Identification of a Genetic Variation in ERAP1 Aminopeptidase that Prevents Human Cytomegalovirus miR-UL112-5p-Mediated Immunoevasion

A icle Iden i ica ion o a Gene ic Va ia ion in ERAP1 Aminopep idase ha P e en s Human Cy omegalo i us miR-UL112-5p-Media ed Immunoe asion G aphical Abs ac Highligh s dHCMV miR-112-5p a ge s ERAP1, p e en ing he p esen a ion o i al epi opes o CTLs dThe s17481334 a ian G p ese es ERAP1 om miR-UL112- 5p-media ed deg ada ion dHCMV-in ec ed GG cells a e mo e e icien ly lysed by HCMV- pep ide-speci ic CTLs dDec eased HCMV se oposi i i y is de ec ed among mul iple scle osis GG pa ien s Au ho s Paolo Romania, Lo edana Ci aldi, Benede a Pignoloni, ..., F anco Loca elli, Mau o D’Ama o, Do iana F uci Co espondence do iana. [email protected] In B ie Romania e al. iden i y a single-nucleo ide polymo phism in ERAP1 gene, which ep esen s a esis ance mechanism o HCMV miR-UL112-5p-based immune e asion s a egy, wi h po en ial implica ions o indi idual suscep ibili y o i al in ec ion and o he diseases. Romania e al., 2017, Cell Repo s 20, 846–853 July 25, 2017 ª2017 The Au ho (s). h p://dx.doi.o g/10.1016/j.cel ep.2017.06.084 Cell Repo s A icle Iden i ica ion o a Gene ic Va ia ion in ERAP1 Aminopep idase ha P e en s Human Cy omegalo i us miR-UL112-5p-Media ed Immunoe asion Paolo Romania, 1 Lo edana Ci aldi, 1 Benede a Pignoloni, 2 Nadia S a c, 1 Vale io D’Alicand o, 1 Omb e a Melaiu, 1 Giuseppina Li Pi a, 1 Ezio Gio da, 3 Rosalba Ca ozzo, 4 Monika Be g all, 5 Tomas Be gs o ¨m, 6 La s Al edsson, 7,8 Tomas Olsson, 9 Ing id Kockum, 9 Ilkka Seppa ¨la ¨, 10 Te ho Leh ima ¨ki, 10 Mikko A. Hu me, 11 Ha mu Hengel, 12 Angela San oni, 2 C is ina Ce boni, 2 F anco Loca elli, 1,13 Mau o D’Ama o, 14,15,16 and Do iana F uci 1,16,17, * 1 Paedia ic Haema ology/Oncology Depa men , Ospedale Pedia ico Bambino Gesu `, IRCCS, 00146 Rome, I aly 2 Depa men o Molecula Medicine, Is i u o Pas eu -Fondazione Cenci Bologne i, ‘‘Sapienza’’ Uni e si y o Rome, 00161 Rome, I aly 3 Uni o Flow Cy ome y, Ospedale Pedia ico Bambino Gesu `, IRCCS, 00146 Rome, I aly 4 Uni o Muscula and Neu odegene a i e Diseases, Labo a o y o Molecula Medicine, Ospedale Pedia ico Bambino Gesu `, IRCCS, 00146 Rome, I aly 5 Depa men o Biosciences and Nu i ion, Ka olinska Ins i u e , 171 77 S ockholm, Sweden 6 Depa men o In ec ious Diseases, Sec ion o Clinical Vi ology, Ins i u e o Biomedicine, Uni e si y o Go henbu g, 41345 Go ¨ ebo g, Sweden 7 Ins i u e o En i onmen al Medicine, Ka olinska Ins i u e , 171 77 S ockholm, Sweden 8 Cen e o Occupa ional and En i onmen al Medicine, S ockholm Coun y Council, 171 77 S ockholm, Sweden 9 Depa men o Clinical Neu oscience and Cen e o Molecula Medicine, Ka olinska Ins i u e , 171 77 S ockholm, Sweden 10 Depa men o Clinical Chemis y, Fimlab Labo a o ies and Finnish Ca dio ascula Resea ch Cen e -Tampe e, Facul y o Medicine and Li e Sciences, Uni e si y o Tampe e School o Medicine, 33014 Tampe e, Finland 11 Depa men o Mic obiology and Immunology, FimLab Labo a o ies and Facul y o Medicine and Li e Sciences, Uni e si y o Tampe e School o Medicine, 33014 Tampe e, Finland 12 Ins i u e o Vi ology, Medical Cen e , and Facul y o Medicine, Uni e si y o F eibu g, 79104 F eibu g, Ge many 13 Uni e si y o Pa ia, 27100 Pa ia, I aly 14 Depa men o Medicine Solna, Ka olinska Ins i u e , 171 77 S ockholm, Sweden 15 BioDonos ia Heal h Resea ch Ins i u e San Sebas ian and IKERBASQUE, Basque Founda ion o Science, 48013 Bilbao, Spain 16 These au ho s con ibu ed equally 17 Lead Con ac *Co espondence: do iana.[email p o ec ed] h p://dx.doi.o g/10.1016/j.cel ep.2017.06.084 SUMMARY He ein, we demons a e ha HCMV miR-UL112-5p a ge s ERAP1, he eby inhibi ing he p ocessing and p esen a ion o he HCMV pp65 495-503 pep ide o speci ic CTLs. In addi ion, we show ha he s17481334 G a ian , na u ally occu ing in he ERAP1 30UTR, p ese es ERAP1 om miR-UL112- 5p-media ed deg ada ion. Speci ically, HCMV miR- UL112-5p binds he 30UTR o ERAP1 A a ian , bu no he 30UTR o ERAP1 G a ian , and, acco dingly, ERAP1 exp ession is educed bo h a RNA and p o- ein le els only in human ib oblas s homozygous o he A a ian . Consis en ly, HCMV-in ec ed GG ib oblas s we e mo e e icien in imming i al an igens and being lysed by HCMV-pep ide-speci ic CTLs. No ably, a signi ican ly dec eased HCMV se oposi i i y was de ec ed among GG indi iduals su e ing om mul iple scle osis, a disease model in which HCMV is nega i ely associa ed wi h adul - onse diso de . O e all, ou esul s iden i y a esis- ance mechanism o HCMV miR-UL112-5p-based immune e asion s a egy wi h po en ial implica ions o indi idual suscep ibili y o in ec ion and o he diseases. INTRODUCTION Human cy omegalo i us (HCMV) is a b-he pes i us ha causes widesp ead, pe sis en in ec ion, and i is o en le hal in subjec s wi h impai ed immuni y (S e n-Ginossa e al., 2012). HCMV has e ol ed mul iple s a egies o in e e e wi h inna e and