research article identification of susceptibility genes...

Research ArticleIdentification of Susceptibility Genes for PeritonealOvarian and Deep Infiltrating Endometriosis Usinga Pooled Sample-Based Genome-Wide Association Study

Bruno Borghese123 Joumlrg Tost4 Magalie de Surville4 Florence Busato4

Frank Letourneur12 Franccediloise Mondon12 Daniel Vaiman12 and Charles Chapron123

1 Institut Cochin Universite Paris Descartes Sorbonne Paris Cite CNRS (UMR 8104) 75014 Paris France2Inserm U1016 75014 Paris France3Universite Paris Descartes Sorbonne Paris Cite Service de Gynecologie Obstetrique 2 et Medecine de la ReproductionHopital Cochin Hopitaux Universitaires Paris Centre AP-HP 74014 Paris France4Laboratoire Epigenetique et Environnement Centre National de Genotypage Institut de GenomiqueCEA 91000 Evry France

Correspondence should be addressed to Bruno Borghese brunoborghesecchaphpfr

Received 3 September 2014 Revised 17 December 2014 Accepted 15 January 2015

Academic Editor Mariela Bilotas

Copyright copy 2015 Bruno Borghese et alThis is an open access article distributed under theCreative CommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Characterizing genetic contributions to endometriosis might help to shorten the time to diagnosis especially in the most severeforms but represents a challenge Previous genome-wide association studies (GWAS) made no distinction between peritonealendometriosis (SUP) endometrioma (OMA) and deep infiltrating endometriosis (DIE)We therefore conducted a pooled sample-based GWAS and distinguished histologically confirmed endometriosis subtypes We performed an initial discovery step on 10-individual pools (two pools per condition) After quality control filtering aMonte-Carlo simulationwas used to rank the significantSNPs according to the ratio of allele frequencies and the coefficient of variation Then a replication step of individual genotypingwas conducted in an independent cohort of 259 cases and 288 controls Our approachwas very stringent but probablymissed a lot ofinformation due to theMonte-Carlo simulation which likely explained why we did not replicate results from ldquoclassicrdquo GWAS Fourvariants (rs227849 rs4703908 rs2479037 and rs966674) were significantly associated with an increased risk of OMA Rs4703908located close to ZNF366 provided a higher risk of OMA (OR= 222 95CI 126ndash392) andDIE especially with bowel involvement(OR = 209 95 CI 112ndash391) ZNF366 involved in estrogen metabolism and progression of breast cancer is a new biologicallyplausible candidate for endometriosis

1 Introduction

Endometriosis is an inflammatory estrogen-driven condi-tion defined as misplaced endometrium outside of theuterine cavity and causing chronic pelvic pain and infertility[1 2] Endometriosis is a major womenrsquos health concernthat dramatically impairs the quality of life With a preva-lence reaching up to 10 of women of reproductive ageendometriosis has a strong socioeconomic impact [3] Itmakes sense to consider endometriosis as a public health pri-ority and to assign substantial human and financial resourcesto improve the management of patients [4]

Endometriosis is held as a complex heritable trait withadditive genetic effects accounting for about one half of thevariance [5] In recent years two genome-wide associationstudies (GWAS) have been conducted in individuals fromJapan Australia and UK The first Japanese study reporteda significant association of endometriosis with rs10965235located in CDKN2BAS in 9p21 and with rs16826658m inthe linkage disequilibrium block including WNT4 in 1p36[6] CDKN2BAS regulates P16 a tumor suppressor genesrepressed in endometriosis [7] possibly by promoter hyper-methylation [8] WNT4 has a role in the development of thegenitourinary system steroidogenesis and folliculogenesis

Hindawi Publishing CorporationBioMed Research InternationalVolume 2015 Article ID 461024 8 pageshttpdxdoiorg1011552015461024

2 BioMed Research International

[9 10] The second UKAustralian GWAS found an asso-ciation of endometriosis with rs12700667 located in 7p152in an intergenic region upstream of NFE2L3 HOXA10 andHOXA11 [11]The role ofHOX genes in endometriosis-relatedinfertility has been largely debated [7] After pooling thedata from the two studies rs12700667 remained significantlyassociated with endometriosis and also with stages IIIIV[12] Consistent with these associations the latest GWASimplicated a 150 kb region around WNT4 that also includeLINC00339 and CDC42 [13] An independent set of 305laparoscopically proven endometriosis patients and 2710controls confirmedWNT4 CDKN2BAS and FN1 as the firstidentified common loci for endometriosis [14]

With odds ratios below 15 these associations arenonetheless not strong enough to suggest a causal relationor to consider a potential clinical application [15] This couldbe due to different genetic backgrounds across populationsrs10965235 for instance is monomorphic in Caucasians Mostlikely this reflects the heterogeneity of endometriosis [16]Indeed endometriosis has inconstant symptoms spanningfrom no symptoms to severe chronic pelvic pain and infer-tility [1 2] Three distinct clinical forms of endometriosishave been identified (i) Peritoneal superficial endometrio-sis (SUP) consists of lesions lying on the surface of theperitoneum or the ovaries (ii) Endometrioma (OMA) isan ovarian endometriotic ldquochocolaterdquo cyst (iii) Deeply infil-trating endometriosis (DIE) comprises lesions infiltratingthe muscularis propria of structures surrounding the uterus(vagina bladder bowel or ureters) [17] The phenotypicdistinction between SUP OMA and DIE may also be raisedin terms of their sex hormone responsiveness SUP OMAand DIE have different responsiveness to progesterone [18]differential gene expressions [19 20] and specific putativevariants that may influence one form and not the others [21ndash23]

Therefore we initiated a pooled sample-based GWASwith a distinction between SUP OMA and DIE As thiswas not a classic GWAS approach we conducted an initialdiscovery step on 10-individual pools in biological duplicatesusing the Affymetrix GenChip 250K Nsp chip After qualitycontrol filtering and SNPs ranking based on a false discoveryrate (FDR) below 5 we performed a replication step ofindividual genotyping to validate the top-ranked SNPs in anindependent cohort of 259 endometriosis and 288 controlsEndometriosis was confirmed histologically in cases andinvalidated surgically in controls

2 Materials and Methods

21 Study Population The data were obtained from 627unrelated women of Caucasian origin treated in our tertiaryreferral center for endometriosis and gynecologic conditionsThe 319 endometriotic patients underwent a complete sur-gical resection of all visible lesions allowing confirmationof the diagnosis by expert pathologists in all the cases Weclassified the patients into SUP OMA and DIE consideringthe most severe lesion DIE was histologically defined whenendometriotic lesions involved the muscularis of the vagina

the bladder the intestine or the ureter [17] As these threetypes of lesionswere frequently concomitant [24 25] patientswere arbitrarily classified as per the worst findings [26]Hence a patient presenting with SUP lesions associated withOMA and DIE was classified as DIE As well a patientwith OMA and concomitant SUP lesions was classified asOMAwhereas a patient classified as SUP only presented SUPlesions The 308 controls (CTR) consisted of women withoutany lesion suggestive of endometriosis as checked during acomprehensive surgical exploration Indications for surgeryin these patients were infertility work-up symptomatic uter-ine fibroids benign ovarian cysts or chronic pelvic pain

The discovery group was composed of 60 cases (20 SUP20 OMA and 20 DIE) and 20 CTR randomly selected fromthe entire cohort for DNA-pooling experiments The DIEgroup included the most severe patients that is those pre-senting at least one lesion infiltrating the bowel and associatedbilateral OMAs more than 3 cm This was done intentionallyin order to have a homogeneous group since DIE mightencompass a large variety of lesions The replication cohortwas composed of the remaining 259 cases (42 SUP 121 OMAand 96 DIE) and 288 CTR The ethical review board of ourcenter (Comite Consultatif de Protection des Personnes dansla Recherche Biomedicale de Paris-Cochin) approved thestudy design All subjects provided written informed consentbefore entering the study

