phth

diFa

pitalc

Article history:Accepted 27 May 2013Available online 7 June 2013

Keywords:ObesityLeptin genePolymorphismFBAT

Gene 527 (2013) 7581

Contents lists available at SciVerse ScienceDirect

Gen

j ourna l homepage: www.e1. Introduction

Obesity is dened as an imbalance between food intake and ener-gy expenditure resulting in the storing of energy as fat, primarily inadipose tissue. It is associated with an increased risk for type 2 diabe-tes and cardiovascular diseases (Kopelman, 2000; Poirier et al., 2006).Obesity has become an important public health problem in both de-veloping and developed countries (Flegal et al., 1998; Gallus et al.,2006). The prevalence of overweight has increased from 47.9 to

57.6% in men and from 37.3 to 45.7% in women between 1960 and1994 in the United States (Flegal et al., 1998). In Tunisia, the preva-lence of obesity was estimated to be 6.1% in men and 18.3% inwomen (Kamoun et al., 2008).

Obesity is a multifactorial disorder characterized by an interactionbetween genetic and environmental factors. Twin, adoptee and familystudies demonstrated genetic involvement in the etiology of obesitywith heritability ranging from 30% to 80% (Liu et al., 2005; Sorensenet al., 1998; Stunkard et al., 1986). To date more than 100 candidateAbbreviations: LEP gene, Leptin gene; SNP, single nufamily-based association test; TFBS, transcription factor Correspondence to: M. Fourati, Laboratory of Human

Medicine, University of Sfax, Tunisia. Tel.: +216 74 241 Corresponding author. Tel.: +216 74 241 888; fax:

E-mail addresses: mounafourati@gmail.com (M. Foufaiza.fakhfakh@gmail.com (F. Fakhfakh).

0378-1119/$ see front matter 2013 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.gene.2013.05.064with plasma leptin level as a quantitative trait. It suggested that H1328084 and A19G have an important rolein regulating plasma leptin level.

2013 Elsevier B.V. All rights reserved.Association studyPlasma leptin level

H1328082, rs10487506, H1328081, H1328080, G-2548A and A19G) were evaluated by PCR-RFLP and directsequencing methods. Single SNP association and haplotype association analyses were performed using thefamily-based association test (FBAT). To determine allele frequencies of these SNPs in general population,52 unrelated individuals from the general Tunisian population were also analyzed.Results: Two SNPs showed signicant associations with plasma leptin level (H1328084: A > G, Z = 2.058,p = 0.039; A19G: G > A, Z = 2.058, p = 0.039). When haplotypes were constructed with these two-markers, the risk AA haplotype (frequency 57.1%) was positively associated with plasma leptin level (Z =2.058, p = 0.039). Moreover, SNPs H1328084 and A19G are predicted to modify transcription-factor bindingsites.Conclusions: Our study provided that two functional variants in 5 regulatory region of LEP gene are associateda b s t r a c t

Introduction: Leptin (LEP) gene is one of the most promising candidate genes for obesity. Previous studieshave tested the association of polymorphisms in LEP gene with obesity and obesity-related metabolic bio-markers (anthropometric variables, glucose, insulin level, leptin level and lipid prole). However, the resultsof these studies were still controversial. To determine whether LEP gene is associated with obesity in Tunisianpopulation, we performed a family-based association study between LEP polymorphisms and obesity andobesity-related metabolic biomarkers.Methods: Seven single nucleotide polymorphisms (SNPs) in 5 region of LEP gene were genotyped in threeconsanguineous families including 33 individuals. The previously reported LEP SNPs (H1328084,a r t i c l e i n f oBioinformatics and Signalling Group, Centre of Biotechnology of Sfax, TunisiaAssociation between Leptin gene polymorlevel in three consanguineous families wi

Mouna Fourati a,b,, Mouna Mnif b, Najla Kharrat c, NSalwa Sessi b, Mohamed Abid b, Ahmed Rebai c, Faizaa Laboratory of Human Molecular Genetics, Faculty of Medicine, University of Sfax, Tunisiab Unit of Obesity and Metabolic Syndrome, Department of Endocrinology, Hedi Chaker Hoscleotide polymorphism; FBAT,binding sites.Molecular Genetics, Faculty of888; fax: +216 74 46 14 03.+216 74 46 14 03.rati),

rights reserved.isms and plasma leptinobesity

a Char b, Mahdi Kammoun b, Nourhne Fendri a,khfakh a,

, Sfax, Tunisia

e

l sev ie r .com/ locate /genegenes were identied, each contributes with small effect to the devel-opment of obesity.

LEP gene is a candidate gene of obesity mapped in chromosome7q31.3 and comprises 3 exons spanning about 20 kb (Isse et al.,1995). It encodes a 16-kDa protein called leptin, which is mainly pro-duced by adipose tissue and plays a critical role in the regulation ofbody weight by inhibiting food intake and stimulating energy expen-diture (Campeld et al., 1996). Plasma leptin level is found to be

Table 1Primers and restriction enzymes used for genotyping.

Marker SNP dbSNP rs number Position Primer sequen

H1328080 C/A rs12535747 5 region F: ACCCATGTGR: CTTGAACCA

G-2548A G/A rs7799039 Promoter F: GCTTTCTAAR: GCTCTTTTT

A19G G/A rs2167270 Exon 1 F: CTCTGGAGGR: CGGGATCCA

SNP, single nucleotide polymorphism; F, forward; R, reverse.