adap i e immune esponses. Vi ually, e e y single s ep wi hin he MHC class I an igen p ocessing and p esen a ion pa hway is a ge ed by i al p o eins ( an de Weije e al., 2015), hus indica ing ha ecogni ion o i al pep ides by CD8 + T cells is a key e en in he elimina ion o HCMV-in ec ed cells. In addi ion o i al p o eins, HCMV also deploys i ally en- coded mic oRNAs (miRNAs) o manipula e hos immune e- sponses. Cu en ly, 24 ma u e miRNAs encoded by HCMV ha e been iden i ied o a ge bo h i al and hos genes in ol ed in immune de ense, ch oma in emodeling, cell cycle egula ion, apop osis, signal ansduc ion, and esicula a icking (G ey e al., 2007, 2010; Hook e al., 2014a, 2014b; Kim e al., 2015; Lee e al., 2012; Ti abassi e al., 2011). In pa icula , h ee i al 846 Cell Repo s 20, 846–853, July 25, 2017 ª2017 The Au ho (s). This is an open access a icle unde he CC BY license (h p://c ea i ecommons.o g/licenses/by/4.0/). miRNAs ha e been ound o a ec componen s o he immune sys em: miR-UL112-1 a ge s MICB, a s ess-induced ligand o he na u al kille (NK) cell ac i a ing ecep o NKG2D, which is c ucial o NK cell-media ed killing o i al-in ec ed cells (S e n-Ginossa e al., 2007); miR-UL148D a ge s he chemo- kine CCL5, which limi s i al in ec ion by ec ui ing immune cells o he si e o in ec ion (Kim e al., 2012); and miR-US4-1 a ge s ERAP1, a key componen o he an igen p ocessing ha ims p ecu so s in o pep ides o he co ec leng h o bind MHC class I molecules, esul ing in an impai ed p oduc ion o many i al epi opes and a educed an i- i al CD8 + T cell esponse (Kim e al., 2011). S udies in ERAP1-de icien mice ha e demons a ed ha ERAP1 is equi ed o p esen a ion o se e al i al epi opes, and i s absence a ec s an i- i al CD8 + T cell esponses (Blan- cha d e al., 2010; Fi a e al., 2007; Sa eanu e al., 2005; Yan e al., 2006). In humans, DNA sequence a ia ion in ERAP1 has been associa ed o gene ic- isk e ec s in a numbe o immune-media ed diseases, including ankylosing spondyli is, pso iasis, Behc¸ e ’s disease, and mul iple scle osis (MS) (S a i- kos e al., 2014). Func ional s udies ha e es ablished ha hese a ian s a ec ERAP1 ac i i y, esul ing in changes in he epe - oi e o an igenic pep ides (Ree es e al., 2014), which is c i ical o shaping inna e and adap i e immune esponses (Ci aldi e al., 2011, 2015; James e al., 2013). The impo ance o ERAP1 imming o immune ecogni ion o i us-in ec ed cells is demons a ed by se e al indings, including he e olu ion o escape mu a ions impai ing ERAP1 ac i i y and esul ing in dec eased CD8 + T cell esponse (D aene e al., 2004; Tenze e al., 2009). Func ional gene ic a ia ions in miRNA p ocessing machine y, genes, and a ge si es, named miR-single-nucleo ide polymo - phisms (miR-SNPs), ha e a ac ed special a en ion due o hei in ol emen in se e al ypes o cance (Ryan e al., 2010). These a ian s can a ec he ansc ip ion o a ge genes, p e-miRNA p ocessing, and/o modula e miRNA-mRNA in e ac ions by a ec ing miRNA binding o a ge si es. I is he e o e possible o en isage a complex scena io whe e ERAP1 unc ional poly- mo phisms may be ele an o indi idual suscep ibili y o bo h HCMV in ec ion and (di ec ly o indi ec ly) au oimmuni y, h ough gene-en i onmen (exposu e) in e ac ions. In his s udy, we demons a e ha HCMV miR-UL112-5p spe- ci ically a ge s ERAP1, he eby inhibi ing he p ocessing and p esen a ion o he HCMV pp65 495-503 pep ide o speci ic cy o- oxic CD8 + T cells. In addi ion, we ound a na u ally occu ing a ian in he 30UTR o he ERAP1 gene (SNP s17481334), which p e en s ERAP1 a ge ing and deg ada ion by HCMV miR-UL112-5p. RESULTS s17481334 G Dis up s miR-UL112-5p Binding o ERAP1 mRNA 30UTR Recen deep-sequencing analyses o small RNAs om HCMV- in ec ed human ib oblas cells led o he iden i ica ion o no el i al miRNA p ecu so s and e inemen o miRNA anno a ions (Meshesha e al., 2012; S a k e al., 2012). In pa icula , se- quences ob ained o miR-US4-1 (now miR-US4-5p) esul ed in being shi ed by 5 bp a he 50end (Figu e S1A), he eby complica ing he in e p e a ion o he seed-media ed ERAP1 30 UTR a ge ing desc ibed so a (Kim e al., 2011). Rega dless o he co ec sequence o miR-US4-1, ERAP1 p o ein dec eased du ing in ec ion wi h wild- ype HCMV, whe eas i emained un- changed in cells in ec ed wi h a mu an HCMV s ain ca ying a de ec i e p ocessing o he miR-US4-1 p ima y ansc ip , sug- ges ing a key ole o ha egion o ERAP1 30UTR in i al-medi- a ed immune e asion. An A- o-G polymo phism mapping o he 30UTR o he ERAP1 mRNA (SNP s17481334), and co esponding o one o he mu a ed nucleo ides in ol ed in he s udy by Kim e al. (Fig- u e S1A), was p edic ed o dis up he consensus sequence o miR-US4-1 binding. Because s17481334 did no ma ch he sequence o ma u e miR-US4-5p, we looked o o he HMCV miRNAs a ge ing he egion encompassing his gene ic a ian . The RNAhyb id algo i hm (Rehmsmeie e al., 2004) p edic ed he miR-UL112-5p o di ec ly a ge ERAP1 30UTR in