.N. Sambrook, P.W. Kleyn, T.D.Spector, Linkage and potential association of obesity-related phenotypes with twogenes on chromosome 12q24 in a female dizygous twin cohort, Eur. J. Hum.Genet. 14 (2006) 340–348.

[52] D.W. Bowden, M. Sale, T.D. Howard, A. Qadri, B.J. Spray, C.B. Rothschild, G. Akots, S.S.Rich, B.I. Freedman, Linkage of genetic markers on human chromosomes 20 and 12to NIDDM in Caucasian sib pairs with a history of diabetic nephropathy, Diabetes 46(1997) 882–886.

[53] J.T. Shaw, P.K. Lovelock, J.B. Kesting, J. Cardinal, D. Duffy, B. Wainwright, D.P.Cameron, Novel susceptibility gene for late-onset NIDDM is localized to humanchromosome 12q, Diabetes 47 (1998) 1793–1796.

[54] C.M. Lindgren, M.M. Mahtani, E. Widen, M.I. McCarthy, M.J. Daly, A. Kirby, M.P.Reeve, L. Kruglyak, A. Parker, J. Meyer, P. Almgren, M. Lehto, T. Kanninen, T.Tuomi, L.C. Groop, E.S. Lander, Genomewide search for type 2 diabetes mellitus sus-ceptibility loci in Finnish families: the Botnia study, Am. J. Hum. Genet. 70 (2002)509–516.

[55] F. Dai, E.D. Keighley, G. Sun, S.R. Indugula, S.T. Roberts, K. Aberg, D. Smelser, J. Tuitele,L. Jin, R. Deka, D.E. Weeks, S.T. McGarvey, Genome-wide scan for adiposity-relatedphenotypes in adults from American Samoa, Int. J. Obes. 31 (2007) 1832–1842.

[56] D. Shmulewitz, S.C. Heath, M.L. Blundell, Z. Han, R. Sharma, J. Salit, S.B. Auerbach, S.Signorini, J.L. Breslow, M. Stoffel, J.M. Friedman, Linkage analysis of quantitativetraits for obesity, diabetes, hypertension, and dyslipidemia on the island of Kosrae,Federated States of Micronesia, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 3502–3509.

[57] C.S. Fox, Y. Liu, C.C. White, M. Feitosa, A.V. Smith, N. Heard-Costa, K. Lohman, A.D.Johnson, M.C. Foster, D.M. Greenawalt, P. Griffin, J. Ding, A.B. Newman, F. Tylavsky,I. Miljkovic, S.B. Kritchevsky, L. Launer, M. Garcia, G. Eiriksdottir, J.J. Carr, V.Gudnason, T.B. Harris, L.A. Cupples, I.B. Borecki, Genome-wide association for ab-dominal subcutaneous and visceral adipose reveals a novel locus for visceral fat inwomen, PLoS Genet. 8 (2012) e1002695.

[58] T. Zemunik, M. Boban, G. Lauc, S. Jankovic, K. Rotim, Z. Vatavuk, G. Bencic, Z. Dogas,V. Boraska, V. Torlak, J. Susac, I. Zobic, D. Rudan, D. Pulanic, D. Modun, I. Mudnic, G.Gunjaca, D. Budimir, C. Hayward, V. Vitart, A.F. Wright, H. Campbell, I. Rudan,Genome-wide association study of biochemical traits in korcula island, CroatiaCroat Med. J. 50 (2009) 23–33.

















































































































































































[59] B. Janssens, S. Goossens, K. Staes, B. Gilbert, J. van Hengel, C. Colpaert, E. Bruyneel, M.Mareel, F. van Roy, alphaT-catenin: a novel tissue-specific beta-catenin-binding pro-tein mediating strong cell-cell adhesion, J. Cell. Sci. 114 (2001) 3177–3188.

[60] A.G. Comuzzie, S.A. Cole, S.L. Laston, V.S. Voruganti, K. Haack, R.A. Gibbs, N.F. Butte,Novel genetic loci identified for the pathophysiology of childhood obesity in the his-panic population, PLoS One 7 (2012), e51954.