76 M. Fourati et al. / Gene 527 (2013) 7581increased in obese individuals and is proportional to body adiposity(Considine et al., 1996).

Several case/control studies have reported that variants G-2548Aand A19G in 5 regulatory region of LEP gene are associated with obe-sity and leptin level (Ben Ali et al., 2009; Cieslak et al., 2011;Constantin et al., 2010; Gaukrodger et al., 2005; Hager et al., 1998;Hinuy et al., 2008; Karvonen et al., 1998; Le Stunff et al., 2000;Lucantoni et al., 2000; Mammes et al., 1998, 2000; Mattevi et al.,2002; Portoles et al., 2006; Ren et al., 2004; Wang et al., 2006;Yiannakouris et al., 2003). Moreover, it has been suggested that veSNPs (H1328084, H1328083, H1328082, H1328081, and H1328080)positioned 2 kb beyond the previously dened promoter regionshowed strong association in single-marker and multiple-markerhaplotype analysis with BMI (Jiang et al., 2004). However, the resultsfrom these studies are still conicting. Hence, additional studies onthe relationships between LEP variants and obesity and obesity-related metabolic biomarkers (anthropometric variables, glucose, in-sulin level, leptin level and lipid prole) are still needed.

In this study, we tested the association of the previously reportedSNPs in the 5 region of LEP gene with obesity and obesity-relatedmetabolic biomarkers in three consanguineous Tunisian families.

2. Materials and methods

2.1. Subjects

Three unrelated consanguineous Tunisian families were recruitedfor the present study at the Department of Endocrinology, CHU HediChaker (Sfax, Tunisia). Obesity was dened as a BMI 30 kg/m2according to the WHO criteria (World Health Organisation, 1995).In addition, 52 unrelated nonobese subjects, considered as controlgroup, were also recruited. Informed consent was obtained from allstudy participants before they entered the study.

All anthropometric measurements including weight, height, waistcircumference (WC) and hip circumference (HC) were taken for eachindividual. Body mass index (BMI) was calculated as weight (kg) di-vided by height squared (m2). Waist-to-hip ratio (WHR) was also cal-culated as WC divided by HC. Type 2 diabetes was dened based onthe criteria of the WHO classication as plasma >7.0 mmol/l orunder antidiabetic treatment. Hypertension was dened as systolicblood pressure (SBP) >130 mm Hg and/or diastolic blood pressure

(DBP) >90 mm Hg or the use of antihypertensive drug treatment.

Table 2Primers used for direct DNA sequencing.

Marker SNP dbSNP rsnumber

Position Primer sequence (53) Product(bp)

H1328084 G/A rs1349419 5region F: GGGTTGGGTAATCTCCCAGTR: TGGCTGGCTATCAGCTAAGAT

206

H1328082 C/A rs12535708 5region 299rs10487506 G/A rs10487506 5region F: CATTAGGAAGCCTCGAATGC

R: ATTGGGCCATCAACTGTTTCH1328081 T/C rs11770727 5region

SNP, single nucleotide polymorphism; F, forward; R, reverse.2.2. Biochemical measurement

After 12 h of fasting, a venous blood sample was collected in tubescontaining EDTA to measure plasma glucose, insulin, leptin, triglycer-ides (TG), total cholesterol (TCHO), plasma creatinine and uric acid.

Insulinwas quantied using the immunoenzymetric assay INS-EASIAkit (Cat. no.: KAP1251; DIAsource Europe S.A.; Nivelles, Belgium).

The homeostatic index of insulin resistance (HOMA IR) was calculat-ed according to the homeostasis model of assessment (Haffner et al.,1997;Matthews et al., 1985) as follows: fasting insulin (U/ml) fastingglucose (mmol/l) / 22.5. A level of HOMA IR >2.5 indicates insulinresistance.

Fasting plasma leptin concentration (expressed as ng/ml) wasmeasured using the DIAsource Leptin EASIA kit (Cat. no.: KAP2281;DIAsource Europe S.A.; Nivelles, Belgium), a solid phase enzyme am-plied sensitivity immunoassay (EASIA) on a microtiter plate. Thisassay uses monoclonal anti-human leptin antibodies.

2.3. Molecular analysis

2.3.1. DNA extractionGenomic DNA was obtained from peripheral blood using phenol

chloroform standard procedures (Lewin and Stewart-Haynes, 1992).

2.3.2. Genotyping and sequencingSeven polymorphisms spanning the LEP gene were genotyped;

three of which (H1328080, LEP G-2548A and LEP A19G) were typedby polymerase chain reaction-restriction fragment length polymor-phism (PCR-RFLP) analysis. The information of primers and PCR-RFLP analysis is given in Table 1. The PCR amplication wasperformed in a 25 l volume containing 5 buffer (PROMEGA),25 mM of dNTP, 10 M of each primer, 1.25 U of Taq DNA polymerase(GO Taq FLEXI), and 100 ng of the genomic DNA. The conditions usedfor PCR amplication were an initial denaturation phase at 94 C for5 min, followed by 35 cycles at 94 C for 30 s, annealing at 5560 C for 3045 s, and extension at 72 C for 30 s, followed by anal extension phase at 72 C for 7 min. A 10 l aliquot of the PCRproduct mixtures was completely digested with 0.5 U of restrictionenzyme overnight. Digestion products were visualized throughethidium bromide staining after electrophoresis in 3% agarose gels.