ha posi- ion (Figu e S1B). This in e ac ion was alida ed in i o using ERAP1 30UTR dual-luci e ase assays in 293T cells ec opically exp essing miR-UL112-5p o a mu a ed e sion o miR-UL112- 5p wi h a single base subs i u ion co esponding o he SNP (Fig- u e S1B). A signi ican dec ease in luci e ase ac i i y was obse ed o he cells exp essing he s17481334 A a ian o ERAP1 30UTR and he wild- ype miR-UL112-5p, as well as o hose exp essing he s17481334 G a ian and he mu a ed miR-UL112-5p (Figu e 1A). These da a sugges ha miR- UL112-5p di ec ly a ge ed ERAP1 in he consensus si e and ha his in e ac ion is a ec ed by he SNP. To es whe he his di e en ial binding could modula e ERAP1 exp ession, endogenous ERAP1 mRNA and p o ein le els we e measu ed in ib oblas s isola ed om indi iduals homozygous o ERAP1 s17481334 A o G alleles equally ans ec ed wi h wild- ype o mu a ed miR-UL112-5p (Figu e S1C). In hese ex- pe imen s, miR-UL112-5p induced 57% down egula ion o ERAP1 mRNA exp ession in ib oblas s om he s17481334 AA indi idual, bu had no e ec in s17481334 GG ib oblas cells. Con e sely, he mu a ed miR-UL112-5p had no e ec in AA cells, whe eas i induced 40% down egula ion o ERAP1 mRNA exp ession in s17481334 GG cells (Figu e 1B). O no e, ERAP1 mRNA was educed a simila le els in ib oblas s om bo h indi iduals by a siRNA a ge ing a coding egion o ERAP1 (siERAP1) (88% and 91% in s17481334 AA and GG i- b oblas s, espec i ely) (Figu e 1B). Vi ually iden ical esul s we e ob ained in immunoblo expe imen s unde he same con- di ions (Figu e 1C). HCMV In ec ion Fails o Down egula e ERAP1 Exp ession in s17481334 GG Fib oblas s Fib oblas s om s17481334 AA and GG homozygous indi id- uals we e in ec ed wi h he HCMV labo a o y s ain AD169 a wo di e en mul iplici y o in ec ions (MOI). A 3 days pos -in ec- ion (dpi), ib oblas s we e in ec ed a a simila le el, as e al- ua ed by exp ession o i al immedia e-ea ly an igens (IE1 and IE2) and MHC class I down egula ion (Figu e 2A). HCMV-in- ec ed cells so ed o educed MHC class I exp ession (Fig- u e 2B; Figu e S2) showed a educed o e all amoun o ERAP1 p o ein eaching a maximal dec ease o 60% in ib oblas s Cell Repo s 20, 846–853, July 25, 2017 847 om AA indi iduals a MOI 5, as compa ed o unin ec ed cells (Figu e 2C). Con e sely, no changes o ERAP1 p o ein exp es- sion we e de ec ed in HCMV-in ec ed ib oblas s om he GG in- di idual (Figu e 2C), sugges ing ha he p esence o G a he s17481334 30UTR si e p e en s ERAP1 down egula ion du ing HCMV in ec ion. In e es ingly, he o e all amoun o ERAP1 p o- ein emained unchanged a di e en ime poin s in cells in ec ed wi h a modi ied HCMV AD169 s ain (HCMVDUL112) dele ed in he UL114 gene encompassing he miR-UL112-5p (S e n-Ginos- sa e al., 2007)(Figu e 2D). s17481334 A ec s he ERAP1-Dependen HCMV pp65 495-503 CD8 + T Cell Epi ope P esen a ion du ing HCMV In ec ion To di ec ly demons a e he unc ional ele ance o s17481334 du ing HCMV in ec ion, we es ed he capabili y o ib oblas s om AA and GG homozygous indi iduals o p esen he HLA- A*0201- es ic ed immunodominan HCMV pp65 495-503 epi ope, which is one o he well-known i al epi opes equi ing ERAP1 ac i i y o i s in i o gene a ion (U ban e al., 2012). Thus, we isola ed ib oblas s om AA and GG homozygous HLA- A*0201 + indi iduals exp essing simila le els o ERAP1 and cell su ace HLA-A2 molecules, wi h a simila abili y o p ocess and p esen HCMV pp65 495-503 pep ide o speci ic cy o oxic T lym- phocy es (CTLs) (Figu es S3A–S3G). Fib oblas s wi h hese ea- u es we e in ec ed wi h wild- ype HCMV o HCMVDUL112 a MOI 1, and hei suscep ibili y o HLA-A*0201- es ic ed/ pp65 495-503 -speci ic CTLs was e alua ed a 3 dpi. Al hough i- b oblas s om ei he geno ype we e simila in ela ion o IE1/2 exp ession and MHC class I down egula ion (Figu e 3A), cells om he GG indi idual in ec ed wi h wild- ype HCMV and exp essing unchanged le els o ERAP1 (Figu e S3H), we e killed mo e e icien ly han hose om he AA indi idual, a all e ec o / a ge a ios es ed (Figu e 3B). This di e en suscep ibili y o CTL-media ed killing appea ed o depend on miR-UL112-5p a ge ing o ERAP1, because i was no e iden in ib oblas cells om AA and GG indi iduals in ec ed wi h HCMVDUL112 (Figu e 3B; Figu e S3H). Consis en ly, GG ib oblas s s ably knocked down o ERAP1 we e killed less e icien ly han con ol cells when in ec ed wi h wild- ype HCMV (Figu es 3C and 3D; Figu e S4). s17481334 GG Geno ype Associa es wi h HCMV-Se onega i e P o ile in MS Pa ien s The po en ial impac o he s17481334 polymo phism in hos - i us in e ac ions was s udied by looking a HCMV se ology da a in ela ion o geno ype, conside ing he GG homozygous combina ion he mos likely o show biological e ec s. The G allele occu s a mino allele equency (MAF) o 0.121 in Caucasians, wi h GG ca ie s ep esen ing 3% o he gene al popula ion (ensembl.o g). Inspec ion o genome-wide associa- ion s udy (GWAS) da a om a Finnish gene al