Upstream Transcription Factor 1 (USF1) allelic variants regulate lipoprotein metabolism in women and USF1 expression in atherosclerotic plaque

Ups eam T ansc ip ion Fac o 1 (USF1) allelic a ian s egula e lipop o ein me abolism in women and USF1 exp ession in a he oscle o ic plaque Yue-Mei Fan 1 , Jussi He nesniemi 1 , Niku Oksala 1,9 , Ma i Le ula 1 , Emma Rai oha ju 1 , Auni Collings 1 , Nina Hu i-Ka ¨ho ¨nen 2 , Ma kus Juonala 3 , Jukka Ma niemi 4 , Leo-Pekka Lyy ika ¨inen 1 , Ilkka Seppa ¨la ¨ 1 , A iMennande 5 , Ma i Ta kka 5 , An i J. Kangas 6,7 ,PasiSoininen 6,7 ,JuhaPekkaSalenius 9 , No man Klopp 10 , Thomas Illig 10 , Tomi Lai inen 11 , Mika Ala-Ko pela 6,7,8,12 , Reijo Laaksonen 1 , Jo ma Viika i 13 , Mika Ka ¨ho ¨nen 14 , Olli T. Rai aka i 15 & Te ho Leh ima ¨ki 1 1 Depa men o Clinical Chemis y, Fimlab Labo a o ies, Tampe e Uni e si y Hospi al and School o Medicine a he Uni e si y o Tampe e, Tampe e, 2 Depa men o Pedia ics, Uni e si y o Tampe e and Tampe e Uni e si y Hospi al, Tampe e, 3 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, 4 Depa men o Heal h and Func ional Capaci y, Popula ion Resea ch Labo a o y, Na ional Public Heal h Ins i u e, 5 Hea Cen e , Depa men o Ca diac Su ge y, Tampe e Uni e si y Hospi al, Finland, 6 Compu a ional Medicine, Ins i u e o Heal h Sciences, Uni e si y o Oulu, Oulu, Finland, 7 NMR Me abolomics Labo a o y, School o Pha macy, Uni e si y o Eas e n Finland, Kuopio, Finland, 8 Oulu Uni e si y Hospi al, Oulu, Finland, 9 Di ison o Vascula Su ge y, Depa men o Su ge y, Tampe e Uni e si y Hospi al and Uni e si y o Tampe e, Finland, 10 Ins i u e o Epidemiology, Helmhol z Cen e Munich, Munich, 85764, Ge many, 11 Depa men o Clinical Physiology and Nuclea Medicine, Kuopio Uni e si y Hospi al and Uni e si y o Eas e n Finland, Finland, 12 Compu a ional Medicine, School o Social and Communi y Medicine & Medical Resea ch Council In eg a i e Epidemiology Uni , Uni e si y o B is ol, B is ol, Uni ed Kingdom, 13 Depa men o Medicine, Uni e si y o Tu ku and Di ision o Medicine, Tu ku Uni e si y Hospi al, Tu ku, Finland, 14 Depa men o Clinical Physiology, Uni e si y o Tampe e and Tampe e Uni e si y Hospi al, Tampe e, 15 Resea ch Cen e o Applied and P e en i e Ca dio ascula Medicine, Uni e si y o Tu ku and Depa men o Clinical Physiology, Uni e si y o Tu ku and Tu ku Uni e si y Hospi al, Tu ku, Finland. Ups eam ansc ip ion ac o 1 ( USF1 ) allelic a ian s signi ican ly in luence u u e isk o ca dio ascula disease and o e all mo ali y in emales. We in es iga ed sex-speci ic e ec s o USF1 gene allelic a ian s on se um indices o lipop o ein me abolism, ea ly ma ke s o asymp oma ic a he oscle osis and hei changes du ing six yea s o ollow-up. In addi ion, we in es iga ed he cis - egula o y ole o hese USF1 a ian s in a e y wall issues in Caucasians. In he Ca dio ascula Risk in Young Finns S udy, 1,608 pa icipan s (56% women, aged 31.9 64.9) wi h lipids and cIMT da a we e included. Fo unc ional s udy, whole genome mRNA exp ession p o iling was pe o med in 91 his ologically classi ied a he oscle o ic samples. In emales, se um o al, LDL choles e ol and apoB le els inc eased g adually acco ding o USF1 s2516839 geno ypes TT ,CT ,CC and s1556259 AA ,AG ,GG as well as acco ding o USF1 H3 (GCCCGG) copy numbe 0 ,1,2. Fu he mo e, he ca ie s o mino alleles o s2516839 (C) and s1556259 (G) o USF1 gene had dec eased USF1 exp ession in a he oscle o ic plaques (P 50.028 and 0.08, espec i ely) as compa ed o non-ca ie s. The gene ic a ia ion in USF1 in luence USF1 ansc ip exp ession in ad anced a he oscle osis and egula es le els and me abolism o ci cula ing apoB and apoB-con aining lipop o ein pa icles in sex-dependen manne , bu is no a majo de e minan o ea ly ma ke s o a he oscle osis. The ups eam ansc ip ion ac o 1 (USF1) is a ubiqui ously exp essed ansc ip ion ac o egula ing an- sc ip ion o many genes om lipid and glucose me abolism pa hways 1 . The USF1 con ains a helix-loop-helix mo i , which binds an E-box mo i in he p omo e egion o i s a ge genes and leads o ansc ip ion ac i a ion and/o enhanced gene exp ession 2 . The e a e e y ew da a ega ding USF1 ansc ip le els in human issues. In wo s udies o adipose issue, USF1 ansc ip le els did no di e in ela ion o USF1 alleles 1,3 . Acco ding o ou knowledge USF1 exp ession o he e ec o i s gene ic a ia ion on gene exp ession in a e y wall has no been epo ed in any p e ious in es iga ions. OPEN SUBJECT AREAS: GENE EXPRESSION GENETIC ASSOCIATION STUDY Recei ed 20 Janua y 2014 Accep ed 26 Ma ch 2014 Published 11 Ap il 2014 Co espondence and eques s o ma e ials should be add essed o Y.-M.F. ([email p o ec ed]) SCIENTIFIC REPORTS | 4 : 4650 | DOI: 10.1038/s ep04650 1 The USF1 gene is localized on ch omosome 1q23, consis s o 11 exons and ex ends o 6.73 kb (Na ional Cen e o Bio echnology (NCBI), gene ID: 7391). Ini ially his gene was localized in 1998 4 and hen iden i ied as he i s amilial combined hype lipidemia (FCHL) gene in a e Finnish pedig ee wi h mul iple a ec ed indi i- duals ha ing a g ea ly inc eased isk o ca dio ascula disease (CVD) 1 . This inding was apidly eplica ed in Mexican amilies 5 . Since hen USF1 and i s gene ic a ia ion has also been associa ed wi h he me abolic synd ome and ype II diabe es 6–8 . In ea lie s udies we showed ha some o he USF1 gene a ian s a e associa ed wi h he su ace a ea o a he oscle o ic lesions measu ed di ec ly om co ona ies a e au opsy in men 9 and epo ed ha wo USF1 SNPs ( s3737787 and s2516838) and USF1 haplo ype a e associa ed wi h ca o id in ima-media hickness (cIMT) 10 . In a s udy wi h wo la ge p ospec i e Finnish coho s, he e ec o USF1 allelic a ian s on CVD isk and o e all mo ali y was seen in emales only 11 . Gi en he di e ences in CVD e en a e, li e-expec - ancy and mo ali y be ween men and women, he gende -speci ic e ec s o USF1 a e o ob ious in e es . Ano he s udy pe o med using mouse models showed ha o e -exp ession o human USF1 in mice in luenced me abolic ai pheno ypes in sex-dependen manne 12 . The e a e also ea lie exis ing sex-speci ic esul s conce n- ing USF1 polymo phisms and lipids pa ame e s 1,13,14 . Any di e ence in he USF1 gene e ec s on CVD isk ac o s o lipop o ein me abo- lisms be ween