Long-term environmental metal exposure is associated with hypomethylation of CpG sites in NFKB1 and other genes related to oncogenesis

S epanyane al. Clinical Epigene ics (2023) 15:126 h ps://doi.o g/10.1186/s13148-023-01536-3 RESEARCH Open Access © The Au ho (s) 2023, co ec ed publica ion 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 eco mmons. o g/ licen ses/ by/4. 0/. The C ea i e Commons Public Domain Dedica ion wai e (h p:// c ea i eco mmons. o g/ publi cdoma in/ 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. Clinical Epigene ics Long- e m en i onmen al me al exposu e isassocia ed wi hhypome hyla ion o CpG si es inNFKB1 ando he genes ela ed ooncogenesis Ani S epanyan1*, Anna Pe acko a2, Si as Hakobyan1, Jakub Sa a a2,3, Su en Da i a yan1, E a K iego a2 and A sen A akelyan1 Abs ac Backg ound Long- e m en i onmen al exposu e o me als leads o epigene ic changes and may inc ease isks o human heal h. The ela ionship be ween he ype and le el o me al exposu e and epigene ic changes in subjec s exposed o high concen a ions o me als in he en i onmen is no ye clea . The aim o ou s udy is o ind he pos- sible associa ion o en i onmen al long- e m exposu e o me als wi h DNA me hyla ion changes o genes ela ed o immune esponse and ca cinogenesis. We in es iga ed he associa ion o plasma le els o 21 essen ial and non- essen ial me als de ec ed by ICP-MS and he me hyla ion le el o 654 CpG si es loca ed on NFKB1, CDKN2A, ESR1, APOA5, IGF2 and H19 genes assessed by a ge ed bisul i e sequencing in a coho o 40 subjec s li ing nea me al mining a ea and 40 unexposed subjec s. Linea eg ession was conduc ed o ind di e en ially me hyla ed posi ions wi h adjus men o gende , age, BMI class, smoking and me al concen a ion. Resul s In he me al-exposed g oup, i e CpGs in he NFKB1 p omo e egion we e hypome hyla ed compa ed o unexposed g oup. Fou di e en ially me hyla ed posi ions (DMPs) we e associa ed wi h mul iple me als, wo o hem a e loca ed on NFKB1 gene, and one each on CDKN2A gene and ESR1 gene. Two DMPs loca ed on NFKB1 (ch 4:102500951, associa ed wi h Be) and IGF2 (ch 11:2134198, associa ed wi h U) a e associa ed wi h speci ic me al le els. The me hyla ion s a us o he se en CpGs loca ed on NFKB1 (3), ESR1 (2) and CDKN2A (2) posi i ely co ela ed wi h plasma le els o se en me als (As, Sb, Zn, Ni, U, I and Mn). Conclusions Ou s udy e ealed me hyla ion changes in NFKB1, CDKN2A, IGF2 and ESR1 genes in indi iduals wi h long- e m human exposu e o me als. Fu he s udies a e needed o cla i y he e ec o en i onmen al me al exposu e on epigene ic mechanisms and pa hways in ol ed. Keywo ds Toxic me al, En i onmen al exposu e, DNA me hyla ion, NFKB1, CDKN2A, IGF2, ESR1, U anium *Co espondence: Ani S epanyan a_s [email protected] Full lis o au ho in o ma ion is a ailable a he end o he a icle Page 2 o 12 S epanyane al. Clinical Epigene ics (2023) 15:126 Backg ound Toxic me als and me alloids a e one o he mos dange - ous g oups o en i onmen al con aminan s wi h ad e se e ec s on li ing o ganisms. Because o hei bioaccu- mula ion p ope ies, hey en e he ood chain and pass h ough ophic le els in he ecosys em. The e is g ow- ing e idence ha ch onic exposu e o essen ial and non-essen ial me als may inc ease he isk o cance , enal, au oimmune, neu ological, neu odegene a i e, hema ological and ca dio ascula diseases a e long- e m en i onmen al and occupa ional human exposu e [1–3]. Binding o p o eins, al e a ion o hei s uc u e and unc ion and gene a ion o eac i e oxygen species, which damage lipids, p o eins and DNA, a e he p ima y ocus o s udies o molecula mechanisms associa ed wi h ha m ul me als oxici y [4–6]. Toxic me al expo- su e leads o oxida i e/ni osa i e s ess and supp esses cell in insic an ioxidan de ense, which is mani es ed by he deple ion o an ioxidan enzymes, al e a ions o DNA me hyla ion and DNA damage [6–9]. Oxida i e s ess-de i ed epigene ic dys egula ion a gene p omo e s is c i ical o gene exp ession and leads o al e a ions in me abolic, gene ic and signal ansduc- ion pa hways and hei unc ions [8, 10, 11]. I has been shown ha oxic me als induce dys egula ion o an- sc ip ion ac o s ha change si e-speci ic DNA me h- yla ion pa e ns ia al e a ion o DNA accessibili y o DNA me hyla ion machine y [12]. Se e al ecen epi- genome-wide