Direct DNA sequencing was used for genotyping the other four

ce (53) Product (bp) Restriction enzyme Allele (bp)

TTCTTGGCACTGGGAACACACA

300 EcoNI C:222A:78

GCCAAGGCAAACAAGGTGCACTG

500 CFOI G:320/180A: 500

GACATCAAGGAGAGTTGTGTG

386 HPYCH4III G: 386A: 296/90SNPs: H1328084 (rs1349419), H1328082 (rs12535708), rs10487506and H1328081 (rs11770725) and the three previously reported mu-tations Met110Val, Arg105Trp and Asp103Lys. The information ofprimers is given in Table 2. After PCR amplication, PCR productwas treated with 10 U of exonuclease enzyme (FERMENTAS), se-quenced by the ABI prism big dye terminator cycle sequencingready reaction kit (ABI Prism/PEBiosystems), and the products wereresolved on ABI Prism3100-Avant. Blast searches were performedusing the NCBI database.

To determine the minor allele frequencies of SNPs in general Tuni-sian population, 52 unrelated individuals from the general Tunisianpopulation were also analyzed.

pula

subj

3.61.20.91.38.40.00.50.20.40.479.12.6.35.71.3

sentpop

77M. Fourati et al. / Gene 527 (2013) 75812.4. Statistical analysis

Haploview program (Barrett et al., 2005) was used to estimate thepattern of linkage disequilibrium (LD) and haplotype frequencies ofthe analyzed SNPs in the Tunisian families and compare with datapertaining to the CEU (Utah residents with ancestry from northernand western Europe) population (HapMap Public Release #27 phaseII + III, Feb 09). Tests for HardyWeinberg equilibrium (HWE) werealso carried out using the program haploview.

The software package family-based association test (FBAT)v2.0.3 (Horvath et al., 2001) was used to perform the family-basedassociation test to search the association between each individualpolymorphism or haplotype with obesity and obesity-related meta-bolic biomarkers (anthropometric variables, glucose, insulin level,leptin level, and lipid prole) in the three studied families. Analyseswere carried out under the additive model in the bi-allelic mode(with respect to the minor allele). Before analysis, quantitative vari-ables (age, BP, BMI, WC, TCHO, HDL-c, TG, glucose, creatinine, uricacid, insulin level, leptin level) with abnormal distribution are logtransformed.

Table 3Characteristics of study subjects genotyped in the 3 pedigrees and general Tunisian po

Characteristics Family study

Obese subjects(N = 27)

Nonobese(N = 6)

Age (years) 45.44 15.906 18.5 Systolic blood pressure (mm Hg) 13.61 2.207 11.67 Diastolic blood pressure (mm Hg) 8.15 1.261 7.167 Body mass index (kg/m2) 36.428 6.201 24.311 Waist circumference (cm) 120.19 11.796 86.5 Waist-to-hip ratio 0.976 0.066 0.874 Total cholesterol (mmol/l) 4.962 0.940 3.652 HDL-cholesterol (mmol/l) 1.066 0.292 1.06 Triglycerides (mmol/l) 1.892 1.090 1.086 Glucose (mmol/l) 6.646 2.931 4.983 Uric acid (mol/l) 284.901 128.778 257.333 Creatinine (mol/l) 77.21 23.897 66.5 Leptin (g/ml) 20.981 16.281 6.818 Insulin (IU/ml) 22.768 17.111 14.174 HOMA-IR 6.541 4.687 3.198

ND, not determined. Number of individuals in parentheses. Continuous variables are prepa: obese subjects versus nonobese subjects, pb: obese subjects versus general Tunisian2.5. Bioinformatic analysis

Transcription element search system (TESS) was used to searchpredicting transcription factor binding sites (TFBS) in DNA sequences.This program can identify binding sites using TRANSFAC database. Foreach polymorphism, TFBS was determined by comparing the alterna-tive sequence of the two SNP alleles.

3. Results

3.1. Clinical and biochemical characteristics

The members of the three studied families were classied into twogroups, namely: 1) affected individuals (27 obese) and 2) unaffectedindividuals (6 nonobese).

The anthropometric and biochemicalmeasurementswere performedin the members of the three families and 52 individuals of the generalTunisian population. Results are shown in brief in Table 3. A comparisonbetween obese andnonobese subjects revealed a signicant difference inBMI and leptin level. In fact in the studied families, obese subjects hadlarger BMI values than nonobese subjects (36.428 6.201 versus24.311 1.397, pa b 0.001). In addition, plasma leptin levels in obesesubjects were signicantly higher than those in nonobese subjects(20.981 16.281 versus 6.818 6.388, pa = 0.046). BMI and plasmaleptin levels were also signicantly higher in obese subjects from thestudied families than those in general population (36.428 6.201 ver-sus 21.808 1.93, pb b 0.001 for BMI; 20.981 16.281 versus7.017 5.79, pb b 0.001 for leptin levels).