22 DNA Extraction and Quantification Pool Constructionand SNP Allelotyping Five millimeters of venous blood werecollected from individuals into EDTA tubes Genomic DNAwas subsequently extracted with the MagNa Pure CompactNucleic Acid Isolation Kit (Roche Applied Science Indi-anapolis IN USA) DNA was quantified using a spectropho-tometer (NanoDrop 1000 Thermo Scientific WilmingtonDE USA) and diluted to a concentration of 50 ng120583L After-wards each DNA was quantified using fluorimetry (Quant-It DNA Assay Kit Invitrogen Carlsbad CA USA) andchecked for quality using 1 agarose gel electrophoresis Eachpool was composed of 10 samples of 100 ng DNA aliquotsfrom the same category (SUP OMA DIE and CTR) In allwe built eight 10-patient DNA pools two per category inbiological replicates (2 SUP 2 OMA 2 DIE and 2 CTR)DNA quantification and quality were checked several timesafter the pooling procedure to make sure each pool had thesame amount of DNA Genotyping analysis with GenChipHuman Mapping 250K Nsp Array Set (Affymetrix SantaClara CA USA) was performed for each pool followingthe manufacturerrsquos guidelines Briefly genomic DNAs wererestricted with NspI NspI adaptors were then ligated torestricted fragments and subjected to PCRusing the universalprimer PCR002 provided by the kit PCR fragmentswere thenpurified and 90 120583g used for fragmentation and end-labelingwith biotin using Terminal Transferase Labeled targets werethen hybridized overnight to Genechip human 250K NspIarray (Affymetrix) at 49∘C Chips were washed on the fluidicstation FS450 following specific protocols (Affymetrix) andscanned using theGCS3000 7GThe imagewas then analyzedwith the GCOS software to obtain raw data

BioMed Research International 3

Quality controls were performed usingAffymetric GTypesoftware and the MPAM algorithm (Modified PartitioningAround Medoids) All samples had a call rate gt94 and adetection rate gt99

23 Analysis of Microarrays Raw Data Raw data presentingas fluorescence intensities for each 25-base perfect matchprobe (PM) and mismatch probe (MM) were reported andanalyzed using Excel 2008 (Microsoft Redmond WA) For agiven SNP we computed the amount of fluorescence (119891) ofeach allele (119860 and 119861) as 119891 = 119875119872 minus 119872119872 Allele frequencyfor allele 119860 was estimated by the formula 119865

119860= 119891119860(119891119860+ 119891119861)

and reciprocally for allele 119861 119865119861= 119891119861(119891119860+ 119891119861) Then we

calculated the ratio of allele frequencies (119877) between the cases(SUP OMA or DIE) and the controls (CTR) For a givencategory since we had 2 pools per category (SUP

1and SUP

2

OMA1and OMA

2 DIE1and DIE

2 and CTR

1and CTR

2) we

obtained 4 ratios of allele frequencies As instance for DIE1198771= 119865DIE1119865CTR1 1198772 = 119865DIE1119865CTR2 1198773 = 119865DIE2119865CTR1 and 1198774

= 119865DIE2119865CTR2 Finally we calculated the mean of the 4 ratiosof allele frequencies and the coefficient of variation (definedas the ratio of the standard deviation to the mean)

24 Genotypes Calling SNPs were sorted according to themean of the four ratios of allele frequencies in each categoryWe selected the SNPs where the mean of the ratios of allelefrequencies was gt20 Then for each selected SNP on a givenchromosome we submitted the mean of the four ratiosof allele frequencies and the coefficient of variation to aMonte-Carlo simulation [27] Briefly we simulated artificialchromosomes with a number of SNPs identical to that of thechip For each SNP we produced random allele frequenciesfor cases and controls (two of each) These frequencies wereused to generate amean of the four ratios of allele frequenciesand a coefficient of variation that were compared to theactual values of each real SNP This procedure was repeatedthousand times enabling to define a probability of randomoccurrence of a given pair (mean and coefficient of variation)When this 119875 value was lt005 corresponding to a FDR of5 the SNP was considered as significantly biased betweencases and controls For instance for chromosome 1 fourfrequencies corresponding to 19865 SNPs were randomlysimulated and used to generate ratios means and coefficientsof variation along the lines of what was done with the realdataset Then for a real given SNP we performed a test tocheck if its values (mean and coefficient of variation) couldoccur and howmany times in the artificial database followingits random generation This Monte-Carlo procedure allowedus to eliminate the SNPs whose coefficient of variation wastoo high and to select the SNPs with a FDR 119875 value of 005After this procedure we obtained a list of 280 SNPs

25 Individual Genotyping To validate high-ranking SNPsfrom the GWAS we made another selection by using dif-ferent criteria (1) We first selected SNPs from gene regionscontaining at least 2 SNPs positioned among the top-280SNPs andor common to at least two endometriosis subtypesIn this way we selected 16 SNPs that were individually

genotyped by pyrosequencing in 188 cases (42 SUP 50 OMAand 96 DIE) and 96 controls from the replication cohort(2) We also selected the top-5 SNPs those with a 119875 valueafter Monte-Carlo testing below 00001 These SNPs weregenotyped in 71 OMA and 192 controls For technical reasons(degraded DNA) samples used for genotyping the top-5SNPs were different from those used to test the 16 SNPsfirst selected Primers for PCR amplification and pyrose-quencingwere purchased fromBioTeZ (Buch Germany) (seeSupplemental Table 1 in Supplementary Material availableonline at httpdxdoiorg1011552015461024) Regions ofinterest were amplified using 20 ng of human genomic DNAand 5 pmol of forward and reverse primer one of thembeing biotinylated Reaction conditions were 10x PlatinumTaqDNAPolymeraseHigh Fidelity buffer supplementedwith05mM MgCl

2 200mM dNTPs and 15U Platinum Taq

DNA Polymerase High Fidelity (Invitrogen Cergy PontoiseFrance) in a 25 120583L volume The PCR program consisted ofa denaturing step of 4min at 95∘C followed by 50 cycles of30 s at 95∘C 30 s at the respective annealing temperature and15 s at 72∘C with a final extension of 4min at 72∘C Ten 120583Lof each amplification product were purified and renderedsingle-stranded on a pyrosequencing workstation (QiagenUppsala Sweden) Beads were released into 12 120583L annealingbuffer containing 4 pmol of the respective sequencing primerPrimers were annealed to the target by incubation at 80∘Cfor 2min Genotyping was carried out on a PyroMark MDsystem with the PyroGold SQA Reagent Kit (Pyrosequenc-ing) and results were analyzed using the PyroMark MDsoftware (Pyrosequencing) A genetic association study wasperformed Statistical analysis was performed using chi-square statistics for assessing single SNP association A 119875value of lt005 was considered statistically significant

3 Results

31 Baseline Characteristics of the Cohorts Three hundredand nineteen (319) endometriosis cases and 308 disease-free controls were recruited for the study Basic clinicalcharacteristics of the discovery group (119873 = 60) and thereplication cohort (119873 = 567) are described in SupplementalTables 2 and 3 All patients were of Caucasian origin andpresented similar mean age BMI parity gravidity andinfertility In the discovery groupDIE patients (119873 = 20) werechosen among the patients having a bowel involvement withassociated bilateral OMAs in order to improve the clinicalhomogeneity of the patients Indeed patients with bowel DIEare considered to have themost severe formof endometriosisIn the replicative cohort DIE group was comprised ofpatients having lesions of the uterosacral ligaments (119899 = 24)vagina (119899 = 23) bladder (119899 = 12) or intestinal (119899 = 37)lesions Most of the DIE patients had a bowel involvement(385)

32 Pooled Sample-Based GWAS Eight DNA pools of 10samples each (two per endometriosis subtype and two forcontrols in biological duplicate) were hybridized to the


Affymetrix 250K chip For the analysis of pooled sample-based array data we used Monte-Carlo simulations of themean of ratios of allelic frequencies and coefficient of vari-ation to rank the 262000 array SNPs Two hundred andeighty SNPs were significantly associated with at least oneendometriosis groupThe complete list of top-ranked SNPs isprovided in Supplemental Table 4 The number of significantSNPs was not different according to the subtype (102 108and 104 for SUP OMA and DIE resp) Thirty-two (114)SNPs were common to at least two subtypes (Table 1) Two(07) SNPs (rs3746192 and rs776108) were common to thethree subtypes Fifteen (54) SNPs were common to SUPand OMA six (21) to SUP and DIE and nine (32)to OMA and DIE Among the 280 top-ranked SNPs SUPhad 79 (774) specific SNPs OMA 82 (759) and DIE 87(837) This distribution although not significant (1205942 test119875 = 0347) suggests that DIE with bowel involvement maybe more homogeneous than other forms of the disease