men and women may p o ide u he aluable insigh in o he biology o he inc eased suscep ibili y o women o CVDs. These p e ious esul s gi e jus i ied a ionale o addi ional sex-spe- ci ic analyses. Ea ly indices o a he oscle osis namely cIMT, ca o id a e y dis- ensibili y (Cdis ), and b achial a e y low-media ed dila a ion (FMD) ha e all been shown o p edic u u e ca dio ascula e en s 15–20 . The p e ious epo s o USF1 alleles on hese ma ke s a e inconclusi e in he sex-speci ic e ec s 10 . In he p esen s udy, we used he ongoing p ospec i e Ca dio- ascula Risk in Young Finns S udy (YFS) ollow-up (2001 and 2007) and Tampe e Vascula S udy (TVS) ma e ials 21–23 in o de o in es iga e he e ec s o USF1 gene allelic a ian s on se um indices o lipop o ein me abolism, ea ly ma ke s o asymp oma ic a he o- scle osis and hei changes du ing six yea s ollow-up sepa a ely in men and women. Since he di ec con ibu ion o hese USF1 gene a ian s on he gene unc ion in a e y wall issues has no been p e iously es ed we also in es iga ed he cis- egula o y ole o s udied USF1 alleles in a e y wall issues. Resul s Cha ac e is ics o YFS and TVS subjec s.Table 1 shows he cha ac e is ics o 1608 s udy subjec s in YFS in 2001. O e all, men had mo e un a o able ca dio ascula isk ac o p o iles compa ed wi h women. In addi ion, men had highe cIMT alues, lowe b achial a e y FMD, as well as lowe Cdis . Supplemen a y Table 1 shows he clinical cha ac e is ics o 91 pa ien s in TVS. Case and con ol subjec s had simila isk ac o p o iles wi h an excep ion ha he cases had lowe BMI han he con ols. USF1 polymo phisms and haplo ypes in YFS.The USF1 allele and haplo ype equencies oge he wi h he six s udied SNPs a e shown in Table 2. A o al o i e common USF1 haplo ypes we e iden i ied, accoun ing o 97.9% o all a ia ion in he USF1 gene. The e we e no s a is ically signi ican di e ences in haplo ype equencies be ween men and women (da a no shown). All USF1 geno ype dis ibu ions we e in acco dance wi h he Ha dy-Weinbe g equilib ium o he en i e popula ion and he subg oups di ided by sex (da a no shown). USF1 polymo phisms, haplo ypes and hei sex in e ac ions wi h se um lipid and apolipop o ein measu emen s.The e we e s a is ically signi ican geno ype ( s2516839, s1556259) and haplo ype (H3) di e ences by sex in e ac ion in ela ion o se um o al- and LDL-choles e ol as well as apoB le els in 2007 (Table 3). In emales, se um o al-, LDL-choles e ol and apoB le els inc eased g adually acco ding o USF1 s2516839 geno ype TT ,TC ,CC and s1556259 AA ,AG ,GG as well as acco ding o USF1 haplo ype 3 (H3, GCCCGG) copy numbe 0 ,1,2, cons an ly in bo h 2001 and 2007 (Table 3). The e was a simila end o bo h yea s 2001 and 2007 o he e ec o hese USF1 polymo phisms and haplo ypes, al hough he end ended o be s a is ically s onge and consis en in 2007 a aged 30–45, when he subjec s we e an a e age 6-yea olde han in 2001 (aged 24–39 yea s) (Table 3). The e we e also o he associa ions be ween s udied USF1 poly- mo phisms, haplo ypes and some o he classical lipid ma ke s in 2001 and in 2007 (see Supplemen a y Tables 2–5), bu hese associa- ions we e no as consis en as hey we e o o al-, LDL-choles e ol and apoB le els. The e o e, his s udy was ocused on mo e de ailed second s age analyses (by using p o on-NMR spec oscopy) on mainly o hose lipop o ein subclass measu es wi h mos consis en gene ic associa ions in a sex-speci ic manne . In emales, bo h s2516839 and s1556259 geno ypes associa ed signi ican ly wi h he longi udinal change om 2001 o 2007 in LDL choles e ol alues (P 50.044 and P 50.009, espec i ely, epea ed- measu emen RANCOVA main e ec o geno ype, age and BMI a 2001 as co a ia es). Simila ly, he USF1 s1556259 geno ypes assoc- ia ed signi ican ly wi h he longi udinal change in o al choles e ol le els (P 50.021). The emales ca ying he haplo ype 3 had signi i- can ly highe o al- and LDL choles e ol concen a ions h oughou he six-yea ollow-up pe iod compa ed wi h nonca ie s o his haplo ype (P 50.008 and P 50.006, espec i ely). These e ec s we e sex-speci ic and we e no ound in males. USF1 polymo phisms, haplo ypes and lipid me abolism a subclass le el in women.To a oid unnecessa y mul iple es ing he h ee mos signi ican gene ic ma ke s associa ed wi h classical lipids ( esul s om abo e), s2516839, s1556259 and haplo ype H3 we e selec ed o mo e de ailed and ocused wo s age s a is ical analysis in ela ion o selec ed indices o choles e ol and lipid me abolism o e di e en lipop o ein subclasses in women. We in es iga ed he e ec s o USF1 geno ypes and haplo ypes o e se um lipop o ein subclass choles e ol ac ions ( o al choles e ol, choles e ol es e , and ee choles e ol when a ailable) and o al Table 1 | Cha ac e is ics o he Ca dio ascula Risk in Young Finns S udy popula ion in 2001 Va iable men women No. o subjec s 706 902 Age, yea s 31.9 65.0 31.9 64.9 Body mass index, kg/m 2 25.6 63.8 24.2 64.3 Sys olic blood p essu e, mm Hg 129 614 116 612 Dias olic blood p essu e, mm Hg 75 697169 To al choles e ol, mmol/L 5.22 60.98 5.03 60.87 LDL choles e ol, mmol/L 3.42 60.88 3.13 60.75 HDL choles e ol, mmol/L 1.17 60.27 1.40 60.30 T iglyce ides, mmol/L 1.42 60.82 1.13 60.53 Apolipop o ein A1, g/L 1.40 60.21 1.56 60.26 Apolipop o ein B, g/L 1.12 60.26 0.99 60.23 Glucose, mmol/L 5.21 60.87 4.90 60.72 Insulin, mU/L 7.46 65.88 7.61 65.63 CRP, mg/L 1.43 63.42 2.18 64.33 Daily smoking, % 46.6 36.9 IMT 2001, mm 0.59 60.10 0.57 60.09 FMD 2001, % 6.83 64.05 8.75 64.49 Cdis 2001, %/10 mmHg 2.02 60.67 2.33 60.77 Values a e mean 6SD o pe cen age o subjec s. All compa isons ( es s) be ween men and women P,0.001, excep o age and insulin (P.0.6). www.na u e.com/scien i ic epo s SCIENTIFIC REPORTS | 4 : 4650 | DOI: 10.1038/s ep04650 2 lipid concen a ions in women. We show he associa ions be ween he abo e pa ame e s and h ee gene ic ma ke s in women in 2007 in Figu e 1. The e we e signi ican geno ype and haplo ype e ec s on he o al choles e ol le el o small VLDL, o al lipid le el o e y small VLDL, o al choles e ol and lipid le els and ee choles e ol le el o IDL, and all he h ee subclasses o LDL in 2007 (Figu e 1 and Supplemen a y Table 6). These esul s we e eplicable and he e was a simila end o bo h ollow-up yea