Activation of telomerase rna gene promoter activity by NF-Y, Sp1, and the retinoblastoma protein and repression by Sp3

Ac i a ion o Telome ase RNA Gene P omo e Ac i i y by NF-Y, Sp1, and he Re inoblas oma P o ein and Rep ession by Sp3 1 Jiang-Qin Zhao*, Rosalind M. Glasspool*, S acey F. Hoa e*, Alan Bilsland*, Is an Sza ma i y and W. Nicol Kei h* *CRC Depa men o Medical Oncology, Uni e si y o Glasgow, CRC Bea son Labo a o ies, Ga scube Es a e, Swi chback Road, Bea sden, Glasgow G61 1BD, UK and y Dep o Biochemis y and Molecula Biology, Uni e si y Medical School o Deb ecen, Hunga y Abs ac Exp ession o he human elome ase RNA componen gene, hTERC is essen ial o elome ase ac i i y. The hTERC gene is exp essed du ing emb yogenesis and hen down egula ed du ing no mal de elopmen , lea ing mos adul soma ic cells de oid o hTERC exp ession. Du ing oncogenesis, howe e , hTERC is e-exp essed consequen ly con ibu ing o he un es ic ed p oli e a- i e capaci y o many human cance s. Thus he iden i i- ca ion o he molecula basis o he egula ion o he elome ase RNA componen gene in no mal cells and i s de egula ion in cance cells is o immedia e in e es . We ha e p e iously cloned he hTERC p omo e and in his s udy ha e iden i ied se e al ansc ip ion ac o s ha modula e he exp ession o hTERC. We demons a e ha NF-Y binding o he CCAAT egion o he hTERC p omo e is essen ial o p omo e ac i i y. Sp1 and he e inoblas oma p o ein (pRb) a e ac i a o s o he hTERC p omo e and Sp3 is a po en ep esso . These ac o s appea o ac in a species-speci ic manne . Whe eas Sp1 and Sp3 ac on he human, bo ine, and mouse TERC p omo e s, pRb ac i a es only he human and bo ine p omo e , and NF-Y is only essen ial o he human TERC gene. Neoplasia (2000) 2, 531±539. Keywo ds: elome ase, NF - Y, hTERC, e inoblas oma, hTR. In oduc ion Telome es a e nucleop o ein complexes a he end o linea ch omosomes and a e conside ed essen ial o main aining ch omosomal in eg i y [1]. In he e en o c i ical elome e- sequence loss du ing cell di ision, a cell-cycle a es known as eplica i e senescence is induced [1,2]. Telome ase is a ibonucleop o ein e e se ansc ip ase ha has he abili y o syn hesize elome ic sequence and hus coun e balance elome e a i ion. Human elome ase ac i i y in i o can be econs i u ed by he RNA subuni , hTERC, and he ca aly ic p o ein componen encoded by he hTERT gene, al hough i is likely hese co e subuni s will be augmen ed by addi ional ac o s o unc ion op imally [1,2]. hTERC gene exp ession is egula ed du ing no mal human de elopmen , wi h he high le els o exp ession ound du ing emb yogenesis dec easing as issue di e en ia ion occu s. By he 10 h pos na al week, he adul pa e n o hTERC exp ession is e iden wi h only p ima y spe ma ocy es e aining high le els o exp ession [3]. Al hough hTERC exp ession is absen om mos adul soma ic cells, low hTERC exp ession can be de ec ed in some cells such as he egene a i e egions o epi helium and ac i a ed lymphocy es [3±5]. Howe e , du ing ca cino- genesis he ep ession o hTERC in he cance cell is los o de egula ed, hus con ibu ing o he almos uni e sal eac i a ion o elome ase in cance cells [2,6]. Recen s udies ha e indica ed ha ansc ip ional egula ion may ha e a ole o play in he con ol o hTERC gene exp ession [7,8]. Thus he iden i ica ion o he molecula basis o hTERC gene egula ion in no mal cells and i s de egula ion in cance cells is o immedia e in e es . I is also an eme ging heme in ansc ip ional con ol ha egula ion esul s om he unc ional in e ac ions be ween g oups o ansc ip ion ac o s. Thus, he aim o he p esen s udy was o iden i y mul iple ansc ip ion ac o s ha egula e he hTERC gene p omo e . Ma e ials and Me hods Si e-Di ec ed Mu agenesis and Cons uc s The human p omo e luci e ase cons uc s, hP om876, hP om176, hP om120, and mouse TERC p omo e con- s uc mP om628 we e made by inse ing PCR p