Estradiol deficiency and skeletal muscle apoptosis : Possible contribution of microRNAs

This is a sel -a chi ed e sion o an o iginal a icle. This e sion may di e om he o iginal in pagina ion and ypog aphic de ails. Au ho (s): Ti le: Yea : Ve sion: Copy igh : Righ s: Righ s u l: Please ci e he o iginal e sion: CC BY-NC-ND 4.0 h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/ Es adiol de iciency and skele al muscle apop osis : Possible con ibu ion o mic oRNAs © 2021 he Au ho s Published e sion Ka inen, Si a; Juppi, Hanna-Kaa ina; Le, Gengyun; Cabelka, Ch is ine A.; Made , Ta a L.; Lowe, Dawn A.; Laakkonen, Eija K. Ka inen, S., Juppi, H.-K., Le, G., Cabelka, C. A., Made , T. L., Lowe, D. A., & Laakkonen, E. K. (2021). Es adiol de iciency and skele al muscle apop osis : Possible con ibu ion o mic oRNAs. Expe imen al Ge on ology, 147, A icle 111267. h ps://doi.o g/10.1016/j.exge .2021.111267 2021 Expe imen al Ge on ology 147 (2021) 111267 A ailable online 4 Feb ua y 2021 0531-5565/© 2021 The Au ho s. Published by Else ie Inc. This is an open access a icle unde he CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/). Es adiol de iciency and skele al muscle apop osis: Possible con ibu ion o mic oRNAs Si a Ka inen a , * , Hanna-Kaa ina Juppi a , Gengyun Le b , Ch is ine A. Cabelka b , c , Ta a L. Made b , d , Dawn A. Lowe b , Eija K. Laakkonen a a Ge on ology Resea ch Cen e , Facul y o Spo and Heal h Sciences, Uni e si y o Jy ¨ askyl¨ a, Jy ¨ askyl¨ a, Finland b Di isions o Rehabili a ion Science and Physical The apy, Depa men o Rehabili a ion Medicine, Medical School, Uni e si y o Minneso a, Minneapolis, MN, USA c Depa men o Physical The apy, The College o S . Scholas ica, Dulu h, MN, USA d Heal h, Physical Educa ion, and Exe cise Science Depa men , Di ision o P o essional S udies, Augsbu g Uni e si y, Minneapolis, MN, USA ARTICLE INFO Sec ion Edi o : We ne Zwe schke Keywo ds: Menopause O a iec omy Muscle mass Caspase Cy och ome C ABSTRACT Backg ound: Menopause leads o es adiol (E 2 ) de iciency ha is associa ed wi h dec eases in muscle mass and s eng h. He e we s udied he e ec o E 2 de iciency on mic oRNA (miR) signaling ha a ge s apop o ic pa hways. Me hods: C57BL6 mice we e di ided in o con ol (no mal es ous cycle, n =8), OVX (E 2 de iciency, n =7) and OVX +E 2 g oups (E 2 -pelle , n =4). Six weeks ollowing he OVX su ge y, mice we e sac i iced and RNA isola ed om gas ocnemius muscles. miR-p o iles we e s udied wi h Nex -Gene a ion Sequencing (NGS) and candida e miRs e i ied using qPCR. The a ge p o eins o he miRs we e ound using in silico analysis and measu ed a mRNA (qPCR) and p o ein le els (Wes e n blo ). Resul s: O he apop osis-linked miRs p esen , ele en (miRs-92a-3p, 122-5p, 133a-3p, 214-3p, 337-3p, 381-3p, 483-3p, 483-5p, 491-5p, 501-5p and 652-3p) indica ed di e en ial exp ession be ween OVX and OVX +E 2 mice in NGS analysis. In qPCR e i ica ion, muscle om OVX mice had lowe exp ession o all ele en miRs compa ed wi h OVX +E 2 (p <0.050). Acco dingly, OVX had highe exp ession o cy och ome C and caspases 6 and 9 compa ed wi h OVX +E 2 a he mRNA le el (p <0.050). A he p o ein le el, OVX also had lowe an i- apop o ic BCL-W and g ea e p o-apop o ic cy och ome C and ac i e caspase 9 compa ed wi h OVX +E 2 (p < 0.050). Conclusion: E 2 de iciency down egula ed se e al miRs ela ed o apop o ic pa hways hus eleasing hei a ge s om miR-media ed supp ession, which may lead o inc eased apop osis and con ibu e o educed skele al muscle mass. 1. In oduc ion Le els o es ogens decline du ing menopause and his is associa ed wi h dec eases in muscle mass and s eng h (Juppi e al., 2020; Mal ais e al., 2009; Bonda e e al., 2020). We and o he s ha e shown ha es ogen eplacemen he apy pa ially o se s hese un a o able changes in skele al muscle mass and unc ion in pos menopausal women (Sipila e al., 2001; Taa e e al., 2005; Ronkainen e al., 2009; G eising e al., 2009). Simila ly, es ogen de iciency in animal models has been shown o media e dec emen s in muscle s eng h (G eising e al., 2009; Mo an e al., 2006; Collins e al., 