Sex bias in celiac disease: XWAS and monocyte eQTLs in women identify TMEM187 as a functional candidate gene

He nangomez‑Lade ase al. Biology o Sex Di e ences (2023) 14:86 h ps://doi.o g/10.1186/s13293‑023‑00572‑1 RESEARCH Sex bias inceliac disease: XWAS andmonocy e eQTLs inwomen iden i y TMEM187 asa unc ional candida e gene Alba He nangomez‑Lade as1,2, A iadna Cille os‑Po e 1,2, Sil ia Ma ínez Velasco2,3, Se gi Ma í1,2, Ma ía Lega da2,3, Bá ba a Paola González‑Ga cía1,2, Ca los Tu au2,3, I aia Ga cía‑San is eban1,2, Iñaki I as o za2,3, No a Fe nandez‑Jimenez1,2* and Jose Ramon Bilbao1,2,4* Abs ac Backg ound Celiac disease (CeD) is an immune‑media ed diso de ha de elops in gene ically p edisposed indi‑ iduals upon glu en consump ion. HLA isk alleles explain 40% o he gene ic componen o CeD, so he e ha e been con inuing e o s o unco e non‑HLA loci ha can explain he emaining he i abili y. As in mos au oimmune diso ‑ de s, he p e alence o CeD is signi ican ly highe in women. He e, we in es iga ed he possible in ol emen o he X ch omosome on he sex bias o CeD. Me hods We pe o med a X ch omosome‑wide associa ion s udy (XWAS) and a gene‑based associa ion s udy in women om he CeD Immunochip (7062 cases, 5446 con ols). We also cons uc ed a da abase o X ch omo‑ some cis‑exp ession quan i a i e ai loci (eQTLs) in monocy es om uns imula ed (n = 226) and lipopolysaccha ide (LPS)‑s imula ed (n = 130) emale dono s and pe o med a Summa y‑da a‑based MR (SMR) analysis o in eg a e XWAS and eQTL in o ma ion. We in e oga ed he exp ession o he po en ially causal gene (TMEM187) in pe iphe al blood mononuclea cells (PBMCs) om celiac pa ien s a onse , on a glu en‑ ee die , po en ial celiac pa ien s and non‑celiac con ols. Resul s The XWAS and gene‑based analyses iden i ied 13 SNPs and 25 genes, espec i ely, 22 o which had no been p e iously associa ed wi h CeD. The X ch omosome cis‑eQTL analysis ound 18 genes wi h a leas one cis‑eQTL in naï e emale monocy es and 8 genes in LPS‑s imula ed emale monocy es, 2 o which we e common o bo h si u‑ a ions and 6 we e unique o LPS s imula ion. SMR iden i ied a po en ially causal associa ion o TMEM187 exp ession in naï e monocy es wi h CeD in women, egula ed by CeD‑associa ed, eQTL‑SNPs s7350355 and s5945386. The CeD‑ isk alleles we e co ela ed wi h lowe TMEM187 exp ession. These esul s we e eplica ed using eQTLs om LPS‑s im‑ ula ed monocy es. We obse ed highe le els o TMEM187 exp ession in PBMCs om emale CeD pa ien s a onse compa ed o emale non‑celiac con ols, bu no in male CeD indi iduals. Conclusion Using X ch omosome geno ypes and gene exp ession da a om emale monocy es, SMR has iden i ied TMEM187 as a po en ially causal candida e in CeD. Fu he s udies a e needed o unde s and he implica ion o he X ch omosome in he highe p e alence o CeD in women. Open Access © The Au ho (s) 2023. Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License, which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons licence, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his a icle a e included in he a icle’s C ea i e Commons licence, unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons licence and you in ended use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/licenses/by/4.0/. The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecom‑ mons.o g/publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed in a c edi line o he da a. Biology o Sex Di e ences *Co espondence: No a Fe nandez‑Jimenez [email p o ec ed] Jose Ramon Bilbao [email p o ec ed] Full lis o au ho in o ma ion is a ailable a he end o he a icle Page 2 o 10 He nangomez‑Lade ase al. Biology o Sex Di e ences (2023) 14:86 Backg ound Celiac disease (CeD) is an immune-media ed en e opa hy ha de elops in gene ically p edisposed indi iduals as a eac ion o glu en inges ion [1]. The global p e alence o CeD is 1.4% acco ding o se ological diagnosis, and 0.7% based on biopsy con i ma ion [2]. As in o he au o- immune diseases, he p e alence o CeD is signi ican ly highe in women [3]. Almos all CeD pa ien s ca y he Human Leucocy e An igen (HLA) alleles ha encode he HLA-DQ2 and/o HLA-DQ8 molecules. Howe e , HLA is necessa y, bu no su icien o de elop he dis- ease, and only explains 40% o he o e all gene ic isk [4, 5]. GWAS and ine-mapping s udies like he Immunochip p ojec ha e iden i ied mo e han 40 non-HLA loci asso- cia ed wi h CeD [6–8]. Ne e heless, mos o he SNPs loca ed in hese loci ei he map o non-coding egions o he genome, a away om genes, o a e in s ong link- age disequilib ium (LD) wi h o he associa ed a ian s, making i di icul o iden i y he genes ha a e unc ion- ally in ol ed in he disease [9]. Al oge he , HLA and non- HLA a ian s iden i ied so a explain a ound 50% o CeD he i abili y [10]. The missing gene ic he i abili y hypo h- esis sugges s ha addi ional, uniden i ied gene ic and en i onmen al ac o s a e in ol ed in he de elopmen o CeD. In his sense, he X ch omosome has been his o i- cally igno ed in mos GWAS, o has been analyzed as i i we e ano he au osome, wi hou accoun ing o male hemizygosi y and emale X ch omosome inac i a ion (XCI), wi h only e y ew s udies ha ake hese consid- e a ions in o accoun [11–13]. Se e al s udies ha e shown a ela ionship be ween he isk o di e en au oimmune diseases, including sys emic lupus e y hema osus (SLE), Sjög en synd ome, ype 1 diabe es melli us and CeD, and X ch omosome ane- uploidies like Kline el e , T iple X o Tu ne synd omes [14–17]. Addi ionally, many o he app oxima ely 1100 genes on he X ch omosome a e hough o be ela ed o he immune unc ion [18, 19]. These indings sugges a ole o he X ch omosome in he biased sex-p e alence o hese condi ions. In he case o CeD, mos isk loci a e loca ed on he au osomes bu X ch omosome genes ha e also been iden i ied, including TLR7 and TLR8, HCFC1, TMEM187 and IRAK1 [6, 7]. Monocy es a e a undamen al pa o he inna e immune de ense agains mic oo ganisms [20]. Di e - en s udies ha e also ela ed his ype o immune cells o CeD, and gliadin pep ides s imula e hei p oduc ion o IL-8 and TNF-α, especially in celiac pa ien s [21, 22]. I Highligh s • The XWAS and gene‑based associa ion s udy iden i ied 13 gene ic a ian s and 25 genes signi ican ly associa ed wi h CeD in women, bu no in men. • eQTL analyses in naï e and LPS‑s imula ed emale monocy es iden i ied 16 unique genes wi h a leas one cis‑eQTL in he naï e condi ion, 6 in he LPS condi ion and 2 genes, namely ZNF185 and TMEM187, common o bo h condi ions. • SMR iden i ied h ee SNPs ( s7350355, s5945386, s80208125) associa ed wi h gene TMEM187. The CeD isk alleles we e nega i ely co ela ed wi h TMEM187 exp ession. • PBMCs om ac i e emale CeD pa ien s showed signi ican ly highe TMEM187 exp ession han con ols, bu no di e ences we e obse ed in men. Keywo ds Celiac disease, XWAS, Mendelian andomiza ion, TMEM187, Monocy es, eQTLs Plain language summa y Celiac disease (CeD) is an immune‑ ela ed condi ion igge ed by glu en consump ion in gene ically suscep ible indi iduals. Women p esen highe p e alence o CeD han men, bu he biological explana ion o such di e ence has no been elucida ed. In his s udy, we in es iga ed whe he speci ic gene ic a ia ions on he X ch omosome we e associa ed wi h CeD in each sex. Su p isingly, we ound 13 gene ic a ian s and 25 genes signi ican ly linked o CeD in women, bu no in men. Addi ionally, we iden i ied gene ic a ian s on he X ch omosome associa ed wi h gene exp ession o monocy es, a ype o immune cells ha is ac i a ed in CeD a e glu en in ake. In eg a ing hese da a wi h ou p e ious indings, we ound ha lowe exp ession o a gene e med TMEM187 migh be associ‑ a ed wi h a po en ial inc ease in CeD isk in women. Finally, alida ion expe imen s con i med highe TMEM187 le els in blood cells om emale CeD pa ien s compa ed o non‑celiac women, while no such di e ence was seen in males. In summa y, ou s udy sugges s ha he X‑ch omosome gene TMEM187 may play a key ole in CeD de elopmen , p o iding insigh s in o he highe p e alence o CeD in emales. Page 3 o 10 He nangomez‑Lade ase al. Biology o Sex Di e ences (2023) 14:86 has been sugges ed ha he esponse igge ed by glia- din in monocy es is simila o ha induced by lipopoly- saccha ide (LPS) h ough ecep o s such as TLR4 [23]. Fai ax e al. analyzed how gene ic a ian s shape gene exp ession in monocy es, unde di e en in i o s im- uli, including LPS, and showed ha mo e han hal o he monocy e eQTLs a e speci ic o he en i onmen al s imulus, bu he X ch omosome was no included in he analysis [24]. Mendelian andomiza ion (MR), and mo e speci ically he Summa y-da a-based MR (SMR), in eg a es GWAS and eQTL summa y s a is ics in o de o de ec he unc- ional in ol emen o genes unde he GWAS peaks [25]. Pa icula ly, associa ion hi s a e ansla ed in o po en- ially causal ela ionships be ween exp ession le els o candida e genes and complex ai s, in ele an issues, cell ypes and con ex . Again, p e ious analyses in CeD ha ha e made use o his me hod o combine GWAS and eQTL summa y s a is ics ha e omi ed he X ch o- mosome [26, 27]. In his s udy, we hypo hesized ha he X ch omosome could ha bo addi ional suscep ibili y loci ha could help explain bo h he missing he i abili y and he highe p e alence in women. The e o e, we aimed o iden- i y genes on he X ch omosome ha migh pa icipa e in he pa hogenesis and also con ibu e o he sex bias in CeD, h ough hei speci ic ansc ip ional p o ile in monocy es. Fo ha pu pose, we pe o med an X-ch o- mosome associa ion s udy (XWAS) in women om he Immunochip p ojec , and cons uc ed a da abase o X ch omosome cis-eQTLs in emale monocy es. Finally, we combined he wo da ase s using SMR in o de o ind monocy e-speci ic unc ional candida es on he X ch o- mosome in ol ed in CeD. Ma e ials andme hods Immunochip da a andX ch omosome associa ion analyses The CeD Immunochip da ase was il e ed o include only X ch omosome a ian s wi h geno yping a e > 95%, mino allele equency (MAF) > 1% and P- alue om Ha dy Weinbe g equilib ium (PHWE) es > 1 × 10–6 using PLINK1.9 [28]. We emo ed indi iduals wi h call a e < 97% and he e ozygosi y de ia ing mo e han 4 s anda d de ia ions om he mean (> 4SD). Geno ypes we e impu ed a he Michigan Impu a ion Se e [29] using HRC 1.1 2016 (GRCh37/hg19) as a e e ence panel, using only he Eu opean popula ion, Minimac4 as he impu a ion so wa e and phasing wi h Eagle 2.4. Impu ed SNPs wi h an R2 impu a ion