oduc s in o pGL-3 Basic (P omega Sou hamp on, U.K.). De ails o he p ime s used can be ob ained om he au ho s. O ien a ion and sequence we e con i med by sequencing. The (hP om176h10m1) cons uc was gene a ed using he si e-di ec ed mu agenesis ki (QuikChange, S a agene Camb idge, UK) o in oduce a ou -bp mu a ion a 58/ 55. The p ime used o cons uc ion o his si e- eplaced mu an is shown in Figu e 1b(h10m1). I was subcloned Neoplasia .Vol. 2, No. 6, 2000, pp. 531 ± 539 www.na u e.com/neo 531 Abb e ia ions: EMSA, elec opho e ic mobili y shi assay Add ess all co espondence o: W. Nicol Kei h, CRC Depa men o Medical Oncology, Uni e si y o Glasgow, CRC Bea son Labo a o ies, Alexande S one Building, Ga scube Es a e, Swi chback Road, Bea sden, Glasgow G61 1BD, UK. E-mail: [email p o ec ed] 1 This wo k was suppo ed by he Cance Resea ch Campaign ( UK ) and Glasgow Uni e si y. Recei ed 1 Sep embe 2000; Accep ed 22 Sep embe 2000. Copy igh #2000 Na u e Ame ica, Inc. All igh s ese ed 1522-8002/00/$15.00 RESEARCH ARTICLE in o he pGL3 luci e ase epo e ec o and sequenced o con i m he mu a ion. P epa a ion o Nuclea P o ein Ex ac s HeLa cells we e lysed in TMS (5 mM T is±HCl pH 7.5, 2.5 mM MgCl 2 , 125 mM suc ose plus 0.25% T i on X-100), and nuclei we e ha es ed by cen i uga ion. Nuclei we e esuspended in 5 ml o TMS and 0.1 olumes o 4 m NaCl. A e cen i uga ion, ammonium sul a e was added and he p ecipi a e pelle ed and edissol ed in E50 bu e (50 mM ammonium sulpha e, 20 mM KOH±Hepes, pH 7.9, 5 mM MgCl 2 , 0.1 mM EDTA, 0.1% ( / ) B ij 35, 20% ( / ) glyce ol, 1 mM DTT). Elec opho esis Mobili y Shi Assay Elec opho esis mobili y shi assays (EMSA) we e pe o med by using he EMSA ki (P omega). HeLa nuclea p o eins we e incuba ed in 15 l o eac ion con aining 4% glyce ol, 1 mM MgCl 2, 0.5 mM di hio h ei ol (DTT), 0.5 mM EDTA, 50 mM NaCl, 10 mM T is±HCl, pH 7.5, and 2.0 g poly(dI±dC) wi h o wi hou 100- old mola excess o unlabeled DNA compe i o s on ice o 10 min, ollowed by addi ion o he adiolabeled p obe. Fo supe shi assays, an ibodies agains Sp1, Sp3, Ap-2, E s2, c/EBP,NF-1, NF-YC (San a C uz Bio echnology San a C uz, CA), NF- YA, and NF-YB, we e added o he eac ion mix u e 25 minu es be o e he addi ion o he p obe. All DNA±p o ein complexes we e esol ed by elec opho esis on 5% na i e polyac ylamide. The c/EBP, CBF, and NF-1 sequences used in EMSA a e om San a C uz Bio echnology: sc-2525, sc- 2591, and sc-2553. The NF-Y sequence is om Geneka Bio echnology Inc. (Mon eal, Canada). T ans ec ion Assays T ans ec ions we e ca ied ou using Supe ec ans ec- ion eagen (Qiagen Wes Sussex, UK) as p e iously desc ibed, [8]. To s anda dize he amoun o hTERC plasmid DNA in ans ec ion assays, pu i ied plasmid DNA was i s quan i ied by UV spec opho ome y. Then 1.0 g o plasmid DNA was diges ed wi h XhoI/HindIII and agmen s esol ed on a 6% PAGE gel o de ec he 176-bp p omo e agmen . The amoun o each DNA sample was analyzed by Phospho Image wi h ImageQuan so wa e analysis (Mole- cula Dynamics Bio-Rad, Hemel Hemps ead, UK). Fo co ans ec ions, 3 g o hP om176 o 1 g o hP om876 epo e plasmids was co ans ec ed wi h 0.01 o 3 g o pNF- YAm29, pNF-Y13, pCMV-Sp1, pCMV-Sp3, pSV40-Rb exp ession ec o s. The hP om176 cons uc was used h oughou his s udy bu he hP om867 cons uc was also used o ex end he analysis o he minimal human p omo e cons uc and es whe he he maximal p omo e esponded o modi ie genes in a simila way o he minimal p omo e . Figu e 1. ( a ) hTERC p omo e sequence and po en ial ansc ip ion ac o ecogni ion si es. A o al o 176 bp o he human p omo e sequence is shown and po en ial egula o y mo i s, iden i ied by compu e analysis, a e unde lined on he sequence. The ansc ip ional s a si e is ma ked as ``+ 1.'' Fou po en ial Sp1 si es a e shown and e med Sp1.1 o Sp1.4. Sp1.4 con ains wo o e lapping Sp1 consensus si es. The names o he oligonucleo