Twenty-eight genetic loci associated with ST-T wave amplitudes of the electrocardiogram

ASSOCIATION STUDIES ARTICLE Twen y-eigh gene ic loci associa ed wi h ST-T-wa e ampli udes o he elec oca diog am Niek Ve weij1,11,23, I ene Ma eo Leach1, Aa on Isaacs4,24, Dan E. A king6, Joshua C. Bis7, Tune H. Pe s10,11, Ma en E. Van Den Be g5, Leo-Pekka Lyy ikäinen12, Phil Ba ne 13, Xinchen Wang14, Li eLines Coho S udy, Elsayed Z. Soliman15, Co nelia M. Van Duijn4, Mika Kähönen16, Di k J. Van Veldhuisen1, Jan A. Ko s5, Olli T. Rai aka i17,18, Claudia T. Sil a4, Te ho Leh imäki12, Hans L. Hillege1,2, Joel N. Hi schho n10,11,19, Lau ie A. Boye 14, Wiek H. Van Gils 1, Al a o Alonso20, Nona So oodehnia8,9, Ma k Eijgelsheim5, Rudol A. De Boe 1, Paul I. W. De Bakke 11,21, Lude F anke3and Pim Van De Ha s 1,3,22,* 1 Depa men o Ca diology, 2 T ial Coo dina ion Cen e and 3 Depa men o Gene ics, Uni e si y o G oningen, Uni e si y Medical Cen e G oningen, Hanzeplein 1, 9713 GZ G oningen, The Ne he lands, 4 Depa men o Epidemiology, Gene ic Epidemiology Uni and 5 Depa men o Medical In o ma ics, E asmus Uni e si y Medical Cen e , Ro e dam, The Ne he lands, 6 McKusick-Na hans Ins i u e o Gene ic Medicine, Johns Hopkins Uni e si y School o Medicine, Bal imo e, MD, USA, 7 Depa men o Medicine, Ca dio ascula Heal h Resea ch Uni , 8 Di ision o Ca diology, Depa men o Medicine and 9 Depa men o Medicine, Ca dio ascula Heal h Resea ch Uni , Uni e si y o Washing on, Sea le, WA, USA, 10 Di ision o Endoc inology, Cen e o Basic and T ansla ional Obesi y Resea ch, Bos on Child en’s Hospi al, Bos on, USA, 11 P og am in Medical and Popula ion Gene ics, B oad Ins i u e o MIT and Ha a d, 301 Binney S ee , Camb idge, MA 02142, USA, 12 Depa men o Clinical Chemis y, Fimlab Labo a o ies and Uni e si y o Tampe e School o Medicine, Tampe e 33520, Finland, 13 Depa men o Ana omy, Emb yologyand Physiology, Academic Medical Cen e , Uni e si yo Ams e dam, 1105 AZ Ams e dam, The Ne he lands, 14 Depa men o Biology, Massachuse s Ins i u e o Technology, 77 Massachuse s A enue, Camb idge, MA 02139, USA, 15 Di ision o Public Heal h Sciences, Epidemiological Ca diology Resea ch Cen e (EPICARE), Wake Fo es School o Medicine, Wins on Salem, NC, USA, 16 Depa men o Clinical Physiology, Tampe e Uni e si y Hospi al and Uni e si y o Tampe e School o Medicine, Tampe e 33521, Finland, 17 Depa men o Clinical Physiology and Nuclea Medicine, Tu ku Uni e si y Hospi al, Tu ku 20520, Finland, 18 Resea ch Cen e o Applied and P e en i e Ca dio ascula Medicine, Uni e si y o Tu ku, Tu ku 20520, Finland, 19 Depa men o Gene ics, Ha a d Medical School, Bos on, USA, 20 Di ision o Epidemiology and Communi y Heal h, School o Public Heal h, Uni e si y o Minneso a, Minneapolis, MN, USA, 21 Depa men o Medical Gene ics, Uni e si y Medical Cen e U ech , Uni e si ei sweg 100, 3584 CG U ech , The Ne he lands, 22 Du e Cen e o Ca diogene ic Resea ch, ICIN-Ne he lands Hea Ins i u e, 3511 GC U ech , The Ne he lands, Recei ed: Oc obe 5, 2015. Re ised: Feb ua y 11, 2016. Accep ed: Feb ua y 17, 2016 © The Au ho 2016. Published by Ox o d Uni e si y P ess. This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion Non-Comme cial License (h p://c ea i ecommons.o g/ licenses/by-nc/4.0/), which pe mi s non-comme cial e-use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. Fo comme cial e-use, please con ac [email p o ec ed] doi: 10.1093/hmg/ddw058 Ad ance Access Publica ion Da e: 8 Ma ch 2016 Associa ion S udies A icle Human Molecula Gene ics, 2016, Vol. 25, No. 10 –2093 2103 2093 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Oc obe 19, 2016h p://hmg.ox o djou nals.o g/Downloaded om |Human Molecula Gene ics, 2016, Vol. 25, No. 10 23 Ca dio ascula Resea ch Cen e and Cen e o Human Gene ic Resea ch, Massachuse s Gene al Hospi al, Bos on, Massachuse s and 24 CARIM School o Ca dio ascula Diseases, Maas ich Cen e o Sys ems Biology (MaCSBio), and Depa men o Biochemis y, Maas ich Uni e si y, Maas ich , The Ne he lands *To whom co espondence should be add essed a : Depa men o Ca diology, Uni e si y Medical Cen e G oningen Uni e si y o G oningen, Hanzeplein 1, 9713 GZ G oningen, The Ne he lands. Tel: +31 503615340; Fax: +31 503614391; Email: [email p o ec ed] Abs ac The ST-segmen and adjacen T-wa e (ST-T wa e) ampli udes o he elec oca diog am a e quan i a i e cha ac e is ics o ca diac epola iza ion. Repola iza ion abno mali ies ha e been linked o en icula a hy hmias and sudden ca diac dea h. We pe o med he fi s genome-wide associa ion me a-analysis o ST-T-wa e ampli udes in up o 37977 indi iduals iden i ying 71 obus geno ype–pheno ype associa ions clus e ed wi hin 28 independen loci. Fi y- ou genes we e p io i ized as candida es unde lying he pheno ypes, including genes wi h es ablished oles in he ca diac epola iza ion phase (SCN5A/SCN10A,KCND3, KCNB1,NOS1AP and HEY2) and o he s wi h as ye undefined ca diac unc ion. These associa ions may p o ide insigh s in he spa io empo al con ibu ion o gene ic a ia ion influencing ca diac epola iza ion and p o ide no el leads o u u e unc ional ollow-up. In oduc ion Du a ion o ca diac epola iza ion has been p e iously s udied by genome-wide associa ion s udies (GWAS) and led o he disco - e y o 35 associa ed loci (1). Howe e , abno mali ies o epola iza- iona e no limi ed ochangesindu a ion, bu a ealsocap u edby changes in ampli udes (2). Ven icula epola iza ion occu s in a geome icpa e nac oss hehea .Thisp ocessen ailsdi e ences in imings o eco e y ac oss he en icula wall and be ween egions o he le en icle a e en icula depola iza ion. The cellula basis o he o igins o he ST-segmen and adjacen Twa e (ST-Twa e) ampli udes is s illunde deba e, bu is gene al- ly hough o a ise as a consequence o elec ical he e ogenei ies ha exis wi hin he en icula myoca dium (3). I is hough ha he ST-T wa e begins when he epica dialcells s a o eco e aheado he midwallandendoca dialcells.Elec icalcu en sflow om he midmyoca dial and endoca