s 2001 and 2007 o he e ec o hese polymo phisms and haplo ypes, al hough he end ended o be s a is ically s onge in 2007 a aged 30–45 (Figu e 1 and Supplemen a y Table 6) han in 2001 (aged 24–39 yea s) (Supple- men a y Table 7). These geno ype and haplo ype e ec s a e consis- en wi h he esul s om he classical lipid measu emen s. USF1 polymo phisms, haplo ypes, and ma ke s o subclinical a he oscle osis.In emales, he USF1 s2516839 geno ypes asso- cia ed signi ican ly wi h he cIMT in longi udinal analysis (P 5 0.012, epea ed-measu emen ANOVA main e ec o geno ype). Those wi h CC geno ype had highes cIMT in bo h 2001 and 2007. Howe e , his e ec disappea ed a e adjus ing o classical isk ac o s o CAD and again was no seen in males. We did no ind any signi ican di e ences in cIMT, b achial a e y FMD, o Cdis be ween o he USF1 SNP geno ype g oups, haplo ype g oups in longi udinal analyses ei he among men o among women (da a no shown). USF1 exp ession in he a he oscle o ic issue.We compa ed he USF1 mRNA exp ession examined wi h GWEA be ween a he o- scle o ic plaque samples and non-a he oscle o ic in e nal ho acic a e ies, as well as USF1 exp ession be ween di e en essel ypes. The USF1 exp ession le el (iso o m 2) was signi ican ly lowe in a he oscle o ic plaque specimens (N 568) han in con ol issue (N 523, P 50.049, Mann-Whi ney U es ). The USF1 gene exp ession was signi ican ly educed in he ca o id plaques (N 5 29, P 50.05, Mann-Whi ney U es ), bu no in he emo al plaques (N 524, P 50.10, Mann-Whi ney U es ) and ao a (N 515, P 50.32, Mann-Whi ney U es ) al hough hey ollowed he same end. The e was no di e ence in he USF1 exp ession be ween women and men. Rela ion o USF1 polymo phisms and haplo ypes o exp ession o USF1 and known USF1 a ge genes.Six y-nine subjec s (aged 40– 91 yea s, males 71%) had comple e da a conce ning he six USF1 SNPs and gene exp ession da a. The six s udied USF1 SNPs o med i e majo haplo ypes. These haplo ypes accoun ed o 97% o all a ia ion in he USF1 gene. Because no in e ac ion be ween he SNP o haplo ype and subjec s a us (case s. con ol) was ound o be signi ican , we p oceeded o analyze he e ec o he haplo ypes o SNPs wi hin all samples in o de o inc ease he s a is ical powe . We ound ha mino homozygo es (CC and GG ca ie s) o bo h SNPs, s2516839 and s1556259, had lowe USF1 exp ession han he T allele and A allele ca ie s (P 50.028 and P 5 0.08, espec i ely, Mann-Whi ney U es ). The GCCCGG haplo ype was he only haplo ype ca ying bo h he C allele o he s2516839 polymo phism and G allele o he s1556259 polymo phism. The ca ie s o he abo e haplo ype ended o ha e lowe USF1 exp ession han he non-ca ie s (P 50.06, Mann-Whi ney U es ). O he 47 known USF1 a ge genes 24,25 , he chemokine (C-X-C mo i ) ecep o 4 (CXCR4) was up- egula ed in CC ca ie s o s2516839 (P 50.013, Mann-Whi ney U es ) and haplo ype GCCCGG ca ie s (P 50.04, Mann-Whi ney U es ). The hemo- globin be a (HBB) was down- egula ed in CC ca ie s o s2516839 (P 50.01, Mann-Whi ney U es ) and haplo ype GCCCGG ca ie s (P 50.05, Mann-Whi ney U es ). Discussion We ound a emale-speci ic associa ion o USF1 a ian s and haplo- ype wi h se um le els o bo h o al lipids and lipop o ein subclasses in YFS. Fu he mo e, we ound he associa ion o hese USF1 a ian s and haplo ype wi h USF1 exp ession in he ca o id a e y plaque in TVS. To ou knowledge, his s udy is he i s showing ha he USF1 gene exp ession is down- egula ed in a he oscle o ic lesions and USF1 a ian s (SNP s2516839 and s1556259) a e associa ed wi h he down- egula ion. The esul s o he p esen s udy imply ha USF1 a ian s a e likely o ha e causal e ec s due o he gene ic in luence o he a ian s on gene exp ession. The biological impo ance o he USF1 gene has been implied in p e ious s udies, which we e mos ly conduc ed on s udy subjec s wi h speci ic selec ion c i e ia, such as p esence o FCHL 1,5 , CVD 14 , diabe es 7,26 , me abolic synd ome 26 o obesi y 27 . The e we e li le di ec da a om in i o s udies as o USF1 ansc ip le els in human issues. In wo s udies o adipose issue, USF1 ansc ip le els did no di e in ela ion o USF1 alleles 1,3 . One s udy showed clea sex- ela ed di e ences in ai exp ession in mouse models and he gene exp ession pa e ns we e also no ably di e en be ween he sexes 12 . Ou esul s a e consis en wi h s udies done using emale mice, in which he en ichmen o me abolism- ela ed ca ego ies was demons a ed 12 . The associa ion be ween he USF1 a ian s and lipid le els a e inconsis en wi h p e ious publica ions. Ou indings a e in ag ee- men wi h a popula ion based da a se om he Finns ha he mino allele (C) o s2516839 was associa ed wi h inc eased lipid alues among s udy subjec s wi h ca dio ascula disease 11 . Con adic o y o s udies whe e he common allele T o s2516839 associa ed wi h a mo e un a o able isk p o ile 6,26,28,29 , he mino allele C o s2516839 o ou sample associa ed wi h inc eased LDL choles e ol le els. The majo alleles o s2516839 and s1556259 we e associa ed wi h Table 2 | De ails o he loci and haplo ypes in he USF1 gene Us 1s1 Us 1s8 Us 1-4530 Us 1s7 Us 1s9 Us 1-81 dbSNP ID s3737787 s2516838 s10908821 s2516839 s1556259 s2774276 Allele G/AC/GC/GT/CA/GC/G MAF 0.357 0.271 0.136 0.371 0.131 0.233 haplo ype F equency (%) H1 AC C TAG 35.4 H2 G GC TAG 26.3 H3 G C C C G G 13.2 H4 G C G C A C 13.1 H5 G C C C A C 10.5 MAF, mino allele equency. The mino allele o each SNP is unde lined. www.na u e.com/scien i ic epo s SCIENTIFIC REPORTS | 4 : 4650 | DOI: 10.1038/s ep04650 3 Table 3 | The associa ion o se um o al- and LDL-choles e ol and apolipop o ein B le els wi h USF1 geno ypes and haplo ype 3 (H3) acco ding o gende and ollow-up yea . The Ca dio ascula Risk in Young Finns S udy Follow-up Yea SNP o haplo ype Geno ypeo copy o haplo ype N (women) Mean 6SD P N (men) Mean 6SD P N (all) Mean 6SD P To al choles e ol (mmol/L) 2001 s2516839 TT 349 5.01 60.85 0.67 282 5.27 61.00 0.536 631 5.12 60.93 0.82 TC 419 5.04 60.89 326 5.18 60.94 745 5.10 60.91 CC 123 5.09 60.90 96 5.22 61.05 219 5.14 60.97 s1556259 AA 665 5.00 60.86 0.041 536 5.23 60.99 0.848 1201 5.10 60.93 0.335 AG 218 5.10 60.90 150 5.18 60.95 368 5.13 60.92 GG 12 5.57 61.03 13 5.19 60.65 25 5.37 60.86 H3 0 670 5.00 60.86 0.047 543 