ides used in elec opho esis mobili y shi assays ( EMSA ) a e indica ed unde hei espec i e si es. ( b ) Oligonucleo ides used in EMSA and mu agenesis s udies o he hTERC p omo e . The sequence o he wild - ype oligonucleo ides co e ing po en ial ansc ip ion ac o ecogni ion si es a e shown. Oligonucleo ide h9 co e s Sp1.1 and he adjacen TATA - box, h10 co e s he CCAAT box, h4 co e s Sp1.2 and h11 co e s he closely opposed Sp1.3 and Sp1.4 si es. Mu a ions, h10m1 and h10m2, we e in oduced in o he wild- ype oligonucleo ide h10 and used in EMSA and in he cons uc ion o he mu an epo e cons uc ( hP om176h10m1 ) . 532 hTERC P omo e Regula ion Zhao e al. Neoplasia .Vol. 2, No. 6, 2000 Bo h human cons uc s esponded in co ans ec ion expe i- men s in iden ical ways. The G ea EscAPe SEAP Sys em (Clon ech Basing- s oke, UK) was used o con ol o ans ec ion e iciency in expe imen s wi h he mu an p omo e cons uc s, he pSV40-Rb ec o , he pNF-Y13 and pNF-YAm29 exp es- sion ec o s. In keeping wi h o he s we ound ha he Sp1 and Sp3 exp ession ec o s modula ed he ac i i y o he SV40 p omo e used in he pSEAP2-Con ol ec o [9,10]. Fo hese expe imen s semiquan i a i e PCR o he ans- ec ed plasmid was ca ied ou using luci e ase-speci ic p ime s and p oduc s adjus ed ela i e o inpu genomic DNA. The means and s anda d de ia ions o duplica e samples om ep esen a i e ans ec ions a e shown and all expe i- men s we e ca ied ou a leas h ee imes in duplica e. The expe imen s shown in Figu e 4bused e inoblas oma p o ein (pRb) a ian s de eloped by Selle s e al. [16]. Cloning o bTERC A PCR p oduc encompassing he 320 bp o he bTERC p omo e was ampli ied using he p ime s bTR 1 GCGCTCGAGCCGTACCTGGCTTTTAAGAG and bTR 1 CGAAGCTTAGATGAGAAATGGCTGCCAC and bo ine genomic DNA as a empla e (Clon ech). P ime s we e designed om he bTERC gene sequence [11]. This was hen cloned in o luci e ase epo e cons uc pGL3 Basic (P omega). O ien a ion and sequence o he inse was checked by sequencing. The bTERC p omo e sequence has been submi ed o GenBank, accession numbe AF176663. Resul s We ha e p e iously iden i ied he p oximal egion o he hTERC p omo e necessa y o di ec hTERC ansc ip ion [8]. DNase I oo p in ing iden i ied a 176-bp egion wi hin his which con ained ou p o ec ed egions (da a no shown). Compu e analysis e ealed ha hese co e- sponded o ou pu a i e Sp1 binding si es and a CCAAT box (Figu e 1: h4, h9, h11, h10). In he p esen s udy a combina ion o gel shi assays, mu a ion analysis, and ans ec ion assays ha e been used o iden i y key egula- o s o hTERC gene exp ession. The NF-Y T ansc ip ional Complex In e ac s wi h he hTERC CCAAT Box and This In e ac ion Is Cen al o hTERC T ansc ip ional E iciency EMSA was used o iden i y whe he cellula p o eins in e ac wi h he CCAAT box ha is ound ups eam o he ansc ip ional s a si e. P o ein ex ac om he HeLa cell line was used wi h a adiolabeled oligonucleo ide co e- sponding o he wild- ype hTERC CCAAT box sequence (h10). A p o ein±DNA complex was iden i ied (Figu e 2a, lane 1) and speci ici y o binding con i med by compe i ion s udies. Compe i ion by he wild- ype oligonucleo ide i sel and by an oligonucleo ide wi h a mu a ion ou side he CCAAT box (h10m2) abolished he complex (Figu e 2a, lanes 2 and 4). In con as he oligonucleo ide wi h a mu a ion wi hin he CCAAT box (h10m1) was unable o compe e o p o ein binding (Figu e 2a,lane 3). To asce ain he likely componen s o he hTERC CCAAT-binding complex, h ee candida e CCAAT-binding complexes we e in es iga ed; NF-Y (also known as CBF and CP1), c/EBP, and NF-1. Two consensus oligonucleo- ides o NF-Y ( e med NF-Y and CBF) we e able o compe e o hTERC CCAAT binding (Figu e 2a,lanes 5 and 6), whe eas consensus oligonucleo ides o c/EBP and NF- 1 we e no (lanes 7 and 8). An ibody supe shi analysis e ealed ha he hTERC CCAAT-binding complex con- ained wo subuni s o he NF-Y