Epigenetically altered miR-193b targets cyclin D1 in prostate cancer

1417 In oduc ion P os a e cance (PC) is he second mos equen ly diagnosed cance ype and he six h leading cause o cance - ela ed dea hs in males wo ldwide [1]. Localized PC can be cu ed wi h p os a ec omy and/o adia ion he apy. Howe e , he e is no cu a i e ea men o ad anced and cas a ion- esis an disease [2]. The use o p os a e- speci ic an igen (PSA) o he sc eening o asymp oma ic men o p os a e cance is known o educe he disease- speci ic mo ali y, bu sc eening is associa ed wi h o e diagnosis [3]. The e o e, he e is a s ong in e es o ind addi ional ma ke s and he apeu ic a ge s o he diagnosis and ea men o p os a e can- ce . Mic oRNAs (miRNAs) could se e as such ma ke s o as d ug a ge s [4]. ORIGINAL RESEARCH Epigene ically al e ed miR- 193b a ge s cyclin D1 in p os a e cance Ki si M. Kaukoniemi1,2,a, Hanna E. Rauhala1,a, Mau o Sca a illi1,2, Leena La onen1,2, Ma i Annala1, Robe L. Vessella3, Ma i Nyk e 1, Teu o L. J. Tammela4 & Tapio Visako pi1,2 1Ins i u e o Biosciences and Medical Technology - BioMediTech, Uni e si y o Tampe e, Tampe e, Finland 2Fimlab Labo a o ies, Tampe e Uni e si y Hospi al, Tampe e, Finland 3Depa men o U ology, Uni e si y o Washing on, Sea le, Washing on 4Depa men o U ology, Uni e si y o Tampe e and Tampe e Uni e si y Hospi al, Tampe e, Finland © 2015 The Au ho s. Cance Medicine published by John Wiley & Sons L d. This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. Keywo ds Cyclin D1, mic o-RNA, p os a e cance Co espondence Tapio Visako pi, Ins i u e o Biosciences and Medical Technology, Uni e si y o Tampe e, Tampe e FI-33014, Finland. Tel: +358 50 318 5829; Fax: +358 3 364 1291; E-mail: [email p o ec ed] Funding In o ma ion The esea ch leading o hese esul s was unded by Tampe e Uni e si y Doc o al P og amme in Biomedicine and Bio echnology. In addi ion, g an suppo has been ecei ed om he Sig id Juselius Founda ion, he Finnish Cul u al Founda ion, Pi kanmaa Regional und, he Academy o Finland, he Cance Socie y o Finland, he Medical Resea ch Fund o Tampe e Uni e si y Hospi al, and he Eu opean Communi y’s Se en h F amewo k P og amme P ospeR (FP7/2007-2013) unde g an ag eemen no. HEALTH-F2-2007-201438. Recei ed: 14 Janua y 2015; Re ised: 13 May 2015; Accep ed: 27 May 2015 Cance Medicine 2015; 4(9):1417–1425 doi: 10.1002/cam4.486 aEqually con ibu ing au ho s. Abs ac Mic o- RNAs (miRNA) a e impo an egula o s o gene exp ession and o en di e en ially exp essed in cance and o he diseases. We ha e p e iously shown ha miR- 193b is hype me hyla ed in p os a e cance (PC) and supp esses cell g ow h. I has been sugges ed ha miR- 193b a ge s cyclin D1 in se e al malignancies. He e, ou aim was o de e mine i miR- 193b a ge s cyclin D1 in p os a e cance . Ou da a show ha miR- 193b is commonly me hyla ed in PC samples compa ed o benign p os a e hype plasia. We ound educed miR- 193b exp ession (P < 0.05) in s age pT3 umo s compa ed o pT2 umo s in a coho o p os a ec omy specimens. In 22R 1 PC cells wi h low endogenous miR- 193b exp ession, he o e exp ession o miR- 193b educed CCND1 mRNA le els and cyclin D1 p o ein le els. In addi ion, he exogenous exp ession o miR- 193b dec eased he phospho yla ion le el o RB, a a ge o he cyclin D1- CDK4/6 pa hway. Mo eo e , acco ding o a epo e assay, miR- 193b a ge ed he 3’UTR o CCND1 in PC cells and he CCND1 ac i i y was escued by exp essing CCND1 lacking i s 3’UTR. Immunohis ochemical analysis o cyclin D1 showed ha cas a ion- esis an p os a e cance s ha e signi ican ly (P = 0.0237) highe exp ession o cyclin D1 compa ed o ho mone- naï e cases. Fu he mo e, he PC cell lines 22R 1 and VCaP, which exp ess low le els o miR- 193b and high le els o CCND1, showed signi ican g ow h e a da ion when ea ed wi h a CDK4/6 inhibi o . In con as , he inhibi o had no e ec on he g ow h o PC- 3 and DU145 cells wi h high miR- 193b and low CCND1 exp ession. Taken oge he , ou da a demons a e ha miR- 193b a ge s cyclin D1 in p os a e cance . Cance Medicine Open Access 1418 © 2015 The Au ho s. Cance Medicine published by John Wiley & Sons L d. K. M. Kaukoniemi e al.miR-193b Ta ge s Cyclin D1 in PC miRNAs a e small, ~22 nucleo ide long noncoding RNA molecules ha we e i s disco e ed in Caeno habdi is elegans [5]. Since hei disco e y, miRNAs ha e been ound o be impo an egula o s o gene exp ession in o he o ganisms as well [5, 6]. miRNAs modula e gene exp es- sion by binding o a complemen a y sequence in he 3’ un ansla ed egion (3’UTR) o a ge messenge RNAs (mRNAs). The binding o a miRNA leads o he deg ada- ion o he mRNA molecule o al e na i ely, o he sup- p ession o mRNA ansla ion, depending on he le el o complemen a y binding [7]. Because one miRNA can ha e mul iple a ge s and one mRNA molecule can be a ge ed by many miRNAs, miRNAs o m a complex le el o