[61] G. Vanpoucke, S. Goossens, B. De Craene, B. Gilbert, F. van Roy, G. Berx, GATA-4 andMEF2C transcription factors control the tissue-specific expression of the alphaT-catenin gene CTNNA3, Nucleic Acids Res. 32 (2004) 4155–4165.

[62] M. Fardaei, M.T. Rogers, H.M. Thorpe, K. Larkin, M.G. Hamshere, P.S. Harper, J.D.Brook, Three proteins, MBNL, MBLL and MBXL, co-localize in vivo with nuclearfoci of expanded-repeat transcripts in DM1 and DM2 cells, Hum. Mol. Genet. 11(2002) 805–814.

[63] T.H. Ho, B.N. Charlet, M.G. Poulos, G. Singh, M.S. Swanson, T.A. Cooper, Muscleblindproteins regulate alternative splicing, Embo J. 23 (2004) 3103–3112.

[64] T. Omland, J.A. de Lemos, M.S. Sabatine, C.A. Christophi, M.M. Rice, K.A. Jablonski, S.Tjora, M.J. Domanski, B.J. Gersh, J.L. Rouleau, M.A. Pfeffer, E. Braunwald, A sensitivecardiac troponin T assay in stable coronary artery disease, N. Engl. J. Med. 361(2009) 2538–2547.

[65] S. Paul, W. Dansithong, D. Kim, J. Rossi, N.J. Webster, L. Comai, S. Reddy, Interactionof muscleblind, CUG-BP1 and hnRNP H proteins in DM1-associated aberrant IRsplicing, Embo J. 25 (2006) 4271–4283.

[66] J.W. Ryder, M. Gilbert, J.R. Zierath, Skeletal muscle and insulin sensitivity: patho-physiological alterations, Front. Biosci. 6 (2001) D154–D163.

[67] J.R. Friedman, K.H. Kaestner, The Foxa family of transcription factors in developmentand metabolism, Cell. Mol. Life Sci. 63 (2006) 2317–2328.

[68] A.J. Richard, J.M. Stephens, The role of JAK-STAT signaling in adipose tissue function,Biochim. Biophys. Acta 1842 (2014) 431–439.

[69] A. Aspatwar, M.E. Tolvanen, C. Ortutay, S. Parkkila, Carbonic anhydrase related pro-teins: molecular biology and evolution, Subcell. Biochem. 75 (2014) 135–156.

[70] I. Nishimori, D. Vullo, T. Minakuchi, A. Scozzafava, C. Capasso, C.T. Supuran, Restoringcatalytic activity to the human carbonic anhydrase (CA) related proteins VIII, X andXI affords isoforms with high catalytic efficiency and susceptibility to anion inhibi-tion. Bioorg. Med. Chem. Lett. 23 (2013) 256–260.

[71] C.T. Supuran, Carbonic anhydrases as drug targets–an overview, Curr. Top. Med.Chem. 7 (2007) 825–833.

[72] R.L. Arechederra, A. Waheed, W.S. Sly, C.T. Supuran, S.D. Minteer, Effect of sulfon-amides as carbonic anhydrase VA and VB inhibitors on mitochondrial metabolicenergy conversion, Bioorg. Med. Chem. 21 (2013) 1544–1548.

[73] C.T. Supuran, A. Di Fiore, G. De Simone, Carbonic anhydrase inhibitors as emergingdrugs for the treatment of obesity, Expert. Opin. Emerg. Drugs 13 (2008) 383–392.

[74] A. Scozzafava, C.T. Supuran, F. Carta, Antiobesity carbonic anhydrase inhibitors: aliterature and patent review, Expert. Opin. Ther. Pat. 23 (2013) 725–735.

[75] C.K. Kramer, C.B. Leitao, L.C. Pinto, L.H. Canani, M.J. Azevedo, J.L. Gross, Efficacy andsafety of topiramate on weight loss: a meta-analysis of randomized controlled trials,Obes. Rev. 12 (2011) e338–e347.