p o ein complex, NF-YA and NF-YB. Nuclea ex ac s we e p eincuba ed wi h an ibodies speci ic o he h ee componen s o he NF-Y complex, NF-YA, NF-YB and NF-YC, and wi h an ibodies speci ic o o he CCAAT-binding complexes c/EBP and NF- 1. Dis up ion o he CCAAT complex and supe shi was only seen wi h an ibodies o he NF-Y complex componen s NF- YA and NF-YB (lanes 10 and 11). An ibodies o he NF-YC subuni o NF-Y did no dis up he complex (lane 12). This has been obse ed by o he s and is possibly due o inaccessibili y o he C subuni o he an ibody on complex o ma ion [12]. An ibodies o NF-1 and c/EBP had no e ec on complex o ma ion (lanes 13 and 14). To e alua e he unc ional signi icance o NF-Y complex o ma ion on he hTERC p omo e , we used a dominan nega i e mu an o NF-YA (NF-YAm29) in ans ec ion expe imen s [13]. As shown in Figu e 2b, co ans ec ion o NF-YAm29 speci ically ep esses hTERC p omo e ac i i y i e- o se en- old in a dose-dependen manne , whe eas co ans ec ion o he wild- ype NF-YA subuni had no e ec on p omo e ac i i y. These esul s con i m ha NF-Y is he majo ac o ha ansac i a es he hTERC p omo e h ough he CCAAT box. The unc ional signi icance o he CCAAT box and i s in e ac ion wi h NF-Y was also con i med by in oducing a mu a ion in o he basal p omo e egion. This was cloned in o a luci e ase epo e cons uc (hP om176- h10m1). Simila ly a p omo e cons uc wi h 56-bp dele ion o he 5 0 end, including he CCAAT box egion, was gene a ed (hP om120). P omo e ac i i ies o he mu an and dele ed CCAAT box cons uc s we e assayed by ans ec ion in o he bladde ca cinoma cell line 5637 and compa ed o he wild- ype p omo e ac i i y. Mu a ion o dele ion o he CCAAT box sequence elemen comple ely abolishes hTERC p omo e ac i i y (Figu e 2c), sugges ing ha his elemen is he si e o ac ion o an ac i a o o hTERC p omo e ac i i y and cen al o hTERC ansc ip- ional e iciency. The Sp1 and Sp3 T ansc ip ion Fac o s In e ac wi h Mul iple Si es wi hin he hTERC P omo e and Can Modula e P omo e Ac i i y To iden i y cellula p o eins in e ac ing wi h he ou pu a i e Sp1 si es in he hTERC p omo e , EMSA we e pe o med wi h h ee oligonucleo ides (Figu e 1b: h9, h4, and h11) spanning he ou si es (Figu e 1aand b). P o ein complexes cha ac e is ic o binding o Sp1 and Sp3 [14] we e seen wi h all h ee oligonucleo ides (Figu e 3a,lanes 1, Neoplasia .Vol. 2, No. 6, 2000 hTERC P omo e Regula ion Zhao e al. 533 Figu e 2. ( a ) Iden i ica ion o nuclea p o eins binding o he CCAAT box. HeLa nuclea ex ac was mixed wi h adiolabeled oligonucleo ide p obe and analyzedby EMSA. Speci ic DNA ±p o ein complexes a e indica ed by he a ow on he le . Complexes supe shi ed by p eincuba ion wi h an ibodies a e shown on he igh by a ows ma ked ``ss.'' The oligonucleo ide h10 was used as a p obe and he oligonucleo ides used as compe i o s a e indica ed a he op o lanes 2 o 8. The an ibodies used in supe shi s a e indica ed abo e lanes 9 o 14. ( b ) Inhibi ion o hTERC p omo e ac i i y by a dominan - nega i e mu an o NF - YA e med NF - YAm29. 5637 cells we e ans ec ed wi h ixed amoun o he hTERC - luci e ase plasmid hP om176 and inc easing concen a ions o he NF - YAm29 dominan - nega i e ec o ( black columns ) o wild - ype NF - YA ec o ( ligh g ey columns ) . Fo each ans ec ion he mean and s anda d de ia ion o duplica e samples is shown. ( c ) Func ional analysis o he CCAAT box. P omo e ac i i ies o a mu an CCAAT box cons uc ( hP om176h10m1 ) and sho e p omo e cons uc wi h dele ion o he CCAAT box ( hP om120 ) we e assayed by ans ec ion in o he bladde ca cinoma cell line 5637 and compa ed o he wild - ype p omo e ac i i y ( hP om176 ) . P omo e ac i i ies o he cons uc s a e shown as pe cen age luci e ase ac i i y o he wild - ype p omo e alone. The pSEAP2 - Con ol ec o was used as an in e nal con ol o ans ec ion e iciency. 