2019a) and ea men wi h es adiol (E 2 ) e e ses o p e en s s eng h loss (G eising e al., 2009; Mo an e al., 2007; Schneide e al., 2004). Fu he mo e, i was ound ha o a ian ho mones ha e a c i ical ole in he eg ow h o a ophied skele al muscle and in he egula ion o muscle s em cell unc ion and egene a ion in emale mice (Collins e al., 2019a; Si nick e al., 2006; McClung e al., 2006; Moo e e al., 2019). Ye he mechanis ic ole o es ogens in he loss o muscle mass has no been es ablished. Abb e ia ions: AIF, Apop osis inducing ac o ; BCL2, B-cell lymphoma-2 egula o p o ein; BCL-XL, B-cell lymphoma-ex a-la ge egula o p o ein; BCL-W, B-cell lymphoma-like p o ein 2; CASP, Caspase; cy C, Cy och ome C; E 2 , Es adiol; FasL, Fas ligand; GAPDH, Glyce aldehyde 3-phospha e dehyd ogenase; HSP, Hea shock p o ein; miR, mic oRNA; NGS, Nex -gene a ion sequencing; OVX, O a iec omy. * Co esponding au ho a : P.O. Box 35 (VIV148), FI-40014 Uni e si y o Jy ¨ askyl¨ a, Finland. E-mail add ess: [email p o ec ed] (S. Ka inen). Con en s lis s a ailable a ScienceDi ec Expe imen al Ge on ology jou nal homepage: www.else ie .com/loca e/expge o h ps://doi.o g/10.1016/j.exge .2021.111267 Recei ed 28 Augus 2020; Recei ed in e ised o m 25 Janua y 2021; Accep ed 27 Janua y 2021 Expe imen al Ge on ology 147 (2021) 111267 2 The degene a i e loss o skele al muscle mass, quali y, and s eng h associa ed wi h ageing is e med sa copenia (Dohe y, 2003). Bo h muscle mass and ibe numbe dec ease signi ican ly wi h ageing (B own e al., 1992; E ans, 1995). P og ammed cell dea h e med apop osis is a highly coo dina ed signaling cascade leading o elimina- ion o cells. Apop osis is a key mechanism in no mal de elopmen o mul icellula o ganisms and is in ol ed in cell u no e and on ogenesis (S elle , 1995). Apop osis has been p oposed o be a key signaling ou e also in skele al muscle homeos asis, including muscle ageing and sa - copenia (Di ks and Leeuwenbu gh, 2002; Ma ze i and Leeuwenbu gh, 2006). The e a e wo majo apop o ic signaling pa hways: he ex insic (dea h ecep o ) and in insic (mi ochond ial) apop o ic pa hways (Fig. 1) (Hassan e al., 2014). The ex insic pa hway o apop osis is media ed ia he umo nec osis ac o (TNF) amily ecep o s, such as Fas ligand (FasL), ini ia ed by ex e nal s imuli ha leads o ac i a ion o caspases (CASP) 8 and 10 (Hassan e al., 2014; Locksley e al., 2001). Consecu i ely, he in insic apop o ic pa hway can be igge ed by a a ie y o in acellula s imuli, such as oxida i e s ess and DNA damage (Wu and B a on, 2013). In he in insic pa hway, apop osis is induced ia elease o cy och ome C (cy C) om mi ochond ia in o he cy osol leading o assembly o a mul ip o ein complex e med apop osome (G een and K oeme , 2004). Apop osome is composed o p ocaspase 9, apop o ic p o ease ac i a ing ac o 1 and cy C. In addi ion, se e al B cell lymphoma-2 (BCL2) amily membe s including BCL2, BCL-W and B- cell lymphoma-ex a-la ge (BCL-XL) ac as an i-apop o ic p o eins by con olling he elease o cy C o he cy osol and egula ing mi ochond ial memb ane pe meabiliza ion (G een and K oeme , 2004). Downs eam, bo h in insic and ex insic apop o ic pa hways a e media ed h ough CASPs, a amily o p o eases ha p o ide c i ical links in cell signaling ne wo ks and cell dea h (McIlwain e al., 2015). In addi ion, apop osis can be ini ia ed independen o CASPs ia apop osis inducing ac o (AIF) ha igge s DNA agmen a ion (Joza e al., 2009). Ups eam, umo p o ein p53 se es as a egula o ha can modula e key con ol poin s in bo h he ex insic and in insic apop o ic pa hways (F idman and Lowe, 2003). Es ogen has been shown o p o ec agains apop osis in non-skele al muscle issues as well as in muscle p ogeni o cells (Hou e al., 2010; Ruan e al., 2014; La Colla e al., 2013). Fu he mo e, i has been sug- ges ed ha es ogens possess a p o ec i e ole agains cellula damage in hea muscle by inc easing he exp ession o he p o ec i e hea shock p o eins (HSPs) (Knowl on and Ko zick, 2014). In muscle biopsies om human wins, E 2 de iciency has been associa ed wi h cell dea h, apop osis, and cell su i al wi h