egu- la ion o gene exp ession. In addi ion o gene ic changes (ch omosomal ea angemen s, dele ions, ampli ica ions, and mu a ions), epigene ic e en s such as abe an p o- mo e hype me hyla ion, global hypome hyla ion, and pos ansc ip ional his one modi ica ion may also cause he dys egula ion o miRNAs in cance [7]. Se e al miR- NAs a e de egula ed in p os a e cance and ha e been shown o a ec apop osis, cell cycle, in acellula signaling, DNA epai , adhesion/mig a ion, and and ogen signaling ( e iewed in [4]). We p e iously showed ha hsa- miR- 193b- 3p (aka miR- 193b) may unc ion as an epigene ically egula ed umo supp esso in p os a e cance [8]. Using bisul i e sequenc- ing we showed ha miR- 193b is hype me hyla ed in some PC cell lines, wi h he 22R 1 cell line being he mos hea ily me hyla ed esul ing in he loss o miR- 193b ex- p ession in he cells. T ansien ans ec ion o p e- miR- 193b in o 22R 1 cells caused signi ican g ow h educ ion due o a dec eased ac ion o cells in S- phase o he cell cycle. The e a e se e al sugges ed a ge genes o miR- 193b in di e en cance s, o example, CCND1, ETS1, and KIT [9–16]. The cyclin D1- encoding gene, CCND1, is a ge ed by miR- 193b in hepa ocellula ca cinoma [10], melanoma [11], and panc ea ic cance [14]. Cyclin D1 is equen ly abe an ly exp essed in cance . I egula es he exp ession o genes ha a e in ol ed in DNA ep- lica ion and he DNA damage checkpoin . The bes known unc ion o cyclin D1 is mos likely i s ca aly ic unc ion as a egula o y pa ne o he cyclin- dependen kinases (CDKs) 4 and 6. In esponse o cellula s imuli, cyclin D1 ac i a es CDKs 4 and 6, which in u n phos- pho yla e a ious p o eins, he p incipal subs a e being e inoblas oma (RB) p o ein. When phospho yla ed, RB eleases he E2F ansc ip ion ac o , which ac i a es genes ha a e necessa y o DNA syn hesis and cell cycle p og ession om he G1 o S phase. In addi ion o i s ca aly ic unc ion, cyclin D1 also has nonca aly ic unc ions, wi h he majo unc ion being ansc ip ional egula ion [17]. In p os a e cance , cyclin D1 has been shown o unc ion as a co ep esso o and ogen ecep- o (AR) [18–20]. Because cyclin D1 is a sugges ed a ge o miR- 193b in o he cance s, we aimed o s udy whe he i is also a a ge in p os a e cance . In addi ion, ou goal was o con i m he hype me hyla ion o miR- 193b in clinical p os a e cance samples. Ma e ials and Me hods Cell lines, xenog a s, and clinical samples The p os a e cance cell lines 22R 1, PC- 3, LNCaP, and DU145 we e ob ained om he Ame ican Type Cell Collec ion (Manassas, VA). LAPC- 4 and VCaP cell lines we e p o ided by D . Cha les Sawye s (Uni e si y o Cali o nia a Los Angeles, Los Angeles, CA) and D . Jack Schalken (Radboud Uni e si y Nijmegen Medical Cen e , Nijmegen, he Ne he lands), espec i ely. All cell lines we e cul u ed unde ecommended condi ions. The 17 PC LuCaP xenog a s we e p o ided by one o he au ho s (R.L.V.). All clinical samples we e ob ained om Tampe e Uni e si y Hospi al (TAUH, Tampe e, Finland). F esh- ozen issue samples, including 10 benign p os a e hype plasias (BPH), 26 un ea ed p os a ec omy specimens, and nine cas a ion- esis an umo s (CRPC), we e used o s udy miR- 193b me hyla ion, whe eas 78 ho monally un ea ed PC p os a ec omy specimens (Table S1) we e used o s udy miR- 193b exp ession. BPH samples we e ob ained om ansu e h al esec ion (TURP) o (cys o) p os a ec omies om pa ien s wi h BPH o bladde cance . Un ea ed cance samples we e ob ained om p os a ec- omies and CRPC samples om TURP. The samples used in his s udy we e his ologically examined o con ain >70% cance ous o hype plas ic issue. Fo cyclin D1 immuno- his ochemical analysis, a o al o 267 o malin- ixed p os a e cance specimens (198 om p os a ec omies and 69 CRPC samples) we e used o cons uc issue mic oa ays (TMAs; Table S2). E hics Commi ee o Tampe e Uni e si y Hospi al and Na ional Au ho i y o Medicolegal A ai s ha e app o ed he use o clinical umo ma e ial. DNA and RNA ex ac ion om clinical samples Fo he me hyla ion analysis, eshly ozen issue blocks we e cu in o 10 × 20- mic ome e sec ions using a c y- o ome. DNA was isola ed using an AllP ep RNA/DNA miniki (Qiagen, Valencia, CA) acco ding o he manu- ac u e ’s p o ocol. Fo miR- 193b exp ession analysis and a ays, RNA was isola ed using T izol® eagen (In i ogen, Li e Technologies Co po a ion, Ca lsbad, CA) acco ding o he manu ac u e ’s p o ocol. 1419 © 2015 The Au ho s. Cance Medicine published by John Wiley & Sons L d. miR-193b Ta ge s Cyclin D1 in PCK. M. Kaukoniemi e al. Mic oa ays miRNA and mRNA exp ession da a om cell lines was p oduced using Agilen Tehcnologies’ (San a Cla a, CA) miRNA mic oa ay sys em ( e sion 1 a ay con aining 470 human and 64 i al miRNAs) [8] and Whole Human Genome Ki (4 × 44k) (Agilen Tehcnologies, San a Cla a, CA), e- spec i ely. Fo LuCaP xenog a s A yme ix (San a Cla