accu acy abo e 0.8, MAF > 1% and PHWE > 1 × 10–6 we e kep . A e applying hese il e s, 12,508 emale samples (7062 cases and 5446 con ols) and 9474 males (3712 cases and 5762 con ols) and 1611 SNPs we e e ained. We calcula ed he i s en p incipal componen s o he geno ypes using PLINK1.9 o con ol o po en ial popula ion s a i ica ion in down- s eam analyses. The op associa ed SNP was eplica ed in he Dubois e al. s udy GWAS da ase [6]. The CeD XWAS was pe o med sepa a ely o men and women using he newml me hod implemen ed in SNPT- EST ( e sion 2.5.6), assuming a comple e inac i a ion o one ch omosome in emales and equal e ec size in bo h sexes [30]. Speci ically, his me hod uses a logis ic eg es- sion model encoding geno ypes in males as 0/1 and in emales as 0 / ½ / 1. The analysis was pe o med assuming an addi i e gene ic model and he i s en p incipal com- ponen s o he geno ype da a we e included as co a ia es. The P- alue h eshold o s a is ical signi icance was se a P < 8.68 × 10–5 a e Bon e oni co ec ion acco ding o he numbe o independen es s, as de e mined wi h he SimpleM me hod [31]. Resul s we e plo ed on a Man- ha an plo gene a ed wi h he qqman R package [32]. A gene-based associa ion analysis was ca ied ou in women using he Fas BAT me hod a ailable in he Genome-wide Complex T ai Analysis (GCTA) so - wa e package [33, 34]. This me hod in eg a es GWAS summa y s a is ics and LD in o ma ion o calcula e he P- alue o a se o a ian s wi hin a p ese dis ance om a gene. The analysis was pe o med using he de aul se ings sugges ed by GCTA-Fas BAT: gene egions ex ended 50kb away om bo h he 3′ and 5′ UTRs o he genes, and SNPs in s ong LD ( 2 ≥ 0.9) we e p uned. We conduc ed he analysis o 2393 genes and 1611 SNPs and he P- alue signi icance h eshold was se a P < 0.05 a e False Disco e y Ra e (FDR) co ec ion. A egional associa ion plo combining XWAS and gene-based asso- cia ion esul s was gene a ed wi h an open-sou ce R sc ip (h ps:// gi hub. com/ Geeke ics/ Locus Zooms). Monocy e cis‑eQTLs A ca alog o naï e and LPS-s imula ed monocy e cis- eQTLs o he X ch omosome was cons uc ed in women using SNP geno ype and exp ession da a om a gen- e al popula ion s udy by Fai ax e al. [24]. X ch omo- some a ian s and indi iduals we e il e ed and impu ed as desc ibed abo e. Addi ionally, only hose SNPs ha we e homozygous o he mino allele in a leas 5 sam- ples we e e ained. A e he QC, 233 emale samples and 165,648 a ian s emained o subsequen analyses. Monocy e exp ession da a ha had al eady unde gone da a no maliza ion, ans o ma ion wi h Va iance S a- bilizing T ans o ma ion (VST), ba ch e ec co ec ion and emo al o ou lie s, we e subjec ed o addi ional QC s eps using he IluminaHuman 4.db [35] o emo e hose p obes ha ma ched mo e han one locus, hose on he au osomes, Y ch omosome o wi hou ch omo- some epo ed, hose con aining SNPs wi h MAF > 0.1, Page 4 o 10 He nangomez‑Lade ase al. Biology o Sex Di e ences (2023) 14:86 and hose desc ibed as Bad o No ma ch. This esul ed in 1347 p obes mapping o 1250 s able gene IDs in he X ch omosome. We combined geno ype (165,648 SNPs) and gene exp ession (1250 genes) in o ma ion o 226 samples om he naï e monocy e emale da ase using he QTL ools so wa e [36]. The associa ions be ween SNP geno ypes and gene exp ession le els we e es ed wi h simple linea eg essions assuming no mal dis ibu ion o he da a, and he i s 10 p incipal componen s o he geno ype we e included as co a ia es. Only SNPs wi hin a 1-Mb window om he ansc ip ion s a si e (TSS) o a gene we e ana- lyzed. PeQTL alues we e calcula ed wi h he nominal pass op ion and a PeQTL < 5 × 10–8 h eshold was se o iden- i y signi ican esul s. The same analysis was pe o med wi h exp ession da a om LPS-s imula ed monocy es in women (n = 130). SMR analyses SMR was pe o med combining he summa y s a is ics om he XWAS and he eQTLs o naï e and LPS-s im- ula ed monocy es using he SMR so wa e [25]. B ie ly, SMR uses cis-eQTLs as ins umen al a iables, gene exp ession as he exposu e, and CeD as he ou come, o in e genes pleio opically o causally associa ed wi h CeD. SMR esul s we e subjec ed o he he e ogenei y in dependen ins umen s (HEIDI) es o de ec he p es- ence o LD. In his es , a signi ican P- alue sugges s ha he associa ion de ec ed could be he esul o wo gene ic a ian s in s ong LD, whe eas a non-signi ican P- alue indica es ha a single a ian is associa ed wi h bo h gene exp ession and he disease. We used he ollowing de aul pa ame e s sugges ed by SMR: he cis-window was se a 2Mb, he h eshold PeQTL o he SMR analysis was se a 5 × 10–8, he h eshold PeQTL o he HEIDI es was se a 1.57 × 10–3 and SNPs wi h a LD 2 > 0.9 o 2 < 0.05 wi h he op associa ed eQTL we e p uned. SMR esul s wi h an FDR q- alue < 0.05 and PHEIDI > 0.05, we e consid- e ed as pleio opic o causal associa ions. We plo ed he TMEM187 egion (2000kb) on he X ch omosome using he SMRLocusPlo sc ip a ailable in he SMR websi e. Exp ession analysis inpedia ic CeD pa ien s TMEM187 exp ession was quan i ied in pe iphe al blood mononuclea cells (PBMCs) om CeD child en a onse (20 emales, 8 males), pa ien s on glu en- ee die (GFD) (6 emales, 9 males), po en ial CeD pa ien s (5 emales, 6 males) and