5.23 60.99 0.882 1213 5.10 60.93 0.328 1 217 5.11 60.90 148 5.18 60.96 365 5.14 60.92 2 15 5.47 60.97 15 5.22 60.62 30 5.34 60.81 2007 s2516839 TT 349 4.81 60.80 0.013 282 5.20 60.91 0.392 631 4.98 60.87 0.254 TC 419 4.93 60.79 326 5.10 60.89 745 5.01 60.84 CC 123 5.03 60.87 96 5.18 60.90 219 5.10 60.89 s1556259 AA 665 4.85 60.80 0.009 536 5.16 60.91 0.443 1201 4.99 60.86 0.182 AG 218 5.02 60.82 150 5.07 60.85 368 5.04 60.83 GG 12 5.22 60.97 13 5.31 60.55 25 5.26 60.77 H3 0 670 4.85 60.80 0.006 543 5.17 60.92 0.336 1213 4.99 60.87 0.129 1 217 5.02 60.82 148 5.07 60.85 365 5.04 60.83 2 15 5.23 60.93 15 5.34 60.63 30 5.29 60.78 LDL choles e ol (mmol/L) 2001 s2516839 TT 349 3.10 60.76 0.57 282 3.45 60.88 0.567 631 3.26 60.84 0.644 TC 419 3.14 60.73 326 3.38 60.85 745 3.24 60.79 CC 123 3.18 60.78 96 3.46 60.96 219 3.30 60.87 s1556259 AA 665 3.11 60.74 0.011 536 3.42 60.88 0.809 1201 3.25 60.82 0.128 AG 218 3.18 60.76 150 3.37 60.90 368 3.26 60.83 GG 12 3.72 60.88 13 3.46 60.79 25 3.58 60.83 H3 0 670 3.10 60.73 0.015 543 3.43 60.87 0.716 1213 3.25 60.82 0.114 1 217 3.19 60.77 148 3.37 60.90 365 3.26 60.83 2 15 3.61 60.85 15 3.51 60.75 30 3.56 60.79 2007 s2516839 TT 349 2.85 60.69 0.002 282 3.30 60.80 0.619 631 3.05 60.77 0.068 TC 419 2.96 60.69 326 3.24 60.79 745 3.08 60.75 CC 123 3.10 60.73 96 3.31 60.79 219 3.19 60.76 s1556259 AA 665 2.89 60.68 0.003 536 3.27 60.79 0.276 1201 3.06 60.76 0.026 AG 218 3.05 60.74 150 3.22 60.78 368 3.12 60.76 GG 12 3.30 60.75 13 3.57 60.56 25 3.44 60.66 H3 0 670 2.89 60.68 0.003 543 3.28 60.80 0.160 1213 3.07 60.76 0.022 1 217 3.05 60.74 148 3.21 60.78 365 3.12 60.76 2 15 3.26 60.75 15 3.60 60.64 30 3.43 60.71 Apolipop o ein B (g/L) 2001 s2516839 TT 349 0.97 60.23 0.469 282 1.12 60.26 0.924 631 1.04 60.26 0.886 TC 419 0.99 60.23 326 1.12 60.24 745 1.05 60.24 CC 123 1.00 60.23 96 1.11 60.27 219 1.05 60.25 s1556259 AA 665 0.98 60.23 0.309 536 1.12 60.26 0.804 1201 1.04 60.25 0.352 AG 218 0.99 60.24 150 1.11 60.24 368 1.04 60.25 GG 12 1.08 60.25 13 1.15 60.21 25 1.11 60.23 H3 0 670 0.98 60.23 0.33 543 1.12 60.26 0.781 1213 1.04 60.25 0.36 1 217 1.00 60.24 148 1.11 60.25 365 1.04 60.25 2 15 1.06 60.23 15 1.16 60.20 30 1.11 60.22 www.na u e.com/scien i ic epo s SCIENTIFIC REPORTS | 4 : 4650 | DOI: 10.1038/s ep04650 4 inc eased iglyce ides and lowe HDL choles e ol le els in he Aus alian subjec s wi h documen ed CAD 28 . The common allele o he SNP associa ed wi h inc eased choles e ol and iglyce ide le els also in U ah amilies asce ained o ype 2 diabe es melli us 29 . In ano he s udy, he CC geno ype o SNP s2516839 showed sug- ges i e associa ion wi h dec eased isk o me abolic synd ome in Chinese hospi al cases 26 . The mino homozygo es o s1556259 had lowes LDL choles e ol le el in women 6 . Howe e , in hese s ud- ies, he e ec o USF1 was mo e p onounced in subjec s al eady diagnosed wi h CAD, FCHL, diabe es and me abolic synd ome. In ou s udy, he subjec s a e heal hy young Finns which migh explain he lack o associa ion o USF1 SNPs ma ke s wi h lipid measu es. The disc epancies ega ding he mino allele o s2516839 and s1556259 in lipid le els in ou s udy and o he s udies may be explained wi h he complica ed in e ac ions be ween USF1 a ian s and age (heal hy and qui e young in ou s udy), sex, o he genes, and en i onmen al ac o s. A he momen he e we e no unc ional s udies a ailable ha could explain hese di e ences. Thus, u he unc ional s udies o he e ec s o s2516839 as well as o he USF1 a ian s a e needed o sol e his disc epancy. We also ound an associa ion be ween USF1 s2516839 and cIMT du ing six yea s ollow-up in emale al hough his e ec disappea ed a e adjus ing o classical isk ac o s o CAD. The ca ie s o mino allele o s2516839 had highe mean cIMT alues. This is in ag ee- men wi h he same allele ca ie s ha ing highe LDL choles e ol le els. The sex- ela ed di e ences obse ed in ou s udy a e in line wi h p e ious s udies epo ing ha USF1 a ian s associa ed wi h CVD 11,30 and mo ali y 11 among women. Con a y o he p e ious s udies 1,9,14 , ou emale subjec s showed s onge associa ions han males, and he e ec s o s2516839 and s1556259 we e signi ican only in he emale subse . In a single s udy ha included only males 9 , he USF1 isk allele iden i ied was di e en om he isk allele seg ega ing in ou s udy. In 700 Finnish middle-aged men ( he Helsinki Sudden Dea h S udy, HSDS) 9 , he isk o ad anced a he o- scle o ic plaques, calci ica ion, and sudden ca diac dea h was assoc- ia ed wi h he common allele o s2516839, whe eas in he p esen s udy he mino allele o s2516839, was associa ed wi h ele a ed LDL choles e ol le el and cIMT. In e es ingly, in emale U.S. Whi es wi h CAD, he a e allele was associa ed wi h isk, whe eas in males he common allele o s3737787 con e ed isk 13 . Consis en wi h human s udies, he sex speci ici y was also obse ed in a mouse model 12 . The molecula mechanisms unde lying ou sex-speci ic allelic di e ence a e unknown. The ho monal ac o s may con ibu e o he di e ences. USF1 is a ubiqui ous ansc ip ion ac o ha egula es he exp ession o many genes in ol ed in lipid me abolism, immune esponse, endo helial unc ion and aging. I could con ibu e o he de elopmen o a he oscle osis and i s complica ions h ough many di e en pa hways. In ou s udy, we did no ind signi ican USF1 geno ype di e ence o gene exp ession in mos o he p e iously iden i ied USF1 a ge genes excep CXCR4 and HBB. P e ious s ud- ies ha e epo ed ha he CXCR4 le els a e inc eased in pa ien s wi h hea ailu e 31 and ha myoca dial CXCR4 le els a e inc eased 5- old in myoca dium subjec ed o ischemic inju y, compa ed wi h le els in non-inju ed myoca dium in he same hea 32 . O he s udies ha e also epo ed ha he hemoglobin le els a e independen ly associa ed wi h inc eased isk o new ca diac e en s 33,34 . Fu he s udies a e needed o unde s anding he mechanism behind ou indings be ween he USF1 a ian s and he up- egula ed CXCR4 and down- egula ed HBB exp ession. We did no ind any signi ican di e ence o USF1 exp ession be ween male and emale in TVS s udy. These exp ession esul s in TVS samples should be in e p e ed wi h cau ion due o small sample size and low numbe o women, and hus possible lack o s a is ical powe . I would be in