a, CA) HG U133 Plus 2.0 mic oa ay was used. The a ays we e done acco ding o he manu ac u e ’s p o ocols. Cell ans ec ion and co ans ec ion To de e mine he e ec o miR- 193b exp ession on cyclin D1, RB, and pRB le els, 22R 1 cells we e ans ec ed wi h a inal concen a ion o 5 nmol/L p e- miR- 193b o sc am- bled con ol using INTERFERin ans ec ion eagen (Polyplus- ans ec ion, Illki ch, F ance) acco ding o he manu ac u e ’s ins uc ions. The cells we e collec ed 3 days a e ans ec ion. miRNA p ecu so s we e pu chased om Ambion (Applied Biosys ems/Ambion, Aus in, TX). Fo he Cyclin D1 escue and cell cycle expe imen , 22R 1 cells we e co ans ec ed wi h 10 μg o plasmid, pCMV- CCND1 lacking a 3’UTR (Plasmid 19927; Addgene, Camb idge, MA), o he con ol plasmid pCMV6- AC- GFP (O iGene Technologies, Rock ille, MD) along wi h ei he p e- miR- 193b o p e- miR- sc amble a a inal concen a- ion o 10 nmol/L, using je PRIME ans ec ion eagen (Polyplus- ans ec ion). The cells we e collec ed 2 days a e ans ec ion. Flow cy ome ic analysis Cells we e ha es ed by ypsiniza ion, washed wi h phospha e- bu e ed saline (PBS) and ixed in cold 70% E OH. E hanol was aspi a ed, and he cells we e washed and ehyd a ed wi h PBS. S aining was pe o med wi h p opidium iodide (Sigma-Ald ich, S . Louis, MO), and cell cycle p o iles we e analyzed wi h Accu i C6 Flow cy ome e . Luci e ase epo e assay Fo he luci e ase epo e assay, he pSGG- 3UTR plas- mid was pu chased om Swi chGea Genomics (Menlo Pa k, CA). The plasmid con ains a luci e ase gene used wi h he 3’UTR o CCND1. 22R 1 cells we e co ans- ec ed wi h he 3’UTR plasmid, a con ol plasmid- con aining Renilla luci e ase and p e- miR- 193b o p e- miR- sc amble using Lipo ec amine™ 2000 ans ec- ion eagen (In i ogen) acco ding o he manu ac u e ’s ins uc ions. Fi e ly and Renilla luci e ase ac i i ies we e measu ed 24 h a e ans ec ion using he Dual- Glo Luci e ase Assay Sys em (P omega, Madison, WI). Renilla luci e ase alues we e used o da a no maliza ion. The luci e ase assay was pe o med in quad uplica e, and epea ed ou imes. Wes e n Blo Nuclea and cy oplasmic p o eins o RB and pRB wes e n blo s and o al p o ein o cyclin D1 wes e n blo s we e isola ed as desc ibed p e iously [21, 22]. A o al o 20 μg o p o eins we e sepa a ed on 8% (RB, pRB) and 10% (Cyclin D1) SDS- PAGE gels and blo ed o a PVDF mem- b ane (Immobilon- P, Millipo e Co p., Bille ica, MA). The memb anes we e incuba ed wi h p ima y an ibodies agains phospho- Rb (Se 795, 1:1000, Cell Signaling, Dan e s, MA), Rb (C- 15: sc- 50, 1:500; San a C uz Bio echnology, Inc., Dallas, TX) o cyclin D1 (clone SP4, 1:100; Dako, Glos up, Denma k) and wi h an ibodies agains ib illa in (C13C3, 1:4000; Cell Signaling) o ac in (pan AB- 5 clone ACTN05, 1:400; Lab Vision Co p., F emon , CA). A e seconda y an ibody in- cuba ion (an i- abbi - HRP o phospho- RB, RB, cyclin D1 and ib illa in, and an i- mouse- HRP o Ac in; 1:3000, Dako), he p o eins we e isualized by au o adiog aphy. Immunohis ochemis y An ibodies agains cyclin D1 (dilu ion 1:100, clone SP4; Dako) and Ki- 67 (dilu ion 1:500, MM1; Leica Biosys ems Newcas le L d., Newcas le upon Tyne, UK) we e used wi h a Powe Vision+ Poly- HRP IHC ki (ImmunoVision Technologies Co., Hillsbo ough, CA) acco ding o he manu ac u e ’s ins uc ions and as desc ibed by Leinonen e al. [23]. The slides we e scanned wi h an Ape io ScanScope XT scanne (Leica Mic osys ems GmbH, We zla , Ge many). The i ual mic oscope [24, 25] and he ImmunoRa io web applica ion [26] we e used o sco e he cyclin D1- s aining in ensi y (0–1 = nega i e and weak, 2 = mode a e, and 3 = s ong) in a blinded ashion. Quan i a i e eal- ime RT- PCR Fo miR- 193b q- RT- PCR, a TaqMan mic oRNA Assay (Applied Biosys ems, Fos e Ci y, CA) and he CFX96 q- RT- PCR de ec ion sys em (Bio- Rad Labo a o ies Inc., He cules, CA) we e used acco ding o he manu ac u e s’ ecommenda ions. miRNA exp ession was no malized o RNU6B exp ession. Fo CCND1 q- RT- PCR, i s - s and complemen a y DNA syn hesis was pe o med om o al RNA using AMV e e se ansc ip ase (Finnzymes Inc., Espoo, Finland) acco ding o he manu ac u e ’s ins uc- ions. The exp ession o CCND1 was measu ed wi h Maxima SYBR G een (Fe men as Inc., Bu ling on, ON, Canada) and he CFX96 q- RT- PCR de ec ion sys em (Bio- Rad Labo a o ies Inc.) and was no malized o he β - Ac in 1420 © 2015 The Au ho s. Cance Medicine published by John Wiley & Sons L d. K. M. Kaukoniemi e al.miR-193b Ta ge s Cyclin D1 in PC e e ence gene. The p ime sequences used o he CCND1 and β- Ac in q- RT- PCR we e: CyclinD1 o 5’- CCCTCGGTG GGTCCTACTTCAA- 3’, CyclinD1 e 5’- TGGCATTTTGG AGAGGAAGT- 3’, and Bac in 4 5’- TGGGACGACATGGAG AAAAT- 3’, Bac in 4 5’- AGAGGCGTACAGGGATAGCA- 3’. miR- 193b me hyla ion analysis Me hyla ed DNA was en iched using Me hylMine ™ Me hyla ed DNA En ichmen Ki (In i ogen) acco