non-celiac con ols (17 emales, 10 males). CeD was diagnosed a he Pedia ic Gas oen- e ology Uni o C uces Uni e si y Hospi al. The s udy was app o ed by he Clinical Resea ch Boa d o C uces Uni e si y Hospi al. Samples (2.5 ml in EDTA-con aining ubes) we e collec ed a e in o med consen had been ob ained om hei pa en s o gua dians and ans e ed o he Basque Biobank o Resea ch. PBMCs we e iso- la ed using he MACSp ep™ PBMC Isola ion ki (Mil e- nyi Bio ec SL, Mad id, Spain; ca . no. 130–115-169), RNA was pu i ied using he NucleoSpin® RNA mini ki (Mache ey–Nagel, Dü en, Ge many; ca . no. 740955.250 4,392,653) and s o ed a –80°C un il use. The exp ession o TMEM187 was quan i ied by RT- qPCR using he TaqMan RNA- o-C 1-S ep Ki (The mo Fishe Scien i ic Inc., Wal ham, MA, USA; ca . no. 4392653) and a comme cially a ailable TaqMan Gene Exp ession assay (The mo Fishe Scien i ic Inc., ca . no. Hs01920894_s1) on a Bio-Rad CFX Real Time PCR sys- em (Bio Rad Scien i ic, He cules, CA, USA). The house- keeping gene RPLPO was simul aneously measu ed and used as an endogenous con ol. Rela i e exp ession in each sample was calcula ed using he 2−ΔΔC me hod and he e o e, he exp ession o each sample was no malized o bo h he exp ession o RPLPO and he a e age o he con ols. Di e ences in gene exp ession le els we e ana- lyzed wi h Mann–Whi ney U- es using G aphPad P ism .8.0.1. Finally, he exp ession o TMEM187 in di e - en immune cell ypes (B, NK, T cells, and monocy es) was analyzed and sex-s a i ied exp ession plo s we e cons uc ed using he de aul se ings o he Da abase o Immune Cell Exp ession eQTL Epigenomics (DICE, h ps:// dice- da ab ase. o g/) online b owse [37]. Resul s XWAS andgene‑le el associa ion analyses inCeD To de e mine whe he he sex bias in CeD p e alence is ela ed o he X ch omosome, we pe o med he XWAS o CeD in women and men sepa a ely (Addi ional ile1). We iden i ied a single associa ion peak on Xq28 including 13 signi ican SNPs (PXWAS < 8.68 × 10–5), wi h s78237385 (PXWAS = 2.30 × 10–5, OR = 1.20 ± 0.10) as he op SNP (Addi ional ile2). The male XWAS did no de ec any signi ican associa ion (Addi ional ile3). The op SNP was also sugges i ely associa ed wi h CeD in emales om he Dubois e al. s udy (P- alue = 1.86 × 10–4) bu was no signi ican (P- alue = 0.19) in men om he same s udy (Addi ional ile4). We also pe o med a gene-based associa ion analy- sis in women, ha akes in o accoun he agg ega ed e ec o se s o SNPs. We in oduced 2392 genes and 1611 SNPs in he analysis and de ined each gene egion as ± 50 kb om bo h 3′ and 5′ UTRs. As a esul , 276 associa ion es s we e pe o med and 25 candida e genes in he X ch omosome we e iden i ied o be asso- cia ed wi h CeD in women (FDR q- alue < 0.05) (Addi- ional ile5). Ou o he 25 genes, 22 a e no el genes associa ed wi h CeD, al hough ARHGAP4, RENBP, NAA10, AVPR2 o MECP2 had been p e iously Page 5 o 10 He nangomez‑Lade ase al. Biology o Sex Di e ences (2023) 14:86 iden i ied in o he au oimmune diseases [38–40]. XWAS and gene-le el analysis esul s a e summa ized in Fig.1. Cis‑eQTL analysis o  heX ch omosome inmonocy es o women In o de o ob ain an addi ional laye o in o ma ion o in e p e he XWAS signal, we calcula ed he X ch o- mosome eQTLs in monocy es om women. We ana- lyzed 226 samples wi h in o ma i e geno ype and gene exp ession da a om naï e monocy es, and pe o med mul iple linea eg essions be ween 165,648 SNPs and 1250 p obes co esponding o 819 genes. Applying a h eshold PeQTL alue o 5 × 10–8, we iden i ied 1097 cis-eQTLs in ol ing 1054 SNPs, 19 p obes and 18 inde- penden genes (Addi ional ile6). The analysis in 130 LPS-s imula ed emale monocy e samples e ealed 150 cis-eQTLs, co esponding o 94 SNPs, 9 p obes and 8 genes (Addi ional ile 7). Two genes (ZNF185 and TMEM187) we e common o bo h si ua ions, 16 we e unique o naï e emale monocy es, and 6 o emale monocy es a e LPS s imula ion. XWAS andcis‑eQTL SMR andgene exp ession analyses The SMR and HEIDI analyses o he summa y s a is- ics o he emale CeD XWAS and he naï e monocy e cis-eQTL da ase iden i ied wo SNPs ( s7350355 and s5945386) ha we e associa ed wi h wo exp ession p obes (ILMN_2198185 and ILMN_3242211, espec- i ely) co esponding o he same gene, TMEM187, wi h a q- alue < 0.05 and PHEIDI > 0.05. The mino alleles s7350355*G and s5945386*G a e bo h he CeD isk alleles and we e nega i ely co ela ed wi h TMEM187 exp ession (Fig. 2; Addi ional ile 8). The SMR analy- sis was eplica ed wi h cis-eQTLs om LPS-s imu- la ed emale monocy es and e ealed a single SNP ( s80208125) ha was associa ed wi h he same wo TMEM187 p obes (Addi ional ile9; Addi ional ile10). We in es iga ed he exp ession o TMEM187 in PBMCs om emale CeD pa ien s a diagnosis, on GFD, po en ial CeD pa ien s and non-celiac emale con ols. TMEM187 showed a signi ican ly highe exp ession in ac i e CeD pa ien s compa ed o con ols (P- alue = 0.0417) and no di e ences we e obse ed in po en ial CeD and GFD- ea ed indi iduals (Fig.3A). We also s udied TMEM187 exp ession in men bu no Fig. 1 Locus zoom plo 200 kb ups eam and downs eam o he XWAS op SNP. The XWAS op SNP, s78237385 is ep esen ed wi h a pu ple diamond. In he op panel, he colo schema ep esen s he LD be