We examined the expression of the genes correspondingto the 108 top-ranked SNPs in the OMA group accord-ing to our transcriptome analysis of OMA [7] (Table 1Supplemental Table 4) Twenty-seven genes (250) weremodified at the expression level (ie induced or repressedmore than 2-fold) while 168 of all the genes were modifiedin OMA as compared to eutopic endometrium (119875 lt 0001)[7] This suggests that these 108 variants might affect geneexpression in OMA Such an association is quite originaland could give independent clues about the importance ofa given gene in a complex disease Another way to obtainindependent clues is the extensive use of bioinformatics thatcan test whether groups of genes were obtained randomly ornot We performed nonsupervised gene functional clusteringseparately for each subtypes by using DAVID Bioinformat-ics Resources (httpdavidabccncifcrfgovhomejsp) Weobserved a strong bias of endometriosis-related genestowards transmembrane proteins affecting cellcell interac-tions and cell adhesion (Supplemental Table 5)

33 Individual Genotyping For this replication step theaverage genotyping rate was 975 Hardy-Weinberg equilib-rium was confirmed in the control group for all SNPstested We performed individual genotyping of 192 cases (96DIE 50 OMA and 46 SUP) and 96 controls for selectedSNPs We first settled on the 16 SNPs that presented thefollowing characteristics (i) common to the 3 subtypes(rs3746192 rs776108) or (ii) common to 2 subtypes and ingenes containing other significant SNPs (rs988147 rs227849rs10733833 rs322609 rs1884779 rs4703908 rs6946871rs11865033 rs247004 rs11647011 rs7477459 rs5913038rs17056959 and rs16986515) Rs227849 and rs4703908 weresignificantly associated with OMA (ORs = 231 (141ndash378)119875 = 0003 and 222 (126ndash392) 119875 = 0009 resp) (Table 2Supplemental Table 6) Rs227849 introducing A instead ofa G at position 44698708 on chromosome 6 is located nearRUNX2 SUPT3H and CDC5L In the SUP group a smalldifference was observed for rs227849 (119875 = 008) In the DIEgroup rs4703908 was significantly associated with a higherrisk of bowel endometriosis (OR = 209 (112ndash391)119875 = 0012)

(Table 2 Supplemental Table 7) Rs4703908 related to a Gto C change is located on chromosome 5 in an intronicregion near ZNF366 (zinc finger protein 366) The resultsfor DIE+OMA were not significant for any of the SNPstested expect for rs4703907 that was just over the limits ofstatistical significance (119875 = 009) We then focused on OMAwhich is regarded as the most well-defined endometriosisphenotype [28] For this attempt we used evenmore stringentcriteria to select 5 SNPs (rs11865033 rs16913217 rs966674rs7333155 and rs2479037) Those criteria were the following(i) probability of random occurrence in the Monte-Carlosimulation le11000 and (ii) allelic frequency of the SNP incontrols similar to that given by HapMap Rs2479037 andrs966674 were found significantly associated with OMA (OR= 436 (132ndash1443) 119875 = 0005 and 295 (160ndash542) 119875 = 0002resp) (Table 2 Supplemental Table 8)

4 Discussion

In this Caucasian population-based study we performed apooled sample-based GWAS in endometriosis with a distinc-tion between SUP OMA and DIE Selection of the cases wasas rigorous as possible Systematic histologic confirmationwas obtained allowing undoubted classification into one ofthe three groups All controls were surgically checked forthe absence of endometriosis We found four new variants(rs227849 rs4703908 rs2479037 and rs966674) significantlyassociatedwith an increased risk ofOMARs4703908 locatedclose to ZNF366 provided a higher risk of OMA (OR = 222(126ndash392)) and DIE especially with bowel involvement (OR= 209 (112ndash391))

This study demonstrated the feasibility of the pooledsample-basedmethod In ldquoclassicrdquo GWAS the cost associatedwith the purchase of microarrays and dedicated reagents fortens of thousands of patients put such studies beyond thereach of most research groups An alternative approach ofgenotyping pooled instead of individual genomic DNA sam-ples has been developed to reduce the overall cost of GWASOur study confirmed the feasibility of DNA pooling on SNPgenotyping microarrays and that the pooled sample-basedapproach presents an attractive and cost-effective alternativeto classic GWAS like other reports [29ndash32]The study designhas been completed for thousands of dollars whereas individ-ual genotyping requires millions of dollars For this reasonsother research groups could easily initiate replication andvalidation of our results This possibility counterbalances thelimitations associatedwith the pooling strategymainly due tothe small number of DNA samples Using 10-individual poolscarries a risk of population stratification as suggested bythe discordance in allelic frequencies between our study andthe HapMap data for some SNPs However two procedureshave been used in the present study in order to minimizethe biases and remove SNPs for which strong intergroupdifferences were observed in terms of allelic frequencies (i)the analysis in parallel of biological duplicates (two pools of 10patients per condition) and (ii) the Monte-Carlo simulationIn other words our approach was certainly very stringentbut not very sensitive The SNPs we selected were likely


Table1To

prank

edSN

Pssig

nificantly

associated

with

atleasttwoendo

metrio

sissubtypesG

enenameisprovided

ifthegeno

typedSN

Pislocatedin

upstream

31015840

UTR

intronicexon

ic

51015840UTR

ord

ownstre

amregion

softhatgenea

sann

otated

byEn

semblGenom

eBrowserG

RCh37(Feb2009)Interd

enotesthataS

NPislocatedinan

intergenicregion

Chrchrom

osom

eCV

coeffi

ciento

fvariatio

nIn

bold

arefi

guredtheS

NPs

common

toatleasttwoendo

metrio

sissubtypesInd

uctio

nratio

isther

atio

ofgene

expressio

nin

thee

ndom

etrio

ticlesio

nas

compared

tothee

xpressionin

thee

utop

icendo

metriu

m(datafrom

Borghese2008)A

ratio

above2

0deno

tesa

nup

regu

latio

nof

theg

enee

xpressionAratio

below050

deno

tesa

downregulationof

theg

enee

xpression

SNP

Chr

SUP

OMA

DIE

Gene(indu

ctionratio

)Mean

CV119875value

Mean

CV119875value

Mean

CV119875value

rs107540

061

1112

01006

40010

195629

0302

0038

RGS1

(240)

RGS18(10)

rs776108

32797

640137

0003

200537

0258

0035

264901

0201

0016

ROBO

2(158)

rs2035669

3223191

0339

0029

153997

0256

0043

GPR

156

rs2200224

40008

0156

0026

0008

0162

0027

SCFD

2(109)

RASL

11B(221)

rs6850850

471599

006

40013

176564

0292

0047

TMEM

33(049)

rs343862

44814

90031

0013

123633

0162

0031

ARH

GAP2

4rs17089182

40015

006

80023

0016

0054

0016

inter

rs247004

51319

340142

0017

156523

0148

0012

ACSL6(118

)rs4703908

5116

523

0153

0029

1077

130184

0047

ZNF366

(129)

rs17080695

573986

0124

0049

45071

0050

0021

MGAT

1(10

8)rs17056959

60015

0106

004

40014

0078

0024

GPR

63(269)

rs988147

6413289

0314

0005

3192

920341

0020

SUPT

3Hrs227849

63210

930325

0016

333485

0289

0016

RUNX2

(092)

SUPT

3H(120)

CDC5

L(131

)rs6946

871

7172041

0165

0014

173995

0354

004

6SD

K1(022)

rs17138951

7000

40422

0030

0005

0416

0038

ZNF6

80(103)

rs6558435

8236327

046

80041

1813

980150

0011

ERICH1(093)D

LGAP2

(089)

rs10733833

10136688

0205

0033

176217

0269

0027

CTNNA3LR

RTM3

rs6589535

11132232

0277

004

278770

0133

0035

Inter

rs10430848

11254632

0439

0028

148169

0187

0026

TMEM

16C(069)

rs2252178

12194081

0413

0035

168457

0333

004

4Inter

rs1548094

14000

90187

0023

0010

0182

0024

Inter

rs11865033

160043

0025

000

9004

40012

000

0MAF(179)

rs116

47011

1638406

40183

0003

323569

0196

000

0CL

EC16A(078)

rs322609

1674431

000

4000

0110368

0343

004

2CD

H11(170)

rs16949415

160015

0099

0019

0020

0098

0021

WWOX(188)

rs2472694

1782014

0159

0030

81709

0223

0035

GRA

P(135)