e es ing o epea he obse a ion wi h la ge numbe s o subjec s. Table 3 | Con inued Follow-up Yea SNP o haplo ype Geno ypeo copy o haplo ype N (women) Mean 6SD P N (men) Mean 6SD P N (all) Mean 6SD P 2007 s2516839 TT 349 0.91 60.22 0.015 282 1.11 60.25 0.222 631 1.00 60.25 0.46 TC 419 0.94 60.22 326 1.08 60.24 745 1.00 60.24 CC 123 0.98 60.24 96 1.08 60.25 219 1.02 60.25 s1556259 AA 665 0.93 60.22 0.049 536 1.10 60.25 0.638 1201 1.00 60.25 0.388 AG 218 0.96 60.23 150 1.08 60.23 368 1.00 60.24 GG 12 1.05 60.24 13 1.09 60.20 25 1.07 60.22 H3 0 670 0.93 60.22 0.055 543 1.10 60.25 0.621 1213 1.00 60.25 0.421 1 217 0.96 60.23 148 1.08 60.23 365 1.00 60.24 2 15 1.03 60.23 15 1.09 60.19 30 1.06 60.21 Values a e mean 6SD. ANOVA. Signi ican alues wi hin he women s a a a e co ec ing o mul iple es ing (P #0.015) a e in bold. www.na u e.com/scien i ic epo s SCIENTIFIC REPORTS | 4 : 4650 | DOI: 10.1038/s ep04650 5 In conclusion, ou indings suppo he ole o USF1 a ian s and haplo ypes in lipid me abolism in gende -dependen manne and es ablish a new associa ion o USF1 a ian s wi h gene exp ession in ca o id a e ies. Fu he esea ch is needed on he associa ions be ween he USF1 SNPs, haplo ypes and ca dio ascula end poin s. Me hods S udy subjec s.The YFS is an ongoing i e-cen e , p ospec i e coho s udy o a he oscle osis isk ac o s unde lying ca dio ascula disease in child en and young adul s. De ails o he s udy design ha e been p esen ed elsewhe e 22 . In sho , he s udy was launched in 1980 and included 3,596 child en and adolescen s aged 3 o 18 yea s. In 2001, a o al o 2,283 pa icipan s aged 24–39 yea s we e e-examined, and in 2007, we examined 2,204 subjec s aged 30–45 yea s. The p esen s udy included 1,608 subjec s o whom comple e da a on lipids and cIMT in bo h yea 2001 and 2007 we e a ailable ( emale, N 5902; male, N 5706). All subjec s ga e a w i en in o med consen in 2001 and 2007. The s udy was app o ed by he local e hics commi ees. The me hods we e ca ied ou in acco dance wi h he app o ed guidelines. Mo e de ailed in o ma ion abou he coho s and he ollow-up p ocedu es can be ound in he coho desc ip ions o he P ojec (h p:// anha.med.u u. i/ca dio/young innss udy/index.h ml). Clinical cha ac e is ics.Weigh and heigh we e measu ed, and body mass index (BMI) was calcula ed. Blood p essu e was measu ed wi h a andom ze o sphygmomanome e . The a e age o h ee measu emen s was used in he analyses. In o ma ion on smoking, alcohol consump ion, and physical ac i i y was ob ained wi h a ques ionnai e. Those smoking on daily basis we e de ined as smoke s. Biochemical analyses.In 2001 and 2007, enous samples we e aken a e he subjec had as ed o 12 hou s. Se um o al choles e ol le els we e measu ed by he enzyma ic choles e ol es e ase – choles e ol oxidase me hod (Choles e ol eagen , Olympus, I eland). The same eagen was used o es ima ing HDL-choles e ol le els a e p ecipi a ion o apoB-con aining lipop o eins wi h dex an sul a e-Mg 21 . LDL- choles e ol was es ima ed by he F iedewald o mula 35 in subjec s wi h iglyce ides le els ,4.0 mmol/L. The se um iglyce ide concen a ion was assayed using he enzyma ic glyce ol kinase-glyce ol phospha e oxidase me hod (T iglyce ide eagen , Olympus). Se um glucose concen a ion was de e mined by he enzyma ic hexokinase me hod (Glucose eagen , Olympus). Apolipop o ein A1 (ApoA1) and B we e analysed immuno u bidome ically (O ion Diagnos ica, Espoo, Finland). The abo e men ioned analyses we e all pe o med on an AU400-analyze (Olympus, Japan). Se um insulin concen a ion was de e mined by a mic opa icle enzyme immunoassay (IMx insulin eagen , Abbo Diagnos ics, USA) on an IMx ins umen (Abbo ). The me hod has been desc ibed in mo e de ail elsewhe e 36 . Fas ing plasma high sensi i e C- eac i e p o ein (CRP) concen a ions we e analyzed by means o la ex u bidome ic immunoassay (Wako Chemicals GmbH, Neuss, Ge many). Measu emen s o ca o id a e y IMT and b achial a e y FMD.Ul asound examina ions we e pe o med using Sequoia 512 ul asound main ames (Acuson, CA, USA) wi h 13.0 MHz linea a ay ansduce s, as desc ibed p e iously in de ail 37 . Figu e 1 | Fo es plo s o he associa ions o USF1 polymo phisms ( s2516839, s1556259,), haplo ype 3 and se um lipop o ein subclass choles e ol ac ions ( o al choles e ol, choles e ol es e , and ee choles e ol when a ailable) and o al lipid concen a ions in women in 2007. Plo s show he associa ion es ima es (b) and 95% con idence in e als o he lipop o ein subclass le els p esen ed as ba s. C, o al choles e ol; CE, choles e ol es e ; FC, ee choles e ol; IDL, in e media e-densi y lipop o ein; HDL, high-densi y lipop o ein; L, o al lipids; LDL, low-densi y lipop o ein; VLDL, e y-low- densi y lipop o ein. www.na u e.com/scien i ic epo s SCIENTIFIC REPORTS | 4 : 4650 | DOI: 10.1038/s ep04650 6 In sho , o measu e ca o id IMT, he image was ocused on he pos e io ( a ) wall o he le ca o id a e y. A minimum o ou measu emen s o he common ca o id a wall was aken app oxima ely 10 mm p oximal o he bi u ca ion o de i e mean and maximal cIMT alues. To e alua e b achial a e y FMD, he le b achial a e y diame e was measu ed bo h a es and du ing eac i e hype emia, as desc ibed p e iously 38 . The essel diame e in scans a e eac i e hype emia was exp essed as a pe cen age ela i e o he es ing scan alue (%). To assess ca o id a e y elas ici y indices, he bes –quali y ca diac cycle was selec ed om he images and manually analyzed o measu e sys olic and dias olic common ca o id diame e s. The ca o id dis ensibili y (Cdis ) was hen calcula ed using ul asound and concomi an b achial blood p essu e measu emen s, as desc ibed p e iously 39 . These analyses we e pe o med o YFS subjec s in 2001 and in 2007. Lipop o ein subclass analysis by p o on-NMR spec oscopy.Concen a ions o lipop o ein subclasses as well as hei choles e ol con en we e analyzed by p o on NMR spec oscopy in na i e se um samples as desc ibed p e iously 40 . These se um subclasses we e classi ied as ollows: chylomic ons and ex emely la ge VLDL pa icles (a e age pa icle diame e a leas 75 nm); i e di e en VLDL subclasses: e y la ge VLDL (a e age pa icle diame e o 64.0 nm), la ge VLDL (53.5 nm), medium VLDL (44.5 nm), small VLDL (36.8 nm), and