E 2 being he p edic ed ups eam egula o o hese p ocesses (Laakkonen e al., 2017). Howe e , he possible mechanis ic ole o E 2 de iciency in skele al muscle apop osis con ibu ing o sa copenia has no been elucida ed. One possible ou e h ough which E 2 may coo dina e skele al muscle apop osis in ol es mic oRNAs (miRs). miRs a e a small non-coding RNA molecules ha unc ion in pos - ansc ip ional egula ion o gene exp ession ei he by a ge ing mRNAs o deg ada ion o inhibi ing ansc ip ion ini ia ion (K ol e al., 2010). miRs ha e been shown o al e gene exp ession in skele al muscle in esponse o a ious ex e nal s imuli such as exe cise (K ol e al., 2010; Sapp e al., 2017) and es ogen Fig. 1. Schema ic o apop o ic pa hways. Key p o eins o ex insic and in insic apop o ic pa hways ha a e possibly egula ed by miRs a e shown. TNF = umo nec osis ac o , FasL =Fas ligand, CASP =caspase, BCL2 =B-cell lymphoma-2 egula o p o ein, BCL-XL =B-cell lymphoma-ex a-la ge egula o p o ein, BCL- W=B-cell lymphoma-like p o ein 2, BID=P o ein o he B-cell lymphoma-2 egula o amily, AIF =Apop osis inducing ac o , p53 = umo p o ein, cy C =cy o- ch ome C, HSP=Hea shock p o ein, miR =mic oRNA. Figu e was c ea ed in BioRende .com. S. Ka inen e al. Expe imen al Ge on ology 147 (2021) 111267 3 s a us (Oli ie i e al., 2014). To da e se e al miRs ha e been iden i ied o egula e apop osis a many s eps leading o p og ammed cell dea h (Su e al., 2015a). In he apop osis ield, he as majo i y o miR s udies ha e concen a ed on cance biology. These s udies e ealed di e gen oles o miRs along he apop osis pa hways, mainly by ac ing as umo supp esso s by inhibi ing he unc ion o key p o eins inducing cell dea h (Da e al., 2013; Bai e al., 2009; Sun e al., 2017). He e we s udied he e ec o E 2 de iciency on miR signaling coo - dina ing he ini ia ion o skele al muscle apop osis o de e mine i miRs associa e wi h E 2 and ac i a ion o apop o ic signaling cascades in skele al muscle. In he p esen s udy we u ilized C57BL6 mice wi h h ee s udy g oups; con ol (no mal es ous cycle), OVX (E 2 de iciency due o o a iec omy) and OVX +E 2 (E 2 supplemen ed by pelle ). The aim o ou s udy was o de e mine he link be ween E 2 and he ini ia ion o skele al muscle apop osis. Ou hypo hesis was ha E 2 de iciency is associa ed wi h lowe exp ession o apop osis-linked miRs, which leads o inc eased exp ession o p o eins along apop o ic pa hways. 2. Me hods 2.1. Animal expe imen Female C57BL/6J mice aged 3–4 mon hs we e ob ained om The Jackson Labo a o y (Ba Ha bo , ME, USA). Mice we e housed in g oups o ou o i e and had phy oes ogen- ee oden eed (Ha lan-Teklad no. 2019; Indianapolis, IN, USA) and wa e ad libi um. The housing oom was main ained on a 14:10 h ligh :da k cycle wi h con olled empe a u e and humidi y. Mice we e andomly di ided in o h ee g oups: con ol (n =8), OVX (n =7) and OVX +E 2 (n =4). A 4–6 mon h o age su gical p ocedu es we e conduc ed as desc ibed p e iously (Mo an e al., 2007) wi h he excep ion ha 2 h p io o su ge y, mice we e gi en a subcu aneous injec ion o slow- elease bup eno phine (2 mg/kg) in he hindlimb a ea. B ie ly, mice we e hen anes he ized wi h iso lu ane and o a iec omy was pe o med unde asep ic condi ions h ough wo small do sal in- cisions be ween he iliac c es and he lowe ibs. In OVX +E 2 mice, a slow- elease 17β-es adiol-con aining pelle (E 2 pelle ) was placed sub- cu aneously in he neck a ea immedia ely ollowing OVX while s ill unde anes hesia. The E 2 pelle s used in his s udy had 0.18 mg o 17β- es adiol eleased o e a 60-day pe iod (Inno a i e Resea ch o Ame - ica, Sa aso a, FL, USA) esul ing in physiological se um E 2 le els (Mo an e al., 2007; Le e al., 2018). Hal o he mice (n =4) in he con ol g oup unde wen he same su ge y as he OVX mice wi hou emo al o he o a ies (~sham su ge y), while ou mice had no su ge y. Fo all a i- ables measu ed, he e we e no di e ences be ween sham mice and mice ha had no su ge y (p >0.05) and he e o e, da a o hese wo g oups o mice we e collapsed o cons i u e he con ol g oup. Mice we e sac i iced 5–8 weeks la e and a his ime success