associa ion s udies e ealed me hyla ion changes in genes o cellula esponse o DNA damage (Mn, Cs and Cu) and s ess s imulus (Cu and Se), egula- ion o NF-kappa B signaling (Cs and As), immune and in lamma o y esponses (Se, Cd and As), cell dea h and p oli e a ion (C , Pb, Mn and As), and es ogen signal- ing pa hway (As) in esponse o he exposu e o me als and me alloids [13–15]. These da a a e suppo ed by he in es iga ions o blood ansc ip ome esponse o en i- onmen al hea y me al (Cd, Pb and Hg) exposu e which e ealed exp ession changes o mul iple genes ela ed o cance (RAC1, MAPK1, TP53, UBA52 and NFKB1) and in lamma ion (TYK2, JAK3, IGF2, APOA5 and STAT3) [16, 17]. These esul s a e in conco dance wi h he ind- ings o in i o s udies ega ding hea y me al-induced gene exp ession p o ile changes epo ed by us and o h- e s [18–20]. Using publicly a ailable da a, we p e iously showed al e ed exp ession o genes in ol ed in cell p o- li e a ion, mig a ion and cell–cell signaling, as well as EGFR signaling, cell cycle con ol and posi i e egula ion o T- and B-cell p oli e a ion in li e cells in esponse o in i o exposu e o Cd, Ni and As [20]. His o ically, Kapan has been a majo cen e o he p oduc ion o many non- e ous me als, and a min- ing plan has been ope a ing he e since 1846 [21]. Fi s publica ions on hea y me al en i onmen al pollu ion in his egion a e da ed o 2009 ollowed by se e al s udies iden i ied ha ace elemen s en e he local ood chain in he Kapan mining a ea (A menia) [21–28]. Mines, p ocessing plan s, ac i e and abandoned ailing eposi o- ies and mine wa e , which local people use o i iga ion wi hou any ea men , a e signi ican sou ces o ha m- ul me als in he men ioned a ea [22–27]. Pa icula ly, he ag icul u al soils con ain maximum allowed concen- a ion exceeding he con en s o Cu, Mo, Ni, C , Hg, As and Cd [22–24, 26]. On he e i o y o he Kapan ci y, subs an ial excesses o oxic me al elemen s o bo h I (As, Cd, Ni, Pb and C ) and II–III (Cu, Zn and Mo) ca ego ies o haza d we e de ec ed [24]. Ag icul u al c op pollu ion by C , Ni and Pb was also epo ed in his a ea p e iously [24]. Mean concen a ions o Hg and Pb in some ege a- bles and ui s exceed he maximum accep able le els se by in e na ional o ganiza ions [24, 27]. The e alua ion o combined es ima ed daily in akes o he local ege ables and ui s e ealed exceeded e e ence doses o Cu and Mo. Mo eo e , ca cinogenic isk alues o Ni, C , As and Cd exceeded he US En i onmen al P o ec ion Agency se limi s [28]. Howe e , no s udies ha e been pe o med o de e mine he blood concen a ions o essen ial and non-essen ial elemen s om his mining egion esi- den s, and whe he he e is an epigene ic dys egula ion caused by long- e m me al exposu e. Based on he p e ious indings [13–20], genes ela ed o immune esponse (APOA5, IGF2, H19 and ESR1) and ca cinogenesis (NFKB1 and CDKN2A) we e selec ed o s udy possible DNA me hyla ion changes caused by en i- onmen al long- e m exposu e o oxic me als in indi id- uals li ing in he mining egions o A menia. Resul s Chemical elemen s analysis inplasma The blood plasma samples we e analyzed by induc i ely coupled plasma mass spec ome y (ICP-MS) o ob ain concen a ions o he 29 essen ial and oxic chemical ele- men s. No de ec able le els we e ound o Bi, Au and P in analyzed plasma samples. The measu ed le els o Cd, C , Hg, Sn and Ti we e below he limi o quan i i- ca ion (LOQ), and hese elemen s we e excluded om he u he analyzes. The numbe o samples < LOQ o ou me als we e mo e han hal o he o al subjec s ec ui ed in his s udy o he ollowing elemen s: Al (unexposed/exposed; 29/28), Ba (unexposed/exposed; 38/39), Li (unexposed/exposed; 32/33) and Ag (unex- posed/exposed; 37/38). These ou me als we e included in he eg ession model as ca ego ical a iables (< LOQ— absen in plasma and ≥ LOQ—p esen in plasma). The e was a signi ican ly highe plasma U le el in he exposed g oup compa ed o unexposed (p = 0.00049, Page 3 o 12 S epanyane al. Clinical Epigene ics (2023) 15:126 Fig.1). No s a is ically signi ican di e ences we e ound by non-pa ame ic es ing (Mann–Whi ney U) be ween he wo s udied g oups o he emaining 16 elemen s (Table1). In e -elemen al co ela ion analysis e ealed clus e s o elemen s co ela ed wi h each o he in bo h, exposed and unexposed g oups (Addi ional ile1: Figu