ding o he manu ac u e ’s ins uc ions. B ie ly, 2 μg o genomic DNA was agmen ed by sonica ion. Me hyla ed DNA was en iched by binding o magne ic beads coa ed wi h he me hyl- CpG- binding domain o he human MBD2 p o ein (Me hyl- CpG Binding Domain P o ein 2) and elu ed as a single ac ion using 2 M NaCl. Finally, he DNA was e hanol- p ecipi a ed and esuspended in 50 μL o DNase- ee wa e . Fo q- PCR, iQTM SYBR® G een supe mix and CFX96 q- RT- PCR de ec ion sys em (Bio- Rad Labo a o ies Inc.) we e used. The sequences o he p ime s used in he q- PCR we e: miR- 193b_DMR_ 1 5’- TGGCGTTTCTGG TTTCTCTT- 3’ and miR- 193b_DMR_ 2 5’- CGCACCTTTTCTCCTCAT TT- 3’. Each sample was un in duplica e. The me hyla ion s a us o miR- 193b was calcula ed om he elu ion- ac ion signal in ela ion o he o al q- PCR signal. G ow h cu es Fo g ow h analysis wi h he CDK4/6 inhibi o PD0332991 (Selleck Chemicals, Hous on, TX), he 22R 1 cells we e seeded in 24- well pla es in quad uplica e o each inhibi- o concen a ion. The ollowing day, he cells we e imaged (day 0), and hen ea ed wi h 0 nmol/L, 100 nmol/L, 500 nmol/L, o 2000 nmol/L inhibi o . The medium was eplaced e e y o he day wi h esh medium con aining he inhibi o . The cells we e imaged daily wi h an Olympus IX71 mic oscope wi h he OASIS au oma ion con ol sys- em and Su eyo imaging so wa e e sion 5.5.5.26 (Objec i e Imaging L d., Camb idge, UK). Cell g ow h was analyzed by measu ing he cell su ace a ea wi h an in- house mac o and ImageJ so wa e (NIH, Be hesda, MD). The expe imen s we e epea ed h ee imes. S a is ical analysis An unpai ed es was used o calcula e he signi ican di e ence in cell cycle analysis and in luci e ase ac i i y be ween he p e- miR- sc amble and p e- miR- 193b ans- ec ed samples as well as he di e ence in miR- 193b ex- p ession be ween di e en ially s aged p os a ec omy umo s. The pai ed es was used o calcula e he P- alue o he g ow h cu e assays on he las day o he expe i- men . Fishe ’s exac es was used o assess he signi ican di e ence in miR- 193b me hyla ion in clinical samples. Fishe ’s exac , chi- squa e, Mann–Whi ney U, and unpai ed es s we e used o analyze he associa ion be ween clin- icopa hologic a iables. A P- alue <0.05 was conside ed signi ican . Spea man’s co ela ion was used o s udy he co ela ion o miR- 193b and CCND1 exp ession in cell lines and xenog a samples in mic oa ay analysis. Resul s We ha e p e iously shown ha miR- 193b is s ongly hy- pe me hyla ed in 22R 1 cells and mode a ely me hyla ed in VCaP cells bu is no me hyla ed in LAPC- 4, LNCaP, DU145, PC- 3, EP156T, o P EC cells, as de e mined by bisul i e sequencing [8]. In addi ion, we sequenced i e un ea ed and ou CRPC clinical specimens and ound inc eased me hyla ion [8]. He e, we wan ed o con i m hype me hyla ion o miR- 193b in clinical specimens. We ecen ly pe o med genome- wide me hyla ion analysis o me hyla ion by MeDIP- sequencing (unpubl. da a) and ound ha he mos di e en ially me hyla ed genomic egion in p oximi y o miR- 193b is 16:14396975–14397475 (GRCh37), which is loca ed 349 bp ups eam o he miR- 193b gene. He e, we measu ed he me hyla ion o miR- 193b in ha egion in BPH (n = 10), PC (n = 26) and CRPC (n = 9) samples using Me hylMine - qPCR. The samples we e classi ied as me hyla ed i hey showed mo e han 20% me hyla ion. Acco ding o ou analysis, miR- 193b was signi ican ly mo e me hyla ed in cance han in BPH (P < 0.0001; Fig. 1A). Nex , we measu ed he exp ession o miR- 193b in a la ge se (n = 78) o p os a ec omy samples and ound ha he exp ession was lowe (P < 0.05) in s age pT3 umo s han in s age pT2 umo s (Fig. S1). Howe e , we did no ind an associa ion be ween miR- 193b exp ession and he Gleason sco e o p og ession- ee su i al (da a no shown). Because CCND1 is a sugges ed a ge o miR- 193b a ge in hepa ocellula and panc ea ic ca cinoma and melanoma [10, 11, 14], we aimed o de e mine i i is also a a ge in p os a e cance . The exp ession le els o miR- 193b and CCND1 we e ob ained using exp ession mic oa ays o PC cell lines and xenog a samples. The a ay da a showed a nega i e co ela ion be ween miR- 193b and CCND1 exp es- sion in he cell lines (Fig. 1B, = −0.7714) as well as in he xenog a samples (Fig. 1C, = −0.2522). We hen s udied he exp ession o CCND1 mRNA and p o ein le els in 22R 1 cells ansien ly ans ec ed wi h p e- miR- 193b. A ma ked educ ion in he CCND1 exp ession a bo h he mRNA (Fig. 2A) and p o ein (Fig. 2B) le el was de ec ed in cells ans ec ed wi h p e- miR- 193b compa ed o he sc ambled p e- miR ans ec ed cells. Simila o 22R 1 cells, he Cyclin D1 p o ein exp ession was diminished also in 1421 © 2015 The Au ho s. Cance Medicine published by John Wiley & Sons L d. miR-193b Ta ge s Cyclin D1 in PCK. M. Kaukoniemi e al. VCaP cells ansien ly ans ec ed wi h p e- miR- 193b (Fig. S2). In addi ion, a ma ked educ ion