PRPS

AP2

(073)

SLC5

A10

(136)

rs10871642

185414

30111

0038

50092

0108

0039

CCDC1

02B(183)

rs3746192

1993400

0173

000

678952

0149

000

997912

0063

000

0KC

NN1(10

2)rs60

49295

201374

900198

0007

90072

0146

0015

Inter

rs1884779

201415

030362

0031

118669

0360

0035

SULF

2(142)

rs7291901

220091

0030

0032

0078

0029

0022

XPNPE

P3(049)

rs16986515

X11344

30037

000

0248589

0543

0020

AMEL

XARH

GAP6

MSL3L1


Table 2 Distribution of genotype and allele frequencies in cases and controls NA not applicable OR odds ratio 95 CI 95 confidenceinterval CTR controls SUP superficial peritoneal endometriosis OMA endometrioma and DIG deep infiltrating endometriosis (DIE)with intestinal involvement 119875 values are for the Pearsonrsquos chi-square test OR and 95 CI are computed for comparison between OMA orDIG and CTR Phet 119875 value for heterogeneity across discovery and replication steps calculated using the Breslow-Day test Detailed resultsfor the 21 SNPs tested are available as Supplemental Tables 6 and 8 Detailed results for all subtypes of DIE are available as Supplemental Table7

SNP (gene)[chromosome]

Genotypes AllelesOR (95 CI) Phet

CTR (119873) OMA (119873) 119875 value CTR () OMA ()Rs227849(RUNX2 SUPT3H CDC5L) [6]

AA AG GG AA AG GG A G A G39 39 17 7 27 16 0003 616 384 410 590 231 (141ndash378) 022

Rs4703908(ZNF366) [5]

CC CG GG CC CG GG C G C G64 30 1 24 21 5 0009 832 168 690 310 222 (126ndash392) 028

Rs2479037(VTI1A) [10]

CC CT TT CC CT TT C T C T5 25 160 0 3 63 0005 92 908 23 977 436 (132ndash1443) 034

Rs966674 [5] CC CG GG CC CG GG C G C G169 22 1 50 17 3 0002 937 63 836 164 295 (160ndash542) 055

CTR (119873) DIG (119873) 119875-value CTR () DIG ()Rs4703908(ZNF366) [5]


to be valid but we probably missed some SNPs associatedwith endometriosis This point is highlighted by the fact thatwe did not replicate the SNPs reported by ldquoclassicrdquo GWAS[6 11 13 14] Indeed it seems difficult to compare the resultsfrom ldquoclassicrdquo GWAS to our results First the SNPs spotted onthe Affymetrix chip are not exactly the same as those on theIllumina chip used by other groups Some genomic regionsare not covered by the Affymetrix chip and reciprocallyFor example the Affymetrix chip did not cover a genomicregion of gt5 kb around the variant reported by the JapaneseGWAS (rs16826658m) In addition the patients included inthe ldquoclassicrdquo GWAS consisted of all-coming endometriosiswithout distinction between the different phenotypes Evenwhen a distinction between stages III and IIIIV was carriedout the rAFS classification was not systematically correlatedwith the phenotype or the extension of the lesions [33] Inother words it is quite possible that a patient scored stage I orII might have a bowel endometriosis [34] Rationale for usingthe rAFS classification rather than the clinical subtypes (SUPOMA and DIE) is truly questionable [35]

It is noteworthy that the four individually validated poly-morphisms (rs4703908 rs227849 rs966674 and rs2479037)were found in patients suffering from OMA not from SUPor DIE In light of the known heterogeneous character ofendometriosis OMA probably represents the most ldquoclear-cutrdquo phenotype of all endometriosis subtypes Our teamrecently emphasized this point [28] Moreover rs4703908wassignificantly associated with bowel endometriosis a specificform of DIE considered as the most severe one This mightindicate that bowelDIE could bemore genetically driven thanthe other forms This observation reinforced the importanceof the clinical phenotype of endometriosis which is inour opinion a crucial parameter in GWAS It might beinteresting to collect precise and homogeneous phenotypesin a multicentric international effort to conduct targetedGWAS in those populations [17 34 36]

Beyond the above-mentioned limitations we introduceda new tool in the fields of genetics of endometriosisTwo SNPs were common to all subtypes rs3746192 andrs776108 Rs3746192 is located in an intronic region nearKCNN1 (potassium intermediatesmall conductance cal-cium-activated channel subfamily 119873 member 1) on chro-mosome 19 and related to a G to A change KCNN1 is amember of the KCNN family of potassium channel genesthat are thought to regulate neuronal excitability Rs776108related to an A to G change is located in a noncoding regionof chromosome 3 near ROBO2 (roundabout axon guidancereceptor and homolog 2) ROBO2 is a receptor for SLIT2known to function in axon guidance and cell migration[37] Defects in this gene are the cause of vesicoureteralreflux type 2 [38] With regard to their function KCCN1 andROBO2 are good candidate genes for endometriosis sinceneoneurogenesis frequently occurs in endometriosis [39] IntheOMAgroup we have identified four newpolymorphismsrs4703908 rs227849 rs966674 and rs2479037 with robustORs Rs2479037 is close to VTI1A possibly involved incolorectal adenocarcinoma [40] Rs4703908 associated withboth OMA and intestinal DIE is located near ZNF366that plays an important role in regulating the expressionof target genes in response to estrogen [41] ZNF366 couldbe an independent prognostic factor for breast tumors [42]ZNF366 could also act as a tumor suppressor in breast cancerdevelopment [43 44] These findings make it a plausiblecandidate gene for endometriosis which is an estrogen-dependent disease [45]

To conclude the present study clearly indicated someareas of the genome to focus on and confirmed the het-erogeneity of endometriosis The combined use of severalpolymorphisms identified in this study and in others couldeventually permit to define Bayesian probabilities enabling tocategorize the patients and predict the risk of endometriosisin future life as shown on other subjects [46] Such predictive


tool could contribute to a better follow-up of every womanand move towards a person-centered care of infertility andpain associated with endometriosis

Conflict of Interests

The authors declare that they have no conflict of interests

Authorsrsquo Contribution

Bruno Borghese Daniel Vaiman and Charles Chapronmadecontributions to conception and design Bruno BorgheseJorg TostMagalie de Surville Florence Busato Frank Letour-neur Francoise Mondon and Daniel Vaiman contributed toanalysis and interpretation of data Bruno Borghese DanielVaiman and Charles Chapron made draft of the paper orrevised it critically for important intellectual content Allauthors gave final approval of the version to be publishedDaniel Vaiman and Charles Chapron jointly directed thework

Acknowledgments

The authors warmly thank surgeons from our department fortheir expert assistance with data collection The authors alsothankfully acknowledge Mrs Nathalie Girma for unabatedlymanaging the patientsrsquo database

References

[1] P Stratton and K J Berkley ldquoChronic pelvic pain andendometriosis translational evidence of the relationship andimplicationsrdquo Human Reproduction Update vol 17 no 3 pp327ndash346 2011

[2] D de Ziegler B Borghese andC Chapron ldquoEndometriosis andinfertility pathophysiology and managementrdquoThe Lancet vol376 no 9742 pp 730ndash738 2010

[3] X Gao J Outley M Botteman J Spalding J A Simon and CL Pashos ldquoEconomic burden of endometriosisrdquo Fertility andSterility vol 86 no 6 pp 1561ndash1572 2006

[4] S Simoens L Hummelshoj and T DrsquoHooghe ldquoEndometriosiscost estimates and methodological perspectiverdquoHuman Repro-duction Update vol 13 no 4 pp 395ndash404 2007

[5] S A Treloar D TOrsquoConnor VMOrsquoConnor andNGMartinldquoGenetic influences on endometriosis in an Australian twinsamplerdquo Fertility and Sterility vol 71 no 4 pp 701ndash710 1999

[6] S Uno H Zembutsu A Hirasawa et al ldquoA genome-wide asso-ciation study identifies genetic variants in theCDKN2BAS locusassociated with endometriosis in Japaneserdquo Nature Geneticsvol 42 no 8 pp 707ndash710 2010