esS2–S3). The ollowing clus e s o a mode a e posi i e ela ionship be ween elemen s we e ound o he exposed g oup: (Mn, Fe, Zn and Sb); (Mn, Al and Ba); (Cd, Se and Sn) and (As, Sn and Tl). The co ela ion ma ix o he unex- posed g oup con ains wo posi i ely co ela ed clus e s o (Mn, Fe and Zn) and (Al, Ni and Sn) (co ela ion coe i- cien ≥ 0.4) (Addi ional ile1: Figu esS2–S3). Di e en ially me hyla ed posi ions O e all, he me hyla ion le el o he 654 CpG dinucleo- ides loca ed on NFKB1, CDKN2A, ESR1, APOA5, IGF2 and H19 genes was measu ed in his s udy o all sam- ples om exposed and unexposed g oups (Addi ional ile2: Sp eadshee 3). In he linea model adjus ed o pa icipan s’ gende , age, BMI, BMI class and smok- ing, hea y me al exposu e was associa ed wi h hypo- me hyla ion o i e CpG posi ions in he NFKB1 gene (CpGs a 102501059, 102500993, 102501010, 102500986 and 102500966 posi ions on ch omosome 4) (Addi- ional ile 2: Sp eadshee 3). Mo eo e , hese associa- ions emained signi ican a e adding 20 me al le els as co ac o s in his model. The me hyla ion le els o hese co e i e CpGs in NFKB1 gene showed nega i e co ela- ion wi h he le els o u anium in plasma (Addi ional ile1: Figu e S4). In case o Be exposu e- ela ed eg es- sion, coe icien s o CpGs a 102501010, 102500986 and 102500966 we e non-signi ican (Fig.2, Addi ional ile2: Sp eadshee 4). Addi ional CpGs we e ound o be di - e en ially me hyla ed in a eg ession model upon he inclusion o me al plasma le els. Thus, ch 9:21975053 a CDKN2A gene was di e en ially me hyla ed in exposed g oup s unexposed g oup upon inclusion o mul iple me als (Ba, I, Al, As, Ca, Ag, Cu, Fe, Li, Mg, Mn, Mo, U, Ni, Sb, Se, Tl, Zn, Co and V); NFKB1 gene (ch 4:102500935 and ch 4:102500975) was di e en ially me hyla ed upon inclusion o Al, As, Ca, Ag, Cu, Fe, Li, Mg, Mn, Mo, U, Ni, Sb, Se, Tl, Zn, Co and V (Fig.2, Addi- ional ile2: Sp eadshee 4). The CpG a ch 6:151808587 posi ion o ESR1 gene was signi ican ly hypome hyla ed when adjus ed wi h Co and V plasma le els (Fig.2, Addi- ional ile2: Sp eadshee 4). As o DMPs associa ed wi h speci ic me als, NFKB1 gene ch 4:102500951 and IGF2 gene ch 11:2134198 si e CpGs we e hype me hyla ed when adjus ed o Be and U plasma le els, espec i ely (Fig.2, Addi ional ile2: Sp eadshee 4). No associa ion o me hyla ion s a us o APOA5 gene CpG islands wi h hea y me al exposu e was ound in ou s udy (Fig. 2, Addi ional ile2: Sp eadshee 4). Me hyla ions o CpGs associa ed wi hplasma le els o chemical elemen s O e all, he me hyla ion le els o he se en CpG si es we e associa ed wi h plasma le els o se en chemical elemen s in all pa icipan s in ol ed in his s udy (Fig. 3, Addi- ional ile2: Sp eadshee 5). Th ee CpGs a NFKB1 gene: ch 4:102501416, ch 4:102501137 and ch 4:102501517 si es we e associa ed wi h As, Sb and Zn plasma le els, espec i ely. Me hyla ion le els o he wo CpG dinucleo- ides a ESR1 gene: ch 6:151808176 and ch 6:151808606 showed posi i e associa ion wi h Ni and U plasma le els, espec i ely. Finally, CDKN2A gene ch 9:21974637 and ch 9:21975132 cy osines we e hype me hyla ed when associa ed wi h he I and Mn le els in plasma, espec i ely (Fig.3, Addi ional ile2: Sp eadshee 5). Fig. 1 Di e ence o u anium plasma le els be ween exposed and unexposed subjec s Page 4 o 12 S epanyane al. Clinical Epigene ics (2023) 15:126 Table 1 Clinical and demog aphic cha ac e is ics o he s udy g oups LOD limi o de ec ion, LOQ limi o quan i ica ion, IQR in e qua ile ange and NA no applicable a Exposed indi iduals a e om he mining a eas: Syunik illage, A s anik illage and Kapan ci y (A menia) b Unexposed indi iduals a e ec ui ed om he Ye e an ci y (A menia) Cha ac e is ics Exposed indi idualsa (n = 40) Unexposed indi idualsb (n = 40) p alue Numbe (%) o median (IQR) Numbe (%) o median (IQR) Age (yea s) 35.5 (32, 38.25) 32 (31, 36) 0.01 Gende Female 25 (62.5%) 22 (55%) 0.53 Smoking Smoke s 13 (32.5%) 14 (35%) 0.88 BMI (kg/m2) 27.2 (22.87, 29.76) 24.25 (21.16, 28.67) 0.06 BMI class 0.05 No mal (18.5–25 kg/m2) 15 (37.5%) 24 (60%) O e weigh (25–30 kg/m2)17 (42.5%) 12 (30%) Obesi y (≥ 30 kg/m2) 8 (20%) 4 (10%) Chemical elemen s in plasma (ppb) Ca 112 × 103(106 × 103,116 × 103)110 × 103(103 × 103, 117 × 103) 0.37 Mg 23 × 103 (22 × 103, 24 × 103)23 × 103 (21 × 103, 24 × 103) 0.18 Co 0.41 (0.30, 0.60) 0.49 (0.37, 0.63) 0.09 Fe 2165 (1655, 2815) 2280 (1760, 2690) 0.95 I 81 (74, 86) 75 (69, 84) 0.06 Mn 2.7 (2.4, 3.2) 2.6 (2.4, 3.3) 0.78 Cu 1240 (1100, 1340) 1170 (1010, 1280) 0.40 Mo 1.15 (0.90, 1.90) 1.24 (1.03, 1.74) 0.26 Se 128.5 (118, 141) 135 (126, 142) 0.18 C < 8.312 (LOQ) NA Zn 1540 (1410, 1780) 1590 (1440, 1670) 0.97 Cd < 0.378 (LOQ) NA As 2 (2, 2) 2 (2, 3) 0.23 Hg < 4.042 (LOQ) NA Sb 8 (7, 9) 8 (7, 9) 0.73 Tl 0.026 (0.015, 0.035) 0.031 (0.021, 0.045) 0.07 Al 260 (240, 315) 260 (230, 310) 0.31 Ba 2 (2, 2) 2 (2, 3) 0.42 Be 0.2 (0.1, 0.2) 0.2 (0.1, 0.2) 0.48 V 4 (4, 4) 4 (4, 4) 0.27 Bi < 0.002 (LOD) NA Au < 0.06 (LOD) NA Li 4 (4, 5) 4 (3, 5) 0.09 Ni 9 (8, 9) 9 (8, 10) 0.52 Sn < 2.558 (LOQ) NA P < 0.003 (LOD) NA Ag 0.14 (0.10, 0.10) 0.19 (0.10, 0.10) 0.86 Ti < 12.259 (LOQ) NA U 0.020 (0.018, 0.023) 0.016 (0.014, 0.021) 0.00049 Page 5 o 12 S epanyane al. Clinical Epigene ics (2023) 15:126 Fig. 2 Di e en ially me hyla ed posi ions in associa ion wi h he plasma le els o chemical elemen s. Sankey diag am isualizes he associa ion be ween me hyla ion s a us o de e en ially me hyla ed CpGs and plasma concen a ions o chemical elemen s. The x-axis ep esen s di e en ially me hyla ed CpGs and chemical elemen s. The nodes on y-axis ep esen he posi ions o CpGs de e en ially me hyla ed in associa ion wi h chemical elemen s, numbe o la e is p opo ional o he size o each node and wid h o each a c Fig. 3 CpGs associa ed wi h plasma le els o chemical elemen s. Sankey diag am isualizes he associa ion be ween me hyla ion s a us o CpGs ( igh ) and plasma concen a ions o chemical elemen s (le ) Page 6 o 12 S epanyane al. Clinical Epigene ics (2023) 15:126 Discussion This s udy explo ed he e ec s o long- e m human en i onmen al exposu e o hea y me als on DNA me h- yla ion o immune esponse and ca cinogenesis- ela ed genes. We iden i ied i e posi ions o CpGs (co e CpGs) in he NFKB1 p omo e egion which we e hypome h- yla ed in he exposed g oup compa ed o unexposed. O e all, we ound ou DMPs associa ed wi h mul iple me als as well as wo DMPs wi h speci ic me al exposu e. Acco dingly, wo o hem a e loca ed on he NFKB1 gene (all me als, excep Be, Ba and I), one on he CDKN2A gene (all me als, excep Be) and one on he ESR1 gene (Co and V). The DMPs associa ed wi h speci ic me al le els a e loca ed on NFKB1 (ch 4:102500951, associa ed wi h Be) and IGF2 (ch 11:2134198, associa ed wi h U). Addi ionally, we iden i ied CpGs a which DNA me h- yla ion was associa ed wi h me al plasma le els o all g oups in ol ed in his s udy. In gene al, he me hyla ion s a us o he se en CpGs loca ed on NFKB1 (3), ESR1 (2) and CDKN2A (2) was posi i ely co ela ed wi h plasma le els o se en me als (As, Sb, Zn, Ni, U, I and Mn). The co e CpGs, hypome hyla ed in he exposed g oup, we e loca ed on he p omo e o NFKB1 gene encoding 105-kD ansc ip ion ac o (TF) p ecu so , which unde - goes co ansla ional p ocessing o o m DNA-binding p50 p o ein [29]. The p50 p o ein ac s in NF-κB signal- ing pa hway as a ansc ip ion ac i a o o ep esso , assembling wi h Rel-p o eins o by homodime iza ion, espec i ely [29]. NF-κB plays a c ucial ole in cell su - i al, g ow h, immune esponse and in lamma ion since i con ols cell esponse o s ess, ee adicals, i a- dia ion and o eign an igens [30, 31]. Se e al p e ious epigenome-wide associa ion s udies (EWAS) epo ed he in ol emen o NF-κB pa hway genes in DNA me h- yla ion changes induced by hea y me al en i onmen al, occupa ional and p ena al exposu e [13, 14, 32]. Also, a ecen s udy on a senic-exposed 396 Bangladeshi adul s e ealed among di e en ially me hyla ed genes signi i- can en ichmen o hose anno a ed o eac i e oxygen species pa hway, in lamma o y esponse and nuclea ac o kappa-B signaling [14]. Zeng e al. epo ed di - e en ial me hyla ion o 125 CpGs mapped o 79 genes in he e-was e-exposed g oup wi h highe concen a ions o Pb, Cd, Mn and C in neona al umbilical co d blood [32]. These genes a e in ol ed in signaling pa hways ela ed o NF-κB ac i a ion, adhe ens junc ion, TGF be a and apop osis [33]. Ano he s udy showed ha p ena al exposu e o 12 me als (As, Cd, C , Cs, Cu, Hg, Mg, Mn, Se and Zn) associa ed wi h DMPs loca ed in genes asso- cia ed wi h he egula ion o NF-κB signaling as well as se e al biological p ocesses such a e neu ode elopmen , in lamma o