was de ec ed in phos- pho- RB p o ein le el in nuclea p o ein ac ions (Fig. 2C). To con i m ha miR- 193b binds o he 3’UTR - egion o he CCND1 gene in p os a e cance cells, we pe o med a luci e ase epo e assay. A pSGG- luci e ase plasmid con aining CCND1 3’UTR a he 3’ end o he luci e ase gene was ans ec ed oge he wi h ei he p e- miR- 193b o p e- miR- sc amble in o 22R 1 cells which exp ess low le els o miR- 193b endogenously. The inhibi ion o luci e ase ac i i y was obse ed when he cells we e ans- ec ed oge he wi h he plasmid and p e- miR- 193b (P < 0.05) bu no wi h p e- miR- sc amble (Fig. 3A), con i ming ha miR- 193b a ge s he 3’UTR o CCND1. To e alua e he signi icance o CCND1 as a miR- 193b a ge , a escue expe imen was pe o med. We used he phospho yla ion o RB as a measu e o cyclin D1 ac i i y. Phospho- RB le els dec eased in cells ans ec ed wi h p e- miR- 193b. When he cells we e co ans ec ed wi h p e- miR- 193b and he pCMV- CCND1 plasmid lacking he 3’UTR o CCND1, no such educ ion in phospho- RB le els was obse ed (Fig. 3B). Simila ly, ansien p e- miR- 193b ans ec ion caused a educ ion in he numbe o cells in S and G2/M phase ac ions o cell cycle, whe eas he e was no di e ence in cells co ans ec ed wi h p e- miR- 193b and CCND1 lacking 3’UTR (Fig. S3). Because cyclin D1 egula es he ac i i y o CDK4/6, we ea ed PC cell lines wi h he CDK4/6 inhibi o PD0332991 a di e en concen a ions and measu ed he e ec on p os a e cance cell g ow h. 22R 1 and VCaP cells, which exp ess low le els o miR- 193b and high le els o CCND1, Figu e 1. miR- 193b is me hyla ed in cance samples and has in e se exp ession pa e n compa ed o CCND1. (A) miR- 193b me hyla ion in clinical samples (BPH, PC, and CRPC) was assessed by Me hylMine - qPCR. miR- 193b and CCND1 exp essions we e s udied by miRNA and mRNA mic oa ay (B) in p os a e cance cell lines and (C) xenog a samples. Spea man co ela ion coe iciencies a e gi en. BPH, benign p os a e hype plasias; CRPC, cas a ion- esis an umo s; PC, p os a e cance . AB C Figu e 2. miR- 193b o e exp ession educes he exp ession o CCND1, he le el o Cyclin D1 p o ein and phospho yla ion le el o e inoblas oma (RB). 22R 1 p os a e cance cells we e ans ec ed wi h p e- miR- 193b o p e- miR- con ol. (A) Exp ession le els o CCND1 mRNA was measu ed using q- RT- PCR. Wes e n blo analysis was used o de ec p o ein exp ession o (B) Cyclin D1 om o al p o eins and (C) phospho yla ion le el o RB p o ein om nuclea p o ein ac ion. Ac in and ib illa in an ibodies we e used as loading con ols. A BC 1422 © 2015 The Au ho s. Cance Medicine published by John Wiley & Sons L d. K. M. Kaukoniemi e al.miR-193b Ta ge s Cyclin D1 in PC showed signi ican g ow h supp ession a a 500 nmol/L concen a ion (P < 0.05; Fig. 4A, B), whe eas he a ious concen a ions o he inhibi o had li le e ec on he g ow h o DU145 (Fig. 4C), LNCaP LAPC- 4 o PC- 3 cells, which exp ess high le els o miR- 193b and low le els o CCND1 (Fig. S4A–C). Figu e S4D shows ha he ehicle, dime hylsul oxide (DMSO), has no e ec on he cell g ow h o 22R 1 cells a 0.4% concen a ion. To s udy he exp ession o cyclin D1 in clinical samples, we pe o med an immunohis ochemical analysis on TMAs. The analysis showed ha he CRPC samples exp essed highe le els o cyclin D1 compa ed o PC samples (P = 0.0237; Table 1). In p os a ec omy samples, he cyclin D1 s aining in ensi y was no associa ed wi h Gleason sco e, pT s age, o diagnos ic PSA le els. Howe e , he e was s ong posi i e associa ion be ween cyclin D1 and he p oli e a ion ma ke Ki- 67 (P < 0.0001) in he p os- a ec omy coho . CRPC cells exp essing high le els o cyclin D1 had inc eased Ki- 67 alues, al hough he as- socia ion was no s a is ically signi ican . Discussion He e, we con i med he hype me hyla ion o miR- 193b gene in p os a e cance . The hype me hyla ion o miR- 193b leads o educed exp ession, as we ha e p e iously shown [8]. Because miRNAs egula e he exp ession o p o ein- coding genes, he key ques ion is which p o eins a e a ge ed by miR- 193b in p os a e cance . CCND1 is a a ge in hepa ocellula and panc ea ic ca cinoma, as well as in melanoma [10, 11, 14]. We ha e p e iously shown ha he o e exp ession o miR- 193b educes he p oli e a- ion o p os a e cance cells due o a dec eased numbe o cells in S- phase o he cell cycle [8], sugges ing ha Figu e 4. CDK 4/6 inhibi o PD0332991 supp esses he g ow h o (A) 22R 1 and (B) VCaP cells bu no (C) DU145 cells. Cells we e ea ed wi h 0, 100, 500, and 2000 nmol/L concen a ions o he inhibi o and g ow h was ollowed o 5 o 6 days. Each concen a ion was done in quad uplica es and each expe imen was done in iplica es, a e ages om expe imen s ±SEM a e shown. P- alues o g ow h di e ences be ween di e en concen a ions on day 5 o 6 we e calcula ed using pai ed - es , *P- alue <0.05, **P- alue <0.01. A B C Figu e 3. miR- 193b a ge s