[7] B Borghese F Mondon J-C Noel et al ldquoGene expressionprofile for ectopic versus eutopic endometrium provides newinsights into endometriosis oncogenic potentialrdquo MolecularEndocrinology vol 22 no 11 pp 2557ndash2562 2008

[8] B Borghese S Barbaux F Mondon et al ldquoResearch resourcegenome-wide profiling of methylated promoters in endometri-osis reveals a subtelomeric location of hypermethylationrdquoMolecular Endocrinology vol 24 no 9 pp 1872ndash1885 2010

[9] S VainioMHeikkila A Kispert N Chin andA PMcMahonldquoFemale development in mammals is regulated by Wnt-4signallingrdquo Nature vol 397 no 6718 pp 405ndash409 1999

[10] A Boyer E Lapointe X Zheng et al ldquoWNT4 is required fornormal ovarian follicle development and female fertilityrdquo TheFASEB Journal vol 24 no 8 pp 3010ndash3025 2010

[11] J N Painter C A Anderson D R Nyholt et al ldquoGenome-wide association study identifies a locus at 7p152 associatedwith endometriosisrdquo Nature Genetics vol 43 no 1 pp 51ndash542011

[12] AFS ldquoRevised American Fertility Society classification of endo-metriosis 1985rdquo Fertility and Sterility vol 43 no 3 pp 351ndash3521985

[13] H M Albertsen R Chettier P Farrington and K WardldquoGenome-wide association study link novel loci to endometrio-sisrdquo PLoS ONE vol 8 no 3 Article ID e58257 2013

[14] L Pagliardini D Gentilini P Vigano et al ldquoAn Italian asso-ciation study and meta-analysis with previous GWAS confirmWNT4 CDKN2BAS and FN1 as the first identified susceptibil-ity loci for endometriosisrdquo Journal of Medical Genetics vol 50no 1 pp 43ndash46 2013

[15] N Rahmioglu S A Missmer G W Montgomery and K TZondervan ldquoInsights into assessing the genetics of endometrio-sisrdquo Current Obstetrics and Gynecology Reports vol 1 no 3 pp124ndash137 2012

[16] C Chapron N Chopin B Borghese et al ldquoDeeply infiltratingendometriosis pathogenetic implications of the anatomicaldistributionrdquoHuman Reproduction vol 21 no 7 pp 1839ndash18452006

[17] C Chapron A Bourret N Chopin et al ldquoSurgery for bladderendometriosis long-term Result s and concomitant manage-ment of associated posterior deep lesionsrdquo Human Reproduc-tion vol 25 no 4 pp 884ndash889 2010

[18] I A Brosens and J J Brosens ldquoRedefining endometriosis isdeep endometriosis a progressive diseaserdquo Human Reproduc-tion vol 15 no 1 pp 1ndash3 2000

[19] Y Wu A Kajdacsy-Balla E Strawn et al ldquoTranscriptionalcharacterizations of differences between eutopic and ectopicendometriumrdquo Endocrinology vol 147 no 1 pp 232ndash246 2006

[20] K M Eyster O Klinkova V Kennedy and K A HansenldquoWhole genome deoxyribonucleic acid microarray analysisof gene expression in ectopic versus eutopic endometriumrdquoFertility and Sterility vol 88 no 6 pp 1505ndash1533 2007

[21] J Kitawaki H Koshiba Y Kitaoka et al ldquoInterferon-120574 genedinucleotide (CA) repeat and interleukin-4 promoter region(minus590CT) polymorphisms in Japanese patients with endo-metriosisrdquo Human Reproduction vol 19 no 8 pp 1765ndash17692004

[22] S Kang S Z Li N Wang et al ldquoAssociation between geneticpolymorphisms in fibroblast growth factor (FGF)1 and FGF2and risk of endometriosis and adenomyosis inChinesewomenrdquoHuman Reproduction vol 25 no 7 pp 1806ndash1811 2010

[23] B Borghese J-D Chiche D Vernerey et al ldquoGenetic polymor-phisms of matrix metalloproteinase 12 and 13 genes are impli-cated in endometriosis progressionrdquoHuman Reproduction vol23 no 5 pp 1207ndash1213 2008

[24] E Somigliana P Vercellini U Gattei N Chopin I Chiodo andC Chapron ldquoBladder endometriosis getting closer and closerto the unifyingmetastatic hypothesisrdquo Fertility and Sterility vol87 no 6 pp 1287ndash1290 2007


[25] C Chapron C Pietin-Vialle B Borghese C Davy H Foulotand N Chopin ldquoAssociated ovarian endometrioma is a markerfor greater severity of deeply infiltrating endometriosisrdquo Fertil-ity and Sterility vol 92 no 2 pp 453ndash457 2009

[26] C Chapron A Fauconnier M Vieira et al ldquoAnatomical dis-tribution of deeply infiltrating endometriosis surgical implica-tions and proposition for a classificationrdquoHumanReproductionvol 18 no 1 pp 157ndash161 2003

[27] NMetropolis and S Ulam ldquoTheMonte Carlo methodrdquo Journalof the American Statistical Association vol 44 no 247 pp 335ndash341 1949

[28] B Borghese P Santulli D Hequet et al ldquoGenetic polymor-phisms of DNMT3L involved in hypermethylation of chro-mosomal ends are associated with greater risk of developingovarian endometriosisrdquoAmerican Journal of Pathology vol 180no 5 pp 1781ndash1786 2012

[29] E Meaburn L M Butcher L C Schalkwyk and R PlominldquoGenotyping pooled DNA using 100K SNP microarrays a steptowards genomewide association scansrdquoNucleic Acids Researchvol 34 no 4 article e27 2006

[30] J V Pearson M J Huentelman R F Halperin et al ldquoIden-tification of the genetic basis for complex disorders by use ofpooling-based genomewide single-nucleotide-polymorphismassociation studiesrdquo The American Journal of Human Geneticsvol 80 no 1 pp 126ndash139 2007

[31] S Macgregor Z Z Zhao A Henders M G Nicholas GW Montgomery and P M Visscher ldquoHighly cost-efficientgenome-wide association studies using DNA pools and denseSNP arraysrdquo Nucleic Acids Research vol 36 no 6 article e352008

[32] J E Craig A W Hewitt A E McMellon et al ldquoRapid inex-pensive genome-wide association using pooled whole bloodrdquoGenome Research vol 19 no 11 pp 2075ndash2080 2009

[33] P Vercellini L Fedele G Aimi G Pietropaolo D ConsonniandPGCrosignani ldquoAssociation between endometriosis stagelesion type patient characteristics and severity of pelvic painsymptoms amultivariate analysis of over 1000 patientsrdquoHumanReproduction vol 22 no 1 pp 266ndash271 2007

[34] B Dousset M Leconte B Borghese et al ldquoComplete surgeryfor low rectal endometriosis long-term results of a 100-caseprospective studyrdquo Annals of Surgery vol 251 no 5 pp 887ndash895 2010

[35] C C Laurie K F Doheny D B Mirel et al ldquoQuality controland quality assurance in genotypic data for genome-wideassociation studiesrdquo Genetic Epidemiology vol 34 no 6 pp591ndash602 2010

[36] C Chapron I Chiodo M Leconte et al ldquoSevere ureteralendometriosis the intrinsic type is not so rare after completesurgical exeresis of deep endometriotic lesionsrdquo Fertility andSterility vol 93 no 7 pp 2115ndash2120 2010

[37] M Kim A P Roesener P R F Mendonca and G S MastickldquoRobo1 and Robo2 have distinct roles in pioneer longitudinalaxon guidancerdquo Developmental Biology vol 358 no 1 pp 181ndash188 2011

[38] W Lu A M van Eerde X Fan et al ldquoDisruption of ROBO2is associated with urinary tract anomalies and confers risk ofvesicoureteral refluxrdquoTheAmerican Journal of HumanGeneticsvol 80 no 4 pp 616ndash632 2007

[39] V Anaf P Simon I El Nakadi et al ldquoHyperalgesia nerveinfiltration and nerve growth factor expression in deep adeno-myotic nodules peritoneal and ovarian endometriosisrdquoHumanReproduction vol 17 no 7 pp 1895ndash1900 2002