y esponse, cellula esponse o s ess and DNA damage and apop osis [13]. In addi ion o he CpGs di e en ially me hyla ed in esponse o speci ic me als, he au ho s epo ed DMPs anno a ed o genes in human leukocy e an igen (HLA) egion o e lapping o C , Cs, Cu, Hg, Mg and Mn me als [13]. Impo an ly, he oxico- logical impac o me als is ele an o mix u es and con- cen a ions hey co-exis in he en i onmen [33]. The s udy ound ha in T-cell cul u e, a senic exposu e only s exposu e o a senic and u anium mix u e ac i a es di e en signaling pa hways, including NF-κB ac i a ion in esponse o a senic and u anium co-exposu e [33]. In ou s udy, we ound wo DMPs loca ed in he NFKB1 p omo e egion, which we e associa ed wi h mul iple me als exposu e, and one CpG dinucleo ide which was di e en ially hypome hyla ed in esponse o Be. Though no s udy o da e has epo ed epigene ic dys egula ion ela ed o Be en i onmen al exposu e, be yllium luo- ide in i o on mu ine pe i oneal mac ophages a low concen a ion a ec s he le el o NF-κB and by ele a ing Ca2+ igge s he ac i a ion o p21 as-dependen MAPK signaling cascade [34]. Ou wo k is he i s s udy p o iding he e e ence al- ues o a wide ange o essen ial and non-essen ial ace elemen s in he blood plasma o adul s om he A me- nian popula ion. Compa ed o o he da a p o ided by he s udies done o da e, gene ally, he ob ained concen- a ions we e compa able o o he coun ies. Howe e , he e we e ends o inc eased plasma le els o se e al non-essen ial ace elemen s, such as nickel, an imony and anadium. Thus, ela i ely high Ni and Sb le els we e de ec ed in ou s udy (9 and 8ppb, espec i ely) com- pa ed o hose epo ed o popula ions in Ge many (0.11 and < 0.013 ppb, espec i ely), Aus alia (2 and 4 ppb, espec i ely) and F ance (1.49 and 0.25ppb, espec i ely) [35–37]. Plasma concen a ions o V ound o A meni- ans (4ppb) we e highe han in Ge mans and Aus alians desc ibed ea lie (0.052 and 0.2ppb, espec i ely) [35, 36]. Ob ained popula ion di e ences o ace elemen plasma concen a ions migh be associa ed wi h gene ic as well as en i onmen al, geog aphical, nu i ional and li es yle ea u es o he A menian popula ion [38, 39]. Acco ding o he co ela ion plo s, di e en clus e s o mode a e posi i ely co ela ed me als we e ound in he me al-exposed e sus unexposed g oup. The in e - elemen al co ela ions o exposed subjec s ha e g ea e s eng h and highe numbe o clus e s compa ed o unexposed ones. The ob ained dissimila i y o he co - ela ion ma ices be ween he wo s udy g oups migh e lec he di e en local en i onmen s o he people om he mining egion and unexposed subjec s. Ele a ed le els o u anium in he exposed g oup ound in ou s udy we e associa ed wi h me hyla ion changes o CpG si es in NFKB1, IGF2 and CDKN2A genes. Ve y ew s udies o da e ha e been ca ied ou o in es iga e Page 7 o 12 S epanyane al. Clinical Epigene ics (2023) 15:126 he in luence o u anium exposu e on epigene ic changes in he human genome. U anium-induced abe an global DNA me hyla ion was epo ed by di e en in i o s ud- ies as well as by in es iga ions on epigene ic dys egula- ions caused by deple ed u anium human low-dose exposu e [40–42]. The inac i a ion o CDKN2A umo supp esso gene and MGMT (O6-me hylguanine-DNA me hyl ans e ase) epai gene by al e ed p omo e me hyla ion among Chinese mine s o U has been ecen ly ound [43]. These esul s a e in conco dance wi h ou inding on hypome hyla ion o in onic CpG o CDKN2A gene. CDKN2A gene locus encodes wo umo supp esso p o eins p16/INK4A and P14/ARF, which egula e cell cycle by cyclin D inhibi ion, and i s inac i a- ion leads o a umo igenesis h ough he PI3K/Ak pa h- way [44]. Recen s udies p o ide e idence o PI3K and MAPK pa hways’ in ol emen in u anium-induced oxic- i y [45, 46]. In addi ion, he ac i a ion o he NF-κB/IL-6 pa hway in esponse o INK4a/ARF inhibi ion igge - ing umo g ow h and de elopmen has been desc ibed ecen ly [47]. Fu he , co-exposu e o sodium a seni e and u anyl ace a e al e s gene exp ession in T cells by posi i e egula ion o NF-κB ansc ip ion ac o s [33]. The oxida i e s ess-induced ac i a ion o he NF-κB leads o IGF2 hype me hyla ion in mice models [48]. A highly signi ican inc ease in he exp ession o he IGF2 gene in esponse o exposu e o i e hea y me als (Cd, C , Cu, Pb and Zn) was epo ed in ishe men’s blood [49]. The IGF2 