CCND1 3’UTR. (A) Luci e ase expe imen was pe o med in 22R 1 cells co ans ec ed wi h pSGG- plasmid con aining CCND1 3’UTR, Renilla luci e ase plasmid, and p e- miR- 193b o p e- miR- con ol. Values we e no malized agains Renilla luci e ase ac i i y. The means o ou expe imen s ±SEM a e shown. *P- alue <0.05 (B) Wes e n blo analysis o pRB in 22R 1 cells ans ec ed wi h pCMV- CCND1 plasmid lacking CCND1 3’UTR oge he wi h miR- sc amble o miR- 193b. Fib illa in an ibody was used as loading con ol o nuclea p o eins. A B 1423 © 2015 The Au ho s. Cance Medicine published by John Wiley & Sons L d. miR-193b Ta ge s Cyclin D1 in PCK. M. Kaukoniemi e al. CCND1 could also be a a ge o miR- 193b in p os a e cance . We showed ha he exp essions o miR- 193b and CCND1 a e in e sely co ela ed in p os a e cance cell lines and xenog a s. Subsequen ly, we demons a ed educed mRNA and p o ein exp ession o CCND1 in 22R 1 cells an- sien ly ans ec ed wi h p e- miR- 193b. Using a epo e assay, we con i med ha miR- 193b a ge s he 3’UTR o he CCND1 gene in 22R 1 p os a e cance cells. Conco dan ly, Chen e al. [11]. and Xu e al. [10]. ha e p e iously shown by luci e ase assay ha miR- 193b a ge s he CCND1 3’UTR in Malme- 3M malignan melanoma and HepG2 hepa ocellula ca cinoma cells. In addi ion, we pe o med a escue expe imen , in which we co ans- ec ed miR- 193b in o 22R 1 cells wi h and wi hou he pCMV- CCND1 plasmid lacking he 3’UTR o CCND1. These esul s show ha CCND1 is a bona ide a ge o miR- 193b in p os a e cance cells. To assess he ele ance o cyclin D1 a ge ing by miR- 193b in p os a e cance cells, we s udied he ac i i y o he cyclinD1–RB pa hway in he egula ion o he G1/S ansi ion in he cell cycle. The ans ec ion o 22R 1 cells wi h miR- 193b educed he le el o phospho- RB, in acco dance wi h he known unc ion o cyclin D1 in he egula ion o RB phospho yla ion. In addi ion, he phos- pho yla ed RB p o ein le els we e escued in pCMV- CCND1/miR- 193b cells o hose o he con ol cells. Finally, when p os a e cance cell lines we e ea ed wi h he CDK4/6 inhibi o PD0332991, he cell lines esponded o he ea men acco ding o miR- 193b exp ession/me hyla- ion s a us and cyclin D1 le els. Those wi h low miR- 193b and high cyclin D1 (22R 1, VCaP) esponded o he d ug wi h g ow h inhibi ion, while he o he s did no . These esul s demons a e ha he down egula ion o cyclin D1 by miR- 193b is unc ionally ele an o p os a e cance cell g ow h. To assess he clinical signi icance o cyclin D1, we s ained TMAs and ound ha he exp ession o cyclin D1 is highe in CRPC (n = 69) compa ed o ho mone- naï e PC (n = 198). P e iously, D obnjak e al. [27]. showed in- c eased cyclin D1 s aining in 22 CRPC bone me as ases compa ed o 86 p ima y PC umo s. In ou p os a ec omy samples, he exp ession was s ongly associa ed wi h p o- li e a ion bu no wi h Gleason sco e, pa hological s a us, PSA alue, o age a diagnosis. This is in line wi h p e i- ous s udies showing ha cyclin D1 exp ession is associa ed wi h he p oli e a ion ma ke Ki- 67 bu no wi h o he clinicopa hological a iables [27–31]. Because i is known ha one gene can be a ge ed by se e al miRNAs and ha cyclin D1 is a key egula o o he cell cycle G1/S ansi ion, i is no su p ising ha CCND1 has been sugges ed o be a a ge o mo e han one miRNA in p os a e cance . Bonci e al. epo ed ha miR15- a and 16- 1 in e ac di ec ly wi h he 3’UTR o CCND1 and educe cyclin D1 p o ein exp ession, he eby educing p os a e cell p oli e a ion [32]. Howe e , he unc ion o CCND1 a ge ing by hese h ee miRNAs is al e ed by di e en mechanisms in p os a e cance . miR- 193b is epigene ically egula ed and silenced by me hyla ion [8], whe eas he unc ion o miRs 15- a and 16- 1 is dis- up ed by dele ions in he encoding ch omosomal egion 13q14 [32]. We ha e p e iously epo ed a homozygous dele ion o he miR15- a and 16- 1 locus in p os a e cance , al hough he equency is ela i ely low [33]. In addi ion o CCND1, he e a e se e al o he sugges ed a ge s o miR- 193b, such as YWHAZ, PLAU (aka uPA), and KIT in di e en cance ypes [9–16]. Xie e al. [34]. showed ha he knockdown o cys ic ib osis ansmem- b ane conduc ance egula o (CFTR) led o he supp ession o miR- 193b exp ession. They also showed ha he o ced o e exp ession o miR- 193b comple ely ab oga ed ele a ed u okinase- ype plasminogen ac i a o (uPA) ac i i y a e CFTR- knockdown in PC- 3 cells, sugges ing ha he umo - supp essing e ec o CFTR is media ed h ough he miR- 193b- uPA axis. Howe e , i should be no ed ha miR- 193b exp ession is low in 22R 1 and VCaP cells [8], which do no exp ess uPA [35, 36], sugges ing ha uPA is no he majo a ge o miR- 193b in p os a e cance . In conclusions, we ha e demons a ed ha he o e - exp ession o cyclin D1 in p os a e cance is d i en, a leas pa ly, by he educed exp ession o miR- 193b. The mechanism o miR- 193b supp ession is he Table 1. Associa ion o clinicopa hological a iables wi h Cyclin D1 exp ession. Va iable Cyclin D1 exp ession P Nega i e (0/1) Posi i e (2–3) P os a ec omy specimens, n (%) 74 (37) 124 (63) Locally ecu en CRPCs, n (%)120 (29) 49 (71) 0.0237 P os a ec omy specimens Gleason sco e, n (%)1 <7 24 (33) 46 (37) 7 38 (52) 60 (48) >7 11 (15) 19 (15) 0.8337 pT S age, n (%)2 pT2 54 (75) 88 (72) pT3 18 (25) 35 (28) 0.6216 PSA ng/mL (mean ± SD)320.0 ± 31.5 14.3 ± 11.3 0.9786 Age (mean ± SD)462.6 ± 5.2 63.2 ± 4.9 0.4167 Ki- 67 (mean ± SD) 7.1 ± 7.0 13.6 ± 14.5 <0.0001 Locally ecu en CRPCs, n (%)3 Ki- 67 (mean ± SD) 13.2 ± 9.3 20.7 ± 15.8 0.0950 1Chi- squa e es . 