dial egions owa d he epi- ca dium, esul ingin a g adual ise in cu en , un il he epica dium is ully epola ized and he cu en s ha e eached maximal in en- si y (T- op) (4). Con inuous indices o T-wa e ampli udes a e no only ele an in pa ien s wi h myoca dial in a c ion (5), ischemic ca diomyopa hy (6) and a malignan o m o ER (7), bu also in non-diseased popula ion coho s. In he gene al o heal hy popu- la ion, in e sions o he ST-Twa e a e s onglyassocia ed wi h ad- e se ou come (8–10), bu also mino dep essions o ele a ions in ST-T-wa e ampli udes a e p edic o s o ca dio ascula mo ali y (11–13)anddisease(10,14). De ia ions o ST-T-wa e ampli udes can be indica i e o a a ie y o ca diac pa hologies, including myoca dial ischemia, en icula hype ophy, long-QT synd ome, ER and B ugada synd ome (2,8,12,15–17). Fo clinical use, pheno ypes a e di- cho omized based on op imal sensi i i y and specifici y o p e- dic wo se ou come. Howe e , he e is no e idence ha he unde lying biology o ST-T wa e is uly bina y. The e o e, we hypo hesized ha common gene ic a ia ion in ol ed in he biology o quan i a i e ST-T wa es ai s migh p o ide add- i ional biological insigh s in o he (pa ho)physiological me- chanisms o epola iza ion. GWAS ha e p o en o be a powe ul and unbiased ool o iden i y no el mechanisms and pa hways. He e, we aim o iden i y key gene ic loci associa ed wi h he hea ’s epola iza ion phase du ing he ST-T wa e o he elec oca diog am (ECG) o u he ad ance ou knowledge on biological ac o s egula ing ca diac epola iza ion and he eby p o ide new gene a ge s o u u e s udy. Resul s Genome-wide disco e y analysis and eplica ion We pe o med genome-wide me a-analyses in 15 943 subjec s o Eu opean descen on he ST-T wa e (Supplemen a y Ma e ial, Table S1), wi h up o 2 316 136 di ec ly geno yped and impu ed au osomal single nucleo ide polymo phisms (SNPs). Pheno ypes o he ST-segmen ampli udes and T-wa e ampli udes we e de- i ed om leads used in he clinic o diagnosing B ugada syn- d ome (17)(Sep al,V 1 and V 2 ) and ea ly epola iza ion (ER) (16– 18) (la e al: I, aVL, V 5 ,V 6 and in e io : II, III, aVF). Aiming o cap- u e addi ional in o ma ion a ailable on he ECG du ing ca diac epola iza ion, we also in es iga ed he o he lead eco dings (An e io : V 3 ,V 4 and lead aVR) ha a e p esumed o ha e ana- omical meaning (Supplemen a y Ma e ial, Table S2). This e- sul ed in 10 pheno ypes ep esen ing he ST-T-wa e ampli ude: 5 ST-segmen ampli udes and 5 T-wa e ampli udes (Fig. 1A). We es ima ed ha s udying 10 co ela ed ST-T wa e pheno ypes was equal o s udying eigh independen pheno ypes (see he ‘Ma e ials and Me hods’sec ion); he e o e, he h eshold o gen- ome-wide significance was se o P<5×10 −8 /8 (P<6.25×10 −9 ). We conside ed a h eshold o P<1×10 −6 o be sugges i e o as- socia ion and used his h eshold o p io i ize SNPs o eplica- ion. The genome-wide me a-analysis ac oss he 10 ST-T-wa e pheno ypes iden ified a o al o 36 loci o pass he sugges i e h eshold o associa ion wi h a leas one o he ST-T wa e pheno ypes. The e was no e idence o infla ion o es s a is- ics in he final me a-analysis (Supplemen a y Ma e ial, Table S3) o significan he e ogenei y. O he 36 loci, he e we e mul iple loci we e associa ed wi h one o mo e 10 ST-T-wa e pheno ypes: o aling 95 geno ype–pheno ype associa ions (P<1 ×10 −6 , Supplemen a y Ma e ial, Table S4). Nex , we pe o med eplica ion es ing o he 95 geno ype– pheno ype associa ions (P<1×10 −6 ) in 22 034 independen indi iduals de i ed om 7 independen coho s (Supplemen a y Ma e ial, Table S5), iden i ying 71 geno ype–pheno ype associa- ions in 28 independen loci o be eplica ed (P< 0.01) and become 2094 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Oc obe 19, 2016h p://hmg.ox o djou nals.o g/Downloaded om Human Molecula Gene ics, 2016, Vol. 25, No. 10 | genome-wide significan ly associa ed (Fig. 1B, Table 1and Sup- plemen a y Ma e ial, Table S4 and Fig. S1). The majo i y o loci ha show mul iple geno ype–pheno ype associa ions in he same locus consis o sen inel SNPs ha a e ei he he same SNP o in LD wi h each o he , as indica ed by he LD ( 2 ) column in Table 1. Condi ional analysis wi hin he dis- co e y se e ealed a seconda y signal [ s9851710 condi ioned on s6801957: β(SE) pe mino allele = 0.13 (0.0204), P= 2.38 × 10 −10 ]in he SCN5A/SCN10A locus o be associa ed wi h ST aVR ampli- udes. This SNP has also been iden ified o be associa ed wi h o he ST-T pheno ypes (see Table 1), sugges ing ha he e a e 29 independen signals wi hin he 28 loci. Some loci we e p edominan ly associa ed wi h ST-segmen ampli udes (locus 4, 5, 16, 26 and 28) while o he s we e p edomin- an ly associa ed wi h T-wa e ampli udes (locus 6, 18 and 29). The es s a is ics o e e y combina ion o geno ype–pheno ype asso- cia ion a e summa ized in Supplemen a y Ma e ial, Table S6. O he elec oca diog aphic ai s P e ious genome-wide associa ions ha e s udied o he ECG indi- ces (1,19–23) such as QRS du a ion, PR in e al, hea a e and QT in e al. We in e sec ed hese p e iously iden ified loci wi h ou ST-T-wa e loci and ound o e lap (wi hin 2 MB) in 13 loci. Lead SNPs om p e ious and cu en findings in 4 loci we e in low LD ( 2 < 0.02), sugges ing ha 19 o he cu en loci a e no el asso- cia ions wi h he ECG (Supplemen a y Ma e ial, Table S7). He i abili y es ima es The 28 iden ified sen inel SNPs collec i ely explained be ween 1.6% (T-wa e sep al) and 5.1% (ST-segmen aVR) o he obse ed pheno ypic a iance (Supplemen a y Ma e ial, Table S8). Famil- ial he i abili y es ima es in he E asmus Rucphen Family (ERF) s udy a ied be ween h 2 = 30% (T-wa e in e io ) and h 2 = 42% (ST sep al) sugges ing addi ional gene ic a ian s and o he me- chanisms ha emain o be disco e ed (Supplemen a y Ma e ial, Table S9). The a ia ion in p opo ion o a iance associa ed wi h co a ia es a ied mo e widely; be ween 2 = 0.09 (ST-segmen in- e io ) and 2 = 0.36 (ST-segmen sep al). Iden ifica ion o candida e genes and pa hway analyses