gene epigene ic dys egula ion caused by hea y me als exposu e was ound by di e en g oups in he p e ious s udies as well [50, 51]. In he cu en s udy, we e ealed ha U and Ni le els we e indepen- den ly co ela ed wi h hype me hyla ion o CpGs loca ed in he in on2 and exon3 o he ESR1 gene, espec i ely. Fu he , we ound he in onic DMP in he ESR1 gene o be associa ed wi h Co and V plasma le els and hypo- me hyla ed in he exposed g oup. E idence ob ained omepidemiological, in i oandin i os udies sug- ges s ac i a ion o es ogen ecep o signaling pa hway as well as al e a ions o ESR1 gene me hyla ion s a us in esponse o hea y me als exposu e [14, 52, 53]. Acco d- ing o he ecen EWAS on a senic en i onmen al expo- su e, es ogen ea ly esponse was among pa hways o he anno a ed genes wi h a senic-associa ed CpGs [14]. In addi ion, he e idence o o e exp ession o ESR1 gene esul ing om a senic-induced p omo e egion hype - me hyla ion was obse ed in ac i a ed hepa ocellula p oli e a ion [53]. Fu he , i was shown ha Cd ea - men s imula es b eas cance cell p oli e a ion by ac i- a ing ERα-dependen PI3K-Ak signaling pa hway [52]. D inking wa e wi h a low dose o u anium causes es o- gen ecep o (ER)-dependen esponses in emale mice [52]. One possible explana ion o he g owing e idence o hea y me al-induced ER signaling ac i a ion migh be he abili y o some me als, so-called me alloes ogens, including Ni, Co and V, o mimic es ogen [54]. By bind- ing o ERα, hese me als ac i a e i and lead o bo h di ec ansc ip ion ac i a ion o a ge genes in he nucleus o swi ch apid nongenomic pa hways ia signaling o mi o- gen-ac i a ed p o ein kinases (MAPKs) [55]. Al oge he , wi h suppo s o ou s udy esul s on epi- gene ic changes a NFKB1, IGF2, ESR1 and CDKN2A genes, hese indings elucida e he dys egula ion o PI3K, MAPK and NF-κB pa hways in al e a ions o immune esponse and oncogenesis induced by hea y me al expo- su e [56]. Howe e , u he esea ch is needed o con i m hese esul s and o ind ou he epigene ic mechanisms o he hea y me al-induced al e a ions o a o emen ioned signaling pa hways. No associa ion o me hyla ion s a us o APOA5 gene CpG islands wi h hea y me al exposu e was ound in ou s udy. Howe e , i was p e iously epo ed abou signi i- can non-p omo e -associa ed inc eases in DNA me h- yla ion o APOA5 in blood cells o Pb-exposed women’s g andchild en [57]. The absence o associa ion ound in ou s udy migh be pa ially explained by hype me h- yla ion and lack o mRNA o APOA5 in blood cells [58]. Ano he eason migh be ha Pb blood le el was no in es iga ed in ou s udy. This s udy has se e al limi a ions. The i s limi a ion is he mode a e sample size, based on he s ic selec ion o subjec s li ing long- e m in he mining a ea. In addi ion o me als in es iga ed in his s udy, o he me als and pos- sible oxic agen s/pollu an s and hei combina ions may induce epigene ic changes and should be in es iga ed in he u u e s udies. Conclusions In he cu en s udy, we in es iga ed he e ec o long- e m hea y me al exposu e on me hyla ion changes in human leukocy es. We iden i ied hypome hyla ion o i e CpG si es in he NFKB1 p omo e egion in he min- ing egion esiden s g oup. Addi ionally, me hyla ion changes o CpG si es in NFKB1, CDKN2A, IGF2 and ESR1 genes o be associa ed wi h mul iple and speci ic me al exposu e. The consequences o hese hea y me als long- e m exposu e on gene exp ession and NF-κB pa h- way should be u he in es iga ed. New s udies mus be ca ied ou on la ge sample coho s o eplica e ou esul s. Me hods S udy popula ion Exposed and unexposed subjec s (n = 80) we e selec ed based on he dis ance o he mine as well as epo s on hea y me al le els in he habi a ion a ea [23, 24]. In Page 8 o 12 S epanyane al. Clinical Epigene ics (2023) 15:126 o al, 40 subjec s (M/F; 15/25, age (mean, min–max); 35.6, 29–43), who a e li ing nea he mining egion and egula ly consume ege ables, ui s and dai y wi h ace elemen s concen a ions exceeding he maximum accep - able le els (11—Syunik illage, 11—A s anik illage and 18—Kapan ci y, A menia) we e ec ui ed in a me al- exposed g oup (Addi ional ile1: Figu e S1). In his a ea, he epo ed en i onmen al high le el o hea y me als is caused by an h opogenic p essu e as well as na u al geo- genic sou ce [22–26, 59]. Fo y subjec s (M/F; 18/22, age (mean, min–max); 