2Fishe ’s exac es . 3Mann–Whi ney U- es . 4Unpai ed es . 1424 © 2015 The Au ho s. Cance Medicine published by John Wiley & Sons L d. K. M. Kaukoniemi e al.miR-193b Ta ge s Cyclin D1 in PC hype me hyla ion o he DMR ups eam o he miR- 193b gene. Addi ional s udies a e wa an ed o ansla e hese indings o clinical bene i s. Fo example, he es o a ion o miR- 193b exp ession could heo e ically educe he p oli e a ion o p os a e cance cells. Al e na i ely, he loss o miR- 193b exp ession could indica e he sensi i i y o p os a e cance cells o cyclin D1 inhibi ion. Acknowledgmen s We hank P. Ma ikainen and M. Vakku i o he skill ul echnical assis ance. The esea ch leading o hese esul s was unded by Tampe e Uni e si y Doc o al P og amme in Biomedicine and Bio echnology. In addi ion, g an sup- po has been ecei ed om he Sig id Juselius Founda ion, he Finnish Cul u al Founda ion, Pi kanmaa Regional und, he Academy o Finland, he Cance Socie y o Finland, he Medical Resea ch Fund o Tampe e Uni e si y Hospi al, and he Eu opean Communi y’s Se en h F amewo k P og amme P ospeR (FP7/2007- 2013) unde g an ag ee- men no. HEALTH- F2- 2007- 201438. Con lic o In e es None decla ed. Re e ences 1. Jemal, A., F. B ay, M. M. Cen e , J. Fe lay, E. Wa d, and D. Fo man. 2011. Global cance s a is ics. CA Cance J. Clin. 61:69–90. 2. Knudsen, K. E., and W. K. Kelly. 2011. Ou sma ing and ogen ecep o : c ea i e app oaches o a ge ing abe an and ogen signaling in ad anced p os a e cance . Expe . Re . Endoc inol. Me ab. 6:483–493. 3. Sch ode , F. H., J. Hugosson, M. J. Roobol, T. L. Tammela, M. Zappa, V. Nelen, e al. 2014. Sc eening and p os a e cance mo ali y: esul s o he Eu opean Randomised S udy o Sc eening o P os a e Cance (ERSPC) a 13 yea s o ollow- up. Lance 384:2027–2035. 4. Ca o, J. W., A. Alca az, A. S. Bja ell, R. De Ve e Whi e, C. P. E ans, S. Fussel, e al. 2011. Mic oRNA in p os a e, bladde , and kidney cance : a sys ema ic e iew. Eu . U ol. 59:671–681. 5. Lee, C., and V. Amb os. 2001. An ex ensi e class o small RNAs in Caeno habdi is elegans. Science 294:862–864. 6. Lagos-Quin ana, M., R. Rauhu , W. Lendeckel, and T. Tuschl. 2001. Iden i ica ion o no el genes coding o small exp essed RNAs. Science 294:853–858. 7. Di Le a, G., and C. M. C oce. 2010. Roles o small RNAs in umo o ma ion. T ends Mol. Med. 16:257–267. 8. Rauhala, H. E., S. E. Jala a, J. Iso alo, H. B acken, S. Lehmus aa a, T. L. Tammela, e al. 2010. miR- 193b is an epigene ically egula ed pu a i e umo supp esso in p os a e cance . In . J. Cance 127:1363–1372. 9. Li, X. F., P. J. Yan, and Z. M. Shao. 2009. Down egula ion o miR- 193b con ibu es o enhance u okinase- ype plasminogen ac i a o (uPA) exp ession and umo p og ession and in asion in human b eas cance . Oncogene 28:3937–3948. 10. Xu, C., S. Liu, H. Fu, S. Li, Y. Tie, J. Zhu, e al. 2010. Mic oRNA- 193b egula es p oli e a ion, mig a ion and in asion in human hepa ocellula ca cinoma cells. Eu . J. Cance 46:2828–2836. 11. Chen, J., H. E. Feilo e , G. C. Pa e, X. Zhang, J. G. Pembe on, C. Ga ady, e al. 2010. Mic oRNA- 193b ep esses cell p oli e a ion and egula es cyclin D1 in melanoma. Am. J. Pa hol. 176:2520–2529. 12. Chen, J., X. Zhang, C. Len z, M. Abi-Daoud, G. C. Pa e, X. Yang, e al. 2011. miR- 193b egula es Mcl- 1 in melanoma. Am. J. Pa hol. 179:2162–2168. 13. Gao, X. N., J. Lin, L. Gao, Y. H. Li, L. L. Wang, and L. Yu. 2011. Mic oRNA- 193b egula es c- Ki p o o- oncogene and ep esses cell p oli e a ion in acu e myeloid leukemia. Leuk. Res. 35:1226–1232. 14. Ikeda, Y., E. Tanji, N. Makino, S. Kawa a, and T. Fu ukawa. 2012. Mic oRNAs associa ed wi h mi ogen- ac i a ed p o ein kinase in human panc ea ic cance . Mol. Cance Res. 10:259–269. 15. Lei onen, S. K., A. Rokka, P. Os ling, P. Kohonen, G. L. Co hals, O. Kallioniemi, e al. 2011. Iden i ica ion o miR- 193b a ge s in b eas cance cells and sys ems biological analysis o hei unc ional impac . Mol. Cell P o eomics 10:M110.005322. 16. Hu, H., S. Li, J. Liu, and B. Ni. 2012. Mic oRNA- 193b modula es p oli e a ion, mig a ion, and in asion o non- small cell lung cance cells. Ac a Biochim. Biophys. Sin. (Shanghai) 44:424–430. 