e y small VLDL (31.3 nm); in e media e-densi y lipop o ein (IDL) (28.6 nm); h ee LDL subclasses: la ge LDL (25.5 nm), medium LDL (23.0 nm), and small LDL (18.7 nm); and ou HDL subclasses: e y la ge HDL (14.3 nm), la ge HDL (12.1 nm), medium HDL (10.9 nm), and small HDL (8.7 nm) 40 . This me hodology has ecen ly been applied in a ious ex ensi e epidemiological and gene ics s udies 41–43 wi h consis en indings wi h espec o lipop o ein gene ics 44 . DNA ex ac ion and geno yping o he USF1 polymo phisms.Fo subjec s in YFS, DNA was ex ac ed om pe iphe al blood leukocy es using a comme cially a ailable ki (Qiagen Inc, Hilden, Ge many) in 2001. DNA samples we e geno yped by employing he 59nuclease assay o allelic disc imina ion, using he ABI P ism 7900 HT Sequence De ec ion Sys em (Applied Biosys ems, Fos e Ci y, CA). PCR eac ion con aining genomic DNA, 2 3TaqMan Uni e sal PCR Mas e Mix, 900 nM o each p ime , and 200 nM o each p obe was pe o med in 384-well pla es acco ding o s anda d p o ocol in a o al olume o 5 ml. Wa e con ols and known con ol samples we e un in pa allel wi h unknown samples. A e cycling, end-poin luo escence was measu ed, and USF1 geno ypes ( s3737787, s2073658, s2516838, s10908821, s2516839, s1556259, s2774279, s2774276) calling was ca ied ou by he allelic disc imina ion analysis module. The eigh SNPs we e selec ed om he HapMap da abase and Sea leSNPs da abase and p e ious publica ions. Two o he SNPs selec ed in he p elimina y phase, s2073658 and s2774279, we e le ou om u he analyses because hey we e in almos comple e linkage disequilib ium wi h s3737787 and s2516838, espec i ely. The linkage be ween s2516838 and s2516839 was weak ( 2 50.22). In TVS, genomic DNA was ex ac ed om pe iphe al blood leukocy es using QIAampHDNA Blood Miniki and au oma ed bio obo M48 ex ac ion (Qiagen, Hilden, Ge many). Whole genome geno yping using he Illumina HumanHap660W- Quad BeadChip (Illumina, Inc., San Diego, CA, USA) was ca ied ou om TVS samples acco ding o manu ac u es ecommenda ion as desc ibed in de ail in sup- plemen a y me hods. Vascula sample collec ion, RNA isola ion, exp ession analysis and quali y con ol.The a he oscle o ic plaque samples (male, N 546; emale, N 522) and non- a he oscle o ic con ol samples (male, N 520; emale, N 53, le in e nal ho acic a e ies, LITA) used in his s udy we e collec ed as pa o ongoing TVS 21,23,45–47 .The s udy was app o ed by he E hics Commi ee o Tampe e Uni e si y Hospi al, and he s udy subjec s ga e hei in o med consen . The clinical cha ac e is ics o TVS subjec s a e shown in supplemen a y Table 1 and de ails o sample collec ion in supplemen a y me hods. All open ascula su gical p ocedu es we e pe o med a he Di ision o Vascula Su ge y and con ol samples collec ed a Hea Cen e , Tampe e Uni e si y Hospi al. The s udy was app o ed by he E hics Commi ee o Tampe e Uni e si y Hospi al, and he s udy subjec s ga e hei in o med consen . The ascula samples we e classi ied acco ding o ecommenda ions o Ame ican Hea Associa ion (AHA) 48 by expe ienced pa hologis . All he in e nal ho acic a e y samples ha we e used as con ols we e e i ied o be mic oscopically heal hy. In TVS, ascula enda e ec omy samples cons i u ing he in ima and inne media om ca o id, emo al and ao ic egions we e ob ained. RNA isola ion, genome-wide exp ession analysis, RNA quali y con ol o hese a he oscle o ic plaque and non- a he oscle o ic con ol issue samples, was pe o med as p e iously desc ibed 21,23,45–47 (see de ails in supplemen a y me hods). S a is ical analyses.S a is ical analyses we e pe o med using he PASW S a is ics 18. Non–no mally dis ibu ed iglyce ides, insulin, and CRP concen a ions we e log 10 - ans o med be o e he analyses, bu he esul s a e exp essed as c ude. Da a a e p esen ed as mean 6SD, unless o he wise s a ed. The cha ac e is ics o he s udy subjec s we e compa ed wi h he es o con inuous a iables. Ca ego ical a iables we e compa ed wi h he x 2 – es , which was also used o es he geno ype equencies unde Ha dy–Weinbe g equilib ium. In o de o s udy he possible associa ion be ween USF1 gene polymo phisms and isk ac o s and subclinical ma ke s o a he oscle osis, we applied analysis o a iance (ANOVA) and analysis o co a iance (ANCOVA). The non-pa ame ic Mann-Whi ney U es was used o compa ison o gene exp ession be ween a he oscle o ic and con ol issues, as well as be ween di e en geno ype and haplo ype g oups. F equencies o he mos common haplo ypes and he mos p obable haplo ypes o each s udy subjec we e de e mined using he PHASE p og am (Ve sion 2.0.2) 49 .To s udy he e ec o haplo ypes, we di ided he popula ion in o ca ie s (one copy and wo copies o haplo ype) and non-ca ie s o in o h ee g oups wi h 0, 1, 2 copy o haplo ype. To a oid edundan mul iple es ing in YFS, he s a is ical analyses we e done in sepa a e s ages. We i s did sex-by-geno ype (o -by-haplo ype) in e ac ion analyses in ela ion o con en ional lipids and apolipop o eins by wo-way ANOVA. I sig- ni ican in e ac ion was ound, we s a i ied he subjec s by sex in u he analyses. As hypo hesized, sex-speci ic in e ac ions in ela ion o se um o al- and LDL-choles- e ol and apoB we e ound and he USF1 geno ype/haplo ype ela ed esul s we e s a is ically signi ican in women only. In he second s age o s a is ical analysis, we in es iga ed in women he e ec s o USF1 geno ypes and haplo ypes o e se um lipop o ein subclass choles e ol ac ions ( o al choles e ol, choles e ol es e , and ee choles e ol when a ailable) and o al lipid concen a ions. The subclass measu es we e in e se no mal ans o med o achie e app oxima e no mali y. Associa ions o USF1 SNPs and haplo ypes wi h lipop o ein subclass measu es we e es ed using uni a ia e addi i e gene ic model. S a is ical analyses we e pe o med using he R S a is ical package . 2.11.1 (h p://www. -p ojec .o g). P– alue ,0.05 was conside ed nominally signi ican . Fo Table 3, we calcula ed alse disco e y a e (FDR) mul iple co ec ion calcula ions in he women s a a assuming he e we e 18 independen es s (2 SNPs and 1 haplo ype, 2 ime poin s, 3 lipid measu es), using he calcula ion below and assuming an FDR alue o ,0.05 was accep able. FDR 5p- alue 3numbe o es s/p – alue ank The e o e, P #0.015 in he women s a a we e conside ed s a is ically signi ican a e co ec ing o mul iple es ing. 1. Pajukan a, P. e al. Familial combined hype lipidemia is associa ed wi h ups eam ansc ip ion ac o 1 (USF1). Na Gene 36, 371–6 (2004). 