3.2. Mutation screening for LEP gene

Screening for the three previously reported LEP mutations,Met110Val and Arg105Trp and Asp103 Lys, was performed in thethree Tunisian families by direct sequencing and neither of themwas found in our study.

3.3. SNP genotyping and single SNP association of LEP gene with obesityand obesity-related metabolic biomarkers

Seven SNPs in and upstream LEP gene were genotyped in threeconsanguineous Tunisian families (Fig. 1) and in general Tunisianpopulation. Except A19G polymorphism, which is in the rst exon,the tested markers are in the 5 region of the gene (Fig. 2A). Allele fre-quencies of the genotyped SNPs in general Tunisian population were

tion group.

General Tunisian population group(N = 52)

pa pb

ects

19 26.692 4.921 b0.001 b0.00111 11.62 1.331 0.047 b0.00183 7.153 1.127 0.085 0.00197 21.808 1.93 b0.001 b0.00132 78.89 7.032 b0.001 b0.00143 0.812 0.064 0.003 b0.00175 ND 0.003 85 ND 0.964 12 ND 0.089 53 ND 0.182 726 ND 0.622 342 ND 0.39 88 7.017 5.79 0.046 b0.00159 14.707 10.112 0.238 0.017 ND 0.09

ed as means SD and were compared by t-test. p b 0.05 was considered as signicant.ulation group.determined using the haploview program. The rare allele frequenciesfor those SNPs ranged from 0.30 to 0.49 (Table 4). All genotypes of ex-amined SNPs were tested according to the HardyWeinberg equilibri-um and no signicant deviations were found.

To test if these SNPs in LEP gene are associated with obesity andobesity-relatedmetabolic biomarkers (anthropometric variables, glucose,insulin level, leptin level and lipidprole) in the three studied families,weperformed a family-based association test using the FBAT software. Noneof the SNPs testedwere associatedwith obesity.However, a signicant as-sociation with plasma leptin level was observed for H1328084 and A19Gmarkers with the same Z-score and p-value (Z = 2.058, p = 0.039)under the additive model in the bi-allelic mode. FBAT test showed alsothat the A alleles of H1328084 and A19G markers have a risk effect(Z = 2.058, p = 0.039) and were associated with hyperleptinemia(Table 5). Given that the two SNPs presented the same p-value, weperformed linkage disequilibrium (LD) analysis and result showed thatH1328084 and A19G variants reveal high LD between these two SNPs inthe Tunisian families (r2 = 0.77) (Fig. 2B).

3.4. Haplotype association of LEP gene with plasma leptin level

To determine whether any specic haplotype would confer ahigher risk for hyperleptinemia, the haplotype association test

III

6

ml))

78 M. Fourati et al. / Gene 527 (2013) 7581A19G

H1328082H1328084

rs1048750H1328081H1328080G-2548A

Leptin(g/BMI(kg/m2

60.47546.82

G AC AG AT CC AG AG A

FAMILY 1I

II

1

1 2 3

A19G

H1328082H1328084

rs10487506H1328081H1328080G-2548A

Leptin(g/ml)BMI(kg/m2)for family-based studies was performed. The haplotype containingalleles A of H1328084 and A19G of LEP gene (frequency 57.1%) wasfound to be signicantly and positively associated with plasmaleptin level (Z = 2.058, p = 0.039). Indeed, in the three studiedfamilies some obese patients who are carriers of the risk AA haplo-type of H1328084 and A19G polymorphisms presented high leptinlevels (3 members in family 1, 4 members in family 2 and 8members in family 3). Other obese patients were found to havethe AA haplotype but without increasing leptin levels (1 memberin family 1, 4 members in family 2 and 5 members in family 3)(Fig. 1).

II

A19G

H1328082H1328084

rs10487506H1328081H1328080G-2548A

Leptin(g/ml)BMI(kg/m2)

A19G

H1328082H1328084

rs10487506H1328081H1328080G-2548A

Leptin(g/ml)BMI(kg/m2)

G GT TA AG GA A

G AC AG AT CC AG AG A

G AC AG AT CC AG AG A

4.33631.32

A AA A

36.23149.2

51.74148.58

A19G

H1328082H1328084

rs10487506H1328081H1328080G-2548A

Leptin(g/ml)BMI(kg/m2)

I

II

1

7.91735.5

A19G

H1328082H1328084

rs10487506H1328081H1328080G-2548A

Leptin(g/ml)BMI(kg/m2)

G AC AG AT CC AG AG AA19G

H1328082H1328084

rs10487506H1328081H1328080G-2548A

Leptin(g/ml)BMI(kg/m2)

G AC AG AT CC AG AG A

A19G

H1328082H1328084

rs10487506H1328081H1328080G-2548A

Leptin(g/ml)BMI(kg/m2)

G AC AG AT CC AG AG A

1 2 3

6.9631.56

A AA A

21.17643

2.54230.07

4

A AC AG AT TC AG G

A A

G AC AG AT CC AG AG A

19.87633.2

5

C

2445

A

GTCGA

1 2

G AC AG AT CC AG AG A

2.79824.69

15.71130

G GT TA AG GA A

III

Fig. 1. Genotype distribution of LEP gene SNPs in Tunisian families. Families' members whmales. Obese families' members are represented by black boxes. Analyzed SNPs are reportedbelow the genotypes. The risk alleles A of H1328084 and A19G polymorphism are boxed.FAMILY 2