ul OVX and E 2 ea men we e e i ied by aginal cy ology o u e ine mass. Body mass a he ime o sac i ice was (mean ±SD); con ol =23.9 ±2.0 g, OVX = 26.5 ±2.4 g, and OVX +E 2 =23.8 ±1.4 g. Mice we e sac i iced by an o e dose o pen oba bi al sodium (200 mg/ kg) and skele al muscles om he hind limbs we e ha es ed, snap ozen in liquid ni ogen, and s o ed a −80 ◦C. 2.2. Skele al muscle sample p ocessing 2.2.1. miR and mRNA isola ion miRs and mRNA we e isola ed om ozen gas ocnemius muscles by miRNeasy Mini Ki (Qiagen, ca . no. 217004) acco ding o manu ac- u e ’s ins uc ions. 40–50 mg o he muscle sample was used o he p ocedu e. B ie ly, ozen muscle sample was homogenized in 700 μ l o Qiazol lysis eagen in TissueLyse wi h magne ic beads. Homogena e was incuba ed a oom empe a u e (RT) o 5 min and 140 μ l chlo o- o m was added. Tubes we e shaken 15 s, incuba ed a RT o 3 min and cen i uged o 15 min a 12000 xg a 4 ◦C. The uppe aqueous phase was ans e ed o a new collec ion ube and he p o ocol was ollowed ac- co ding o he manu ac u e ’s ins uc ions un il RNA was elu ed om he column. 2.2.2. P o ein isola ion Remaining gas ocnemius muscle samples we e homogenized by a TissueLyse wi h me allic beads in a lysis bu e ha con ained Tissue ex ac ion Reagen I bu e (TERI, In i ogen, FNN0071) supplemen ed wi h 10 μ l/ml o he ollowing p o ease inhibi o s: Hal P o ease and Phospha ase Inhibi o Cock ail wi h EDTA (Pie ce 78444) and peps a in A (Sigma P5318). 15 μ l/mg o he lysis bu e mix was used o each muscle sample. Samples we e mixed o 30 min in end-o e -end o a ion a +4 ◦C. The ea e , samples we e cen i uged o 10 min a 10000 g a +4 ◦C. To al p o ein o he supe na an was measu ed using a BCA p o ein assay (Pie ce™ BCA P o ein Assay Ki #23227). 2.3. miR analysis by NGS To al RNA samples we e p epa ed o explo a i e NGS analysis wi h pooled samples o he s udied g oups; con ol(pool), OVX(pool) and OVX +E 2 (pool). The con ol(pool) was comp ised o o a y-in ac mice (bo h sham-ope a ed mice and mice ha did no unde go any su ge y). F om each indi idual sample, 1000 ng o RNA was used o making he pool. The ea e , pooled samples om he con ol, OVX and OVX +E 2 g oups we e dilu ed acco ding o manu ac u e ’s ins uc ions o 20 ng/ μ l, o which 5 μ l was used o he miR-lib a y ( o al 100 ng o sample). Fo he miR lib a y, QIAseq miRNA Lib a y Ki (Ca no. 331502) wi h QIAseq miRNA index ki (Ca no 331592) was used. MiR sequencing was pe o med using he Nex Seq 500/550 high Ou pu 2 ki (75 cycles) (FC-404-2005, Illumina) un wi h Nex Seq500 equip- men . The aw da a was p ocessed using he Seconda y QIASeq miRNA Lib a y Ki Da a Analysis So wa e ha analyzes he Unique Molecula Index (UMI) coun s o calcula e changes in he miR exp ession. Fo he analysis, con ol was se as con ol g oup wi h T immed mean selec ed as a no maliza ion me hod. The di e en ial exp ession o miRs be ween OVX(pool) and OVX +E 2 (pool) samples we e epo ed as old change ela i e o con ol g oup. Cu poin o a ge miR selec ion was se a >100 eads (i.e., a leas one g oup had o ha e had >100 eads o he a ge miR). 2.4. In silico analysis o miR a ge genes The sea ch o miR a ge genes on apop o ic pa hways was pe - o med using ou di e en miR a ge da abases: Ta ge Scan, Diana, PicTa and miRWalk as well as Ingenui y Pa hway Analysis (IPA, e sion 57662101, Qiagen). U ilizing hese ou miR da abases oge he wi h IPA enabled analyses o a mo e comp ehensi e a las o p edic ed and alida ed miR a ge s han by using only a single da abase. 2.5. miR analysis by qPCR Exp ession o he miRs selec ed om he NGS miR-lib a y based on in silico analysis we e con i med wi h qPCR using he ollowing p ime s om Qiagen: miR-92a-3p (MS00005971), miR-122-5p (MS00003416), miR-133a-3p (MS00031423), miR-214-3p (MS00031605), miR-337-3p (MS00011844), miR-381-3p (MS00004116), miR-483-3p (MS00007693), miR-483-5p (MS00012264), miR-491-5p (MS00004326), miR-501-5p (MS00032935) and miR-652-3p (MS00010451). Samples we e un as iplica es. qPCR esul s we e analyzed wi h he ΔΔC me hod and a e exp essed as old change ela i e o con ol samples. 