ween he op SNP and he SNPs included in he egion. In he bo om panel, he colo schema ep esen s he P‑ alue om he gene‑based associa ion analysis Page 6 o 10 He nangomez‑Lade ase al. Biology o Sex Di e ences (2023) 14:86 signi ican di e ences we e ound (Fig. 3B). On he o he hand, he exp ession o TMEM187 a ied among di e en immune cell ypes (Addi ional ile11). Discussion The p e alence o CeD is signi ican ly highe in women, as in he case o o he au oimmune diseases [3]. A highe p e alence o immune-media ed diso de s has Fig. 2 SMR locus plo o he esul s o he SMR analysis be ween he CeD XWAS and he naï e emale monocy e eQTLs. In he op panel, g ey do s ep esen − log10(P‑ alues) o he emale XWAS SNPs. Diamonds ep esen − log10(P‑ alues) o p obes om he SMR analysis and illed diamonds show hose ha pass he HEIDI es . In he middle panel, he ed c osses ep esen − log10(P‑ alues) o gene p obes in he eQTL analysis. In he bo om panel, he loca ion o he p obes on he X ch omosome is shown Fig. 3 Resul s o he exp ession analysis. A Exp ession o TMEM187 in PBMCs o emale samples. B Exp ession o TMEM187 in PBMCs o male samples. Bo h emale and male samples we e classi ied in o ou g oups: non‑celiac con ols, celiac pa ien s a onse , po en ial celiac pa ien s and celiac pa ien s on GFD ep esen ed by whi e, ligh g ey, da k g ey and black ci cles, espec i ely Page 7 o 10 He nangomez‑Lade ase al. Biology o Sex Di e ences (2023) 14:86 also been obse ed in indi iduals wi h synd omes ela ed o X ch omosome aneuploidies [14, 17], sug- ges ing an implica ion o he X ch omosome in he isk o au oimmune diseases, p obably also in he isk o CeD. In he cu en s udy, we ha e ocused on gene exp ession in emale monocy es in o de o iden i y genes on he X ch omosome ha a e in ol ed in CeD and could explain he sex bias in he disease p e alence. Fo ha pu pose, we conduc ed a sex-speci ic XWAS in women om he Immunochip da ase and iden i ied 25 genes associa ed wi h CeD in women, o which 22 ha e no been p e iously epo ed. Rema kably, 24 ou o he 25 genes ha we e signi ican in he analysis a e loca ed on ch omosome egion Xq28, which has been p e i- ously associa ed wi h di e en au oimmune condi ions such as SLE [39], heuma oid a h i is [40], sys emic scle osis [38] o CeD [7]. On he o he hand, i has been sugges ed ha gliadin igge s an inna e esponse in monocy es simila o ha p oduced by LPS [23]. We pe o med an X ch omo- some cis-eQTL analysis in emale naï e and LPS-s imu- la ed monocy es and iden i ied 6 eQTL-genes unique o LPS-s imula ed emale monocy es, some o which ha e been p e iously associa ed wi h au oimmune diso de s, including PLXNA3 [41]. We hen in eg a ed he XWAS wi h cis-eQTL da a om emale naï e and LPS-s imula ed monocy es using an MR app oach. To ou knowledge, his is he i s SMR analysis o he X ch omosome in CeD. We iden i- ied wo SNPs ( s7350355 and s5945386) ha egula e he exp ession o TMEM187 in naï e monocy es and a di e en SNP ( s80208125) associa ed wi h TMEM187 in LPS-s imula ed monocy es, sugges ing ha he gene ic backg ound may be mo e impo an han LPS s imula ion in he egula ion o TMEM187 exp ession. s7350355 is a missense a ian loca ed wi hin exon 2 o TMEM187, s5945386 is an in e genic a ian loca ed 21 kb downs eam o TMEM187, and s80208125 is loca ed on he 5’ UTR o TMEM187. All hese a ian s exhibi s ong LD (> 0.87) in he B i ish popula ion in England and Sco land, om he 1000 Genomes P ojec [42]. In all h ee cases, he CeD isk alleles ( s7350355*G, s5945386*G and s80208125*G) we e he mino alleles (MAF a ound 0.2) and we e nega i ely co ela ed wi h TMEM187 exp ession. TMEM187 encodes a mul ipass ansmemb ane p o ein o unknown unc ion [43], and i has been p oposed as a pu a i e candida e gene in CeD oge he wi h IRAK1 and HCFC1, loca ed in he same locus [7]. As a as we know, he only s udy ha obse ed TMEM187 dys egula ion in CeD is he one ca ied ou by Pascual e al. in duodenal biopsies [44]. The TMEM187 locus has also been associa ed wi h o he au oimmune diso de s such as SLE o heuma oid a h i is [39, 45], suppo ing he hypo hesis o a sha ed gene ic back- g ound in au oimmune diso de s. In ou s udy, we in e oga ed he exp ession o TMEM187 in PBMCs om celiac pa ien s a onse , celiac pa ien s on GFD, po en ial celiac pa ien s and non-celiac con ols. PBMCs a e a mix u e o immune cells ha con ain monocy es (10–20%), oge he wi h lympho- cy es (70–90%) and dend i ic cells (1–2%), among o he s [46]. We obse ed a highe exp ession o TMEM187 in PBMCs om emale pedia ic pa ien s a disease onse compa ed o non-celiac child en. These esul s a e con- sis en wi h a s udy published in 2016 by Pascual e al. ha showed an up egula ion o TMEM187 exp ession in biopsies o celiac adul s [44]. The up egula ion o TMEM187 in emale CeD pa ien s was no eplica ed in male PBMCs, sugges ing possible ole o TMEM187 in he sex bias o CeD ha ne e heless needs o be con- i med wi h addi ional in es iga ions. The o e exp ession epo ed o TMEM187 in emale CeD pa ien s a onse is appa en ly con adic o y o he ac ha CeD isk alleles co ela e wi h lowe exp ession. This obse ed di e gence could be due o di e en ea- sons: i s , he CeD isk eQTLs could ha e an e ec a he p o ein le el, aking