[40] A J Bass M S Lawrence L E Brace et al ldquoGenomicsequencing of colorectal adenocarcinomas identifies a recurrentVTI1A-TCF7L2 fusionrdquoNatureGenetics vol 43 no 10 pp 964ndash970 2011

[41] J Lopez-Garcia M Periyasamy R S Thomas et al ldquoZNF366is an estrogen receptor corepressor that acts through CtBP andhistone deacetylasesrdquo Nucleic Acids Research vol 34 no 21 pp6126ndash6136 2006

[42] A M Sieuwerts M Ansems M P Look et al ldquoClinicalsignificance of the nuclear receptor co-regulator DC-SCRIPTin breast cancer an independent retrospective validation studyrdquoBreast Cancer Research vol 12 no 6 article R103 2010

[43] M Ansems S Hontelez MW G Looman et al ldquoDC-SCRIPTnuclear receptor modulation and prognostic significance inprimary breast cancerrdquo Journal of the National Cancer Institutevol 102 no 1 pp 54ndash68 2010

[44] M Ansems S Hontelez N Karthaus P N Span and GJ Adema ldquoCrosstalk and DC-SCRIPT expanding nuclearreceptor modulationrdquo Biochimica et Biophysica ActamdashReviewson Cancer vol 1806 no 2 pp 193ndash199 2010

[45] S E Bulun ldquoEndometriosisrdquo The New England Journal ofMedicine vol 360 no 3 pp 268ndash279 2009

[46] R Ciampolini K Moazami-Goudarzi D Vaiman et al ldquoIndi-vidual multilocus genotypes using microsatellite polymor-phisms to permit the analysis of the genetic variability withinand between Italian beef cattle breedsrdquo Journal of AnimalScience vol 73 no 11 pp 3259ndash3268 1995

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


MEDIATORSINFLAMMATION

of


Behavioural Neurology

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom


[9 10] The second UKAustralian GWAS found an asso-ciation of endometriosis with rs12700667 located in 7p152in an intergenic region upstream of NFE2L3 HOXA10 andHOXA11 [11]The role ofHOX genes in endometriosis-relatedinfertility has been largely debated [7] After pooling thedata from the two studies rs12700667 remained significantlyassociated with endometriosis and also with stages IIIIV[12] Consistent with these associations the latest GWASimplicated a 150 kb region around WNT4 that also includeLINC00339 and CDC42 [13] An independent set of 305laparoscopically proven endometriosis patients and 2710controls confirmedWNT4 CDKN2BAS and FN1 as the firstidentified common loci for endometriosis [14]

With odds ratios below 15 these associations arenonetheless not strong enough to suggest a causal relationor to consider a potential clinical application [15] This couldbe due to different genetic backgrounds across populationsrs10965235 for instance is monomorphic in Caucasians Mostlikely this reflects the heterogeneity of endometriosis [16]Indeed endometriosis has inconstant symptoms spanningfrom no symptoms to severe chronic pelvic pain and infer-tility [1 2] Three distinct clinical forms of endometriosishave been identified (i) Peritoneal superficial endometrio-sis (SUP) consists of lesions lying on the surface of theperitoneum or the ovaries (ii) Endometrioma (OMA) isan ovarian endometriotic ldquochocolaterdquo cyst (iii) Deeply infil-trating endometriosis (DIE) comprises lesions infiltratingthe muscularis propria of structures surrounding the uterus(vagina bladder bowel or ureters) [17] The phenotypicdistinction between SUP OMA and DIE may also be raisedin terms of their sex hormone responsiveness SUP OMAand DIE have different responsiveness to progesterone [18]differential gene expressions [19 20] and specific putativevariants that may influence one form and not the others [21ndash23]

Therefore we initiated a pooled sample-based GWASwith a distinction between SUP OMA and DIE As thiswas not a classic GWAS approach we conducted an initialdiscovery step on 10-individual pools in biological duplicatesusing the Affymetrix GenChip 250K Nsp chip After qualitycontrol filtering and SNPs ranking based on a false discoveryrate (FDR) below 5 we performed a replication step ofindividual genotyping to validate the top-ranked SNPs in anindependent cohort of 259 endometriosis and 288 controlsEndometriosis was confirmed histologically in cases andinvalidated surgically in controls

2 Materials and Methods

21 Study Population The data were obtained from 627unrelated women of Caucasian origin treated in our tertiaryreferral center for endometriosis and gynecologic conditionsThe 319 endometriotic patients underwent a complete sur-gical resection of all visible lesions allowing confirmationof the diagnosis by expert pathologists in all the cases Weclassified the patients into SUP OMA and DIE consideringthe most severe lesion DIE was histologically defined whenendometriotic lesions involved the muscularis of the vagina

the bladder the intestine or the ureter [17] As these threetypes of lesionswere frequently concomitant [24 25] patientswere arbitrarily classified as per the worst findings [26]Hence a patient presenting with SUP lesions associated withOMA and DIE was classified as DIE As well a patientwith OMA and concomitant SUP lesions was classified asOMAwhereas a patient classified as SUP only presented SUPlesions The 308 controls (CTR) consisted of women withoutany lesion suggestive of endometriosis as checked during acomprehensive surgical exploration Indications for surgeryin these patients were infertility work-up symptomatic uter-ine fibroids benign ovarian cysts or chronic pelvic pain

The discovery group was composed of 60 cases (20 SUP20 OMA and 20 DIE) and 20 CTR randomly selected fromthe entire cohort for DNA-pooling experiments The DIEgroup included the most severe patients that is those pre-senting at least one lesion infiltrating the bowel and associatedbilateral OMAs more than 3 cm This was done intentionallyin order to have a homogeneous group since DIE mightencompass a large variety of lesions The replication cohortwas composed of the remaining 259 cases (42 SUP 121 OMAand 96 DIE) and 288 CTR The ethical review board of ourcenter (Comite Consultatif de Protection des Personnes dansla Recherche Biomedicale de Paris-Cochin) approved thestudy design All subjects provided written informed consentbefore entering the study

22 DNA Extraction and Quantification Pool Constructionand SNP Allelotyping Five millimeters of venous blood werecollected from individuals into EDTA tubes Genomic DNAwas subsequently extracted with the MagNa Pure CompactNucleic Acid Isolation Kit (Roche Applied Science Indi-anapolis IN USA) DNA was quantified using a spectropho-tometer (NanoDrop 1000 Thermo Scientific WilmingtonDE USA) and diluted to a concentration of 50 ng120583L After-wards each DNA was quantified using fluorimetry (Quant-It DNA Assay Kit Invitrogen Carlsbad CA USA) andchecked for quality using 1 agarose gel electrophoresis Eachpool was composed of 10 samples of 100 ng DNA aliquotsfrom the same category (SUP OMA DIE and CTR) In allwe built eight 10-patient DNA pools two per category inbiological replicates (2 SUP 2 OMA 2 DIE and 2 CTR)DNA quantification and quality were checked several timesafter the pooling procedure to make sure each pool had thesame amount of DNA Genotyping analysis with GenChipHuman Mapping 250K Nsp Array Set (Affymetrix SantaClara CA USA) was performed for each pool followingthe manufacturerrsquos guidelines Briefly genomic DNAs wererestricted with NspI NspI adaptors were then ligated torestricted fragments and subjected to PCRusing the universalprimer PCR002 provided by the kit PCR fragmentswere thenpurified and 90 120583g used for fragmentation and end-labelingwith biotin using Terminal Transferase Labeled targets werethen hybridized overnight to Genechip human 250K NspIarray (Affymetrix) at 49∘C Chips were washed on the fluidicstation FS450 following specific protocols (Affymetrix) andscanned using theGCS3000 7GThe imagewas then analyzedwith the GCOS software to obtain raw data




119860= 119891119860(119891119860+ 119891119861)