2.6. Gene exp ession analysis by qPCR MiRs a ge ed p o eins on he apop osis pa hways we e measu ed by mRNA exp ession le els using he ollowing p ime s om Qiagen: S. Ka inen e al. Expe imen al Ge on ology 147 (2021) 111267 4 GAPDH (NM_008084), CASP3 (NM_00981), CASP6 (NM_009811), CASP8 (NM_009812), CASP9 (NM_015733), BCL2 (NM_009741), BCL-W (o icial name: Bcl2l2, NM_007537), BCL-XL (o icial name: Bcl2l1, NM_009743), cy C (o icial name: Cycs, NM_007808) and p53 (o icial name: T p53, NM_011640). Samples we e un as iplica es. qPCR e- sul s we e analyzed wi h he ΔΔC me hod and a e exp essed as old change ela i e o OVx +E 2 samples. 2.7. P o ein exp ession analysis by Wes e n blo ing The le el o se e al p o eins on he apop o ic pa hways we e quan- i ied using Wes e n blo . B ie ly, 30 μ g o isola ed p o ein om gas ocnemius muscle was used om each sample. Samples we e i s hea ed o 10 min a 95 ◦C in sample bu e (1:1 in 20:1 Laemmli and 5% β-me cap oe hanol) and un o 35 min in a S ain F ee g adien gel (4–20%, BioRad) a 270 V. P o eins we e blo ed on o ni ocellulose memb ane and o al p o ein was isualized a e p o ein ans e . The ea e , memb anes we e blocked o 2 h a RT in blocking bu e (Odyssey) and incuba ed wi h p ima y an ibody in 1:1 TBS and blocking bu e +4 ◦C o e nigh . The ollowing an ibodies we e used: CASP3 (ac i e) (#9661, 1:1000, Cell Signaling) CASP6 (inac i e and ac i e) (ab185645, 1:1000, Abcam), CASP8 (inac i e) (ab25901, 1:1000, Abcam), CASP9 (ac i e) (#9509, Cell Signaling, 1:1000), CASP9 (inac- i e) (ab202068, 1:2000, Abcam), BCL2 (ab692, 1:500, Abcam), BCL-W (#2724, 1:1000, Cell Signaling), BCL-XL (ab32370, 1:1000, Abcam), cy C (ab90529, 1:2000, Abcam), p53 (o icial name: T p53, ab26, 1:500, Abcam), AIF (Sc-13,116, 1:200, San a C uz), CRYAB (o icial name: Hspb5, #45844S, 1:1000, Cell Signaling), HSP27 (o icial name: HSPB1, CPTC-HSPB1-1, 1:500, Hyb idoma Bank) and HSP60 (o icial name: Hspd1, #4870, 1:500, Cell Signaling). Memb anes we e hen washed and incuba ed wi h seconda y an ibody o 1 h a RT (1:1 TBS-T and blocking bu e ), washed and isualized. P o eins we e quan i ied using Image Lab so wa e. Fi s , he esul s we e no malized o he a e age signal o he memb ane and he ea e o co esponding o al p o ein amoun ob ained om he whole memb ane blo ed om a s ain ee gel (whole p o ein lane). Resul s a e exp essed as old change ela i e o OVx +E 2 samples. 2.8. S a is ical analyses Resul s a e p esen ed as mean and s anda d e o o means (SEM). The no mali y o a iables was assessed using Shapi o-Wilks es s ol- lowed by Le ene’s es o examining he equali y o he a iances. Fi s , he ex eme ou lie s we e excluded om he analysis (>3×in e qua ile ange). When he no mali y c i e ia we e me , di e ences be ween he g oups (OVX and OVX +E 2 ) we e examined using S uden ’s - es (Wes e n blo p o ein exp ession le els). When he no mali y c i e ia we e no ul illed, di e ences be ween he g oups we e examined using K uskal-Wallis es ollowed by Mann-Whi ney U es (qPCR esul s o miRs) o Mann-Whi ney U es (mRNA le els). Da a analyses we e ca - ied ou using IBM SPSS S a is ics so wa e e sion 24 (Chicago, IL, US), and he le el o signi icance was se a p ≤0.050. 3. Resul s 3.1. NGS iden i ica ion o miRs NGS analysis was done o i s e i y he mos abundan miRs and second o explo e he speci ic miRs ha espond o E 2 in skele al muscle. The wo mos abundan miRs in all g oups o mice, miR-1a-3p and miR- 133a-3p, a e muscle speci ic miRs (Table 1). 3.2. E 2 esponsi e miRs ha a ge apop o ic pa hways To in es iga e he possible link be ween E 2 de iciency and miR- media ed egula ion o apop o ic signaling, we ocused on apop osis- linked miRs. In silico analysis wi h ou di e en da abases and IPA showed ha ele en miRs iden i ied o be esponsi e o E 2 (Table 2) ha e se e al a ge s on apop o ic pa hways (Table 3). NGS analysis o he E 2 sensi i e miRs was un compa ing OVX and OVX +E 2 each o he con ol g oup and indica ed di e en ial exp ession pa e ns be ween OVX and OVX +E 2 mice. These apop osis-linked, di e en ially exp essed E 2 - esponsi e miRs we e used as candida e