33.4, 28–42) li ing in Ye e an ci y, A menia, we e ec ui ed in he unexposed g oup since he e is no mining ac i i y a ound he ci y. All subjec s we e o A menian descen and had li ed hei en i e li es in he espec i e loca ion (Table1). The subjec s we e selec ed based on a ques ionnai e ha included in o ma ion abou age, esidence ime in he espec i e locali y, smoking habi , use o d ugs, alco- hol, illnesses and ches X- ay. The ollowing exclusion c i e ia we e used: (a) me abolic and acu e in lamma o y diseases; (b) he use o mine al supplemen s; (c) excessi e alcohol consump ion; (d) he p esence o me al implan s; (e) ege a ianism and o he speci ic ea ing habi s; ( ) su - ge y o exposu e o adia ion o any chemicals a leas a mon h be o e sampling and (g) p egnancy.BMI was calcula ed using he s anda d o mula: The body mass di ided by he squa e o he body heigh . BMI ca ego ies we e de ined acco ding o he s anda ds o he Wo ld Heal h O ganiza ion (WHO) [60]. The de ails o demog aphic and clinical cha ac e is ics o he s udy g oups a e p esen ed in Table1. A he ime o sample collec ion, all subjec s p o ided in o med consen abou he use o he blood o planned analysis. The s udy was app o ed by he E hic Commi ee (IRB/IEC:IRB00004079, App o al#1/2020). Analysis o  hechemical elemen s inplasma Plasma samples o 80 subjec s ec ui ed in his s udy we e sepa a ed om whole blood by cen i uga ion (2000×g, 10min, 4°C) and s o ed a −40°C o u he chemical analysis. P ep ocessing o he plasma samples was pe o med be o e analysis by adding 15 olumes o diluen (1% 1-bu anol, 0.1% T i on × 100 and 0.07% HNO3 in dis illed deionized wa e (18 MΩ·cm) pH = 2.0) o he 0.8mL o plasma as desc ibed ea lie [61]. The concen a ions o 29 essen ial and oxic/po en ially oxic ace elemen s (calcium (Ca), magnesium (Mg), cobal (Co), i on (Fe), iodine (I), manganese (Mn), coppe (Cu), molybdenum (Mo), selenium (Se), ch omium (C ), zinc (Zn), cadmium (Cd), a senic (As), me cu y (Hg), an imony (Sb), hallium (Tl), aluminum (Al), ba ium (Ba), be yllium (Be), anadium (V), bismu h (Bi), gold (Au), li hium (Li), nickel (Ni), in (Sn), pla inum (P ), sil e (Ag), i anium (Ti) and u anium (U)) in plasma samples we e analyzed wi h NexION 300D (Pe kinElme Inc., Shel on, CT 06484, USA) ICP-MS. The Dynamic Reac- ion Cell (DRC) echnology was used o he emo al o a omic in e e ences. Calib a ion o he sys em was done acco ding o he manu ac u e ’s guidelines using S and- a ds Ki s (Pe kinElme Inc., Shel on, CT 06484, USA). The online in e nal s anda diza ion wi h hodium-103 was pe o med o accoun o he incomple e acidi y and iscosi y ma ching be ween calib a ion and sample ma ices. In alabo a o y con ol was pe o med using he e e ence ma e ials ClinCheck Plasma Con ol (REC- IPE Chemicals + Ins umen s GmbH, Ge many). We applied he ollowing quali y con ol measu es: A coe icien o a ia ion (CV) was below 15% o in a- and in e -day p ecision, he eco e y a es o all elemen s we e wi hin he limi o 90–110%. The limi o de ec ion (LOD) and LOQ alues o each elemen we e measu ed (Addi ional ile1: TableS1). In his analysis, we include only elemen s ha ha e ≥ LOQ. DNA ex ac ion andbisul i e ea men Pe iphe al enous blood samples we e collec ed in K3EDTA ubes (4mL) om he me al-exposed and unex- posed indi iduals. Genomic DNA was isola ed om pe iphe al blood cells acco ding o he s anda d sal ing ou me hod wi h a modi ica ion o adding chlo o o m s ep [62]. All DNA samples had A260/A280 a io wi hin 1.8–2.0 ange and a concen a ion o mo e han 50ng/ μL. The EZ DNA Me hyla ion-Gold™ Ki (Zymo Resea ch, USA) was used o he bisul i e con e sion o DNA sam- ples. Bisul i e ea men o samples was done acco ding o he manu ac u e ’s ins uc ions wi h some changes a incuba ion and elu ion s ages o educe sal s in inal bisul i e-con e ed DNA samples (BS-DNA) and deg- ada ion o DNA. Thus, he i s incuba ion was done as 90°C o 10min, and elu ion wi h 20 μL o TE bu e was ollowed by cen i uga ion 5min a 2500g a he las s ep. The inpu o DNA was be ween 400 and 500ng, and quan i ica ion o BS-DNA was done by spec opho om- e e wi h RNA se ings. To con ol he success o he con e sion and possible agmen a ion o BS-DNA, he BS-PCR was pe o med wi h p ime s a ge ing di e en leng hs o empla e DNA (Addi ional ile1: TableS2). The p oduc sizes we e ana- lyzed by