popula ion p e iously s udied in ela ion o HCMV se oposi i i y (Kupa inen e al., 2012) did no disclose any signi ican di e ence o s17481334 GG homozygo es (da a no shown). Howe e , as shown in Figu e 4, he same analysis pe o med in indi iduals su e ing om MS, a condi ion whe e HCMV in ec ion has been p oposed o play a p o ec i e ole (Sundq is e al., 2014; Wauban e al., 2011), de ec ed 5- old dec eased odds o HCMV se oposi i i y among GG pa ien s compa ed o all o he indi iduals (p = 0.021; odds a io [OR] = 0.193 [95% con- idence in e al (CI), 0.041–0.913]). Figu e 1. s17481334 Va ian G Abolishes HCMV miR-UL112-5p Binding o he 30UTR o ERAP1 (A) Luci e ase assays o 293T cells ans ec ed wi h dual-luci e ase epo e ec o bea ing he s17481334 A o G a ian o ERAP1 30UTR and ei he a con ol sc ambled sequence (miR-sc ), miR-UL112-5p, o mu a ed miR- UL112-5p (Mu miR-UL112-5p). Da a a e ep esen a i e o i e independen expe imen s. (B and C) Real- ime qPCR (B) and ep esen a i e immunoblo ing analysis (C) o ERAP1 exp ession in ib oblas s wi h AA o GG geno ypes, ans ec ed wi h miR-sc , miR-UL112-5p, Mu miR-UL112-5p, o siERAP1. In (C), an ERp57 an ibody (Ab) is used o no maliza ion. Densi ome ic analysis o ERp57- no malized ERAP1 alues o h ee independen expe imen s is shown below. Mean ±SD; *p < 0.05, **p < 0.001, ***p < 0.0001. 848 Cell Repo s 20, 846–853, July 25, 2017 DISCUSSION In his s udy, we demons a ed ha HCMV miR-UL112-5p a - ge s ERAP1 and ha he s17481334 G a ian na u ally occu - ing in he 30UTR o he ERAP1 gene p ese es ERAP1 om miR-UL112-5p-media ed deg ada ion. Acco dingly, ERAP1 exp ession was educed by miR-UL112-5p o e exp ession bo h a RNA and p o ein le els in human ib oblas s om AA in- di iduals, bu no in ib oblas s om GG indi iduals. Consis en Figu e 2. s17481334 Con ols ERAP1 Exp ession in HCMV-In ec ed Fib oblas s (A) Rep esen a i e low-cy ome ic analysis o MHC class I cell su ace exp ession and in acel- lula IE1/IE2 (IE1/2) i al p o ein exp ession in i- b oblas s wi h AA o GG geno ypes in ec ed wi h HCMV AD169 s ain a he indica ed MOI, 3 days a e in ec ion. The pe cen age o cells in each quad an is indica ed. (B) Ga ing s a egy used o so AD169-in ec ed cells based on he i al-media ed MHC class I down egula ion. MHC class I exp ession in unin- ec ed and in ec ed cells is shown in ligh and da k g ay his og ams, espec i ely. (C) Rep esen a i e immunoblo ing analysis o ERAP1 and MHC class I exp ession in unin ec ed (U) o HCMV AD169-in ec ed ib oblas s wi h AA o GG geno ypes. Densi ome ic analysis o ERp57-no malized ERAP1 alues ela i e o unin- ec ed ib oblas s om h ee independen expe i- men s is shown. (D) Rep esen a i e immunoblo ing analysis o ERAP1 in ib oblas s wi h AA geno ype, ei he unin ec ed (U) o in ec ed wi h wild- ype HCMV AD169 (WT) o HCMVDUL112 (DUL112) a MOI o 5 a 24, 48, and 72 h a e in ec ion. Densi o- me ic analysis o ERp57-no malized ERAP1 alues ela i e o unin ec ed ib oblas s om h ee independen expe imen s is shown. Mean ±SD; *p < 0.05, **p < 0.001. wi h hese esul s, AA ib oblas s in ec ed wi h he wild- ype HCMV, bu no wi h he mu an miR-UL112-5p knockou i- us, showed a lowe amoun o ERAP1 p o ein. HCMV-in ec ed GG ib oblas s we e mo e e icien in imming i al an igens o gene a e immunogenic epi- opes and mo e suscep ible o elimina- ion by HCMV-pep ide-speci ic CTLs. Hence, hese da a demons a e ha he s17481334 GG geno ype hampe s he HCMV miR-UL112-5p-based immunoe- asion s a egy. HCMV has e ol ed o es ablish li elong la en in ec ions in immunocompe en in- di iduals, wi h pe iodic and spon aneous eac i a ion. This asymp oma ic coexis- ence is e lec ed by a p omp esponse o HCMV-speci ic T cells ha domina e he memo y compa men o exposed in- di iduals, accoun ing o app oxima ely 10% o o al CD8 + T cells (Sylwes e e al., 2005). Con e sely, HCMV in ec ion in in- di iduals wi h ei he an imma u e o comp omised immune sys- em (e.g., p e- o pos -na al in an s and ansplan ecipien s, leukemia, o HIV-in ec ed pa ien s, espec i ely) can cause di e en diseases, he se e i y o which depends on he deg ee o immunosupp ession (Halenius and Hengel, 2014). Fo a i us o su i e in an immunocompe en hos , i is essen- ial o a oid he p esen a ion o i al epi opes o CD8 + T cells Cell Repo s 20, 846–853, July 25, 2017 849 (Halenius e al., 2015). In his con ex , HCMV has e ol ed se e al s a egies o a ge key componen s in he MHC class I an igen- p ocessing pa hway, including ERAP1 (Kim e al., 2011). We a - emp ed o explo e his possibili y wi h ou pilo s udy o HCMV se oposi i i y in ela ion o ERAP1 s17481334 geno ype. Unsu - p isingly, no signi ican inding could be de ec ed a he le el o he gene al (Finnish) popula ion, al hough we obse ed a signi - ican associa ion when we in es iga ed he ela ionship be ween GG geno ype and HCMV se ology using MS o ‘‘model’’ hese in- e ac ions. Gi en he a ian allele equency and he size o he coho examined, he