signaling modu- la o s o apop osis. Resul s o in slico analysis we e no comple ely consis en be ween he da abases (Table 3). Fo example, only miRs-133a-3p and 214-3p we e ound o a ge apop osis-linked p o eins in all ou o he miR Table 1 Twen y mos abundan miRs acco ding o eads in NGS analysis in muscle om con ol, OVX and OVX +E2 g oups. Rank Con ol OVX OVX +E2 miR Reads miR Reads miR Reads 1 miR- 1a-3p 21,097,106 miR- 1a-3p 17,434,920 miR- 1a-3p 19,618,500 2 miR- 133a- 3p 2,616,739 miR- 133a- 3p 2,469,748 miR- 133a- 3p 3,176,871 3 miR- 126a- 3p 585,076 miR- 126a- 3p 727,958 le -7 - 5p 738,205 4 le -7 - 5p 574,105 le -7 - 5p 701,060 miR- 126a- 3p 586,756 5 miR- 143-3p 504,622 miR- 16-5p 563,935 miR- 16-5p 500,677 6 miR- 26a-5p 481,154 miR- 26a- 5p 522,538 miR- 26a-5p 490,611 7 miR- 16-5p 443,756 miR- 143- 3p 518,288 le -7a- 5p 476,089 8 le -7a- 5p 316,171 le -7a- 5p 419,319 miR- 143-3p 409,992 9 miR- 125b- 5p 302,611 miR- 125b- 5p 349,376 miR- 125b- 5p 320,767 10 miR- 30a-5p 296,285 miR- 206- 3p 328,095 le -7i- 5p 300,807 11 miR- 206- 3p 259,993 le -7c- 5p 323,725 le -7c- 5p 294,044 12 miR- 1b-5p 251,335 le -7i- 5p 289,413 miR- 206- 3p 288,776 13 le -7i- 5p 238,688 miR- 30a- 5p 254,499 miR- 378a- 3p 269,323 14 le -7c- 5p 233,160 miR- 378a- 3p 236,223 miR- 30a-5p 268,885 15 miR- 378a- 3p 232,999 miR- 1b-5p 231,988 miR- 1b-5p 247,180 16 miR- 29a-3p 211,195 miR- 29a- 3p 209,769 miR- 29a-3p 206,621 17 miR- 26b-5p 164,072 miR- 26b- 5p 186,750 miR- 26b-5p 172,737 18 miR- 126a- 5p 156,207 le -7b- 5p 162,794 le -7b- 5p 159,892 19 miR- 101a- 3p 143,259 miR- 126a- 5p 158,968 miR- 126a- 5p 134,825 20 miR- 199a/ b-3p 127,357 miR- 101a- 3p 145,617 miR- 199a/ b-3p 132,129 Muscle speci ic miRs a e bolded. S. Ka inen e al. Expe imen al Ge on ology 147 (2021) 111267 5 da abases as well as IPA and only miRwalk and IPA ound pu a i e apop osis-linked a ge p o eins o each o he s udied miRs. Table 3 shows also expe imen ally alida ed associa ions be ween miR and i s a ge p o ein. 3.3. qPCR – exp ession le els o pu a i e E 2 sensi i e miRs The exp ession o he po en ial a ge miRs ound om NGS we e con i med by qPCR (Fig. 2). In Fig. 2, he iangles o con ol g oup ep esen he mice ha did no unde go any su ge y and black do s ep esen sham-ope a ed mice. Because da a om sham-ope a ed mice and hose ha had no su ge y did no di e (p ≥0.072), hese sub- popula ions we e combined o o m one con ol (o a y-in ac ) g oup o u he s a is ical analyses. Due o an ex eme alue ha was no de ec ed as a ue ou lie ound in miR-122-5p da a in g oup OVX +E 2 (Fig. 2B), he analysis was un excluding ha highes da a poin . Excluding he one da a poin did no change he esul s be ween he g oups (see Supplemen a y Fig. 1). K uskal-Wallis es showed signi i- can di e ence be ween he g oups o each o he ele en miRs (p ≤ 0.044). Pos hoc analyses showed ha muscles om OVX mice had lowe exp ession o all ele en o he s udied miRs compa ed wi h OVX + E 2 (p ≤0.017, Fig. 2). OVX muscles also had lowe miRs 122-5p and 214-3p exp ession compa ed wi h con ol (p ≤0.028, Fig. 2B, D). OVX +E 2 muscles had highe exp ession o miRs 92a-3p, 133a-3p, 214-3p, 381-3p, 483-3p, 483-5p and 491-5p han con ols (p ≤0.042, Fig. 2A, C-D, F-I). 3.4. qPCR – mRNA exp ession o miR a ge s To concen a e on he ex emes o sys emic E 2 le el, we con inued he analysis wi h he g oup o he lowes sys emic E 2 le el (OVX) and he cons an ly high E 2 le el (OVX +E 2 ) o a oid he possible con ounding e ec o he cyclic na u e o sys emic E 2 le el in he con ol g oup. The mRNA exp ession o p edic ed miR a ge s on apop osis pa hways we e measu ed using qPCR (Figs. 3 and 4). OVX muscles had highe exp es- sion o CASPs 6 and 9 compa ed o OVX +E 2 muscles (p ≤0.023; Fig. 3B, D) bu no CASP3 o CASP8 (p =0.571, p =0.072, espec i ely, Fig. 4A, C). OVX muscles had highe cy C and p53 mRNA exp essions compa ed o OVX +E 2 (p ≤0.038; Fig. 4D-E) whe eas no changes we e obse ed in BCL2, BCL-XL o BCL-W. 3.5. Wes e n blo - p o ein exp ession o miR a ge s The p o ein le els o key CASPs a e p esen ed in Fig. 5. Muscles om mice in he OVX g oup had a lowe