2

1 2 3

21.99435.78

1A AC AG GT TC AG GA A

17.35633.75

G AC CG AT CC CG AG G

4 5 6

3

G AC CG AT CC CG AG G

48.99232.723.5. Bioinformatic analysis of H1328084 and A19G SNPs

H1328084 and A19G SNPs were found to be signicantly associ-ated with leptin level and these two SNPs are in the regulatory re-gion of LEP gene, one in 5 region (H1328084) and the other in 5UTR (A19G). For these reasons, we performed bioinformatic analy-sis using transcription element search system (TESS) to searchwhether sequence variations in H1328084 and A19G SNPs are pre-dicted to modify transcription factor binding sites (TFBS). Wefound that substitution of G allele by A allele of A19G variant in-duces a putative TFBS called C-Myb. Moreover, the modication

G AC AG GT CC AG GG G

51.01142.27

18.75641.88

G AC CG AT CC CG AG A

G AC CG AT CC AG AG A

36.51535.11

I

1 2 3 4

C C

C AG A

G AG A

24.29740.8

G A

T CC A

C CA A

A AG A

10.19331.13

G G

T CC C

C AG A

G AG A

8.33930.74

G A

T CC A

C CA A

A AG G

5.60324.45

G G

C CC C

C CG A

G AG A

3.79530.14

G A

T CC C

G G

G GG A

10.625

G A

T CC C

C AG A

G AG A

3.31822.73

G A

T CC A

G AC AG AT CC AG AG A

G AC AG AT CC AG AG A

FAMILY 3

2

18.02235.32

A AC AG GT TC AG GG A

6

A

.809.09

A

GTAGA

A A

14.59340.03

A A

G GT TA AG GA A

7

A A

8.8631.91

A A

G GT TA AG GA A

8

C A

3.32533.21

A A

G GT TC AG GA A

A A

17.88726.31

A A

G GT TA AG GA A

9 10 11

A A

14.12836.28

A A

G GT TA AG GA A

0.70422.69

14.55330.13

Obese male

Obese femme

Unaffected male

Unaffected femme

Consanguinous marriage

Death

ose DNAs were studied are represented. Circles represent females. Squares representin the map order from 5 to 3 regions of the LEP gene. Leptin level and BMI are indicated

(B)

3

(A)

H1328084

5

H1328082

H1328081H1328080

G-2548A

A19G

79M. Fourati et al. / Gene 527 (2013) 7581rs13

4941

9

rs12

5357

08

rs10

4875

06

rs11

7707

25

rs12

5357

47

rs77

9903

9

rs21

6727

0

Block 1 (4 kb)1

46 35 46 35 35 58

2 3 4 5 6 7of the G allele to A allele of SNP H1328084 results in the loss of NF-1binding site.

4. Discussion

To investigate the role of LEP gene in association with obesity andobesity-related metabolic biomarkers, we performed a family-basedassociation study in Tunisian population. In the present study, we an-alyzed seven SNPs (H1328084, H1328082, rs10487506, H1328081,H1328080, G-2548A and A19G) in LEP gene in three consanguineousTunisian families. The results demonstrated that two SNPs, H1328084

75

75

46

75 60

35

77

58

58

35 35 46 60

Fig. 2. Gene map and linkage disequilibrium (LD) of LEP gene. (A) Gene map of single nucleoindicates coding sequence. (B) Linkage disequilibrium (LD) of LEP gene in the studied familieach box indicate the pair wise r2 as a percentage. The color schema of white-to-black gradtween rs1349419 and rs2167270 (r2 = 0.77) using haploview program.

Table 4Allele frequencies of the genotyped SNPs in general Tunisian population.

Marker dbSNP rsnumber

SNPa Frequencyb Chromosomalposition

SNP type

H1328084 rs1349419 A/G 0.49 127877213 5 regionH1328082 rs12535708 C/A 0.3 127878098 5 region

rs10487506 G/A 0.38 127878155 5 regionH1328081 rs11770725 C/T 0.49 127878267 5 regionH1328080 rs12535747 C/A 0.3 127878335 5 regionG-2548A rs7799039 G/A 0.4 127878783 PromoterA19G rs2167270 G/A 0.34 127881349 Exon 1

SNP, single nucleotide polymorphism.a The second allele is the minor allele.b Allele frequency for the minor allele.(A > G) and A19G (G > A) were associated with high plasma leptinlevel (Table 5). When haplotypes were constructed with H1328084and A19G SNPs, we found that AA haplotype displayed a positive as-sociation with plasma leptin level (Z = 2.058, p = 0.039). Moreover,these two SNPs are in high LD (r2 = 0.77) and are predicted to

tide polymorphisms (SNPs) typed across the LEP gene on chromosome 7. Black shadinges. LD was measured using data from families in the present study. The numbers withinient reects lower to higher LD. Very high linkage-disequilibrium is demonstrated be-

Table 5Results of FBAT analysis between the seven SNPs in LEP gene and plasma leptin as aquantitative trait.