ob ained esul s, al hough limi ed o small numbe o indi iduals, we e s a is ically signi ican . Al hough he ue (i any) impac o GG geno ype on HCMV in ec ion, and e en ually MS onse , can only be assessed ia la ge-scale longi udinal epidemiological s udies aking in o accoun indi id- ual i al exposu e, i is no ewo hy ha we obse ed a 5- old dec eased likelihood (OR = 0.193) o se oposi i i y among GG indi iduals wi h MS ( ha is, in a disease con ex whe e HCMV in ec ion has been p oposed o play a p o ec i e ole) compa ed o o he pa ien s. Hence, al hough lacking mechanis ic ela ion o causali y, hese esul s a e in iguing because hey sugges ha impo an in o ma ion may be gained by assessing mul iple ac o s a once and open up no el lines o in es iga ion, whe e he in e play be ween HCMV exposu e, geno ype-d i en im- mune esponses, and ul ima ely disease (au oimmuni y/cance ) can be explo ed owa d ailo ed p e en i e app oaches. Consis- en ly, Li e al. (2011) iden i ied a no el link be ween HCMV in ec- ion and essen ial hype ension, in which miR-UL112 was inde- penden ly associa ed wi h an inc eased isk o hype ension. O no e, ERAP1 is also known o play a ole in he egula ion o blood p essu e h ough i s in ol emen in he enin-angio ensin sys em (Ha o i e al., 2000). Fu he s udies a e wa an ed o con i m and ex end hese indings, also in ela ion o he p esence o mul iple SNP-exp ession quan i a i e ai loci (eQTLs) a he ERAP1 locus, including s17481334 whe e he G a ian appea s o associa e wi h lowe ERAP1 exp ession (h ps://g expo al.o g/). In addi ion o he pp65 495–503 epi ope, ERAP1 is known o im many o he immunogenic HCMV epi opes (Kim e al., 2011). Expe imen s in he ERAAP / mouse model ha e demons a ed ha ERAP1 con ols he magni ude o CD8 + T cell esponse du ing mu ine cy omegalo i us (mCMV) in ec ion (Blancha d e al., 2010). Al hough nei he all HCMV no mCMV-de i ed epi- opes equi e ERAP1 ac i i y o hei gene a ion, ecen mass spec ome y analyses ha e demons a ed ha ERAP1 has a Figu e 3. s17481334 A ec s pp65 495-503 Epi ope P esen a ion (A) Rep esen a i e low-cy ome ic analysis o cell su ace MHC class I exp ession and in acellula IE1/2 i al p o ein exp ession o ib oblas s wi h AA o GG geno ypes bo h unin ec ed (U) o in ec ed a MOI 1 wi h wild- ype HCMV AD169 (WT) o HCMVDUL112 (DUL112), 3 days a e in ec ion. The pe cen age o cells in each quad an is indica ed. (B) 51 C - elease assay o HCMV-in ec ed ib oblas s o (A) co-cul u ed wi h HLA-A2- es ic ed/pp65 495-503 -speci ic CTLs a he indica ed e ec o / a ge (E:T) a io. (C) Rep esen a i e low-cy ome ic analysis o cell su ace MHC class I and in acellula IE1/2 i al p o ein exp ession o shERAP1- o shCTRL- ansduced GG ib oblas s in ec ed wi h wild- ype HCMV a MOI 1, 3 days a e in ec ion. The pe cen age o cells in each quad an is indica ed. (D) 51 C - elease assay o HCMV-in ec ed GG ib oblas s o (C) co-cul u ed wi h HLA-A2- es ic ed/pp65 495-503 -speci ic CTLs a he indica ed e ec o / a ge (E:T) a io. All da a a e ep esen a i e o h ee independen expe imen s. Mean ±SD; *p < 0.05, **p < 0.001. 850 Cell Repo s 20, 846–853, July 25, 2017 signi ican in luence on he o e all pep idome (Lo ´pez de Cas o e al., 2016; Naga ajan e al., 2016). In conclusion, we demons a e ha a a ian in he 30UTR o ERAP1, a key componen o MHC class I an igen p ocessing, impai s miR-UL112-5p-media ed mechanism o HCMV immu- noe asion, wi h po en ial implica ions o he con ol o he i al in ec ion i sel and, e en ually, indi ec e ec s on he isk o HCMV-associa ed diseases. EXPERIMENTAL PROCEDURES 30UTR Luci e ase Assay HEK293T cells we e co- ans ec ed wi h 1.6 mg o pEZX-MT01 ec o s and 80 nmol/L miR-UL112-5p, mu a ed miR-UL112-5p (Mu miR-UL112-5p), o sc ambled con ol (Ambion) using Lipo ec amine 2000 (In i ogen). A e 48 h , luci e ase ac i i y was measu ed wi h a LucPai Duo-Luci e ase Assay Ki (Genecopoeia). Fi e ly luci e ase ac i i y was no malized o Renilla luci - e ase ac i i y. T ans ec ion, Vi al In ec ions, and Pep ide Pulsing Fib oblas s we e ans ec ed wi h 200 nM o ei he miR-sc amble, miR-UL112- 5p, Mu miR-UL112-5p (Ambion), o a siRNA a ge ing a coding egion o ERAP1 (Sigma-Ald ich) (Table S1) wi h Lipo ec amine 2000. As con ol, cells we e ans ec ed wi h he same amoun o an oligo-FITC (In i ogen) in o de o assess he e iciency a 24 h pos - ans ec ion by luo escence-ac i a ed cell so ing (FACS) analysis. Fib oblas s we e in ec ed wi h HCMV s ains a 80%–90% con luence in se um- ee medium. A e 2–3 h o i us adso p ion a 37C, esh g ow h medium was eplaced (day 0). A 3 dpi, cells we e ha es ed and analyzed. In some expe imen s, ib oblas s we e in ec ed ei he wi h a len i i us con aining pp65-GFP and so ed o GFP exp ession, o wi h a non- a ge shRNA con ol ec o (SHC002), o he ERAP1 shRNA (TRCN0000060542) (Sigma-Ald ich) and selec ed wi h 2 mg/mL pu omycin o 3 days. Fo pep ide pulsing, cells we e seeded in 96-well la -bo om pla es in iplica e and pulsed wi h 5 mg/mL pp65 495-503 pep ide. A e 2 h , esh g ow h medium was eplaced. P oli e a ion Assay and Ch omium-Release Assay Fo p oli e a ion assays, in ec ed o pep ide-pulsed cells we e seeded in 96-well la -bo om pla es in iplica e and incuba ed wi h HLA-A2- es ic ed/ pp65 495-503 -speci ic CD8 + T cells. A e 72-h incuba ion a 37C, cells we e pulsed wi h 3 H- hymidine (1 mg Ci/well; speci ic ac i i y, 6.7 Ci/mmol; Pe kinElme ) and ha es ed a e 18 h . 