inac i e CASP8 le el (p =0.028, Fig. 5B) and highe ac i e CASP9 compa ed wi h OVX +E 2 muscles (p = 0.046, Fig. 5F). The e we e no signi ican indings in he o he CASPs measu ed (p ≥0.140). The Wes e n blo esul s om addi ional apop osis-linked p o eins a e p esen ed in Fig. 6. OVX muscles had lowe BCL-W and highe cy C le els compa ed o OVX +E 2 (p =0.010, p =0.023, espec i ely; Fig. 6C, F). The e we e no signi ican indings in he o he p o eins s udied (p ≥0.108). The Wes e n blo esul s om HSPs a e p esen ed in Fig. 7. OVX muscles had lowe le els o CRYAB and HSP60 and a highe le el o HSP27 compa ed o OVX +E 2 (p ≤0.008; Fig. 7). 4. Discussion Ou esul s we e in line wi h ou hypo hesis, ha E 2 de iciency was associa ed wi h lowe exp ession o apop osis-linked miRs. Conse- quen ly, E 2 de iciency led o highe mRNA exp ession o p o-apop o ic cy C, p53 and CASPs 6 and 9 and g ea e p o ein abundance o cy C and ac i e CASP9. Howe e , we also disco e ed lowe p o ein abun- dance o an i-apop o ic BCL-W, CRYAB and HSP60 and g ea e abun- dance o HSP27. Ou da a sugges s ha E 2 de iciency may play a ole in Table 2 Es adiol- esponsi e, apop osis-linked miRs ha showed di e en ial exp ession be ween muscle om OVX and OVX +E 2 mice (exp essed as old change ela i e o con ol). miR ID OVX OVX +E 2 miR-92a-3p 0.99 1.42 miR-122-5p 0.78 2.64 miR-133a-3p 0.85 1.21 miR-214-3p 0.91 1.28 miR-337-3p 0.95 1.63 miR-381-3p 0.80 2.18 miR-483-3p 0.78 2.53 miR-483-5p 1.04 2.11 miR-491-5p 1.37 1.95 miR-501-5p 1.39 0.48 miR-652-3p 1.20 0.94 Table 3 In silico analysis o miR a ge genes ia ou miR a ge da abases, IPA and expe imen ally alida ed associa ions be ween miR and a ge p o ein. miR Ta ge Scan Diana miRwalk PicTa IPA Expe imen ally shown miR- a ge associa ions miR-92a- 3p FasL – CASP3, CASP8, CASP9, HSP60, Cy C – p53 CASP3, BCL2 (Li e al., 2019) miR-122- 5p – – CASP3, CASP8, BCL-XL, BCL-W, HSP60, Cy C BCL-W BCL-W CASP8, BCL2, BCL-XL, p53 (Yin e al., 2011; Zhang e al., 2017) miR- 133a- 3p BCL-XL, BCL- W BCL- XL CASP3, CASP8, CASP9, BCL2, BCL-XL, BCL-W, HSP60, Cy C BCL-XL, BCL-W BCL-XL, BCL- W CASP3, CASP9, BCL-XL (Chen e al., 2016; Ji e al., 2013) miR-214- 3p HSP27, Casp2 HSP27 CASP3, CASP6, CASP8, BCL-W BCL-XL, HSP60 HSP27 BCL-W, HSP70 (Bai e al., 2020; Fan and Wu, 2017) miR-337- 3p – – CASP9, BCL-XL, BCL-W – BCL-XL, BCL- W CASP3, CASP7, BCL2 (Xia e al., 2019; Pa k e al., 2018) miR-381- 3p – – CASP3, CASP8, CASP9, BCL2, BCL-XL, BCL-W, Cy C – CASP6 CASP3, CASP8, CASP9, BCL2 (Zhao e al., 2020; Qiao e al., 2019; Shang e al., 2019) miR-483- 3p – – CASP3, CASP6, CASP8, BCL-XL, BCL-W, HSP60, Cy C – FasL CASP3, BCL2, p53 (Ve onese e al., 2010; Lu e al., 2020) miR-483- 5p – – CASP3, CASP6, CASP8, CASP9, BCL-XL, BCL-W, Cy C – CASP3, CASP8 CASP3, BCL2 (Wu e al., 2016; Liu e al., 2019) miR-491- 5p – – CASP3, CASP6, CASP8, Cy C, FasL, BCL2, BCL-W – p53, CRYAB BCL-XL, p53 (Guo e al., 2012) miR-501- 5p – – CASP3, CASP6, CASP8, CASP9, BCL2, BCL-XL, BCL-W, Cy C – p53, HSP60, cy C BCL2 (Sanches e al., 2018) miR-652- 3p – – CASP3, CASP8, CASP9, BCL2, BCL-XL, BCL-W, Cy C – p53, CASP3, CASP6 CASP3, BCL2 (Wang e al., 2017) S. Ka inen e al. Expe imen al Ge on ology 147 (2021) 111267 6 Fig. 2. miR exp ession in skele al muscle measu ed by qPCR. OVX mice had lowe exp essions o all o he s udied miRs compa ed wi h OVX +E 2 g oup (A-K). OVX mice had also lowe miR-122-5p and miR-214-3p exp ession compa ed wi h con ol (B, D). Compa ed wi h con ol, OVX +E 2 g oup had highe exp ession o miRs 92a-3p, 133a-3p, 214-3p, −381, 483-3p, 483-5p and 491-5p (A, C, D, F, G, H, I). The con ol g oup is comp ised o o a y-in ac mice wi h iangles ep esen ing mice ha did no unde go any su ge y and black do s ep esen ing sham-ope a ed mice. Resul s a e exp essed as mean ±SEM. *p <0.050, **p <0.010. S. Ka inen e al. Expe imen al Ge on ology 147 (2021) 111267 7 he ini ia ion o skele al muscle apop osis h ough miRs coo dina ing he in insic apop o ic pa hway al hough down egula ion o miRs was also measu ed con ibu ing o a concomi an up egula ion o some o an i- apop o ic agen s. Ou da a sugges s ha E 