In o al, we p io i ized 54 candida e genes in he 28 gene ic loci (Table 1) ha could play a causal ole in ST-T-wa e ampli udes based on se e al c i e ia, as desc ibed in he ‘Ma e ials and Figu e 1. (A) We conduc ed genome-wide analyses o ST-segmen ampli udes a 80 ms a e J-poin and T-wa e ( op) ampli udes eflec ing empo al pa e ns in he ca diac cycle du ing he epola iza ion phase. In o al, 12 pheno ypes we e defined by aking he sum o he ST-T-wa e ampli udes in he la e al (I, aVL, V 5 and V 6 ), in e io (II, III and aVF), sep al (V 1 and V 2 ), an e io (V 3 and V 4 ) leads and lead aVR. These lead g oups co e he combina ion o leads wi h p esumed ana omical meaning o he hea , and hose ha a e used in he clinic o diagnosing B ugada synd ome (17)andER(16–18). (B) Genome-wide associa ion analyses o all ST-T wa e ai s iden ified 71 significan geno ype–pheno ype associa ions in 28 gene ic loci (2 MB). The x-axis ep esen s he ch omosomal posi ion o each SNP, which was assigned he lowes P- alue ac oss he 12 ai s; he y-axis ep esen s he −log10 o he P- alue o associa ion. Twen y-se en loci we e significan o ST-T-wa e ampli udes. (C) Fou genes o e lapping ou loci a e di ec ly in ol ed in he ca diac ac ion po en ial; SNPs nea SCN5A we e associa ed wi h ST-segmen and T- op ampli udes whe eas he loci con aining po assium channel-coding genes KCND3,KCNA7 and KCNB1 we e p ima ily associa ed wi h ST-segmen ampli udes. 2095 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Oc obe 19, 2016h p://hmg.ox o djou nals.o g/Downloaded om |Human Molecula Gene ics, 2016, Vol. 25, No. 10 Table 1. All sen inel SNP associa ions pe ST-T wa e pheno ype No. Region SNP T ai LD ( 2 ) wi h Sen inel SNP (†) in Locus EA (FRQ)/NEA β(SE) P me a NGene 1 1p36.33 s260505 T.an †A (0.49)/C 0.05 (0.01) 1.30E-11 37775 SKI n 2 1p36.12 s2072944 T.sep 0.37 T (0.55)/C 0.05 (0.01) 1.70E-13 37821 LUZP1 n , KDM1A n , WNT4 e 2 1p36.12 s2298632 ST.sep †T (0.47)/C −0.07 (0.01) 4.40E-18 37753 TCEA3 ne 3 1p31.3 s2207792 T.aVR –A (0.66)/G −0.06 (0.01) 3.30E-13 37799 NFIA nd 3 1p31.3 s2207792 T.in †A (0.66)/G 0.07 (0.01) 3.00E-20 37810 NFIA nd 4 1p13.2 s12145374 ST.an –A (0.76)/C 0.11 (0.01) 5.90E-33 37753 KCND3 ng , FAM212B d 4 1p13.2 s12145374 ST.sep –A (0.76)/C 0.07 (0.01) 1.20E-14 37753 KCND3 ng , FAM212B d 4 1p13.2 s12145374 ST.aVR –A (0.76)/C −0.12 (0.01) 1.40E-31 37784 KCND3 ng , FAM212B d 4 1p13.2 s12145374 ST.la †A (0.76)/C 0.12 (0.01) 8.20E-34 37771 KCND3 ng , FAM212B d 4 1p13.2 s12145374 ST.in –A (0.76)/C 0.09 (0.01) 6.00E-18 37770 KCND3 ng , FAM212B d 5 1q22 s10908505 ST.aVR †A (0.66)/T −0.05 (0.01) 1.70E-11 37784 MEF2D ngd 6 1q23.3 s12567315 T.an 0.42 A (0.19)/G −0.06 (0.01) 4.60E-11 37775 NOS1AP n 6 1q23.3 s4657178 T.in †T (0.24)/C −0.06 (0.01) 1.90E-11 37810 NOS1AP n 7 1q24.2 s545833 ST.an †T (0.26)/C −0.06 (0.01) 7.00E-18 37753 DPT n 7 1q24.2 s511187 ST.in 0.79 A (0.72)/G 0.06 (0.01) 2.60E-12 37770 DPT n 7 1q24.2 s511187 ST.aVR 0.79 A (0.72)/G −0.07 (0.01) 3.10E-17 37784 DPT n 7 1q24.2 s511187 ST.la 0.79 A (0.72)/G 0.07 (0.01) 7.20E-18 37771 DPT n 8 2p15 s7576036 ST.sep †T (0.54)/C 0.05 (0.01) 2.80E-10 37753 XPO1 n 9 2q33.1 s1866666 T.aVR †T (0.49)/C −0.05 (0.01) 1.80E-11 37799 PLCL1 nced , MARS2 e , RFTN2 e , MOB4 d 10 3p25.1 s4684185 T.aVR †T (0.35)/C −0.05 (0.01) 7.10E-10 37799 LSM3 n , TMEM43 d 11 3p22.2 s7638909 ST.sep 0.02 T (0.76)/G −0.06 (0.01) 2.00E-11 37753 SCN5A nd , ACVR2B d 11 3p22.2 s9851710* T.in 0.03 A (0.66)/C 0.08 (0.01) 9.30E-19 37810 SCN10A n , SCN5A d , ACVR2B d 11 3p22.2 s7633988 T.an 0.02 A (0.71)/T 0.06 (0.01) 3.80E-11 37775 SCN10A n , SCN5A d , ACVR2B d 11 3p22.2 s6783110 T.la 0.94 A (0.42)/G 0.06 (0.01) 3.70E-12 30699 SCN10A nc , SCN5A d , ACVR2B d 11 3p22.2 s4076737 T.aVR 0.93 T (0.57)/G 0.07 (0.01) 1.20E-20 37799 SCN10A nc , SCN5A d , ACVR2B d 11 3p22.2 s4076737 ST.in 0.93 T(0.57)/G −0.07 (0.01) 8.30E-20 37770 SCN10A nc , SCN5A d , ACVR2B d 11 3p22.2 s6801957 ST.aVR †T (0.43)/C −0.08 (0.01) 1.10E-25 37784 SCN10A nc , SCN5A d , ACVR2B d 11 3p22.2 s6801957 ST.la –T (0.43)/C 0.06 (0.01) 8.50E-16 37771 SCN10A nc , SCN5A d , ACVR2B d 12 6p21.31 s7756236 ST.an †A (0.76)/G −0.06 (0.01) 1.30E-14 37753 CDKN1A ned 13 6q22.2 s210966 ST.la –C (0.28)/G 0.07 (0.01) 5.10E-16 37771 ROS1 n , VGLL2 d 13 6q22.2 s210966 ST.aVR †C (0.28)/G −0.07 (0.01) 4.00E-18 37784 ROS1 n , VGLL2 d 13 6q22.2 s172409 ST.in 0.41 A (0.54)/C −0.05 (0.01) 6.40E-11 30708 ROS1 n , VGLL2 d 14 6q22.31 s9388451 ST.an †T (0.52)/C −0.06 (0.01) 2.00E-18 37753 HEY2 ned 14 6q22.31 s9388451 T.an –T (0.52)/C −0.05 (0.01) 3.60E-13 37775 HEY2 ned 15 8p23.1 s1458942 ST.aVR †A (0.60)/G 0.06 (0.01) 1.10E-13 37784 TNKS nd , SGK223 d , XKR6 d , PPP1R3B d 16 8p23.1 s7011924 ST.aVR †A (0.40)/G 0.07 (0.01) 1.30E-19 37784 DEFB136 n , NEIL2 e 16 8p23.1 s7011924 ST.in –A (0.40)/G −0.07 (0.01) 3.90E-19 37770 DEFB136 n , NEIL2 e 17 12p13.32 s7953399 ST.sep 0.95 A (0.61)/G −0.04 (0.01) 2.10E-09 37753 GALNT8 n 17 12p13.32 s2286582 ST.an †T (0.40)/C 0.06 (0.01) 2.40E-16 37753 GALNT8 n 18 12p12.1 s10842350 T.la –A (0.54)/G −0.09 (0.01) 4.60E-37 37759 SOX5 ng 18 12p12.1 s10842350 T.aVR †A (0.54)/G 0.09 (0.01) 1.30E-37 37799 SOX5 ng 19 12q24.21 s10850409 T.sep †A (0.27)/G 0.07 (0.01) 4.30E-17 34881 TBX3 n 20 13q22.1 s728926 T.la †T (0.38)/C −0.07 (0.01) 2.00E-18 37759 KLF12 n 20 13q22.1 s728926 T.aVR –T (0.38)/C 0.06 (0.01) 1.50E-16 37799 KLF12 n 21 15q26.3 s7174918 ST.an 0.74 T (0.29)/C −0.06 (0.01) 6.90E-16 37753 IGF1R nd 21 15q26.3 s3803476 T.sep –A (0.37)/G −0.05 (0.01) 2.00E-14 37821 IGF1R nd 21 15q26.3 s3803476 ST.sep †A (0.37)/G −0.07 (0.01) 1.80E-23 37753 IGF1R nd 22 16p13.3 s7192150 T.aVR †A (0.40)/G 0.05 (0.01) 4.30E-13 37799 LMF1 ne , SOX8 gd 23 16p13.13 s7191330 T.aVR 0.95 T (0.46)/C −0.08 (0.01) 3.10E-16 30731 LITAF n 23 16p13.13 s735951 ST.an –A (0.47)/G 0.06 (0.01) 9.50E-17 37753 LITAF n 23 16p13.13 s735951 ST.aVR –A (0.47)/G −0.07 (0.01) 1.70E-18 37784 LITAF n 23 16p13.13 s735951 ST.la †A (0.47)/G 0.07 (0.01) 1.20E-18 