3 H- hymidine inco po a ion was measu ed by s anda d p ocedu es wi h Mic obe a T ilux 1450 ins umen (Pe kinElme ). Fo cy o oxic ac i i y, 51 C -labeled in ec ed o pep ide-pulsed a ge cells (5 310 3 pe well) we e mixed wi h HLA-A2- es ic ed/ pp65 495-503 -speci ic CD8 + T cells in 96-well pla es in iplica e a a ious a ios and incuba ed a 37C. A e 5 h o incuba ion, supe na an was collec ed, and he 51 C elease was measu ed by a TopCoun NXT bde ec o (Pe kinElme ). All expe imen al g oups we e analyzed in iplica e, and he speci ic lysis was de e mined as ollows: 100 3(expe imen al elease spon aneous elease)/ ( o al elease spon aneous elease). Spon aneous elease lowe han abou 5% o he o al elease was obse ed in all assays. MS Pa ien Coho Samples om 949 MS pa ien s wi h ages be ween 16 and 70 yea s om he Epidemiological In es iga ion o Mul iple Scle osis (EIMS) we e included in his s udy. EIMS is a na ionwide popula ion-based case-con ol s udy o inciden cases o MS in Sweden (Heds o ¨me al.,2011). All o he cases ul illed he McDonald c i e ia o MS (Polman e al., 2005) and we e examined and diag- nosed by a neu ologis a one o 30 ec ui ing cen e s in Sweden. An i-HCMV se ological s a us was de e mined in plasma using ELISA as desc ibed p e i- ously (Sundq is e al., 2014). The HCMV an igen was p epa ed om emb yonic ib oblas s in ec ed wi h s ain AD169 un il comple e cy opa hic e ec . Human plasma samples we e analyzedin duplica es, a a dilu ion o 1/200. Alkaline phos- pha ase-conjuga ed, pu i ied F(ab0) 2 agmen goa an i-human IgG was used as conjuga e (Jackson ImmunoResea ch). The pla es we e ead a 405 and 650 nmin a Mul iskan FC eade (The moFishe Scien i ic). Cu o was calcula ed by adding 0.2 op ical densi y (OD)uni s o he mean abso bance alue o 160 uns o nega i e con ol samples. Plasma samples wi h OD alues a cu o we e un in 2- old dilu ions (1/100–1/12,800) and hen judged as posi i e o nega i e. E hics S a emen E hical app o al o gene a ing and analyzing human ib oblas s om skin bi- opsies was ob ained om he Ins i u ional Re iew Boa d o he Hospi al Bambino Gesu `. W i en in o med consen was ob ained om adul pa ien s, o he gua dians o pa ien s who we e child en. The EIMS s udy has been app o ed by he Regional E hical Re iew Boa d in S ockholm, Sweden (h p://www.epn.se). W i en in o med consen om all s udy pa icipan s o hei pa en s was ob ained. In es iga ions we e ca ied ou acco ding o guidelines om he Decla a ion o Helsinki. S a is ical Analysis Da a a e p esen ed as mean, and e o ba s indica e he SD. Digi al images o wes e n blo s we e analyzed by ImageJ (h ps://imagej.nih.go /ij/index.h ml), and s a is ical signi icance was assessed by he wo-way ANOVA es . c 2 es was used o analyze he di e en dis ibu ion o s17481334 geno ype e- quency in HCMV-posi i e and -nega i e Swedish MS cases. SUPPLEMENTAL INFORMATION Supplemen al In o ma ion includes Supplemen al Expe imen al P ocedu es, ou igu es, and one able and can be ound wi h his a icle online a h p:// dx.doi.o g/10.1016/j.cel ep.2017.06.084. AUTHOR CONTRIBUTIONS D.F., F.L., and M.D. designed he s udy. D.F. supe ised he wo k. P.R., L.C., B.P., N.S., V.D., O.M., G.L.P., E.G., and C.C. pe o med he expe imen s. R.C. and H.H. p o ided c i ical eagen s. M.B., T.B., L.A., T.O., I.K., I.S., T.L., M.A.H., and M.D. pe o med analysis o mul iple scle osis. P.R., L.C., C.C., M.D., and D.F. analyzed and in e p e ed he da a. P.R., M.D., F.L., and D.F. w o e he manusc ip . A.S. and C.C. edi ed he manusc ip . ACKNOWLEDGMENTS We hank D . S e ano Pe occhi o kindly p o iding genomic DNA and P o . Bodo Plach e and Pe e an Ende o pcDNA6-pp65 plasmid and ERAP1 an ibody, espec i ely. This wo k was suppo ed by g an s om he I alian Min- is y o Heal h (PE-2011-02351866 o D.F.; RF-2010-2316606 o F.L.), he As- sociazione I aliana Rice ca sul Canc o (AIRC) (18495 o D.F. and Special P oj- ec 5x1000 9962 o F.L.), PRIN 2010 ( o F.L.), PRIN 2015-W729WH ( o C.C.), and Co don de Vie ( o F.L.). The EIMS s udy was suppo ed by g an s om he Swedish Resea ch Council, he Swedish Resea ch Council o Heal h and Wo king Li e and Wel a e, and he Knu and Alice Wallenbe g Founda ion. Figu e 4. s17481334 Geno ype and HCMV Se ology in MS Pa ien s Se ological an i-HCMV Ab p o iles a e epo ed o MS pa ien s s a i ied acco ding o s17481334 geno ype. Cell Repo s 20, 846–853, July 25, 2017 851 Recei ed: Augus 16, 2016 Re ised: Feb ua y 6, 2017 Accep ed: June 28, 2017 Published: July 25, 2017 REFERENCES Blancha d, N., Kanaseki, T., Escoba , H., Delebecque, F., Naga ajan, N.A., Reyes-Va gas, E., C ocke , D.K., Raule , D.H., Delgado, J.C., and Shas i, N. (2010). Endoplasmic e iculum aminopep idase associa ed wi h an igen p ocessing de ines he composi ion and s uc u e o MHC class I pep ide epe oi e in no mal and i us-in ec ed cells. J. Immunol. 