in o accoun ha s7350355 is a mis- sense a ian ha could al e he unc ion o TMEM187, ega dless o mRNA quan i y. On he o he hand, we a e unable o de ini i ely asce ain which o he eQTLs is he causal SNP, gi en he s ong LD. Ano he eason could be ha PBMCs con ain a ela i ely modes p opo ion o monocy es, and he SNP could ha e di e en e ec s on he gene exp ession in o he cell ypes, he e o e explain- ing he appa en con adic ion. The highly a iable exp ession o TMEM187 in di e en immune cell ypes wa an s u he esea ch on i s ole in he immune sys- em. In addi ion, we ha e o bea in mind ha he p esen exp ession analyses ha e been ca ied ou in a pedia ic coho o diagnosed celiac child en, while emale dono s in he monocy e exp ession s udy we e non-celiac adul s. I has been well epo ed ha disease and immunogenic insul can some imes su pass he geno ypic e ec , and lead o his kind o appa en ly con adic o y si ua ions [47]. Finally, Pascual e  al. obse ed di e ences in he gene exp ession p o ile o suscep ibili y genes in CeD be ween child en and adul s, including TMEM187 [44]. Las bu no leas , i is wo h men ioning ha he main objec i e in he p esen wo k was no o iden i y SNPs wi h a unc ional in ol emen in CeD, no o asce ain he mechanism by which hey exe hei unc ion, bu o highligh po en ially causal genes ha pa icipa e in he pa hogenesis o he disease h ough hei exp ession in monocy es. We ca ied ou bo h he XWAS and he eQTL calcula ions wi h he aim o ob aining ins umen s o pe o m downs eam analyses such as SMR, being Page 8 o 10 He nangomez‑Lade ase al. Biology o Sex Di e ences (2023) 14:86 awa e ha in ou esul s SNPs will lose hei ele ance and will be eplaced by unc ional candida e genes. One limi a ion o his s udy is ha bo h he lack o associa ion as well as he absence o signi ican di e en- ial exp ession be ween CeD cases and con ols obse ed in men could be due o he smalle sample size o he male coho . This is pa ly a consequence o he highe incidence o CeD in women and an impo an ac o o conside . Howe e , we ha e s udied he associa ion o he op SNP o ou emale XWAS ( s78237385) in an independen da ase [6] and he P- alues in men and women a e 0.19 and 1.86 × 10–4, espec i ely. Addi ion- ally, he SNPs wi h signi ican esul s in he SMR analysis ( s5945386, s7350355 and s80208125) ha e P- alues in women o 8.29 × 10–5, 4.66 × 10–3 and 3.34 × 10–3, espec- i ely, while in men, hey show P- alues o 0.50, 0.49 and 0.57, espec i ely. We conside hese P- alues unlikely o become signi ican e en wi h a highe numbe o male samples, and his lack o a signi ican associa ion in men could imply a possible di e gen mechanism o pa hogen- esis be ween sexes, ha could explain he inc eased p e - alence o CeD, and o he au oimmune diseases obse ed in women [48]. Pe spec i es andsigni icance This is he i s SMR app oach in he X ch omosome in CeD. We ha e iden i ied TMEM187 as a candida e gene in CeD in monocy es and alida ed i s di e en ial exp es- sion in PBMCs om emale CeD pa ien s a onse . The ac ha bo h he gene ic associa ion and he di e en ial gene exp ession a e no ound in male pa ien s sugges s a ole o TMEM187 in he sex bias obse ed in CeD. SMR appea s as a use ul app oach o iden i y po en ially causal genes unde associa ion peaks, including he X ch omo- some. Fu he s udies a e needed o iden i y he unc ion o TMEM187 and unde s and i s beha io in di e en cell ypes and disease s a us, and o cla i y i s ole in CeD pa hogenesis and he sex bias. Supplemen a y In o ma ion The online e sion con ains supplemen a y ma e ial a ailable a h ps:// doi. o g/ 10. 1186/ s13293‑ 023‑ 00572‑1. Addi ional ile1. Sex‑speci ic summa y s a is ics om Immunochip XWAS in women and men. BP: posi ion o he a ian ; OA: o he allele; EA: e ec allele; F: equency o he e ec allele; B: be a alue, SE: s anda d e o o he be a, P: P‑ alue. Addi ional ile2. Manha an plo o he CeD XWAS in women. The op SNP s78237385 (P‑ alue = 2.30 × 10−5) is shown as a ed ci cle. The blue line ep esen s he signi ican h eshold acco ding o he Bon e oni co ‑ ec ion o he numbe o independen es s (P‑ alue = 8.68 × 10−5). Addi ional ile3. Manha an plo o he CeD XWAS o CeD in men. The blue line ep esen s he signi ican h eshold acco ding o he Bon e oni co ec ion o he numbe o independen es s (P‑ alue = 8.68 × 10−5). Addi ional ile4. Nominal P‑ alues o he op Immunochip emale XWAS SNP ( s78237385) in he di e en da ase s s a i ied by sex. EAF: equency o he e ec allele in cases; EAF: equency o he e ec allele in con ols; OR: odd a io. Addi ional ile5. CeD candida e genes on he X ch omosome iden i ied by gene‑based associa ion analysis a q‑ alue < 0.05. Top associa ed SNP: he op associa ed XWAS SNP in he egion; Top PXWAS: P‑ alue o he op associa ed XWAS SNP in he egion; P as BAT: gene‑based es P‑ alue; q‑ alue: FDR adjus ed gene‑based es P‑ alue. Addi ional ile6. cis‑eQTLs iden i ied on he X ch omosome by cis‑eQTL analysis o naï e emale monocy es a nominal P‑ alue < 5 × 10–8. A o al o 1,097 cis‑eQTLs, implica ing 1,054 SNPs, 19 p obes and 18 genes we e iden i ied on he X ch omosome o naï e emale monocy es. n_ a _in_cis: he o al numbe o a ian s es ed in cis; dis _gene_ a : he dis ance be ween he gene and he es ed