37771 LITAF n 24 16q21 s4784938 ST.la 0.7 A (0.74)/G 0.08 (0.01) 4.80E-21 37771 GINS3 n 24 16q21 s4784939 ST.an –T (0.36)/C −0.09 (0.01) 4.80E-41 37753 GINS3 n 24 16q21 s4784939 T.an –T (0.36)/C −0.06 (0.01) 1.00E-17 37775 GINS3 n 24 16q21 s4784939 T.sep –T (0.36)/C −0.08 (0.01) 4.10E-29 37821 GINS3 n 24 16q21 s4784939 ST.sep †T (0.36)/C −0.11 (0.01) 3.00E-57 37753 GINS3 n 24 16q21 s8057901 T.aVR 0.93 A (0.63)/G −0.06 (0.01) 4.90E-11 30731 NDRG4 n 24 16q21 s8057901 ST.aVR 0.93 A (0.63)/G −0.08 (0.01) 1.60E-21 30713 NDRG4 n Table con inues 2096 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Oc obe 19, 2016h p://hmg.ox o djou nals.o g/Downloaded om Human Molecula Gene ics, 2016, Vol. 25, No. 10 | Me hods’sec ion. Thi y- ou genes we e p io i ized based on he p oximi y c i e ia, 5 genes con ained one o mo e non- synonymous SNPs in high LD wi h a lead SNP (Supplemen a y Ma e ial, Table S10), 12 genes had a op-eQTL SNP in high LD wi h a lead SNP (Supplemen a y Ma e ial, Table S11), 65 genes we e selec ed by he da a-d i en exp ession p io i ized in eg a- ion o complex ai (DEPICT) analysis o which 25 we e wi hin genome-wide significan loci (a alse disco e y a e ≤5%; Sup- plemen a y Ma e ial, Table S12) and 3 genes a e based on li e a- u e mining using GRAIL (Supplemen a y Ma e ial, Table S13). The op fi e keywo ds e ie ed om GRAIL we e ‘muscle’, ‘channel’,‘ ansc ip ion’,‘hea ’and ‘channels’. The DEPICT me hod (24) was applied o subsequen ly o e in- sigh in o he biological pa hways unde lying changes in ST-T- wa e ampli udes o he ECG. DEPICT iden ified 56 econs i u ed gene se s, which g ouped in o 9 dis inc clus e s. The clus e con- aining he mos significan ly en iched econs i u ed gene se (‘VCL p o ein complex’,ap o ein–p o ein in e ac ion ne wo k cen e ed on he Vinculin gene p oduc , P=6.00×10 −6 ) ep e- sen ed a ious p o ein complexes likely ela ed o Fascia Adhe- ens, while he second mos significan gene clus e ‘Abno mal Ca dio ascula Sys em Physiology’con ains econs i u ed gene se known o be impo an o ca diac epola iza ion (Fig. 2and Supplemen a y Ma e ial, Table S14). Role o egula o y DNA and 1000 genomes impu a ion ST-T-wa e associa ed SNPs we e 3- old en iched in Dnase1 Hype sensi i i y si es (DHSs) om human e al hea (Fig. 2), which was significan ly highe compa ed wi h 337 o he cell ypes and issues (P= 0.0004, Z-sco e s a is ics) in line wi h ea lie obse a ions (25,26). We hen pe o med fine mapping o all 71 genome-wide significan geno ype–pheno ype associa ions (Table 1) in p e en ion o enal and ascula end-s age disease (P e end) and Li elines impu ed wi h 1000 genomes (1000G), co e ing mo e o he known common a ian s in humans. We obse ed ha 64 (o 71) associa ions we e assigned o a di e en lead SNP, ha 63 o hese associa ions we e mo e significan in he 1000G impu ed da a and e ec es ima es we e highe o 58 o hese associa ions (Supplemen a y Ma e ial, Table S16). To a- cili a e u u e unc ional expe imen s owa d he iden ifica ion o causal a ian s and hei unde lying biological mechanisms, we p io i ized po en ial causal SNPs using he p obabilis ic amewo k o P obabilis ic Anno a ion INTeg a OR (PAINTOR) (27). PAINTOR de e mined he significance o each anno a ion (Fig. 2) and used he fi e mos significan anno a ions (conse a- ion sco es, DHS o e al hea , enhance s o he le en icle and Tbx3 bound egions) o p io i ize po en ial causal SNPs in he 28 loci (27). This yielded 315 SNPs in he 99% confidence se and 96 SNPs in he 95% confidence se (Supplemen a y Ma e ial, Table S17). Discussion Ampli udes o he ST segmen and T wa e a e impo an ai s ha a e associa ed wi h abno mal hea hy hm, conduc ion dis- u bances and en icula a hy hmias. In his s udy, we pe - o med he fi s GWAS o ST-T-wa e ampli udes o he ECG in up o 37 977 indi iduals. T ai s we e defined acco ding o he combina ion o ECG leads p esumed o ha e ana omical meaning o used in he clinic o diagnosing B ugada synd ome and ea ly epola iza ion (16–18). ST-T-wa e ai s we e mode a ely he i - able (h 2 =30–42%) while he p opo ion o a iance associa ed wi h co a ia es (gende , body mass index and age) a ied consid- e ably (9–36%) s ongly suppo ing he gene ic backg ound o hese ai s. We iden ified 28 genome-wide significan loci o ST-T-wa e ampli udes and a se o 54 candida e genes. A ecen GWAS on B ugada synd ome e ealed he associ- a ion o wo loci ha a e sha ed by ou ST-T-wa e loci (28). One o hese signals in he SCN5A/SCN10A locus is in comple e LD wi h ou sen inel SNP ( s10428132, 2 = 0.96 wi h s6801957) and s9388451 nea HEY2, is a ST-T-wa e sen inel SNP; sugges ing ha B ugada synd ome suscep ibili y loci sha e a common gen- e ic backg ound wi h ST-T-wa e ai s. One o he s onges asso- cia ed GWAS signals o all ampli udes o he ST segmen was in he KCND3 gene (locus 4). KCND3 encodes he K 4.3 α-subuni ha conduc s he ca diac as ansien ou wa d K+ cu en (I To, ). This cu en is p ominen in Phase 1 o he ac ion po en ial Table 1. Con inued No. Region SNP T ai LD ( 2 ) wi h Sen inel SNP (†) in Locus EA (FRQ)/NEA β(SE) P me a NGene 24 16q21 s8057901 ST.in 0.93 A (0.63)/G 0.07 (0.01) 1.30E-16 30708 NDRG4 n 24 16q21 s9940062 T.in 0.94 T (0.68)/G 0.07 (0.01) 1.20E-10 30738 NDRG4 n 25 18q12.1 s8083566 ST.la –A (0.07)/C −0.09 (0.01) 4.30E-10 37771 CDH2 nd 25 18q12.1 s8083566 ST.in –A (0.07)/C −0.11 (0.02) 1.70E-12 37770 CDH2 nd 25 18q12.1 s8083566 ST.aVR †A (0.07)/C 0.12 (0.01) 3.60E-15 37784 CDH2 nd 26 19q13.33 s11673003 ST.la †A (0.90)/G −0.09 (0.01) 5.10E-14 37771 KCNA7 nc , NTF4 n , GYS1 d , HRC d 27 20q11.22 s6087666 ST.in †A (0.20)/G −0.07 (0.01) 3.10E-15 37770 TRPC4AP ne , EDEM2 ce , MYH7B cd ,NCOA d 27 20q11.22 s6088738 ST.aVR 0.55 A (0.21)/G 0.06 (0.01) 5.90E-13 37784 EDEM2 ne , PROCR e , MYH7B cd , NCOA d 28 20q13.13 s11907908 ST.an 0.83 T (0.07)/C −0.12 (0.01) 8.00E-18 37753 ZNFX1 (−AS1) n , STAU1 d 28 20q13.13 s6019750 ST.la 1 C (0.93)/G 0.13 (0.01) 4.30E-23 37771 KCNB1 n ,STAU1 d 28 20q13.13 s2202261 ST.in –A (0.93)/G 0.12 (0.01) 4.40E-18 37770 KCNB1 n ,STAU1 d 28 20q13.13 s2202261 ST.aVR †A (0.93)/G −0.14 (0.01) 5.60E-26 37784 KCNB1 n ,STAU1 d The e a e 71 genome-wide significan geno ype–pheno ype associa ions clus e ed in 28 gene ic loci (based on a 2 MB locus defini ion, see he ‘Ma e ials and Me hods’ sec ion). Gene supe sc ip s indica e he me hod o iden ifica ion: n, nea es gene o nea by gene (wi hin 10 kb o he SNP); g, G ail’d, Depic ; e, eQTL; c, coding SNP (non-synonymous). s9851710 deno ed by an as e isk (*) indica es ha his SNP is an independen , seconda y, associa ion o ST.aVR ( s9851710 condi ioned on s6801957: β(SE) pe mino allele = 0.13 (0.0204), P= 2.38 × 10 −10 ). 