534 hTERC P omo e Regula ion Zhao e al. Neoplasia .Vol. 2, No. 6, 2000 5, and 10). To con i m he iden i y o he p o eins binding he Sp1 si es, nuclea ex ac s we e p eincuba ed wi h an i- bodies o he ansc ip ional egula o s Sp1 and Sp3 and wi h con ol an ibodies speci ic o E s2 and Ap2. Speci ic dis up ion o complex o ma ion was only seen when p eincuba ion occu ed wi h an ibodies agains Sp1 and Sp3 (Figu e 3a,lanes 2, 3, 6, 7, 11, 12, and 13). This sugges s ha he Sp amily o ansc ip ion ac o s is in ol ed in he egula ion o hTERC gene exp ession. To es his hypo hesis, exp ession ec o s o Sp1 and Sp3 we e co ans ec ed in o 5637 cells wi h he hTERC-luci e ase cons uc . Figu e 3bdemons a es dose-dependen ac i a- ion o he p omo e by Sp1 and dose-dependen ep ession by Sp3. The Re inoblas oma Gene P oduc Ac i a es he hTERC P omo e The pRb is conside ed o ha e impo an oles in no mal ansc ip ional egula ion and in he p og ession o a no mal mo al cell o an immo al cance cell [2,15]. The abili y o pRb o modula e he hTERC p omo e was he e o e assessed. The pRb/p53 null bladde cell line, 5637, was co ans ec ed wi h a ixed amoun o he hTERC-luci e ase plasmid and inc easing amoun s o a pRb exp ession ec o . Figu e 4ashows ha exp ession o he pRb is a dose- dependen ac i a o o he hTERC p omo e . To in es iga e he ansc ip ional ac i a ion o hTERC by pRb in mo e de ail, a se ies o pRb a ian s de eloped by Selle s e al. [16] was used. Va ian s 663 and 661W a e unable o bind E2F and he e o e do no cause cell-cycle a es , ye hey e ain he abili y o ac i a e gene ansc ip ion. Va ian 657 is a null mu an o pRb, i is unable o bind E2F and canno ac i a e gene ansc ip ion [16]. As can be seen om Figu e 4b, ansc ip ional ac i a ion o he hTERC p omo e by pRb is no dependen on E2F binding as i is obse ed wi h a ian s ha e ain ansc ip ional ac i a ion capabili ies bu which a e unable o bind E2F (663 and 661W). The null mu an o pRb (657), which is unable o bind E2F o o ac i a e gene ansc ip ion in he assays used by Selle s e al., is simila ly unable o ac i a e he hTERC p omo e . The speci ici y o pRb o ac i a e he hTERC p omo e was es ed by compa ing he human and mouse elome ase RNA gene p omo e s, which, a he sequence le el, show no signi ican homology [8]. As can be seen om Figu e 4a, he mouse p omo e does no espond o he pRb hus demons a ing a unc ional di e ence in he egula ion o he Figu e 3. ( a ) Iden i ica ion o nuclea p o eins binding o he Sp1 si es. HeLa nuclea ex ac was mixed wi h adiolabeled oligonucleo ide p obe and analyzedby EMSA. The p o eins bound o he p obe we e iden i ied by p eincuba ion o nuclea ex ac s wi h an ibodies speci ic o Sp1, Sp3, o con ol an ibodies speci ic o E s2 and Ap2. Complexes speci ic o Sp1 and Sp3 a e indica ed o he le o he panel. The an ibodies used a e indica ed a he op o he igu e. ( b ) The Sp1 ansc ip ion ac o up egula es and Sp3 ansc ip ion ac o down egula es he hTERC p omo e . The wild - ype hTERC - luci e ase plasmid was co ans ec ed wi h inc easing amoun s o ei he an Sp1 o Sp3 exp ession ec o ( 0.25 g o 2.0 g ) in o he 5637 cell line. The mean luci e ase ac i i y no malized o p o ein and ans ec ion e iciency is shown wi h he s anda d de ia ion o duplica e samples. Neoplasia .Vol. 2, No. 6, 2000 hTERC P omo e Regula ion Zhao e al. 535 mouse and human elome ase RNA genes. In addi ion, he speci ici y o hTERC ac i a ion by pRb was con i med by co ans ec ion o he hTERC p omo e cons uc con aining he mu a ion in he CCAAT box (hP om120) wi h pRb. As shown in Figu e 2C, mu a ion o he CCAAT box abolishes p omo e ac i i y. The exp ession o pRb is unable o ac i a e hTERC p omo e ac i i y in he p esence o he mu a ed CCAAT box, hus con i ming he unc ional speci- ici y o he pRb ac i a ion on he wild- ype p omo e (da a no shown). Func ional Di e gence in he T ansc ip ional Regula ion o TERC Genes be ween Species Despi e conse