he mic ochip elec opho esis sys em (MCE™- 202 Mul iNA, Shimadzu, Japan). Lib a y p epa a ion andsequencing The egions o in e es in he ESR1, IGF2, H19, APOA5, NFKB1 and CDKN2A genes we e selec ed based on Page 9 o 12 S epanyane al. Clinical Epigene ics (2023) 15:126 sequences om UCSC genome b owse (Genome Re - e ence Conso ium Human Build 38) [63] and li e a u e e iew. By design, a ge s we e en iched o di e en- ially me hyla ed egula o y egions and TF binding si es, which p edominan ly ca y H3K27ac and H3K4me3 anno a ions, mainly loca ed in p omo e s, CpG Islands (Addi ional ile2: Sp eadshee 1). The p ime design o a ge loca ions was pe o med using BS-DNA as a em- pla e wi h he online ool BiSea ch [64] and he ollowing es ic ions: (a) leng h o PCR p oduc up o 290bp; (b) he simila Tm o p ime pai s and among all p ime s, p e e ably ± 2°C and (c) leng h o p ime s: om 15 up o 35 bases. To e alua e he designed p ime s, he BS-PCR was pe - o med wi h genomic DNA and wa e as a nega i e con- ol as well as wo quali y con ol p ime s. The p oduc sizes we e analyzed by he Mul iNA mic ochip elec o- pho esis sys em o exclude non-speci ic ampli ica ion. Fu he , 54 pai s o bisul i e-speci ic p ime s (Addi ional ile1: TableS2) we e selec ed and used o a ge en ich- men applying o he 48.48 Access A ay In eg a ed Fluidic Ci cui (Fluidigm, USA). In o al, 54 indexed amplicons we e syn hesized o each sample and deep- sequenced on MiSeq (Illumina, USA) as epo ed p e i- ously [65]. Da a p ocessing BiSul i e Bol (BSBol ) was used o he alignmen o sequencing eads o a e e ence o hg38 genome [66]. To dis inguish ue C/T om unme hyla ed cy osines, BS-SNPe was used o mo e accu a e quan i ica ion o me hyla ion le els [67]. Me hyla ion le els in CpG and CHH (H = A, C o T) con ex wi h a minimum quali y sco e o 20 we e calcula ed and subjec ed o he ollow- ing quali y con ol s eps: (a) The samples wi h mo e han 98% con e sion e iciency conside ing me hyla ion le els a CHH con ex we e included in he analysis (Addi ional ile2: Sp eadshee 2); (b) PCR a i ac s and non-speci ic PCR p oduc s we e excluded om analysis and (c) co e - age cu o o 1000 eads, CpGs wi h less han 1000 eads we e excluded om downs eam analysis. S a is ical analysis Medians and in e qua ile anges (IQRs) o con inu- ous a iables and equencies/p opo ions o ca - ego ical a iables we e calcula ed o he clinical and demog aphic cha ac e is ics o he s udy g oups. The di - e ences be ween me al-exposed and unexposed g oups we e assessed by applying he Mann–Whi ney es o con inuous a iables and he Chi-squa ed es o ca - ego ical a iables. Spea man co ela ion was used o analyze ela ionships be ween concen a ions o me - als in exposed and con ol g oups. The M- alue was calcula ed o each CpG posi ion as he log2 a io o he eads con aining me hyla ed e sus unme hyla ed cy o- sine o a gi en CpG posi ion [68]. Linea eg ession was conduc ed o ind DMPs using he R package limma [69] wi h adjus men o gende , age, BMI class, smoking and chemical elemen s concen a ion ( o each elemen sepa- a ely). The Benjamini–Hochbe g alse disco e y a e (FDR) me hod o adjus men o p alues o mul iple compa isons (FDR < 0.05 is e ained as signi ican ) was used. Abb e ia ions Ag Sil e Al Aluminum As A senic Au Gold Ba Ba ium Be Be yllium Bi Bismu h BS-DNA Bisul i e-con e ed DNA samples Ca Calcium Cd Cadmium Co Cobal C Ch omium Cu Coppe CV Coe icien o a ia ion DMP Di e en ially me hyla ed posi ion ER Es ogen ecep o EWAS Epigenome-wide associa ion s udies FDR False disco e y a e Fe I on Hg Me cu y HLA Human leukocy e an igen I Iodine ICP-MS Induc i ely coupled plasma mass spec ome y IQR In e qua ile ange Li Li hium LOD Limi o de ec ion LOQ Limi o quan i ica ion MAPK Mi ogen-ac i a ed p o ein kinase Mg Magnesium Mn Manganese Mo Molybdenum Ni Nickel P Pla inum Sb An imony Se Selenium Sn Tin TF T ansc ip ion ac o Ti Ti anium Tl Thallium U U anium V Vanadium WHO Wo ld Heal h O ganiza ion Zn Zinc 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/ s13148- 023- 01536-3. Addi ional ile1: This ile con ains in o ma ion abou he de ec ion limi s o chemical elemen s measu ed in plasma (Table S1), bisul i e-speci ic p ime s and a ge egions spanning CpG islands (Table S2), map o he pollu ed egion whe e exposed subjec s a e li ing (Figu e S1), co ela ion