2097 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Oc obe 19, 2016h p://hmg.ox o djou nals.o g/Downloaded om |Human Molecula Gene ics, 2016, Vol. 25, No. 10 and con ibu es o he ‘no ch’o he ca diomyocy e’sac ion po en ial (2). Mu a ions in KCND3 ha e been implica ed wi h in- c eased isk o sudden ca diac dea h (29,30), as well as mu a ions in SCN5A (2)andSCN10A (31). We specifically iden ified loci con aining po assium channels KCND3 (locus 4), KCNB1 (locus 28) and KCNA7 (locus 26) o be p edominan ly associa ed wi h ampli udes o he ST segmen and no wi h he T wa e, sugges - ing ha hese po assium channels a e ac i a ed du ing he ST Figu e 2. (A) DEPICT iden ified 56 significan ly en iched gene-se s ele an o ST-T-wa e ampli udes. (B) SNPs we e significan ly mo e en iched in DHSs o e al hea issue (n= 12) compa ed wi h o he issue and cells (n= 337), ac oss he ull ange o P- alues o he disco e y me a-analyses (genome wide), sugges ing ha unc ionali y o egula o y DNA elemen s may unde lie some o he associa ions. (Cand D) Nex , we pe o med a me a-analysis o he 28 iden ified ST-T wa e loci using 1000 Genomes impu ed da a, o his 1000G a ian s needed o be in LD 2 > 0.1 wi h he HapMap sen inel SNP. Subsequen p io i iza ion o po en ial causal anno a ions in hese loci also sugges ed ha egions o DHS in e al hea a e possibly unde lying he associa ions as well as ca diac ansc ip ion ac o s, conse ed egions (exonic), ac i e and weak enhance s. While egions ha a e ansc ibed, igh ly packed (he e och oma in) o unc ion as p omo e s in he en icles may be less impo an o he biological mechanisms o gene ic a ian s ha a e associa ed wi h ST-T-wa e ampli udes. Sub le di e ence a e p esen be ween he en icles and e al hea which could sugges ha p omo e s ha o e lap po en ial causal ST-T wa e SNPs may be ac i e in he e al hea bu ep essed in he en icles. Pe cen ages be ween pa en heses indica e he amoun o SNPs in he 28 loci o e lapping wi h he anno a ion. Conse a ion (GERP and 29 mammals), DHS o e al hea , enhance s o he le en icle and Tbx3 bound egions we e used o p io i ize po en ial causal SNPs in he 28 loci. 2098 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Oc obe 19, 2016h p://hmg.ox o djou nals.o g/Downloaded om Human Molecula Gene ics, 2016, Vol. 25, No. 10 | segmen o he epola iza ion phase and may ha e clinical ele- ance (16,32,33). Fu u e s udies a e equi ed o examine he po en ial ole o he o he ST-T-wa eSNPs o inc eased a hy h- mogenic isk and sudden dea h h ough an al e ed ca diac epola iza ion. In addi ion o HEY2, a ansc ip ional egula o o ca diac elec- ical unc ion in he igh en icula ou flow ac (28), we iden- ified six loci con aining genes wi h s ong e idence o being di ec ly in ol ed in ca diac issue de elopmen ia ansc ip ion- al egula o y pa hways. Locus 5 con ains a myocy e enhance ac o -2 (MEF2D) impo an o ca diac muscle mo phogenesis and hea looping (34,35). Loci 18 and 22 con ain wo ansc ip- ion ac o s om he SOX amily, SOX5 and SOX8. SOX5 is impo - an o a unc ioning hea and in ol ed in co ec Wn signaling (36). This locus has p e iously been associa ed wi h he PR in e - al and es ing hea a e, bu he iden ified a ian in he cu en s udy is in low LD ( 2 = 0.01), sugges ing mul iple molecula -gen- e ic mechanisms a his locus o influence hea unc ion. T an- sc ip s o SOX8 a e concen a ed in he subendo helial mesenchym o he whole ou flow ac (37) in which N-Cadhe in (38)(encodedbyCDH2, locus 25), he ansc ip ion ac o Tbx3 (39)(TBX3, locus 19) and WNT4 (40) (specific o he endoca dial endo helial cushion de elopmen , pa o he Wn pa hway) in locus 2 also play essen ial de elopmen al oles, among o he ca diac cell lineages. O he compelling ca diac genes a e: SKI, o which mu a ions (1p36 dele ion- and Sp in zen–Goldbe g synd ome) a e cha ac e ized by hea de ec s, b ain abno mal- i ies and muscle one (hypo onia) in in ancy (41); TMEM43 (locus 10) in which mu a ions cause a hy hmogenic igh en icula ca diomyopa hy and Eme y–D ei uss muscula dys- ophy; he sa come ic gene MYH7B and miR499 (locus 27), which egula es a mul i ude o ca diac mRNA and mic oRNAs and p omo es en icula specifica ion; and GALNT1, a glycosyl- ans e ases ha is equi ed o no mal hea al e de elopmen and ca diac unc ion by egula ing he ex acellula ma ix and al e ing conse ed signaling pa hways ha egula e cell p oli e - a ion du inghea de elopmen (42). We also iden ified a numbe o less well-cha ac e ized genes, which should be explo ed in u he esea ch. Fi e o he 28 ST-T-wa e sen inel SNPs we e in LD wi h SNPs associa ed wi h QT du a ion (1), sugges ing ha ST-T-wa e ampli- udes p o ides addi ional in o ma ion on ca diac epola iza ion. The known unc ions o he candida e genes and ou pa hway analyses sugges ha he ST-T-wa e ampli udes a e influenced and egula ed by a wide a ie y o molecula mechanisms, includ- ing ion-channels, s uc u al p o eins and ca diac ansc ip ion ac o s. Howe e , he finding ha p o ein complexes ela ed o Fascia Adhe ens a e mos en iched in ou pa hway analyses hin s ha some o he biology unde lying ST-T wa es is less well cap u ed in well-es ablished pa hways and be e ep esen ed by da a-d i en (and no manually cu a ed) pa hways. Ven icula epola iza ion is a complex p ocess. To cap u e his p ocess, we chose o apply a selec