SNP Allele freq Z p values

rs1349419(H1328084)

A: 0.619 2.058 0.039G: 0.381 2.058 0.039

rs12535708(H1328082)

C: 0.551 1.565 0.117A: 0.449 1.565 0.117

rs10487506G: 0.682 1.152 0.249A: 0.318 1.152 0.249

rs11770725(H1328081)

T: 0.656 0.384 0.700C: 0.344 0.384 0.700

rs12535747(H1328080)

C: 0.551 1.565 0.117A: 0.449 1.565 0.117

rs7799039(G-2548A)

G: 0.669 0.717 0.473A: 0.331 0.717 0.473

rs2167270(A19G)

G: 0.412 2.058 0.039A: 0.588 2.058 0.039

FBAT, family-based association test; allele freq, allele frequency.Z-statistic of the low-frequency allele, positive Z-statistics are indicative of a high-riskallele and negative values are indicative of a protective allele.Signicant p values (p b 0.05) are in boldface.

80 M. Fourati et al. / Gene 527 (2013) 7581modify TFBS. Direct sequencing revealed no functional mutationwithin the protein coding region of the LEP gene in our samplesuggesting that mutations are rare in the general population in agree-ment with previous studies (Karvonen et al., 1998; Strobel et al.,1998).

Serum leptin level, especially in obese individuals, exhibits a highdegree of variability. 40% of the variation in leptin levels could beexplained by adiposity (Lonnqvist et al., 1997). It has been suggestedthat genetic and environmental factors may have considerable impor-tance in the regulation of leptin concentrations. For these reasons, sev-eral investigators have focused to study genetic variants in theregulatory region of LEP gene. However, the results from these studiesare still conicting (Ben Ali et al., 2009; Cieslak et al., 2011; Constantinet al., 2010; Gaukrodger et al., 2005; Hager et al., 1998; Hinuy et al.,2008; Karvonen et al., 1998; Le Stunff et al., 2000; Lucantoni et al.,2000; Mammes et al., 1998, 2000; Mattevi et al., 2002; Portoles et al.,2006; Ren et al., 2004; Wang et al., 2006; Yiannakouris et al., 2003).

In previous studies, anAmerican group analyzes the association of 29SNPs spanning 240 kb across the LEP region in 82white families. Amongthese, ve SNPs (H1328084, H1328083, H1328082, H1328081, andH1328080) positioned 2 kb beyond thepreviously dened promoter re-gion, showed strong associationwith obesity in men (Jiang et al., 2004).In a second study, they reported signicant associations between LEPpolymorphisms (including H1328083 and G-2548A) and hypertensionin postmenopausal women only but no relationship was detected be-tween LEP variants and leptin level (Ma et al., 2009). In our study, wedidn't nd signicant association of H1328084, H1328082, H1328081,and H1328080 markers with obesity (p > 0.05). However, a positivesignicant association was obtained for A allele of H1328084 and leptinlevel (p = 0.039) in the three studied consanguineous Tunisianfamilies.

Also, our results show a lack of association of LEPA19G polymorphismwith obesity and are in agreement withmany other reports (Gaukrodgeret al., 2005; Karvonen et al., 1998; Lucantoni et al., 2000). Nevertheless,using FBAT analysis we found that A allele of LEP A19G was associatedwith hyperleptinemia (p = 0.039). This nding is in contrast withother previous studies that reported a lack of association between leptinlevel and LEP A19G polymorphism (Cieslak et al., 2011; Gaukrodger etal., 2005; Lucantoni et al., 2000).

In addition, AA haplotype of H1328084 and A19G markers re-vealed an association with high plasma leptin level (Z = 2.058,p = 0.039). Indeed, in the studied families some obese patientswho are carriers of two risk alleles of these two SNPs have high leptinlevels but others have the two risk alleles without increasing leptinlevels (Fig. 1). Disconcordance between leptin levels and carriers ofrisk alleles may be explained by the fact that obesity is a multifactorialdisease and that other genes can inuence phenotype expression andleptin level but environmental factors are also important to consider.

Our results show also that H1328084 and A19G polymorphisms arein high LD (r2 = 0.77) (Fig. 2B). The effect observed for A19G polymor-phismmay be attributed to strong LD between this SNP and H1328084.

In agreement with the study of Jiang et al., SNPs H1328084 andA19G are predicted to modify transcription factor binding sites(Jiang et al., 2004). For the H1328084 variant, the use of TESS showsthat the modication of A allele to G allele results in the apparitionof the nuclear factor-1 (NF-1) site.

NF-1 is a transcription factor expressed in adipose tissue; it acts asrepressor in the regulation of transcription of GLUT4 (glucose trans-porter type 4 isoform) gene (Seung-Soon et al., 2007). Therefore, car-riers of A allele may have elevated leptin levels (Fig. 1). A19G SNP islocated in the 5UTR of the rst exon of LEP gene. The modicationof the G allele to A allele results in the apparition of the C-Myb site.

C-Myb is a transcription factor belonging to the Myb family oftranscription activators that comprises two other members A andB-Myb. C-Myb plays an important role in controlling differentiation

in hemopoietic precursor cells (Weston, 1998).Recently, the functional role of A19G polymorphism is demon-strated. It was reported that the overexpression of C-Myb could en-hance adipogenic differentiation in mesenchymal stem cells (hMSCs).hMSCs are pluripotent and in addition to adipocytes, they can giverise to several distinct cell lineages, such as osteoblasts, chondrocytes,myocytes, and even neurons under appropriate conditions (YilanChen et al., 2011).