1and SUP

2

OMA1and OMA

2 DIE1and DIE

2 and CTR

1and CTR

2) we








3 Results








4 Discussion




Table1To

prank

edSN

Pssig

nificantly

associated

with

atleasttwoendo

metrio

sissubtypesG

enenameisprovided

ifthegeno

typedSN

Pislocatedin

upstream

31015840

UTR

intronicexon

ic

51015840UTR

ord

ownstre

amregion

softhatgenea

sann

otated

byEn

semblGenom

eBrowserG


enotesthataS

NPislocatedinan

intergenicregion

Chrchrom

osom

eCV

coeffi

ciento

fvariatio

nIn

bold

arefi

guredtheS

NPs

common

toatleasttwoendo

metrio

sissubtypesInd

uctio

nratio

isther

atio

ofgene

expressio

nin

thee

ndom

etrio

ticlesio

nas

compared

tothee

xpressionin

thee

utop

icendo

metriu

m(datafrom

Borghese2008)A

ratio

above2

0deno

tesa

nup

regu

latio

nof

theg

enee

xpressionAratio

below050

deno

tesa

downregulationof

theg

enee

xpression

SNP

Chr

SUP

OMA

DIE

Gene(indu

ctionratio

)Mean

CV119875value

Mean

CV119875value

Mean

CV119875value

rs107540

061

1112

01006

40010

195629

0302

0038

RGS1

(240)

RGS18(10)

rs776108

32797

640137

0003

200537

0258

0035

264901

0201

0016

ROBO

2(158)

rs2035669

3223191

0339

0029

153997

0256

0043

GPR

156

rs2200224

40008

0156

0026

0008

0162

0027

SCFD

2(109)

RASL

11B(221)

rs6850850

471599

006

40013

176564

0292

0047

TMEM

33(049)

rs343862

44814

90031

0013

123633

0162

0031

ARH

GAP2

4rs17089182

40015

006

80023

0016

0054

0016

inter

rs247004

51319

340142

0017

156523

0148

0012

ACSL6(118

)rs4703908

5116

523

0153

0029

1077

130184

0047

ZNF366

(129)

rs17080695

573986

0124

0049

45071

0050

0021

MGAT

1(10

8)rs17056959

60015

0106

004

40014

0078

0024

GPR

63(269)

rs988147

6413289

0314

0005

3192

920341

0020

SUPT

3Hrs227849

63210

930325

0016

333485

0289

0016

RUNX2

(092)

SUPT

3H(120)

CDC5

L(131

)rs6946

871

7172041

0165

0014

173995

0354

004

6SD

K1(022)

rs17138951

7000

40422

0030

0005

0416

0038

ZNF6

80(103)

rs6558435

8236327

046

80041

1813

980150

0011

ERICH1(093)D

LGAP2

(089)

rs10733833

10136688

0205

0033

176217

0269

0027

CTNNA3LR

RTM3

rs6589535

11132232

0277

004

278770

0133

0035

Inter

rs10430848

11254632

0439

0028

148169

0187

0026

TMEM

16C(069)

rs2252178

12194081

0413

0035

168457

0333

004

4Inter

rs1548094

14000

90187

0023

0010

0182

0024

Inter

rs11865033

160043

0025

000

9004

40012

000

0MAF(179)

rs116

47011

1638406

40183

0003

323569

0196

000

0CL

EC16A(078)

rs322609

1674431

000

4000

0110368

0343

004

2CD

H11(170)

rs16949415

160015

0099

0019

0020

0098

0021

WWOX(188)

rs2472694

1782014

0159

0030

81709

0223

0035

GRA

P(135)

PRPS

AP2

(073)

SLC5

A10

(136)

rs10871642

185414

30111

0038

50092

0108

0039

CCDC1

02B(183)

rs3746192

1993400

0173

000

678952

0149

000

997912

0063

000

0KC

NN1(10

2)rs60

49295

201374

900198

0007

90072

0146

0015

Inter

rs1884779

201415

030362

0031

118669

0360

0035

SULF

2(142)

rs7291901

220091

0030

0032

0078

0029

0022

XPNPE

P3(049)

rs16986515

X11344

30037

000

0248589

0543

0020

AMEL

XARH

GAP6

MSL3L1







Rs4703908(ZNF366) [5]


Rs2479037(VTI1A) [10]















Acknowledgments


References





















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of



















119860= 119891119860(119891119860+ 119891119861)



1and SUP

2

OMA1and OMA

2 DIE1and DIE

2 and CTR

1and CTR

2) we








3 Results








4 Discussion




Table1To

prank

edSN

Pssig

nificantly

associated

with

atleasttwoendo

metrio

sissubtypesG

enenameisprovided

ifthegeno

typedSN

Pislocatedin

upstream

31015840

UTR

intronicexon

ic

51015840UTR

ord

ownstre

amregion

softhatgenea

sann

otated

byEn

semblGenom

eBrowserG


enotesthataS

NPislocatedinan

intergenicregion

Chrchrom

osom

eCV

coeffi

ciento

fvariatio

nIn

bold

arefi

guredtheS

NPs

common

toatleasttwoendo

metrio

sissubtypesInd

uctio

nratio

isther

atio

ofgene

expressio

nin

thee

ndom

etrio

ticlesio

nas

compared

tothee

xpressionin

thee

utop

icendo

metriu

m(datafrom

Borghese2008)A

ratio

above2

0deno

tesa

nup

regu

latio

nof

theg

enee

xpressionAratio

below050

deno

tesa

downregulationof

theg

enee

xpression

SNP

Chr

SUP

OMA

DIE

Gene(indu

ctionratio

)Mean

CV119875value

Mean

CV119875value

Mean

CV119875value

rs107540

061

1112

01006

40010

195629

0302

0038

RGS1

(240)

RGS18(10)

rs776108

32797

640137

0003

200537

0258

0035

264901

0201

0016

ROBO

2(158)

rs2035669

3223191

0339

0029

153997

0256

0043

GPR

156

rs2200224

40008

0156

0026

0008

0162

0027

SCFD

2(109)

RASL

11B(221)

rs6850850

471599

006

40013

176564

0292

0047

TMEM

33(049)

rs343862

44814

90031

0013

123633

0162

0031

ARH

GAP2

4rs17089182

40015

006

80023

0016

0054

0016

inter

rs247004

51319

340142

0017

156523

0148

0012

ACSL6(118

)rs4703908

5116

523

0153

0029

1077

130184

0047

ZNF366

(129)

rs17080695

573986

0124

0049

45071

0050

0021

MGAT

1(10

8)rs17056959

60015

0106

004

40014

0078

0024

GPR

63(269)

rs988147

6413289

0314

0005

3192

920341

0020

SUPT

3Hrs227849

63210

930325

0016

333485

0289

0016

RUNX2

(092)

SUPT

3H(120)

CDC5

L(131

)rs6946

871

7172041

0165

0014

173995

0354

004

6SD

K1(022)

rs17138951

7000

40422

0030

0005

0416

0038

ZNF6

80(103)

rs6558435

8236327

046

80041

1813

980150

0011

ERICH1(093)D

LGAP2

(089)

rs10733833

10136688

0205

0033

176217

0269

0027

CTNNA3LR

RTM3

rs6589535

11132232

0277

004

278770

0133

0035

Inter

rs10430848

11254632

0439

0028

148169

0187

0026

TMEM

16C(069)

rs2252178

12194081

0413

0035

168457

0333

004

4Inter

rs1548094

14000

90187

0023

0010

0182

0024

Inter

rs11865033

160043

0025

000

9004

40012

000

0MAF(179)

rs116

47011

1638406

40183

0003

323569

0196

000

0CL

EC16A(078)

rs322609

1674431

000

4000

0110368

0343

004

2CD

H11(170)

rs16949415

160015

0099

0019

0020

0098

0021

WWOX(188)

rs2472694

1782014

0159

0030

81709

0223

0035

GRA

P(135)

PRPS

AP2

(073)

SLC5

A10

(136)

rs10871642

185414

30111

0038

50092

0108

0039

CCDC1

02B(183)

rs3746192

1993400

0173

000

678952

0149

000

997912

0063

000

0KC

NN1(10

2)rs60

49295

201374

900198

0007

90072

0146

0015

Inter

rs1884779

201415

030362

0031

118669

0360

0035

SULF

2(142)

rs7291901

220091

0030

0032

0078

0029

0022

XPNPE

P3(049)

rs16986515

X11344

30037

000

0248589

0543

0020

AMEL

XARH

GAP6

MSL3L1







Rs4703908(ZNF366) [5]


Rs2479037(VTI1A) [10]