ion o composi e ECG pa ame e s wi h he aim o gaining mo e insigh in o he bio- logical p ocesses unde lying he ST-T wa e. These su ace ECG pa ame e s a e no specific enough o iden i y he exac ana om- ical egion o he hea ; e.g. he sep al leads (V 1 and V 2 ) may also include aspec s o he igh en icula wall ac i i y. Howe e , he e a e no da a a ailable o mo e p ecise measu emen s such as could be de i ed om mo e sophis ica ed su ace ECG equip- men o in a-ca diac ECG measu emen s. In summa y, we p esen a la ge numbe o genome-wide significan loci obus ly associa ed wi h ca diac epola iza ion pa ame e s, some wi h compelling biological basis o hei associa ion and a numbe o loci no p e iously implica ed in ca - diac unc ion. The iden ified loci and selec ed genes ha e he po- en ial o aid u u e s udies ha a e ocused on isk s a ifica ion o on molecula mechanisms unde lying ca diac epola iza ion and diseases o ca diac epola iza ion; o acili a e hese s udies, we ha e made ou esul s (including he genome-wide associ- a ion) publicly a ailable. Ma e ials and Me hods S udy popula ions The disco e y phase included pa icipan s o he P e end s udy and he Li elines s udy coho . Bo h a e communi y-based coho s udies om he no he n pa o he Ne he lands. The eplica- ion phase included pa icipan s o he A he oscle osis Risk in Communi ies s udy (ARIC), Ca dio ascula Heal h S udy (CHS), Young Finns S udy (YFS), ERF s udy, Ro e dam s udy I, II and III (RS I, II and III). De ailed desc ip ions o each coho a e gi en in he supplemen a y ma e ial. Cha ac e is ics o pa ici- pan s a e summa ized in Supplemen a y Ma e ial, Tables S1 and S5. De ailed In o ma ion on geno yping me hods, quali y con ol o SNPs, impu a ion and s a is ical analysis o each coho is summa ized in Supplemen a y Ma e ial, Table S18. Pheno ype modeling Summing co ela ed a iables when expec ing ha gene ic a - ian s a e associa ed o mul iple o hese a iables will inc ease he powe o de ec ion; wi h gene ic a ian (Snp), and ai s (Ampi ude i ), conside he co ela ion: Co (Snp, Ampi ude 1 + Ampi ude 2 )=Co (Snp, Ampi ude 1 )+Co (Snp, Ampi ude 2 )+2×Co (Ampi ude 1 , Ampi ude 2 )>Co (Snp, Ampi ude 1 )+Co (Snp, Ampi ude 2 ), hence i Co (Ampi ude 1 , Ampi ude 2 )>0, he ai s Ampi ude i meas- u e he same la en a iable (L). The e o e, pheno ypes o he ST- segmen ampli udes 80 ms a e J-poin and T-wa e ampli udes we e defined by aking he sum o he la e al (I, aVL, V 5 and V 6 ), in- e io (II, III and aVF), sep al (V 1 and V 2 ), an e io (V 3 and V 4 )leads and lead aVR. Indi iduals we e excluded o bundleb anch blocko QRS du a ion> 120 ms, a ial fib illa ion, flu e , his o y o myoca - dial in a c ion o elec onic pacemake hy hm and when a ail- able, hea ailu e and ECG al e ing medica ion. Also pa icipan s wi h ex eme measu emen s (mo e han ±4SD om mean) we e excluded on a pe pheno ype basis. S a is ical analyses To con ol o mul iple es ing o he 10 pheno ypes while accoun ing o he co ela ion be ween hem, we pe o med an eigen alue decomposi ion o he co ela ion ma ix (Supplemen- a y Ma e ial, Table S2) o he pheno ypes. The a iance o he eigen alues [Va (λ obs ) = 2.36] was used o es ima e he e ec i e numbe o independen pheno ypes es ed (43). Ou findings indica e ha s udying 10- ela ed ST-T-wa e ai s is equi alen o analysis o 8 independen pheno ypes. We, he e o e, adop α a =5×10 −8 /8 = 6.25 × 10 −9 as h eshold o decla ing genome- wide significance in o de o co ec o he e ec i e numbe o independen pheno ypes s udied. Residuals o ST-T-wa e ampli udes we e calcula ed using gene al linea eg ession models o adjus o age, gende and body mass index and s anda dized o a mean o ze o and a s and- a d de ia ion o one. GWAS analyses in P e end and Li elines o 2 316 136 geno yped o impu ed SNPs (Supplemen a y Ma e ial, Table S18) we e pe o med on he s anda dized esiduals using an addi i e gene ic model in PLINK ( .1.07). Pheno ypes we e 2099 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Oc obe 19, 2016h p://hmg.ox o djou nals.o g/Downloaded om |Human Molecula Gene ics, 2016, Vol. 25, No. 10 no mally dis ibu ed (Supplemen a y Ma e ial, Fig. S2). Tes s a- is ics om each coho we e hen co ec ed o hei espec i e genomic con ol infla ion ac o o adjus o esidual popula ion sub-s uc u e and me a-analyzed using he in e se- a iance me hod. SNPs wi h mino allele equency < 1% (weigh ed a e - age ac oss coho s) we e emo ed. Va ian s we e conside ed o be independen i he pai -wise LD ( 2 )was<0.1andi hey we e sepa a ed by a leas 1 MB; his was defined a ‘locus’. We se- lec ed one sen inel SNP ( he mos -significan ly associa ed SNP in a locus) o each geno ype–pheno ype combina ion. Fo condi ional analyses, we epea ed he p ima y associa ion analysis o each ai while condi ioning on he ai -specific genome-wide significan sen inel SNPs by adding he SNP geno- ypes as co a ia es. Associa ion esul s o each s udy we e again combined by in e se a iance weigh ing. Replica ion o he significan geno ype–pheno ype associa- ions (P<6.25×10 −9 ) and associa ions ha did no exceed his h eshold, bu we e sugges i e (6.25 × 10 −9 <P<1 × 10 −6 ), was pe - o med in he RS I, II and III, ERF, ARIC, CHS and YFS and com- bined using fixed-e ec s me a-analysis by in e se a iance weigh ing (Supplemen a y Ma e ial, Table S18). The p e-specified s a is ical significance h eshold o he e ogenei y by Coch an’s Q(P he )wasP he < 0.0007 o accoun o mul iple es ing. In e se a iance weigh ing was used o de e mine me a-P- alues o combined disco e y and eplica ion da a. An associa ion was conside ed eplica ed i he di ec ion o e ec was conco dan wi h disco e y, eplica ion P< 0.01 and me a-P< 6.25 × 10 −9 . He i abili y es ima es we e calcula ed in he ERF s udy using he ‘ dis ’ unc ion in he SOLAR so wa e, including gende , age and body mass index as co a ia es. Da a-d i en exp ession p io i ized in eg a ion o complex ai s DEPICT sys ema ically iden ifies he mos likely causal gene a a gi en associa ed