a ion in elome ase enzyme unc ion ac oss species, he e may be undamen al di e ences in elome ase egula ion be ween mouse and human [8,17]. To gain a clea e insigh in o he egula ion o elome ase RNA genes, he esponse o he human, mouse, and bo ine elome ase RNA gene p omo e s o he ansc ip ional egula o s pRb, NF-Y, Sp1, and Sp3 we e compa ed in ansien ans ec ion expe imen s. The bo ine elome ase RNA gene p omo e was included in his s udy as i is mo e closely ela ed o he human gene in e ms o sequence and may he e o e be o alue in dissec ing p omo e unc ion [11]. The human wild- ype p omo e cons uc shown in Figu e 4is he ull-leng h 867-bp plasmid, hP om867 [8] o allow compa ison o he ull-leng h p omo e o each species. The mouse elome ase (mTe c) p omo e luci e - ase cons uc used was mP om628 [8], and a luci e ase epo e cons uc con aining bo ine elome ase RNA gene p omo e sequences (bTERC) was cons uc ed as de ailed in he Ma e ials and Me hods sec ion. Signi ican unc ional di e gence was obse ed be ween species; he human and bo ine p omo e s a e ac i a ed by pRb whe eas he mouse p omo e is no (Figu e 4c). Sp1 ac i a es all h ee p omo e s and Sp3 ep esses all h ee p omo e s. In e es ingly, NF-Y egula ion appea s o be speci ic o he human p omo e . Figu e 4. ( a ) Ac i a ion o hTERC p omo e ac i i y by pRb. The pRb / p53 nega i e cell line 5637 was ans ec ed wi h a ixed amoun o he hTERC - luci e ase plasmid and inc easing concen a ions o pRb exp ession ec o . In addi ion, he mouse elome ase ( mTe c ) p omo e - luci e ase cons uc mP om628 was also analyzed o i s esponse o pRb [ 8 ]. The da a is exp essed as old induc ion o luci e ase ac i i y ela i e o he p omo e alone. The pSEAP2 - Con ol ec o was used as an in e nal con ol o ans ec ion e iciency. ( b ) Di e en ial ac i a ion o hTERC p omo e ac i i y by wild - ype pRb and h ee a ian s, 663, 661W, and 657. The pRb/ p53 nega i e cell line 5637 was ans ec ed wi h 3 g o he hTERC - luci e ase plasmid ( phP om176 ) and 0.5 g o each pRb exp ession ec o . The da a is exp essed as old induc ion o luci e ase ac i i y ela i e o he exp ession ec o alone ( emp y ec o ) . The pSEAP2 -Con ol ec o was used as an in e nal con ol o ans ec ion e iciency. ( c ) Func ional di e gence in he ansc ip ional egula ion o TERC genes be ween species. P omo e - luci e ase cons uc s o he human (hP om867 ) , mouse ( mP om628 ) and bo ine TERC genes we e compa ed o hei egula ion by Sp1, Sp3, NF - YAm29, and pRb. The ac i i y o he p omo e s in esponse o co ans ec ion wi h he ansc ip ional egula o s is shown as old induc ion ela i e o he p omo e s alone. The pRb / p53 nega i e cell line 5637 was used in all expe imen s and he mean luci e ase ac i i y no malized o p o ein and ans ec ion e iciency is shown wi h he s anda d de ia ion o duplica e samples. 536 hTERC P omo e Regula ion Zhao e al. Neoplasia .Vol. 2, No. 6, 2000 Discussion Reac i a ion o elome ase is an almos uni e sal e en in ca cinogenesis. Though o he p o eins a e in ol ed in op imal unc ion o he enzyme, he RNA subuni and he ca aly ic p o ein componen s a e su icien o econs i u e enzyme ac i i y. Consis en wi h his, bo h componen s a e exp essed a high le els in he majo i y o cance s. In his s udy, se e al ac o s capable o modula ing hTERC p omo e ac i i y in cell cul u e ha e been iden i ied hus p o iding an insigh in o he possible molecula mechanisms o hTERC gene egula ion in no mal cells and i s de egula- ion in cance cells. The mechanisms egula ing no mal exp ession and ep ession may, howe e , di e om hose in ol ed in loss o ep ession and ac i a ion o exp ession in oncogenesis. The p esen s udy shows ha he CCAAT box is essen ial o basal p omo e unc ion. We ha e demons a ed ha his can be ecognized by he NF-Y complex and inhibi ion o NF-Y ac i i y ab oga es p omo e ac i i y. NF-Y is a he e ome ic p o ein composed