184, 3033–3042. Ci aldi, L., Lo Monaco, E., Fo loni, M., Gio da, E., Lo enzi, S., Pe ini, S., T em- an e, E., Pende, D., Loca elli, F., Giacomini, P., and F uci, D. (2011). Na u al kille cells e icien ly ejec lymphoma silenced o he endoplasmic e iculum aminopep idase associa ed wi h an igen p ocessing. Cance Res. 71, 1597– 1606. Ci aldi, L., Romania, P., Falco, M., Lo enzi, S., Meazza, R., Pe ini, S., An- d eani, M., Pende, D., Loca elli, F., and F uci, D. (2015). ERAP1 egula es na - u al kille cell unc ion by con olling he engagemen o inhibi o y ecep o s. Cance Res. 75, 824–834. D aene , R., Le Gall, S., P a e o , K.J., Leslie, A.J., Che y, P., B ande , C., Holmes, E.C., Chang, S.C., Feeney, M.E., Addo, M.M., e al. (2004). Immune selec ion o al e ed an igen p ocessing leads o cy o oxic T lymphocy e escape in ch onic HIV-1 in ec ion. J. Exp. Med. 199, 905–915. Fi a , E., Sa eanu, L., Aichele, P., S aeheli, P., Huai, J., Gaedicke, S., Nil, A., Besin, G., Kanzle , B., an Ende , P., and Niede mann, G. (2007). The ole o endoplasmic e iculum-associa ed aminopep idase 1 in immuni y o in ec- ion and in c oss-p esen a ion. J. Immunol. 178, 2241–2248. G ey, F., Meye s, H., Whi e, E.A., Spec o , D.H., and Nelson, J. (2007). A hu- man cy omegalo i us-encoded mic oRNA egula es exp ession o mul iple i al genes in ol ed in eplica ion. PLoS Pa hog. 3, e163. G ey, F., Ti abassi, R., Meye s, H., Wu, G., McWeeney, S., Hook, L., and Nelson, J.A. (2010). A i al mic oRNA down- egula es mul iple cell cycle genes h ough mRNA 50UTRs. PLoS Pa hog. 6, e1000967. Halenius, A., and Hengel, H. (2014). Human cy omegalo i us and au oimmune disease. BioMed Res. In . 2014, 472978. Halenius, A., Ge ke, C., and Hengel, H. (2015). Classical and non-classical MHC I molecule manipula ion by human cy omegalo i us: so many a - ge s—bu how many a ows in he qui e ? Cell. Mol. Immunol. 12, 139–153. Ha o i, A., Ki a ani, K., Ma sumo o, H., Miyazawa, S., Rogi, T., Tsu uoka, N., Mizu ani, S., Na o i, Y., and Tsujimo o, M. (2000). Cha ac e iza ion o ecombi- nan human adipocy e-de i ed leucine aminopep idase exp essed in Chinese hams e o a y cells. J. Biochem. 128, 755–762. Heds o ¨m, A.K., Sundq is , E., Ba ¨a ¨ nhielm, M., No din, N., Hille , J., Kockum, I., Olsson, T., and Al edsson, L. (2011). Smoking and wo human leukocy e an igen genes in e ac o inc ease he isk o mul iple scle osis. B ain 134, 653–664. Hook, L., Hancock, M., Landais, I., G abski, R., B i , W., and Nelson, J.A. (2014a). Cy omegalo i us mic oRNAs. Cu . Opin. Vi ol. 7, 40–46. Hook, L.M., G ey, F., G abski, R., Ti abassi, R., Doyle, T., Hancock, M., Land- ais, I., Jeng, S., McWeeney, S., B i , W., and Nelson, J.A. (2014b). Cy omeg- alo i us miRNAs a ge sec e o y pa hway genes o acili a e o ma ion o he i ion assembly compa men and educe cy okine sec e ion. Cell Hos Mic obe 15, 363–373. James, E., Bailey, I., Sugiya o, G., and Ellio , T. (2013). Induc ion o p o ec i e an i umo immuni y h ough a enua ion o ERAAP unc ion. J. Immunol. 190, 5839–5846. Kim, S., Lee, S., Shin, J., Kim, Y., E nouchidou, I., Kim, D., Kim, Y.K., Kim, Y.E., Ahn, J.H., Riddell, S.R., e al. (2011). Human cy omegalo i us mic oRNA miR- US4-1 inhibi s CD8 + T cell esponses by a ge ing he aminopep idase ERAP1. Na . Immunol. 12, 984–991. Kim, Y., Lee, S., Kim, S., Kim, D., Ahn, J.H., and Ahn, K. (2012). Human cy o- megalo i us clinical s ain-speci ic mic oRNA miR-UL148D a ge s he human chemokine RANTES du ing in ec ion. PLoS Pa hog. 8, e1002577. Kim, S., Seo, D., Kim, D., Hong, Y., Chang, H., Baek, D., Kim, V.N., Lee, S., and Ahn, K. (2015). Tempo al landscape o mic oRNA-media ed hos - i us c oss- alk du ing p oduc i e human cy omegalo i us in ec ion. Cell Hos Mic obe 17, 838–851. Kupa inen, T., Seppa ¨la ¨, I., Jylha ¨ a ¨, J., Ma ila, S., Ai oniemi, J., Ke unen, J., Viika i, J., Ka ¨ho ¨nen, M., Rai aka i, O., Leh ima ¨ki, T., and Hu me, M. (2012). Genome-wide associa ion s udy does no e eal majo gene ic de e minan s o an i-cy omegalo i us an ibody esponse. Genes Immun. 13, 184–190. Lee, S.H., Kalej a, R.F., Ke y, J., Semmes, O.J., O’Conno , C.M., Khan, Z., Ga cia, B.A., Shenk, T., and Mu phy, E. (2012). BclAF1 es ic ion ac o is neu alized by p o easomal deg ada ion and mic oRNA ep ession du ing hu- man cy omegalo i us in ec ion. P oc. Na l. Acad. Sci. USA 109, 9575–9580. Li, S., Zhu, J., Zhang, W., Chen, Y., Zhang, K., Popescu, L.M., Ma, X., Lau, W.B., Rong, R., Yu, X., e al. (2011). Signa u e mic oRNA exp ession p o ile o essen ial hype ension and i s no el link o human cy omegalo i us in ec- ion. Ci cula ion 124, 175–184. Lo ´pez de Cas o, J.A., Al a ez-Na a o, C., B i o, A., Guasp, P., Ma ı ´n-Es e- ban, A., and Sanz-B a o, A. (2016). Molecula and pa hogenic e ec s o endo- plasmic e iculum aminopep idases ERAP1 and ERAP2 in MHC-I-associa ed in lamma o y diso de s: owa ds a uni ying iew. Mol. Immunol. 