a ian ; a _posi ion: he posi ion o he a ian ; nom_p al: he nominal P‑ alue o he associa ion be ween he a ian and he gene; _squa ed: he co ela ion coe icien o he linea eg ession; slope: he slope (be a) o he linea eg ession; slope_se: he s anda d e o o he be a. Addi ional ile7. cis‑eQTLs iden i ied in he X ch omosome by cis‑eQTL analysis o LPS‑s imula ed emale monocy es a nominal P‑ alue < 5 × 10–8. A o al o 150 cis‑eQTLs, implica ing 94 SNPs, 9 p obes and 8 genes we e iden i ied on he X ch omosome o LPS‑s imula ed emale monocy es. n_ a _in_cis: he o al numbe o a ian s es ed in cis; dis _gene_ a : he dis ance be ween he gene and he es ed a ian ; a _posi ion: he posi ion o he a ian ; nom_p al: he nominal P‑ alue o he associa ion be ween he a ian and he gene; _squa ed: he co ela ion coe icien o he linea eg ession; slope: he slope (be a) o he linea eg ession; slope_se: he s anda d e o o he be a. Addi ional ile8. Summa y o he SMR analysis be ween he CeD XWAS and he naï e emale monocy e eQTLs. A1: he e ec allele; A2: he o he allele; b_XWAS: he e ec size om XWAS; p_XWAS: P‑ alue om he XWAS; b_eQTL: he e ec size om eQTL analysis; p_eQTL: P‑ alue om he eQTL analysis; b_SMR: he e ec size om he SMR analysis; p_SMR: nominal P‑ alue om he SMR analysis; q‑ alue: FDR adjus ed P‑ alue. Resul s in bold indica e s a is ical signi icance a e mul iple es ing co ec ion. Addi ional ile9. SMR locus plo o he esul s o he SMR analysis be ween he CeD XWAS and he LPS‑s imula ed emale monocy e eQTLs. In he op panel, g ey do s ep esen ‑log10(P‑ alues) o he emale XWAS SNPs. Diamonds ep esen ‑log10(P‑ alues) o p obes om he SMR analysis and illed diamonds show hose ha pass he HEIDI es . In he middle panel, he ed c osses ep esen ‑log10(P‑ alues) o gene p obes in he eQTL analysis. In he bo om panel, he loca ion o he p obes on he X ch omosome is shown. Addi ional ile10. Summa y o he SMR analysis be ween he CeD XWAS and he LPS‑s imula ed emale monocy e eQTLs. A1: he e ec allele; A2: he o he allele; b_XWAS: he e ec size om XWAS; p_XWAS: P‑ alue om he XWAS; b_eQTL: he e ec size om eQTL analysis; p_eQTL: P‑ alue om he eQTL analysis; b_SMR: he e ec size om he SMR analysis; p_SMR: nominal P‑ alue om he SMR analysis; q‑ alue: FDR adjus ed P‑ alue. Resul s in bold indica e s a is ical signi icance a e mul iple es ‑ ing co ec ion. Addi ional ile11. TMEM187 exp ession in di e en immune cells. Red and blue boxes ep esen TMEM187 exp ession in he di e en immune cells om emales and males, espec i ely. Acknowledgemen s We acknowledge all he pa ien s and hei amilies, as well as he Basque Biobank o cus ody and p epa a ion o he clinical samples. Au ho con ibu ions JRB and NF‑J designed and supe ised he esea ch; AH‑L pe o med he compu a ional and expe imen al analyses; AH‑L, AC‑P and SM p epa ed he code and o ma ed he da a o he compu a ional analyses, BPG and IG‑S helped o design and pe o m he expe imen al analyses; SMV, ML, CT and Page 9 o 10 He nangomez‑Lade ase al. Biology o Sex Di e ences (2023) 14:86 II diagnosed pa ien s and collec ed biological samples; JRB, NF‑J and AH‑L w o e he i s d a o he manusc ip , and all au ho s con ibu ed o he a icle and app o ed he inal e sion. Funding JRB is unded by Resea ch G an PID2019‑106382RB‑I00 om he MCIN/ AEI/h ps://doi.o g/10.13039/501100011033. AH‑L is a p edoc o al el‑ low suppo ed by g an PRE‑C‑2020‑0091 om he MCIN/AEI/h ps://doi. o g/10.13039/501100011033and by ESF In es ing in you u u e. NFJ is unded by esea ch g an s 2019/111085 om he Basque Depa men o Heal h, and PI21/01491 om he Ins i u o de Salud Ca los III (ISCIII), co‑ unded by he Eu opean Union. A ailabili y o da a and ma e ials The signi ican esul s o XWAS, gene‑based associa ion analysis, eQTL analysis, and SMR analysis a e included a icle and i s addi ional iles. The comple e summa y s a is ics gene a ed du ing he cu en s udy a e a ailable om he co esponding au ho on easonable eques . Monocy e da a we e ob ained om public eposi o ies, acco ding o a Da a T ans e Ag eemen be ween he Uni e si y o he Basque Coun y (UPV/EHU) and he Uni e si y o Ox o d. The geno yping da a we e downloaded om he Eu opean Genome‑phenome A chi e (h ps:// www. ebi. ac. uk/ ega/ da as e s/; expe imen EGAS00000000109) and he gene exp ession mic oa ay da a om he same indi iduals we e downloaded om he EBI A ayExp ess da abase (h ps:// www. ebi. ac. uk/ a ay exp e ss/ expe imen s/; expe imen numbe E‑MTAB‑2232). This s udy makes use o he Immnochip da a gene a ed by he Wellcome T us Case–Con ol Conso ium (WTCC da a se s EGAD00010000246, EGAD00010000248 and EGAD00010000250). A ull lis o he in es iga o s who con ibu ed o he gene a ion o he da a is a ailable om h p:// www. w ccc. o g. uk. Funding o he p ojec was p o ided by he Wellcome T us unde awa ds 076113, 085475 and 090355. Decla a ions E hics app o al and consen o pa icipa e The s udies in ol ing human pa icipan s we e e iewed and app o ed by E hical Commi ee o Clinical Resea ch o C uces Uni e si y Hospi al. W i en in o med consen o pa icipa e in his s udy was p o ided by he pa icipan s’ legal gua dian/nex o kin. Consen o publica ion No applicable. Compe ing in e es s The au ho s decla e ha he esea ch was conduc ed in he absence o any comme cial o inancial ela ionships ha could be cons ued as a po en ial con lic o in e es . Au ho de ails 1 Depa men o Gene ics, Physical An h opology and Animal Physiology, Uni e si y o he Basque Coun y (UPV/EHU), Leioa, Basque Coun y, Spain. 