o h ee subuni s, NF-YA, NF-YB and NF-YC [18]. NF-Y ac i i y is ound o change du ing senescence, and in esponse o inducing agen s [18,19]. I binds and con ols a numbe o cell-cycle egula ed p omo e s and may be di ec ly in ol ed in oncogenesis [18,20,21]. NF-Y may exe i s e ec h ough al e a ions in p omo e a chi ec u e. Ch oma in s uc u e can es ic he access o ansc ip ion-associa ed p o eins o p omo e s and his one ace yla ion is a majo mechanism o an agonism o ch oma in-media ed ep ession. NF-Y is able o associa e wi h nucleosomes, and o inhibi nucleosome o ma ion in in i o models [22]. I can also in e ac wi h he his one ace yl ans e ases, p300, GCN5, and P/CAF, [12,23,24] and may p o ide a means o a ge ing hese o speci ic p omo e s. NF-Y has also been shown o inc ease he a ini y o s uc u ally un ela ed ac i a o s, including Sp1, o hei binding si es [25±27]. I would be o in e es o examine whe he he egula ion o elome ase genes in de elopmen and in mo al and immo al cells is ela ed o ch oma in s uc u e and i so, whe he his is media ed by NF-Y. Sp1 and Sp3 ep esen jus wo membe s o a g owing amily o mammalian K uppel-like ansc ip ion ac o s ha bind GC- ich sequence elemen s, and a e in ol ed in he egula ion o issue speci ic and housekeeping genes. Sp p o eins ha e been implica ed in cell-cycle egula ion, ch oma in modeling, and main enance o me hyla ion- ee islands [14]. In addi ion, Sp1 knockou s show ha Sp1 is equi ed o ea ly mouse de elopmen [14]. T ansc ip ional egula ion by he Sp amily o p o eins is complex, bu in gene al Sp1 ac i a es gene exp ession, whe eas Sp3 ep esses i [14,28]. We ha e iden i ied Sp p o ein-binding si es wi hin he hTERC p omo e . We ha e demons a ed ha Sp1 and Sp3 can bind o hese si es and ha Sp1 ac i a es and Sp3 ep esses gene exp ession. Thus luc ua ions in Sp1/Sp3 a ios o al e a ion in hei ela i e binding a ini ies could modula e exp ession in i o. This is o pa icula signi icance because Sp1 has ecen ly been demons a ed o ac i a e he human elome ase p o ein componen p omo e (hTERT) [29] and we ha e he e o e iden i ied an o e lapping, hough no necessa ily co-coo di- na ed, con ol mechanism. Ou inding ha pRb is an ac i a o o he hTERC p omo e was somewha su p ising gi en i s well-known unc ion as a umo supp esse . P o ein Rb nega i ely egula es ansc ip ion o a numbe o genes ha play a ole in cell-cycle p og ession by o ming complexes wi h membe s o he E2F ansc ip ion amily. Howe e , pRb also has a i al ole in no mal de elopmen , a ime when hTERC is highly exp essed, and he abili y o pRb o coac i a e a numbe o genes including TGF-be a 2 , he e inoblas oma gene i sel , cyclin D1, and he We ne helicase gene is well desc ibed [15,16,30±34]. How pRb ac i a es ansc ip ion is no ully unde s ood bu in keeping wi h he esul s o Selle s e al., he abili y o pRb o ac i a e he hTERC p omo e is independen o i s abili y o bind E2F. P o ein Rb has no been shown o bind di ec ly o he p omo e s ha i ac i a es [35] hough ce ain sequence elemen s ha e been iden i ied as being in ol ed in pRb modula ion in o he p omo e s [30,36]. pRb may ac by seques e ing an inhibi o o o he ansc ip ion ac o s [37]. The iden i ica ion o hTERC as ano he gene ha is ac i a ed by pRb may con ibu e o he in es iga ion o his aspec o pRb unc ion. The unc ional signi icance o he abili y o pRb o ac i a e he hTERC p omo e in oncogenesis is no clea gi en ha dys unc ion o some pa o he pRb signal ansduc ion pa hway is e y common in cance . Howe e , he comple e loss o pRb i sel is ela i ely a e and e en ce ain ge m line mu a ions in RB-1 can gi e ise o a p o ein de ec i e in E2F-binding ye e aining he abili y o ac i a e ansc ip ion [16]. Fu he mo e, we ha e demons a ed ha pRb is a modula o o p omo e ac i i y bu i is no essen ial o ansc ip ion. One migh p edic ha when pRb- ansac i a - ing unc ion is los om he cance