It was reported also that C-Myb regulates genes by epigeneticmechanism: the domain binding DNA (DBD) of C-Myb binds to theN-terminal histone tails of H3 and H3.3, and facilitates histone tailacetylation, which is a prerequisite for the activation of differentia-tion genes in conjunction with CCAAT enhancer-binding proteins(C/EBP) (Mo et al., 2005).

No other functional studies are available in the literature for theother genotyped SNPs.

5. Conclusions

The present study demonstrated a positive association betweentwo functional SNPs in LEP gene, H1328084 and A19G SNPs, withplasma leptin level as a quantitative trait in Tunisian families. Therisk haplotype containing alleles A of A19G and H1328084 of LEPgene was found to be signicantly and positively associated withplasma leptin level. Our results suggested that H1328084 and A19GSNPs have an important role in regulating plasma leptin level.

Conict of interest

The authors declare no conict of interest.

Acknowledgments

The authors would like to thank all the patients and their familiesfor their cooperation and participation. They also wish to expresstheir gratitude to the staff of the Department of Endocrinology, HediChaker Hospital (Sfax, Tunisia) for their help in collecting samples.

References

Barrett, J.C., Fry, B., Maller, J., Daly, M.J., 2005. Haploview: analysis and visualization ofLD and haplotype maps. Bioinformatics 21, 263265.

Ben Ali, S., et al., 2009. Association of G-2548A LEP polymorphism with plasma leptinlevels in Tunisian obese patients. Clin. Biochem. 42, 584588.

Campeld, L.A., Smith, F.J., Burn, P., 1996. The Ob protein (leptin) pathway: a link be-tween adipose tissue mass and central neural networks. Horm. Metab. Res. 12,619632.

Cieslak, J., et al., 2011. Effect of three common SNPs in 5-anking region of LEP andADIPOQ genes on their expression in Polish obese children and adolescents. Mol.Biol. Rep. 11741177.

Considine, R.V., et al., 1996. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N. Engl. J. Med. 334, 292295.

Constantin, A., Costache, G., Sima, A.V., Glavce, C.S., Vladica, M., Popov, D.L., 2010. Lep-tin G-2548A and leptin receptor Q223R gene polymorphisms are not associatedwith obesity in Romanian subjects. Biochem. Biophys. Res. Commun. 391,282286.

Flegal, K.M., Carroll, M.D., Kuczmarski, R.J., Johnson, C.L., 1998. Overweight and obesityin the United States: prevalence and trends, 19601994. Int. J. Obes. Relat. Metab.Disord. 22, 3947.

Gallus, S., et al., 2006. Overweight and obesity in Italian adults 2004, and an overviewof trends since 1983. Eur. J. Clin. Nutr. 60, 11741179.

Gaukrodger, N., et al., 2005. A rare variant of the leptin gene has large effects on bloodpressure and carotid intima-medial thickness: a study of 1428 individuals in 248families. J. Med. Genet. 42, 474478.

Haffner, S.M., Miettinen, H., Stern, M.P., 1997. The homeostasis model in the SanAntonio Heart Study. Diabetes Care 20, 10871092.

Hager, J., et al., 1998. A polymorphism in the 5 untranslated region of the human obgene is associated with low leptin levels. Int. J. Obes. 22, 200205.

Hinuy, H.M., et al., 2008. Leptin G-2548A promoter polymorphism is associated withincreased plasma leptin and BMI in Brazilian women. Arq. Bras. Endocrinol.Metabol. 52, 611616 (366).

Horvath, S., Xu, X., Laird, N., 2001. The family based association test method: strategiesfor studying general genotype phenotype associations. Eur. J. Hum. Genet. 9,

301306.

Isse, N., et al., 1995. Structural organization and chromosomal assignment of thehuman obese gene. J. Biol. Chem. 270, 2772827733.

Jiang, Y., et al., 2004. Common variants in the 5 region of the leptin gene are associatedwith body mass index in men from the National Heart, Lung, and Blood InstituteFamily Heart Study. Am. J. Hum. Genet. 75, 220230.

Kamoun, M., Hajem, S., Imen, S., Achour, N., Slimane, H., 2008. Prevalence of obesityand overweight in Tunisia on 2000. Tunis. Med. 86, 649652.

Karvonen, M.K., et al., 1998. Identication of new sequence variants in the leptin gene.J. Clin. Endocrinol. Metab. 83, 32393242.

Kopelman, P.G., 2000. Obesity as a medical problem. Nature 404, 635643.Le Stunff, C., Le Bihan, C., Schrk, N.J., Bougneres, P., 2000. A common promoter variant

of the leptin gene is associated with changes in the relationship between serumleptin and fat mass in obese girls. Diabetes 49, 21962200.

Lewin, H.A., Stewart-Haynes, J.A., 1992. A simple method for DNA extraction fromleukocytes for use in PCR. Biotechniques 13, 522524.

Liu, Y.J., et al., 2005. Linkage and association analyses of the UCP3 gene with obesityphenotypes in Caucasian families. Physiol. Genomics 22, 197203.