2. Si i o, M. e al. Membe s o he USF amily o helix-loop-helix p o eins bind DNA as homo- as well as he e odime s. Gene Exp 2, 231–40 (1992). 3. Plaisie , C. L. e al. A sys ems gene ics app oach implica es USF1, FADS3, and o he causal candida e genes o amilial combined hype lipidemia. PLoS Gene 5, e1000642 (2009). 4. Pajukan a, P. e al. Linkage o amilial combined hype lipidaemia o ch omosome 1q21-q23. Na Gene 18, 369–73 (1998). 5. Hue as-Vazquez, A. e al. Familial combined hype lipidemia in Mexicans: associa ion wi h ups eam ansc ip ion ac o 1 and linkage on ch omosome 16q24.1. A e ioscle Th omb Vasc Biol 25, 1985–91 (2005). 6. Holzap el, C. e al. Gene ic a ian s in he USF1 gene a e associa ed wi h low- densi y lipop o ein choles e ol le els and inciden ype 2 diabe es melli us in women: esul s om he MONICA/KORA Augsbu g case-coho s udy, 1984– 2002. Eu J Endoc inol 159, 407–16 (2008). 7. Meex, S. J. e al. Ups eam ansc ip ion ac o 1 (USF1) in isk o ype 2 diabe es: associa ion s udy in 2000 Du ch Caucasians. Mol Gene Me ab 94, 352–5 (2008). 8. Au o, K. e al. USF1 gene a ian s con ibu e o me abolic ai s in men in a longi udinal 32-yea ollow-up s udy. Diabe ologia 51, 464–72 (2008). 9. K is iansson, K. e al. Associa ion analysis o allelic a ian s o USF1 in co ona y a he oscle osis. A e ioscle Th omb Vasc Biol 28, 983–9 (2008). 10. Collings, A. e al. Allelic a ian s o ups eam ansc ip ion ac o 1 associa e wi h ca o id a e y in ima-media hickness: he Ca dio ascula Risk in Young Finns s udy. Ci c J 72, 1158–64 (2008). 11. Komulainen, K. e al. Risk alleles o USF1 gene p edic ca dio ascula disease o women in wo p ospec i e s udies. PLoS Gene 2, e69 (2006). 12. Wu, S. e al. Ups eam ansc ip ion ac o 1 in luences plasma lipid and me abolic ai s in mice. Hum Mol Gene 19, 597–608 (2010). 13. Lee, J. C. e al. USF1 con ibu es o high se um lipid le els in Du ch FCHL amilies and U.S. whi es wi h co ona y a e y disease. A e ioscle Th omb Vasc Biol 27, 2222–7 (2007). 14. Coon, H. e al. Ups eam s imula o y ac o 1 associa ed wi h amilial combined hype lipidemia, LDL choles e ol, and iglyce ides. Hum Gene 117, 444–51 (2005). 15. Blache , J. e al. Ca o id a e ial s i ness as a p edic o o ca dio ascula and all- cause mo ali y in end-s age enal disease. Hype ension 32, 570–4 (1998). 16. Bone i, P. O., Le man, L. O. & Le man, A. Endo helial dys unc ion: a ma ke o a he oscle o ic isk. A e ioscle Th omb Vasc Biol 23, 168–75 (2003). 17. Lo enz, M. W., on Kegle , S., S einme z, H., Ma kus, H. S. & Si ze , M. Ca o id in ima-media hickening indica es a highe ascula isk ac oss a wide age ange: p ospec i e da a om he Ca o id A he oscle osis P og ession S udy (CAPS). S oke 37, 87–92 (2006). 18. O’Lea y, D. H. e al. Ca o id-a e y in ima and media hickness as a isk ac o o myoca dial in a c ion and s oke in olde adul s. Ca dio ascula Heal h S udy Collabo a i e Resea ch G oup. N Engl J Med 340, 14–22 (1999). 19. Sch oede , S. e al. Nonin asi e de e mina ion o endo helium-media ed asodila ion as a sc eening es o co ona y a e y disease: pilo s udy o assess he p edic i e alue in compa ison wi h angina pec o is, exe cise elec oca diog aphy, and myoca dial pe usion imaging. Am Hea J 138, 731–9 (1999). www.na u e.com/scien i ic epo s SCIENTIFIC REPORTS | 4 : 4650 | DOI: 10.1038/s ep04650 7 20. Simons, P. C., Alg a, A., Bo s, M. L., G obbee, D. E. & an de G aa , Y. Common ca o id in ima-media hickness and a e ial s i ness: indica o s o ca dio ascula isk in high- isk pa ien s. The SMART S udy (Second Mani es a ions o ARTe ial disease). Ci cula ion 100, 951–7 (1999). 21. Oksala, N. e al. ADAM-9, ADAM-15, and ADAM-17 a e up egula ed in mac ophages in ad anced human a he oscle o ic plaques in ao a and ca o id and emo al a e ies--Tampe e ascula s udy. Ann Med 41, 279–90 (2009). 22. Rai aka i, O. T. e al. Coho p o ile: he ca dio ascula isk in Young Finns S udy. In J Epidemiol 37, 1220–6 (2008). 23. Rai oha ju, E. e al. miR-21, miR-210, miR-34a, and miR-146a/b a e up- egula ed in human a he oscle o ic plaques in he Tampe e Vascula S udy. A he oscle osis 219, 211–7 (2011). 24. Naukka inen, J. e al. USF1 and dyslipidemias: con e ging e idence o a unc ional in onic a ian . Hum Mol Gene 14, 2595–605 (2005). 25. Naukka inen, J. e al. Func ional a ian dis up s insulin induc ion o USF1: mechanism o USF1-associa ed dyslipidemias. Ci c Ca dio asc Gene 2, 522–9 (2009). 26. Ng, M. C. e al. The linkage and associa ion o he gene encoding ups eam s imula o y ac o 1 wi h ype 2 diabe es and me abolic synd ome in he Chinese popula ion. Diabe ologia 48, 2018–24 (2005). 27. Ho s ed , J., Ryde ´n, M., Wah enbe g, H., an Ha melen, V. & A ne , P. Ups eam ansc ip ion ac o -1 gene polymo phism is associa ed wi h inc eased adipocy e lipolysis. J Clin Endoc inol Me ab 90, 5356–60 (2005). 28. Lau ila, P. P. e al. Gene ic associa ion and in e ac ion analysis o USF1 and APOA5 on lipid le els and a he oscle osis. A e ioscle Th omb Vasc Biol 30, 346–52 (2010). 29. Zeggini, E. e al. Va ia ion wi hin he gene encoding he ups eam s imula o y ac o 1 does no in luence suscep ibili y o ype 2 diabe es in samples om popula ions wi h eplica ed e idence o linkage o ch omosome 1q. Diabe es 55, 2541–8 (2006). 30. Silande , K. e al. Gende di e ences in gene ic isk p o iles o ca dio ascula disease. PLoS One 3, e3615 (2008). 31. Damås, J. K. e al. Myoca dial exp ession o CC- and CXC-chemokines and hei ecep o s in human end-s age hea ailu e. Ca dio asc Res 47, 778–87 (2000). 32. Mis a, P. e al. Quan i a ion o CXCR4 exp ession in myoca dial in a c ion using 99mTc-labeled SDF-1alpha. JNuclMed49, 963–9 (2008). 33. Chonchol, M. & Nielson, C. Hemoglobin le els and co ona y a e y disease. Am Hea J 155, 494–8 (2008). 34. Go, A. S. e al. Hemoglobin le el, ch onic kidney disease, and he isks o dea h and hospi aliza ion in adul s wi h ch onic hea ailu e: he Anemia in Ch onic Hea Failu e: Ou comes and Resou ce U iliza ion (ANCHOR) S udy. Ci cula ion 113, 2713–23 (2006). 