locus, es s gene se s o en ichmen in asso- cia ed SNPs, and iden ifies issues and cell ypes in which genes om associa ed loci a e highly exp essed [see Pe s e al. (24) o a de ailed desc ip ion]. Fo his wo k, we an DEPICT on 140 independen ly associa ed loci (associa ion P<10 −5 ;PLINK pa ame e s, ‘–clump-p1 1e-5 –clump-kb 500 –clump- 2 0.05’) e- sul ing in 103 independen , au osomal DEPICT loci con aining 363 genes (loci o e lapping wi h he majo his ocompa ibili y complex egion a e by de aul excluded in DEPICT). We ha e ex ended he locus defini ion used in DEPICT (LD 2 >0.5)wi h 100 kb a ei he side o he loci, because se e al genes ha may be impo an o ca diac epola iza ion we e ou side he de aul DEPICT locus bounda ies (e.g. SCN5A,TMEM43 and VGLL2). The gene se en ichmen esul s o his sligh ly ex ended locus defini ion we e simila o he esul s based on he de aul locus defini ion used in DEPICT (Supplemen a y Ma e ial, Table S19). Iden ifica ion o candida e genes We p io i ized candida e genes based on nea by genes: we consid- e ed he nea es gene and any o he gene loca ed wi hin 10 kb o he sen inel SNP. Coding a ian s: o iden ifica ion o coding a ian s and LD calcula ions, we used he 1000G P ojec da a se (Ma ch 2012 elease) in he Eu opean popula ions. We consid- e ed genes ha ha bo non-synonymous SNPs in LD wi h he ST-T-wa e SNPs a 2 > 0.8. GRAIL analyses: we ca ied ou a li e a u e analysis by employing he GRAIL ex -mining algo- i hm, a s a is ical ool ha iden ifies subse s o genes wi h known unc ional in e ela ionships based on PubMed abs ac s. We ca ied ou he analysis using he 2006 da a se o a oid con- ounding by subsequen GWAS disco e y. The DEPICT me hod (see abo e), and exp ession QTL (eQTL) analyses in cis, we sea ch o eQTLs (sen inel SNPs o SNPs in LD, 2 >0.8,HapMap 27)in an eQTL da a se ha was compiled om he summa y s a is ics o a ious s udies and issues (see Supplemen a y Ma e ial, Table S11). We only conside ed eQTLs ha we e in LD ( 2 > 0.8, HapMap 27) wi h he sen inel SNP and eached a P- alue cu -o o a leas P< 0.05/(28 loci × 17 eQTL s udies). Impu a ion using 1000G Genome posi ions om he P e end and Li elines geno ypes we e con e ed om hg18 o hg19 using he UCSC Li O e ool. Gen- ome-wide geno ype impu a ion was pe o med wi h SHAPEIT ( 2) and IMPUTE2 ( 2.3.0) using he comple e 1000 Genomes 3, Ma ch 2012 haplo ypes Phase I in eg a ed a ian se elease as e e ence panel. Func ional in o ma ion We o e lapped SNPs wi h da a om he ENCODE p ojec (44)and Roadmap Epigenomics P og am (45), conse a ion ac oss mam- mals, a ious ca diac ansc ip ion ac o measu ed by ChIP-Seq and con iguous anno a ions o he human e al hea ,le en icle and igh en icle as de e mined by Ch omHMM (46)(Supplemen- a y Ma e ial). P io i iza ion o po en ially causal a ian s and en ichmen o DNA elemen s Fo he p io i iza ion o gene ic a ian s o he u u e unc ional ollow-up and insigh o he unde lying DNA elemen s ha migh be ele an o causal ST-T-wa e ampli ude a ian s, we employed he PAINTOR amewo k (27). In sho , his me hod al- lows us o p io i ize gene ic a ian s in each o he 28 significan associa ed loci by an in eg a ing s eng h o associa ion and an- no a ions o unc ional DNA elemen s o es ima e he p obabili y o each a ian o be causal, bu also in es iga e which anno a- ions a e po en ially causal. Supplemen a y Ma e ial Supplemen a y Ma e ial is a ailable a HMG online. Acknowledgemen s We hank Beh ooz Alizadeh, Annemieke Boesjes, Ma cel B uinen- be g, Noo je Fes en, Ilja Nol e, Lude F anke, Mi a Valimohammadi o hei help in c ea ing he GWAS da abase, and Rob Bie inga, Joos Kee s, René Oos e go, Rosalie Visse , Judi h Vonk o hei wo k ela ed o da a-collec ion and alida ion. The au ho s a e g a e ul o he s udy pa icipan s, he s a om he Li eLines Co- ho S udy and Medical Biobank No he n Ne he lands, and he pa icipa ing gene al p ac i ione s and pha macis s. Li eLines Scien- ific P o ocol P epa a ion: Rudol de Boe , Hans Hillege, Melanie an de Klauw, Ge jan Na is, Hans O mel, Di kje Pos ma, Judi h Rosma- len, Jo is Slae s, Ronald S olk, B uce Wol enbu el; Li eLines GWAS Wo king G oup: Beh ooz Alizadeh, Ma ike Boezen, Ma cel B uinen- be g, Noo je Fes en, Lude F anke, Pim an de Ha s , Ge jan Na is, Di kje Pos ma, Ha old Sniede , Cisca Wijmenga and B uce Wol enbu el. The au ho s acknowledge he se ices o he Li e- Lines Coho S udy, he con ibu ing esea ch cen es deli e ing da a o Li eLines and all he s udy pa icipan s. 2100 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Oc obe 19, 2016h p://hmg.ox o djou nals.o g/Downloaded om Human Molecula Gene ics, 2016, Vol. 25, No. 10 | Conflic o In e es s a emen . None decla ed. Funding PREVEND: PREVEND gene ics is suppo ed by he Du ch Kidney Founda ion (g an E033), he EU P ojec g an GENECURE (FP-6 LSHM CT 2006 037697), he Na ional Ins i u es o Heal h (g an 2R01LM010098), The Ne he lands O ganiza ion o Heal h Resea ch and De elopmen (NWO-G oo g an 175.010.2007.006, NWO VENI g an 916.761.70, ZonMw g an 90.700.441) and he Du ch In e Uni e si y Ca diology Ins i u e Ne he lands (ICIN). N.V. is suppo ed by he Ne he lands Hea Founda ion (g an NHS2010B280). Li eLines: The Li eLines Coho S udy, and gene - a ion and managemen o GWAS geno ype da a o he Li eLines Coho S udy is suppo ed by he Ne he lands O ganiza ion o Scien ific Resea ch NWO (g an 175.010.2007.006), he Economic S uc u e Enhancing Fund (FES) o he Du ch go e nmen , he Minis y o Economic A ai s, he Minis y o Educa ion, Cul u e and Science, he Minis y o Heal h, Wel a e and Spo s, he No he n Ne he lands Collabo a ion o P o inces (SNN), he P o ince o G oningen, Uni e si y Medical Cen e G oningen, he Uni e si y o G oningen, Du ch Kidney Founda ion and Du ch Diabe es Resea ch Founda ion. L.F. is suppo ed by he Ne he lands O ganiza ion o Scien ific Resea ch (NWO VENI g an 916.10.135) and a Ho izon B eak h ough g an om he Ne he lands Genomics Ini ia i e (g an 92519031. The esea ch leading o hese esul s has