Lonnqvist, F., Wennlund, A., Arner, P., 1997. Relationship between circulating leptinand peripheral fat distribution in obese subjects. Int. J. Obes. 21, 255260.

Lucantoni, R., et al., 2000. The A19G polymorphism in the 5 untranslated region of thehuman obese gene does not affect leptin levels in severely obese patients. J. Clin.Endocrinol. Metab. 85, 35893591.

Ma, D., et al., 2009. Leptin is associated with blood pressure and hypertension inwomen from the National Heart, Lung, and Blood Institute Family Heart Study.Hypertension 53, 473479.

Mammes, O., et al., 1998. Novel polymorphisms in the 5 region of the LEP gene: asso-ciation with leptin levels and response to low-calorie diet in human obesity. Diabe-tes 47, 487489.

Mammes, O., Betoulle, D., Aubert, R., Herbeth, B., Siest, G., Fumeron, F., 2000. Associa-tion of the G-2548A polymorphism in the 5 region of the LEP gene with over-weight. Ann. Hum. Genet. 64, 391394.

Mattevi, V.S., Zembrzuski, V.M., Hutz, M.H., 2002. Association analysis of genes in-volved in the leptin-signaling pathway with obesity in Brazil. Int. J. Obes. Relat.Metab. Disord. 26, 11791185 (358, 15).

Matthews, D.R., Hosker, J.P., Rudenski, A.S., Naylor, B.A., Treacher, D.F., Turner, R.C., 1985.Homeostasis model assessment: insulin resistance and beta-cell function fromfasting plasma glucose and insulin concentrations in man. Diabetologia 28, 412419.

Mo, X., Kowenz-Leutz, E., Laumonnier, Y., Xu, H., Leutz, A., 2005. Histone H3 tail posi-tioning and acetylation by the c-Myb but not the v-Myb DNA-binding SANTdomain. Genes Dev. 19, 24472457.

Poirier, P., et al., 2006. Obesity and cardiovascular disease: pathophysiology, evalua-tion, and effect of weight loss. Arterioscler. Thromb. Vasc. Biol. 26, 968976.

Portoles, O., et al., 2006. Effect of genetic variation in the leptin gene promoter and theleptin receptor gene on obesity risk in a population-based casecontrol study inSpain. Eur. J. Epidemiol. 21, 605612.

Ren, W., Zhang, S.H., Wu, J., Ni, Y.X., 2004. Polymorphism of the leptin gene promoter inpedigrees of type 2 diabetes mellitus in Chongqing, China. Chin. Med. J. 117,558561.

Seung-Soon, Im., Sool-Ki, K., Tae-Hyun, K., Ha-il, K., Ahn, Yong-Ho, 2007. Regulation ofglucose transporter type 4 isoform gene expression in muscle and adipocytes. Life59 (3), 134145.

Sorensen, T.I., Holst, C., Stunkard, A.J., 1998. Adoption study of environmental mo-dications of the genetic inuences on obesity. Int. J. Obes. Relat. Metab. Disord.22, 7381.

Strobel, A., Issad, T., Camoin, L., Ozata, M., Strosberg, A.D., 1998. A leptin missensemutation associated with hypogonadism and morbid obesity. Nat. Genet.1821318215.

Stunkard, A.J., Foch, T.T., Hrubec, Z., 1986. A twin study of human obesity. JAMA 256,5154.

Wang, T.N., et al., 2006. G-2548A polymorphism of the leptin gene is correlated withextreme obesity in Taiwanese aborigines. 354 Obesity 14, 183187 (14).

Weston, K., 1998. Myb proteins in life, death and differentiation. Curr. Opin. Genet. Dev.8, 7681.

World Health Organisation, 1995. Physical status: the use and interpretation anthro-pometry: report of a WHO Expert Committee. WHO Technical Report Series, No.854. WHO, Geneva, pp. 321344.

Yiannakouris, N., Relistas, L., Yannakoulia, M., Mungal, K., Mantzoros, C.S., 2003. The2548G/A polymorphism in the human leptin gene promoter region is associatedwith plasma free leptin levels: interaction with adiposity and gender in healthysubjects. Hormones (Athens) 2, 229236.

Yilan Chen, A., Jin Liu, B., Hong, Xu., Luo, Z., Zhang, C., 2011. Overexpression of the c-Myb but not its leukemogenic mutant DNA-binding domain increased adipogenicdifferentiation in mesenchymal stem cells. Biochem. Biophys. Res. Commun. 407(1), 202206.

81M. Fourati et al. / Gene 527 (2013) 7581

Association between Leptin gene polymorphisms and plasma leptin level in three consanguineous families with obesity1. Introduction2. Materials and methods2.1. Subjects2.2. Biochemical measurement2.3. Molecular analysis2.3.1. DNA extraction2.3.2. Genotyping and sequencing

2.4. Statistical analysis2.5. Bioinformatic analysis

3. Results3.1. Clinical and biochemical characteristics3.2. Mutation screening for LEP gene3.3. SNP genotyping and single SNP association of LEP gene with obesity and obesity-related metabolic biomarkers3.4. Haplotype association of LEP gene with plasma leptin level3.5. Bioinformatic analysis of H1328084 and A19G SNPs

4. Discussion5. ConclusionsConflict of interestAcknowledgmentsReferences