35. F iedewald, W. T., Le y, R. I. & F ed ickson, D. S. Es ima ion o he concen a ion o low-densi y lipop o ein choles e ol in plasma, wi hou use o he p epa a i e ul acen i uge. Clin Chem 18, 499–502 (1972). 36. Raiko, J. R. e al. Follow-ups o he Ca dio ascula Risk in Young Finns S udy in 2001 and 2007: le els and 6-yea changes in isk ac o s. J In e n Med 267, 370–84 (2010). 37. Rai aka i, O. T. e al. Ca dio ascula isk ac o s in childhood and ca o id a e y in ima-media hickness in adul hood: he Ca dio ascula Risk in Young Finns S udy. JAMA 290, 2277–83 (2003). 38. Juonala, M. e al. In e ela ions be ween b achial endo helial unc ion and ca o id in ima-media hickness in young adul s: he ca dio ascula isk in young Finns s udy. Ci cula ion 110, 2918–23 (2004). 39. Juonala, M. e al. Risk ac o s iden i ied in childhood and dec eased ca o id a e y elas ici y in adul hood: he Ca dio ascula Risk in Young Finns S udy. Ci cula ion 112, 1486–93 (2005). 40. Soininen, P. e al. High- h oughpu se um NMR me abonomics o cos -e ec i e holis ic s udies on sys emic me abolism. Analys 134, 1781–5 (2009). 41. S anc ˇa ´ko a ´,A.e al. E ec s o 34 isk loci o ype 2 diabe es o hype glycemia on lipop o ein subclasses and hei composi ion in 6,580 nondiabe ic Finnish men. Diabe es 60, 1608–1616 (2011). 42. Chambe s, J. C. e al. Genome-wide associa ion s udy iden i ies loci in luencing concen a ions o li e enzymes in plasma. Na Gene 43, 1131–8 (2011). 43. In e na ional Conso ium o Blood P essu e Genome-Wide Associa ion S udies e al. Gene ic a ian s in no el pa hways in luence blood p essu e and ca dio ascula disease isk. Na u e 478, 103–9 (2011). 44. Tukiainen, T. e al. De ailed me abolic and gene ic cha ac e iza ion e eals new associa ions o 30 known lipid loci. Hum Mol Gene 21, 1444–55 (2012). 45. Le ula, M. e al. ADAM8 and i s single nucleo ide polymo phism 2662 T/G a e associa ed wi h ad anced a he oscle osis and a al myoca dial in a c ion: Tampe e ascula s udy. Ann Med 41, 497–507 (2009). 46. Oksala, N. e al. Ca bonic anhyd ases II and XII a e up- egula ed in os eoclas -like cells in ad anced human a he oscle o ic plaques-Tampe e Vascula S udy. Ann Med 42, 360–70 (2010). 47. Niinisalo, P. e al. Ac i a ion o indoleamine 2,3-dioxygenase-induced yp ophan deg ada ion in ad anced a he oscle o ic plaques: Tampe e ascula s udy. Ann Med 42, 55–63 (2010). 48. S a y, H. C. e al. A de ini ion o ad anced ypes o a he oscle o ic lesions and a his ological classi ica ion o a he oscle osis. A epo om he Commi ee on Vascula Lesions o he Council on A e ioscle osis, Ame ican Hea Associa ion. Ci cula ion 92, 1355–74 (1995). 49. S ephens, M. & Donnelly, P. A compa ison o bayesian me hods o haplo ype econs uc ion om popula ion geno ype da a. Am J Hum Gene 73, 1162–9 (2003). Acknowledgmen s The s udy was inancially suppo ed by he Finnish Founda ion o Ca dio ascula Resea ch (Y.-M. F., T.L., M.A.-K.), he Al ed Ko delin Founda ion, he compe i i e esea ch Funding o he Tampe e Uni e si y Hospi al (g an 9M048 and 9N035 o T.L.) and Tu ku Uni e si y Hospi al, he Pi kanmaa Regional Fund o he Finnish Cul u al Founda ion, he Emil Aal onen Founda ion (T.L.), he Academy o Finland (g an no. 53392, 34316), he Social Insu ance Ins i u ion o Finland, he Tu ku Uni e si y Founda ion, he Juho Vainio Founda ion, and he Finnish Cul u al Founda ion. The s udy has also been unded by a Eu opean Union 7 h F amewo k P og amme g an numbe 201668 o he A he oRemo P ojec . This wo k was also suppo ed by he Academy o Finland (g an no. 137870 o P.S.), he Responding o Public Heal h Challenges Resea ch P og amme o he Academy o Finland (g an numbe 129429 o M.A.-K.), he S a egic Resea ch Funding om he Uni e si y o Oulu (M.A.-K.), he Finnish Funding Agency o Technology and Inno a ion (M.A.-K.) and he Jenny and An i Wihu i Founda ion (A.J.K.). The Young Finns S udy has been inancially suppo ed by he Academy o Finland: g an s 126925, 121584, 124282, 129378, 117797, and 41071, he Social Insu ance Ins i u ion o Finland, Kuopio, Tampe e and Tu ku Uni e si y Hospi al Medical Funds, Juho Vainio Founda ion, Paa o Nu mi Founda ion, Finnish Founda ion o Ca dio ascula Resea ch and Finnish Cul u al Founda ion, Sig id Juselius Founda ion, Tampe e Tube culosis Founda ion and Emil Aal onen Founda ion. The expe echnical assis ance in da a managemen and s a is ical analyses by I ina Lisinen and Ville Aal o a e g a e ully acknowledged. Au ho con ibu ions Y.-M.F. pa icipa ed in he s udy design, pe o med he mos o he da a analyses and d a ed he manusc ip . J.H., A.C. we e esponsible o pa o da a analyses. N.O., A.M., M.T., J.P.S. con ibu ed o ascula sample collec ion and associa ed da a collec ion. M.L., E.R. did he ascula sample handling and RNA ex ac ion. M.J., J.M., J.V., O.T.R., N.H.K., M.K., we e in ol ed in he ini ial Young Finns S udy and he w i ing. L.P.L., I.S. did pa o da a analyses and p epa ed Figu e 1. N.K., T.I. con ibu ed o exp ession analysis. A.J.K., P.S., M.A.K. designed and pe o med he NMR analyses. T.L. con ibu ed o measu emen s o ca o id a e y IMT and b achial a e y FMD. R.L., T.L. pa icipa ed in he s udy design and e ised he manusc ip c i ically. All in es iga o s con ibu ed o he w i ing o he manusc ip . All au ho s e iewed he manusc ip . Addi ional in o ma ion Supplemen a y in o ma ion accompanies his pape a h p://www.na u e.com/ scien i ic epo s Compe ing inancial in e es s: A.J.K., P.S., and M.A.-K. a e sha eholde s o B ainshake L d, a s a up company o e ing NMR-based me aboli e p o iling. How o ci e his a icle: Fan, Y.-M. e al. Ups eam T ansc ip ion Fac o 1 (USF1) allelic a ian s egula e lipop o ein me abolism in women and USF1 exp ession in a he oscle o ic plaque. Sci. Rep. 4, 4650; DOI:10.1038/s ep04650 (2014). This wo k is licensed unde a C ea i e Commons A ibu ion 3.0 Unpo ed License. Theimagesin hisa iclea eincludedin hea icle’sC ea i eCommonslicense, unless indica ed o he wise in he image c edi ; i he image is no included unde he C ea i e Commons license, use s will need o ob ain pe mission om he license holde in o de o ep oduce he image. To iew a copy o his license, isi h p://c ea i ecommons.o g/licenses/by/3.0/ www.na u e.com/scien i ic epo s SCIENTIFIC REPORTS | 4 : 4650 | DOI: 10.1038/s ep04650 8