Acknowledgments


References





















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of





















4 Discussion




Table1To

prank

edSN

Pssig

nificantly

associated

with

atleasttwoendo

metrio

sissubtypesG

enenameisprovided

ifthegeno

typedSN

Pislocatedin

upstream

31015840

UTR

intronicexon

ic

51015840UTR

ord

ownstre

amregion

softhatgenea

sann

otated

byEn

semblGenom

eBrowserG


enotesthataS

NPislocatedinan

intergenicregion

Chrchrom

osom

eCV

coeffi

ciento

fvariatio

nIn

bold

arefi

guredtheS

NPs

common

toatleasttwoendo

metrio

sissubtypesInd

uctio

nratio

isther

atio

ofgene

expressio

nin

thee

ndom

etrio

ticlesio

nas

compared

tothee

xpressionin

thee

utop

icendo

metriu

m(datafrom

Borghese2008)A

ratio

above2

0deno

tesa

nup

regu

latio

nof

theg

enee

xpressionAratio

below050

deno

tesa

downregulationof

theg

enee

xpression

SNP

Chr

SUP

OMA

DIE

Gene(indu

ctionratio

)Mean

CV119875value

Mean

CV119875value

Mean

CV119875value

rs107540

061

1112

01006

40010

195629

0302

0038

RGS1

(240)

RGS18(10)

rs776108

32797

640137

0003

200537

0258

0035

264901

0201

0016

ROBO

2(158)

rs2035669

3223191

0339

0029

153997

0256

0043

GPR

156

rs2200224

40008

0156

0026

0008

0162

0027

SCFD

2(109)

RASL

11B(221)

rs6850850

471599

006

40013

176564

0292

0047

TMEM

33(049)

rs343862

44814

90031

0013

123633

0162

0031

ARH

GAP2

4rs17089182

40015

006

80023

0016

0054

0016

inter

rs247004

51319

340142

0017

156523

0148

0012

ACSL6(118

)rs4703908

5116

523

0153

0029

1077

130184

0047

ZNF366

(129)

rs17080695

573986

0124

0049

45071

0050

0021

MGAT

1(10

8)rs17056959

60015

0106

004

40014

0078

0024

GPR

63(269)

rs988147

6413289

0314

0005

3192

920341

0020

SUPT

3Hrs227849

63210

930325

0016

333485

0289

0016

RUNX2

(092)

SUPT

3H(120)

CDC5

L(131

)rs6946

871

7172041

0165

0014

173995

0354

004

6SD

K1(022)

rs17138951

7000

40422

0030

0005

0416

0038

ZNF6

80(103)

rs6558435

8236327

046

80041

1813

980150

0011

ERICH1(093)D

LGAP2

(089)

rs10733833

10136688

0205

0033

176217

0269

0027

CTNNA3LR

RTM3

rs6589535

11132232

0277

004

278770

0133

0035

Inter

rs10430848

11254632

0439

0028

148169

0187

0026

TMEM

16C(069)

rs2252178

12194081

0413

0035

168457

0333

004

4Inter

rs1548094

14000

90187

0023

0010

0182

0024

Inter

rs11865033

160043

0025

000

9004

40012

000

0MAF(179)

rs116

47011

1638406

40183

0003

323569

0196

000

0CL

EC16A(078)

rs322609

1674431

000

4000

0110368

0343

004

2CD

H11(170)

rs16949415

160015

0099

0019

0020

0098

0021

WWOX(188)

rs2472694

1782014

0159

0030

81709

0223

0035

GRA

P(135)

PRPS

AP2

(073)

SLC5

A10

(136)

rs10871642

185414

30111

0038

50092

0108

0039

CCDC1

02B(183)

rs3746192

1993400

0173

000

678952

0149

000

997912

0063

000

0KC

NN1(10

2)rs60

49295

201374

900198

0007

90072

0146

0015

Inter

rs1884779

201415

030362

0031

118669

0360

0035

SULF

2(142)

rs7291901

220091

0030

0032

0078

0029

0022

XPNPE

P3(049)

rs16986515

X11344

30037

000

0248589

0543

0020

AMEL

XARH

GAP6

MSL3L1







Rs4703908(ZNF366) [5]


Rs2479037(VTI1A) [10]















Acknowledgments


References





















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















Table1To

prank

edSN

Pssig

nificantly

associated

with

atleasttwoendo

metrio

sissubtypesG

enenameisprovided

ifthegeno

typedSN

Pislocatedin

upstream

31015840

UTR

intronicexon

ic

51015840UTR

ord

ownstre

amregion

softhatgenea

sann

otated

byEn

semblGenom

eBrowserG


enotesthataS

NPislocatedinan

intergenicregion

Chrchrom

osom

eCV

coeffi

ciento

fvariatio

nIn

bold

arefi

guredtheS

NPs

common

toatleasttwoendo

metrio

sissubtypesInd

uctio

nratio

isther

atio

ofgene

expressio

nin

thee

ndom

etrio

ticlesio

nas

compared

tothee

xpressionin

thee

utop

icendo

metriu

m(datafrom

Borghese2008)A

ratio

above2

0deno

tesa

nup

regu

latio

nof

theg

enee

xpressionAratio

below050

deno

tesa

downregulationof

theg

enee

xpression

SNP

Chr

SUP

OMA

DIE

Gene(indu

ctionratio

)Mean

CV119875value

Mean

CV119875value

Mean

CV119875value

rs107540

061

1112

01006

40010

195629

0302

0038

RGS1

(240)

RGS18(10)

rs776108

32797

640137

0003

200537

0258

0035

264901

0201

0016

ROBO

2(158)

rs2035669

3223191

0339

0029

153997

0256

0043

GPR

156

rs2200224

40008

0156

0026

0008

0162

0027

SCFD

2(109)

RASL

11B(221)

rs6850850

471599

006

40013

176564

0292

0047

TMEM

33(049)

rs343862

44814

90031

0013

123633

0162

0031

ARH

GAP2

4rs17089182

40015

006

80023

0016

0054

0016

inter

rs247004

51319

340142

0017

156523

0148

0012

ACSL6(118

)rs4703908

5116

523

0153

0029

1077

130184

0047

ZNF366

(129)

rs17080695

573986

0124

0049

45071

0050

0021

MGAT

1(10

8)rs17056959

60015

0106

004

40014

0078

0024

GPR

63(269)

rs988147

6413289

0314

0005

3192

920341

0020

SUPT

3Hrs227849

63210

930325

0016

333485

0289

0016

RUNX2

(092)

SUPT

3H(120)

CDC5

L(131

)rs6946

871

7172041

0165

0014

173995

0354

004

6SD

K1(022)

rs17138951

7000

40422

0030

0005

0416

0038

ZNF6

80(103)

rs6558435

8236327

046

80041

1813

980150

0011

ERICH1(093)D

LGAP2

(089)

rs10733833

10136688

0205

0033

176217

0269

0027

CTNNA3LR

RTM3

rs6589535

11132232

0277

004

278770

0133

0035

Inter

rs10430848

11254632

0439

0028

148169

0187

0026

TMEM

16C(069)

rs2252178

12194081

0413

0035

168457

0333

004

4Inter

rs1548094

14000

90187

0023

0010

0182

0024

Inter

rs11865033

160043

0025

000

9004

40012

000

0MAF(179)

rs116

47011

1638406

40183

0003

323569

0196

000

0CL

EC16A(078)

rs322609

1674431

000

4000

0110368

0343

004

2CD

H11(170)

rs16949415

160015

0099

0019

0020

0098

0021

WWOX(188)

rs2472694

1782014

0159

0030

81709

0223

0035

GRA

P(135)

PRPS

AP2

(073)

SLC5

A10

(136)

rs10871642

185414

30111

0038

50092

0108

0039

CCDC1

02B(183)

rs3746192

1993400

0173

000

678952

0149

000

997912

0063

000

0KC

NN1(10

2)rs60

49295

201374

900198

0007

90072

0146

0015

Inter

rs1884779

201415

030362

0031

118669

0360

0035

SULF

2(142)

rs7291901

220091

0030

0032

0078

0029

0022

XPNPE

P3(049)

rs16986515

X11344

30037

000

0248589

0543

0020

AMEL

XARH

GAP6

MSL3L1







Rs4703908(ZNF366) [5]


Rs2479037(VTI1A) [10]















Acknowledgments


References





















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of






















Rs4703908(ZNF366) [5]


Rs2479037(VTI1A) [10]















Acknowledgments


References





















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of






















Acknowledgments


References





















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of












































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of
















research article identification of susceptibility genes...

Documents