17. Musg o e, E. A., C. E. Caldon, J. Ba aclough, A. S one, and R. L. Su he land. 2011. Cyclin D as a he apeu ic a ge in cance . Na . Re . Cance 11:558–572. 18. Knudsen, K. E., W. K. Ca enee, and K. C. A den. 1999. D- ype cyclins complex wi h he and ogen ecep o and inhibi i s ansc ip ional ansac i a ion abili y. Cance Res. 59:2297–2301. 19. Reu ens, A. T., M. Fu, C. Wang, C. Albanese, M. J. McPhaul, Z. Sun, e al. 2001. Cyclin D1 binds he and ogen ecep o and egula es ho mone- dependen signaling in a p300/CBP- associa ed ac o (P/ CAF)- dependen manne . Mol. Endoc inol. 15:797–811. 20. Pe e, C. E., Y. B. We he ill, M. Danielsen, and K. E. Knudsen. 2002. Cyclin D1: mechanism and consequence o and ogen ecep o co- ep esso ac i i y. J. Biol. Chem. 277:2207–2215. 1425 © 2015 The Au ho s. Cance Medicine published by John Wiley & Sons L d. miR-193b Ta ge s Cyclin D1 in PCK. M. Kaukoniemi e al. 21. Jala a, S. E., K. P. Po kka, H. E. Rauhala, J. Iso alo, T. L. Tammela, and T. Visako pi. 2009. TCEB1 p omo es in asion o p os a e cance cells. In . J. Cance 124:95–102. 22. U banucci, A., B. Sahu, J. Seppala, A. La jo, L. M. La onen, K. K. Wal e ing, e al. 2012. O e exp ession o and ogen ecep o enhances he binding o he ecep o o he ch oma in in p os a e cance . Oncogene 31:2153–2163. 23. Leinonen, K. A., T. T. Tolonen, H. B acken, U. H. S enman, T. L. Tammela, O. R. Sa amaki, e al. 2010. Associa ion o SPINK1 exp ession and TMPRSS2: ERG usion wi h p ognosis in endoc ine- ea ed p os a e cance . Clin. Cance Res. 16:2845–2851. 24. Tuominen, V. J., and J. Isola. 2009. The applica ion o JPEG2000 in i ual mic oscopy. J. Digi . Imaging 22:250–258. 25. Tuominen, V. J., and J. Isola. 2010. Linking whole- slide mic oscope images wi h DICOM by using JPEG2000 in e ac i e p o ocol. J. Digi . Imaging 23:454–462. 26. Tuominen, V. J., S. Ruo ois enmaki, A. Vii anen, M. Jumppanen, and J. Isola. 2010. ImmunoRa io: a publicly a ailable web applica ion o quan i a i e image analysis o es ogen ecep o (ER), p oges e one ecep o (PR), and Ki- 67. B eas Cance Res. 12:R56. 27. D obnjak, M., I. Osman, H. I. Sche , M. Fazza i, and C. Co don-Ca do. 2000. O e exp ession o cyclin D1 is associa ed wi h me as a ic p os a e cance o bone. Clin. Cance Res. 6:1891–1895. 28. Mu phy, C., M. McGu k, J. Pe ig ew, A. San inelli, R. Mazzucchelli, P. G. Johns on, e al. 2005. Nonapical and cy oplasmic exp ession o in e leukin- 8, CXCR1, and CXCR2 co ela es wi h cell p oli e a ion and mic o essel densi y in p os a e cance . Clin. Cance Res. 11:4117–4127. 29. Aal omaa, S., M. Eskelinen, and P. Lipponen. 1999. Exp ession o cyclin A and D p o eins in p os a e cance and hei ela ion o clinopa hological a iables and pa ien su i al. P os a e 38:175–182. 30. Kallaku y, B. V., C. E. Sheehan, R. A. Amb os, H. A. Fishe , R. P. J Kau man, and J. S. Ross. 1997. The p ognos ic signi icance o p34cdc2 and cyclin D1 p o ein exp ession in p os a e adenoca cinoma. Cance 80:753–763. 31. Coms ock, C. E., M. P. Re elo, C. R. Bunche , and K. E. Knudsen. 2007. Impac o di e en ial cyclin D1 exp ession and localisa ion in p os a e cance . B . J. Cance 96:970–979. 32. Bonci, D., V. Coppola, M. Musumeci, A. Adda io, R. Giu ida, L. Memeo, e al. 2008. The miR- 15a- miR- 16- 1 clus e con ols p os a e cance by a ge ing mul iple oncogenic ac i i ies. Na . Med. 14:1271–1277. 33. Po kka, K. P., E. L. Ogg, O. R. Sa amaki, R. L. Vessella, H. Pukkila, H. Lahdesmaki, e al. 2011. The miR- 15a- miR- 16- 1 locus is homozygously dele ed in a subse o p os a e cance s. Genes Ch omosom. Cance 50:499–509. 34. Xie, C., X. H. Jiang, J. T. Zhang, T. T. Sun, J. D. Dong, A. J. Sande s, e al. 2013. CFTR supp esses umo p og ession h ough miR- 193b a ge ing u okinase plasminogen ac i a o (uPA) in p os a e cance . Oncogene 32:2282–2291, 2291.e1–7. 35. Helenius, M. A., K. J. Sa inainen, G. S. Bo a, and T. Visako pi. 2006. Ampli ica ion o he u okinase gene and he sensi i i y o p os a e cance cells o u okinase inhibi o s. BJU In . 97:404–409. 36. P ensne , J. R., M. K. Iye , O. A. Balbin, S. M. Dhanaseka an, Q. Cao, J. C. B enne , e al. 2011. T ansc ip ome sequencing ac oss a p os a e cance coho iden i ies PCAT- 1, an unanno a ed lincRNA implica ed in disease p og ession. Na . Bio echnol. 29:742–749. Suppo ing In o ma ion Addi ional suppo ing in o ma ion may be ound in he online e sion o his a icle: Table S1. Desc ip ion o clinical samples used o s udy miR- 193b exp ession. Table S2. Desc ip ion o clinical samples used in immunohis ochemis y. Figu e S1. miR- 193b exp ession in clinical samples. Figu e S2. miR- 193b o e exp ession educes he exp es- sion o CCND1 in p o ein le el in VCaP cells. Figu e S3. CCND1 lacking 3’ UTR is able o escue cell cycle e ec s o miR- 193b. Figu e S4. The e ec o CDK 4/6 inhibi o PD0332991 and dissol en dime hylsul oxide (DMSO) on g ow h o (A) LAPC- 4, (B) LNCaP, (C) PC- 3, and (D) 22R 1 cells.