2 Biobizkaia Heal h Resea ch Ins i u e, Ba akaldo, Basque Coun y, Spain. 3 Pedia ic Gas oen e ology Uni , C uces Uni e si y Hospi al, Ba akaldo, Basque Coun y, Spain. 4 Spanish Biomedical Resea ch Cen e in Diabe es and Associ‑ a ed Me abolic Diso de s (CIBERDEM), Mad id, Spain. Recei ed: 28 July 2023 Accep ed: 28 No embe 2023 Re e ences 1. Lind o s K, Ciacci C, Ku ppa K, Lundin KEA, Makha ia GK, Mea in ML, e al. Coeliac disease. Na Re Dis P ime s. 2019;5(1):3. 2. Singh P, A o a A, S and TA, Le le DA, Ca assi C, G een PH, e al. Global p e alence o celiac disease: sys ema ic e iew and me a‑analysis. Clin Gas oen e ol Hepa ol. 2018;16(6):823‑836.e2. 3. Jansson‑Knodell CL, Hujoel IA, Wes CP, Taneja V, P okop LJ, Rubio‑Tapia A, e al. Sex di e ence in celiac disease in undiagnosed popula ions: a sys ema ic e iew and me a‑analysis. Clin Gas oen e ol Hepa . 2019;17(10):1954‑1968.e13. 4. Ka ell K, Louka AS, Moodie SJ, Asche H, Clo F, G eco L, e al. HLA ypes in celiac disease pa ien s no ca ying he DQA1 *05‑DQB1 *02 (DQ2) he ‑ e odime : esul s om he Eu opean gene ics clus e on celiac disease. Hum Immunol. 2003;64(4):469–77. 5. Be an S, Popa S, B aegge CP, Busch D, Odonoghue A, Fal h‑Magnusson K, Fe guson A, e al. Con ibu ion o he MHC egion o he amilial isk o coeliac disease. J Med Gene . 1999;36(9):687–90. 6. Dubois PCA, T ynka G, F anke L, Hun KA, Romanos J, Cu o i A, e al. Mul iple common a ian s o celiac disease in luencing immune gene exp ession. Na Gene . 2010;42(4):295–302. 7. T ynka G, Hun KA, Bocke NA, Romanos J, Mis y V, Szpe l A, e al. Dense geno yping iden i ies and localizes mul iple common and a e a ian associa ion signals in celiac disease. Na Gene . 2011;43(12):1193–201. 8. Ga cia‑E xeba ia K, Jau egi‑Miguel A, Rome o‑Ga mendia I, Plaza‑Izu ie a L, Lega da M, I as o za I, e al. Ances y‑based s a i ied analysis o Immu‑ nochip da a iden i ies no el associa ions wi h celiac disease. Eu J Hum Gene . 2016;24(12):1831–4. 9. Gallaghe MD, Chen‑Plo kin AS. The pos ‑GWAS e a: om associa ion o unc ion. Am J Hum Gene . 2018;102(5):717–30. 10. Gu ie ez‑Achu y J, Zhe nako a A, Puli SL, T ynka G, Hun KA, Romanos J, e al. Fine mapping in he MHC egion accoun s o 18% addi ional gene ic isk o celiac disease. Na Gene . 2015;47(6):577–8. 11. Chang D, Gao F, Sla ney A, Ma L, Waldman YY, Sams AJ, e al. Accoun ing o eXen ici ies: analysis o he X ch omosome in GWAS e eals X‑linked genes implica ed in au oimmune diseases. PLoS ONE. 2014;9(12): e113684. 12. Li YR, Li J, Zhao SD, B ad ield JP, Men ch FD, Maggado i SM, e al. Me a‑ analysis o sha ed gene ic a chi ec u e ac oss en pedia ic au oimmune diseases. Na Med. 2015;21(9):1018–27. 13. Wise AL, Gyi L, Manolio TA. eXclusion: owa d in eg a ing he X ch omosome in genome‑wide associa ion analyses. Am J Hum Gene . 2013;92(5):643–7. 14. Wegiel M, An osz A, Giebu owska J, Szeliga K, Hankus M, G zybowska‑ Chlebowczyk U, e al. Au oimmuni y p edisposi ion in gi ls wi h u ne synd ome. F on Endoc inol. 2019;10:511. 15. Ha is VM, Sha ma R, Ca e J, Ku ien BT, Liu K, Koelsch KA, e al. Kline el e ’s synd ome (47, XXY) is in excess among men wi h Sjög en’s synd ome. Clin Immunol. 2016;168:25–9. 16. Liu K, Ku ien BT, Zimme man SL, Kau man KM, Ta DH, Ko yan LC, e al. X ch omosome dose and sex bias in au oimmune diseases: inc eased p e alence o 47, XXX in sys emic Lupus e y hema osus and Sjög en’s synd ome. A h i is Rheuma ol. 2016;68(5):1290–300. 17. Sco ield RH, B une GR, Namjou B, Kimbe ly RP, Ramsey‑Goldman R, Pe i M, e al. Kline el e ’s synd ome (47, XXY) in male sys emic lupus e y he‑ ma osus pa ien s: suppo o he no ion o a gene‑dose e ec om he X ch omosome. A h i is Rheum. 2008;58(8):2511–7. 18. Fish EN. The X‑ iles in immuni y: sex‑based di e ences p edispose immune esponses. Na Re Immunol. 2008;8(9):737–44. 19. Bianchi I, Lleo A, Ge shwin ME, In e nizzi P. The X ch omosome and immune associa ed genes. J Au oimmun. 2012;38(2–3):J187–92. 20. Se bina NV, Jia T, Hohl TM, Pame EG. Monocy e‑media ed de ense agains mic obial pa hogens. Annu Re Immunol. 2008;26:421–52. 21. Cino a J, Palo á‑Jelínko á L, Smy hies LE, Ce ná M, Pecha o á B, D o ák M, e al. Gliadin pep ides ac i a e blood monocy es om pa ien s wi h celiac disease. J Clin Immunol. 2007;27(2):201–9. 22. Jelínko á L, Tucko á L, Cino á J, Flegelo á Z, Tlaskalo á‑Hogeno á H. Gliadin s imula es human monocy es o p oduc ion o IL‑8 and TNF‑alpha h ough a mechanism in ol ing NF‑kappaB. FEBS Le . 2004;571(1–3):81–5. 23. Palo á‑Jelínko á L, Dáňo á K, D aša o á H, D ořák M, Funda DP, Fundo á P e al. Pepsin diges o whea gliadin ac ion inc eases p oduc ion o IL‑1β ia TLR4/MyD88/TRIF/MAPK/NF‑κB signaling pa hway and an NLRP3 in lammasome ac i a ion. PLoS ONE. 2013; 8(4):e62426. 24. Fai ax BP, Humbu g P, Makino S, Na anbhai V, Wong D, Lau E, e al. Inna e immune ac i i y condi ions he e ec o egula o y a ian s upon mono‑ cy e gene exp ession. Science. 2014;343(6175):1246949.