77, 193–204. Meshesha, M.K., Veksle -Lublinsky, I., Isako , O., Reichens ein, I., Shom on, N., Kedem, K., Zi -Ukelson, M., Ben wich, Z., and A ni, Y.S. (2012). The mic o- RNA ansc ip ome o human cy omegalo i us (HCMV). Open Vi ol. J. 6, 38–48. Naga ajan, N.A., de Ve euil, D.A., S i anganadane, D., Yahyaoui, W., Thibaul , P., Pe eaul , C., and Shas i, N. (2016). ERAAP shapes he pep idome associ- a ed wi h classical and nonclassical MHC class I molecules. J. Immunol. 197, 1035–1043. Polman, C.H., Reingold, S.C., Edan, G., Filippi, M., Ha ung, H.P., Kappos, L., Lublin, F.D., Me z, L.M., McFa land, H.F., O’Conno , P.W., e al. (2005). Diagnos ic c i e ia o mul iple scle osis: 2005 e isions o he ‘‘McDonald C i e ia’’. Ann. Neu ol. 58, 840–846. Ree es, E., Coleba ch-Bou n, A., Ellio , T., Edwa ds, C.J., and James, E. (2014). Func ionally dis inc ERAP1 allo ype combina ions dis inguish indi id- uals wi h Ankylosing Spondyli is. P oc. Na l. Acad. Sci. USA 111, 17594– 17599. Rehmsmeie , M., S e en, P., Hochsmann, M., and Giege ich, R. (2004). Fas and e ec i e p edic ion o mic oRNA/ a ge duplexes. RNA 10, 1507–1517. Ryan, B.M., Robles, A.I., and Ha is, C.C. (2010). Gene ic a ia ion in mic o- RNA ne wo ks: he implica ions o cance esea ch. Na . Re . Cance 10, 389–402. Sa eanu, L., Ca oll, O., Lindo, V., Del Val, M., Lopez, D., Lepelle ie , Y., G ee , F., Schombu g, L., F uci, D., Niede mann, G., and an Ende , P.M. (2005). Conce ed pep ide imming by human ERAP1 and ERAP2 aminopep idase complexes in he endoplasmic e iculum. Na . Immunol. 6, 689–697. S a k, T.J., A nold, J.D., Spec o , D.H., and Yeo, G.W. (2012). High- esolu ion p o iling and analysis o i al and hos small RNAs du ing human cy omegalo- i us in ec ion. J. Vi ol. 86, 226–235. S e n-Ginossa , N., Ele an , N., Zimme mann, A., Wol , D.G., Saleh, N., Bi on, M., Ho wi z, E., P okocime , Z., P icha d, M., Hahn, G., e al. (2007). Hos immune sys em gene a ge ing by a i al miRNA. Science 317, 376–381. S e n-Ginossa , N., Weisbu d, B., Michalski, A., Le, V.T., Hein, M.Y., Huang, S.X., Ma, M., Shen, B., Qian, S.B., Hengel, H., e al. (2012). Decoding human cy omegalo i us. Science 338, 1088–1093. S a ikos, E., S amogiannos, A., Ze oudi, E., and F uci, D. (2014). A ole o na u ally occu ing alleles o endoplasmic e iculum aminopep idases in umo immuni y and cance p e-disposi ion. F on . Oncol. 4, 363. Sundq is , E., Be gs o ¨m, T., Daialhosein, H., Nys o ¨m, M., Sunds o ¨m, P., Hille , J., Al edsson, L., Kockum, I., and Olsson, T. (2014). Cy omegalo i us se oposi i i y is nega i ely associa ed wi h mul iple scle osis. Mul . Scle . 20, 165–173. 852 Cell Repo s 20, 846–853, July 25, 2017 Sylwes e , A.W., Mi chell, B.L., Edga , J.B., Tao mina, C., Pel e, C., Ruch i, F., Slea h, P.R., G abs ein, K.H., Hosken, N.A., Ke n, F., e al. (2005). B oadly a - ge ed human cy omegalo i us-speci ic CD4+ and CD8+ T cells domina e he memo y compa men s o exposed subjec s. J. Exp. Med. 202, 673–685. Tenze , S., Wee, E., Bu ge in, A., S ewa -Jones, G., F iis, L., Lambe h, K., Chang, C.H., Ha ndahl, M., Weime shaus, M., Ge s o , J., e al. (2009). An i- gen p ocessing in luences HIV-speci ic cy o oxic T lymphocy e immunodomi- nance. Na . Immunol. 10, 636–646. Ti abassi, R., Hook, L., Landais, I., G ey, F., Meye s, H., Hewi , H., and Nelson, J. (2011). Human cy omegalo i us US7 is egula ed syne gis ically by wo i ally encoded mic oRNAs and by wo dis inc mechanisms. J. Vi ol. 85, 11938–11944. U ban, S., Tex o is-Taube, K., Reimann, B., Janek, K., Dannenbe g, T., Ebs ein, F., Sei e , C., Zhao, F., Kessle , J.H., Halenius, A., e al. (2012). The e iciency o human cy omegalo i us pp65(495-503) CD8 + T cell epi ope gen- e a ion is de e mined by he balanced ac i i ies o cy osolic and endoplasmic e iculum- esiden pep idases. J. Immunol. 189, 529–538. an de Weije , M.L., Lu eijn, R.D., and Wie z, E.J. (2015). Vi al immune e asion: lessons in MHC class I an igen p esen a ion. Semin. Immunol. 27, 125–137. Wauban , E., Mow y, E.M., K upp, L., Chi nis, T., Yeh, E.A., Kun z, N., Ness, J., Chabas, D., S obe , J., McDonald, J., e al.; US Pedia ic MS Ne wo k (2011). Common i uses associa ed wi h lowe pedia ic mul iple scle osis isk. Neu ology 76, 1989–1995. Yan, J., Pa ekh, V.V., Mendez-Fe nandez, Y., Oli a es-Villago ´mez, D., D ag- o ic, S., Hill, T., Roopenian, D.C., Joyce, S., and Van Kae , L. (2006). In i o ole o ER-associa ed pep idase ac i i y in ailo ing pep ides o p esen a ion by MHC class Ia and class Ib molecules. J. Exp. Med. 203, 647–659. Cell Repo s 20, 846–853, July 25, 2017 853