[76] D.J. Heal, J. Gosden, S.L. Smith, What is the prognosis for new centrally-acting anti-obesity drugs? Neuropharmacology 63 (2012) 132–146.

[77] T. Fuke, T. Yoshizaki, M. Kondo, K. Morino, T. Obata, S. Ugi, Y. Nishio, S. Maeda, A.Kashiwagi, H. Maegawa, Transcription factor AP-2beta inhibits expression and se-cretion of leptin, an insulin-sensitizing hormone, in 3T3-L1 adipocytes, Int. J. Obes.34 (2010) 670–678.

[78] D. Eckert, S. Buhl, S. Weber, R. Jager, H. Schorle, The AP-2 family of transcriptionfactors, Genome Biol. 6 (2005) 246.

[79] M.J. Williams, P. Goergen, J. Rajendran, G. Zheleznyakova, M.G. Hagglund, E. Perland,S. Bagchi, A. Kalogeropoulou, Z. Khan, R. Fredriksson, H.B. Schioth, Obesity-linkedhomologues TfAP-2 and Twz establish meal frequency in Drosophila melanogaster,PLoS Genet. 10 (2014), e1004499.

[80] T. Otoda, T. Takamura, H. Misu, T. Ota, S. Murata, H. Hayashi, H. Takayama, A.Kikuchi, T. Kanamori, K.R. Shima, F. Lan, T. Takeda, S. Kurita, K. Ishikura, Y. Kita, K.Iwayama, K. Kato, M. Uno, Y. Takeshita, M. Yamamoto, K. Tokuyama, S. Iseki, K.Tanaka, S. Kaneko, Proteasome dysfunction mediates obesity-induced endoplasmicreticulum stress and insulin resistance in the liver, Diabetes 62 (2013) 811–824.

[81] S. Costes, C.J. Huang, T. Gurlo, M. Daval, A.V. Matveyenko, R.A. Rizza, A.E. Butler, P.C.Butler, beta-cell dysfunctional ERAD/ubiquitin/proteasome system in type 2 diabe-tes mediated by islet amyloid polypeptide-induced UCH-L1 deficiency, Diabetes60 (2011) 227–238.

[82] M. Muriach, M. Flores-Bellver, F.J. Romero, J.M. Barcia, Diabetes and the brain: oxida-tive stress, inflammation, and autophagy, Oxidative Med. Cell. Longev. 2014 (2014)102158.

[83] A.T. Kraja, H.A. Lawson, D.K. Arnett, I.B. Borecki, U. Broeckel, L. de las Fuentes, S.C.Hunt, M.A. Province, J. Cheverud, D.C. Rao, Obesity-insulin targeted genes in the3p26–25 region in human studies and LG/J and SM/J mice, Metabolism 61 (2012)1129–1141.

[84] M.W. Schwartz, D. Porte Jr., Diabetes, obesity, and the brain, Science 307 (2005)375–379.

[85] K.Wijndaele, G. Beunen, N. Duvigneaud, L. Matton, W. Duquet, M. Thomis, J. Lefevre,R.M. Philippaerts, A continuous metabolic syndrome risk score: utility for epidemi-ological analyses, Diabetes Care 29 (2006) 2329.

[86] B.E. Klein, R. Klein, K.E. Lee, Components of the metabolic syndrome and risk ofcardiovascular disease and diabetes in Beaver Dam, Diabetes Care 25 (2002)1790–1794.

[87] S. Agarwal, D.R. Jacobs Jr., D. Vaidya, C.T. Sibley, N.W. Jorgensen, J.I. Rotter, Y.D. Chen,Y. Liu, J.S. Andrews, S. Kritchevsky, B. Goodpaster, A. Kanaya, A.B. Newman, E.M.Simonsick, D.M. Herrington, Metabolic syndrome derived from principal componentanalysis and incident cardiovascular events: the Multi Ethnic Study of Atherosclero-sis (MESA) and Health, Aging, and Body Composition (Health ABC), Cardiol. Res.Pract. 2012 (2012) 919425.

























































