2 de iciency may play a ole in he ini ia ion o skele al muscle apop osis h ough miRs coo dina ing he in insic apop o ic pa hway. 4.1. Es adiol de iciency in skele al muscle is associa ed wi h dec eased exp ession o se e al miRs linked o apop osis We ound ha he ele en apop osis-linked miRs, miR 92a-3p, 122- 5p, 133a-3p, 214-3p, 337-3p, 381-3p, 483-3p, 483-5p, 491-5p, 501-5p and 652-3p we e less abundan in mice wi h E 2 de iciency (OVX) compa ed o hose mice ea ed wi h E 2 (OVX +E 2 ) (Fig. 2). The miRs in es iga ed in his s udy pa icipa e in pos ansc ip ional egula ion o se e al key p o eins in apop osis and highligh he in insic apop o ic pa hway (Su e al., 2015b). P e iously, ans ec ing cells wi h miR-92a-3p mimic dec eased cell apop osis ia BCL2 and CASP3 pa hways, sugges ing an an i-apop o ic ole in cell homeos asis (Li e al., 2019). In ou esul s E 2 de iciency was associa ed wi h low miR-92a-3p le el in muscle. In u n, miR-122 up egula ion has been shown o induce apop osis in ca diomyocy es and li e cells (Wu e al., 2009; Zhang and Jing, 2018). Suppo ing his obse a ion, miR-122 inhibi ion led o inc eased exp ession o an i- apop o ic p o ein BCL-XL oge he wi h lowe exp ession o p o- apop o ic CASP3 (Zhang and Jing, 2018). Simila ly, up egula ion o miR-133a induced apop osis in os eosa coma cell lines, whe eas i s down egula ion s ongly co ela ed wi h umo p og ession (Ji e al., 2013). Fu he mo e, es o ing miR-133a/b has been p oposed o be a key he apeu ic a ge o gas ic cance ea men in humans (Liu e al., 2015). I seems ha miR-133a ac s as a umo -supp esso in cance cells by a ge ing BCL-XL and CASP9 (Liu e al., 2015; He e al., 2011). Simila ly, wi h miRs 122-5p and 133a-3p, up egula ion o miR-214-3p is associa ed wi h down egula ion o BCL-W and ice e sa, sugges ing a p o-apop o ic ole o miR-214-3p (Fan and Wu, 2017). In ou s udy, E 2 de iciency led o low le els o miRs 122-5p, 133a-3p and 214-3p in muscle, which acco ding o p e ious li e a u e could lead o dec eased apop osis (e.g., [45, 47, 61,]). Howe e , he high le els o hese miRs in OVX +E 2 mice was no associa ed wi h low le els o hei an i-apop o ic a ge p o eins BCL-2 o BCL-XL o a highe CASP8 le el (Table 3), bu lowe p53 ansc ip and highe BCL-W p o ein le els. These esul s indica e ha hese miRs may sha e an an i-apop o ic ole in skele al muscle. Cell expe imen s wi h miR-337-3p ha e yield con adic ing esul s. Expe imen s wi h PANC-1 cells show ha o e exp ession o miR-337-3p down egula es CASPs 3 and 7 (Pa k e al., 2018) ye ec opic exp ession o miR-337 malignan T cells esul ed in inc eased CASP3 and 7 ac i i y while dec easing BCL2 exp ession and inc easing apop osis (Xia e al., 2019). miR-381-3p sha es simila s o y: s udies show ha miR-381-3p ac s as an oncogenic miRNA ha coun e ac s apop o ic signaling pa h- ways in enal cance cells (Zhao e al., 2020). Howe e , in di e en cance cell lines miR-381 mimics ha e ound o inc ease CASP3 and Fig. 3. mRNA exp ession o caspases 3, 6, 8 and 9 in skele al muscle measu ed by qPCR. OVX mice had highe exp ession o CASP6 and CASP9 compa ed wi h OVX +E 2 (B, D). Resul s a e exp essed as mean +SEM. *p <0.050, **p <0.010. S. Ka inen e al. Expe imen al Ge on ology 147 (2021) 111267 8 CASP9 and dec ease BCL2 le els whe eas i s inhibi o s ha e opposing e ec (Qiao e al., 2019; Shang e al., 2019). I seems ha miR unc ion is sensi i e o he cell line used. In ou s udy, E2 de iciency lead o low le el o miR337-3p and 381-3p oge he wi h highe p o ein le el o ac i e CASP9 in muscle, suppo ing he an i-apop o ic unc ion o hese miRs. miR-483-3p inhibi s apop osis in b eas cance as well as in enal malignancy umo cells by egula ing SOX3, BAX and BCL2 exp essions (Che e al., 2019; Cui e al., 2019). miR-483-5p in u n up egula es CASP3 p o ein in a kidney cell line (Liu e al., 2019) whe eas miR-483- Fig. 4. mRNA exp ession o BCL2, BCL-XL, BCL-W, cy C and p53 in skele al muscle measu ed by qPCR. OVX mice had highe exp ession o cy C and p53 compa ed wi h OVX +E 2 (D, E). Resul s a e exp essed as mean +SEM. *p <0.050, **p <0.010. S. Ka inen e al.