ecei ed unding om he Eu opean Communi y’s Heal h Se en h F amewo k P og amme (FP7/ 2007-2013) unde g an ag eemen 259867. RS I, II and III: Akzo Nobel, Alzheime ’s Associa ion, As a Pha maceu ial N.V., As a- Zeneca, Baye AG, Blinden-penning Founda ion, Ams e dam, B ain Founda ion o he Ne he lands, B is ol-Mye s Squibb, Cen e o Medical Sys emsBiology (CMSB), Du ch Diabe es Resea ch Foun- da ion, Du ch Kidney Founda ion, Du ch A h i is Associa ion, Elise Ma hilde Founda ion, Maa n, E asmus Medical Cen e , E as- mus Uni e si y Ro e dam, Eu opean Commission, Founda ion o Helping he Blind, The Hague, Founda ion o he Oph halmic Diseased, Ro e dam, Founda ion G. Ph. Ve hagen, Gene al Elec ic Heal hca e, Glaxo Smi h Kline, In e na ional Founda ion Alzhei- me ’s Resea ch, Inspec o a e o Heal hca e, Jani o Founda ion, K.F. Hein Founda ion, Me ck Sha p and Dohme, Haa lem, Munici- pali y o Ci y o Ro e dam, Na ional Epilepsy Fund, Na ional Heal h Fund aising Founda ion, Na ional Ins i u e on Aging, NIH, Be hesda, MD, USA, Na ional Socie y o he Blind and Visually Im- pai ed (LSBS), Ne he lands Founda ion o he Blind and Visually Handicapped, Ne he lands Hea Founda ion, Ne he lands Ins i- u e o Heal h Sciences(Nihes),Ne he landsOph halmicResea ch Ins i u e, Ne he lands O ganisa ion o Heal h Resea ch and De elopmen (ZonMw),Ne he landsO ganisa ion o Scien ificRe- sea ch (NWO), Ne he lands Socie y o he P e en ion o Blindness, Ne he lands Th ombosis Founda ion, No o No disk, Numico Re- sea ch B.V., OOG Founda ion, The Hague, N.V. O ganon, Oxagen, Op imix Founda ion, Ams e dam, Physico he apeu ic Ins i u e, P inses Bea ix Founda ion, P oc e and Gamble, Resea chIns i u e o Diseases in he Elde ly (RIDE), Ro e dam Founda ion o Oph- halmic Resea ch, Ro e dam Founda ion o he In e es s o he Blind, S Lau ens Ins i u e, Ro e dam, Topcon Eu ope B.V., T us Fund E asmus Uni e si y Ro e dam, Unile e , Van Leeuwen Van Lignac Founda ion, Ro e dam. ERF: The ERF s udy was suppo ed by g an s om he Ne he lands O ganiza ion o Scien ificRe- sea ch (NWO; Pionee g an ), E asmus Medical Cen e , he Cen e o Medical Sys ems Biology (CMSB) and he Ne he lands Kidney Founda ion. We a e g a e ul o all pa ien s and hei ela i es, gen- e al p ac i ione s and neu ologis s o hei con ibu ions and o P. Ve aa o he help in genealogy, Jeanne e Ve gee o he supe ision o he labo a o y wo k and P. Snijde s o his help in da a collec ion. ARIC: The A he oscle osis Risk in Communi ies S udy is ca ied ou as a collabo a i e s udy suppo ed by Na ional Hea , Lung and Blood Ins i u e con ac s (HHSN268201100005C, HHSN268201100006C, HHSN268201100007C, HHSN268201100008C, HHSN268201100009C, HHSN268201100010C, HHSN268201100011C and HHSN268201100012C), R01HL087641, R01HL59367 and R01HL086694; Na ional Human Genome Resea ch Ins i u e con- ac U01HG004402 and Na ional Ins i u es o Heal h con ac HHSN268200625226C. The au ho s hank he s a and pa ici- pan s o he ARIC s udy o hei impo an con ibu ions. In a- s uc u e was pa ly suppo ed by g an numbe UL1RR025005, a componen o he Na ional Ins i u es o Heal h and NIH Roadmap o Medical Resea ch. CHS: Ca dio ascula Heal h S udy: his CHS esea ch was suppo ed by NHLBI con ac s HHSN268201200036C, HHSN268200800007C, N01HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083, N01HC85086 and NHLBI g an s HL080295, HL087652, HL105756, HL103612 and HL120393 wi h addi ional con ibu ion om he Na ional Ins i u e o Neu o- logical Diso de s and S oke (NINDS). Addi ional suppo was p o- ided h ough AG023629 om he Na ional Ins i u e on Aging (NIA). A ull lis o p incipal CHS in es iga o s and ins i u ions can be ound a CHS-NHLBI.o g/. The p o ision o geno yping da a was suppo ed in pa by he Na ional Cen e o Ad ancing T ansla ional Sciences, CTSI g an UL1TR000124 and he Na ional Ins i u e o Diabe es and Diges i e and Kidney Disease Diabe es Resea ch Cen e (DRC) g an DK063491 o he Sou he n Cali o nia Diabe es Endoc inology Resea ch Cen e . The con en is solely he esponsibili y o he au ho s and does no necessa ily ep esen he o ficial iews o he Na ional Ins i u es o Heal h. The Young Finns S udy has been financially suppo ed by he Academy o Finland: g an s 286284 (T.L.), 134309 (Eye), 126925, 121584, 124282, 129378 (Sal e), 117787 (Gendi) and 41071 (Skidi); he Social Insu ance Ins i u ion o Finland; Kuopio, Tampe e and Tu ku Uni- e si y Hospi al Medical Funds (g an X51001 o T.L.); Juho Vainio Founda ion; Paa o Nu mi Founda ion; Finnish Founda ion o Ca - dio ascula Resea ch (T.L.); Finnish Cul u al Founda ion; Tampe e Tube culosis Founda ion (T.L.); Emil Aal onen Founda ion (T.L.) and Y jö Jahnsson Founda ion (T.L.). Funding o pay he Open Access publica ion cha ges o his a icle was p o ided by he Ma ie Sklodowska-Cu ie GF g an (661395) om N. Ve weij. Re e ences 1. A king, D.E., Puli , S.L., C o i, L., an de Ha s , P., Mun oe, P.B., Koopmann, T.T., So oodehnia, N., Rossin, E.J., Mo ley, M., Wang, X. e al. (2014) Gene ic associa ion s udy o QT in e al highligh s ole o calcium signaling pa hways in myoca dial epola iza ion. Na . Gene .,46, 826–836. 2. Ne bonne, J.M. and Kass, R.S. (2005) Molecula physiology o ca diac epola iza ion. Physiol. Re .,85, 1205–1253. 3. An zele i ch, C. (2006) Cellula basis o he epola iza ion wa es o he ECG. Ann. N. Y. Acad. Sci.,1080, 268–281. 4. Goldbe ge , A.L. and Mi is, D.M. (2011) In Bonow, R.O., Mann, D.L., Zipes, D.P. and Libby, P. (eds), B aunwald’s Hea Disease: A Tex book o Ca dio ascula Medicine. Else ie Heal h Sciences, in p ess, US, Philadelphia, pp. 126–167. 5. Hasan, M.A., Abbo , D. and Baume , M. (2013) Bea - o-bea QT in e al a iabili y and T-wa e ampli ude in pa ien s wi h myoca dial in a c ion. Physiol. Meas.,34, 1075–1083. 6. Al-Zai i, S.S., Falla olli a, J.A., Can y, J.M. and Ca ey, M.G. (2015) The p ognos ic alue o disco dan T wa es in lead aVR: a 2101 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Oc obe 19, 2016h p://hmg.ox o djou nals.o g/Downloaded om