cell, compensa ion o he educed hTERC p omo e ac i i y migh occu . Indeed, he hTERC gene can be ampli ied in human cance s [38]. I could be ha his is selec ed o o inc ease hTERC exp ession in he absence o he posi i e egula o y in luences (such as pRb) o o i a e ou nega i e egula o y ac o s (such as Sp3). Telome ase enzyme unc ion is conse ed ac oss species bu he e a e signi ican di e ences in gene exp ession be ween species wi h high le els o exp ession in he mouse compa ed o he human [39,40]. We ha e demons a ed a ma ked di e gence in he egula ion o he human and he mouse TERC gene. This is pe haps no ha su p ising because he wo p omo e s show no signi ican sequence homology [8] and, in pa icula , a sec ion o he human and bo ine p omo e s lying be ween he ansc ip ion s a si e and he empla e egion is absen in he mouse p omo e . Phylogenic analysis o he TERC sequences o se e al species desc ibed by Chen e al. [41] e eals ha his a ea is also absen in o he oden s. Whe he his sec ion o he p omo e is espon- sible o he species a ia ion emains o be in es iga ed. The di e gence in he egula ion o elome ase RNA genes Neoplasia .Vol. 2, No. 6, 2000 hTERC P omo e Regula ion Zhao e al. 537 be ween species needs o be aken in o accoun when s udying elome ase in animal models o de elopmen and disease s a es [17]. Conclusion Ou esul s indica e ha mul iple signals including Sp1, Sp3, pRb, and NF-Y a e able o con ibu e o he egula ion o hTERC gene exp ession. Thus, al e a ions in he ela i e con ibu ions o Sp1, Sp3, pRb, and NF-Y in a cell, o signal- ansduc ion-speci ic manne , may be ele an o hTERC gene exp ession. This in o ma ion p o ides a ocus o s udies designed o unco e how cance cells eac i a e o up egula e hTERC gene exp ession. In addi ion, an unde - s anding o he molecula basis o hTERC gene egula ion may be o alue in he manipula ion o elome ase exp es- sion, o example in age- ela ed disease and o de eloping a ional he apies o ea cance h ough he exploi a ion o hTERC gene ansc ip ion [6,8]. Acknowledgemen s We hank D . William Kaelin, J . and D . Robe Whi e o he pRb exp ession ec o s, D . Robe o Man o ani o NF-Y exp ession ec o s and an ibodies, D . Gun am Suske o he Sp1 and Sp3 exp ession ec o s, and D . L. B yce o c i ical eading o he manusc ip . Re e ences [1] O'Reilly M, Teichmann SA, and Rhodes D ( 1999 ) . Telome ases. Cu Opin S uc Biol 9, 56± 65. [2] Hol SE, and Shay JW ( 1999 ) . Role o elome ase in cellula p oli e a ion and cance . J Cell Physiol 180, 10±18. [3] Yashima K, Mai a A, Roge s BB, Timmons CF, Ra hi A, Pina H, W igh WE, Shay JW, and Gazda AF ( 1998 ) . Exp ession o he RNA componen o elome ase du ing human de elopmen and di e en ia- ion. Cell G ow h Di e 9, 805 ± 813. [4] Weng N, Le ine BL, June CH, and Hodes RJ ( 1997 ) . Regula ion o elome ase RNA empla e exp ession in human T lymphocy e de el- opmen and ac i a ion. J Immunol 158, 3215± 3220. [5] Ogoshi M, Le T, Shay JW, and Taylo RS ( 1998 ) . In si u hyb idiza ion analysis o he exp ession o human elome ase RNA in no mal and pa hologic condi ions o he skin. J In es De ma ol 110, 818 ± 823. [6] Sa es a an J, Going JJ, Mil oy R, Kaye SB, and Kei h WN ( 1999 ) . Is small cell lung cance he pe ec a ge o an i - elome ase ea men ? Ca cinogenesis 20, 1649 ± 1652. [7] Yi X, Tesme VM, Sa e - T ain I, Shay JW, and W igh WE ( 1999 ) . Bo h ansc ip ional and pos ansc ip ional mechanisms egula e human elome ase empla e RNA le els. Mol Cell Biol 19, 3989 ± 3997. [8] Zhao JQ, Hoa e SF, McFa lane R, Mui S, Pa kinson EK, Black DM, and Kei h WN ( 1998 ). Cloning and cha ac e iza ion o human and mouse elome ase RNA gene p omo e sequences. Oncogene 16, 1345±1350. [9] Chen SJ, A le CM, Jimenez SA, and Va ga J (1998) . Modula ion o human alpha 1( I ) p ocollagen gene ac i i y by in e -ac ion wi h Sp1 and Sp3 ansc ip ion ac o s in i o.Gene 215, 101 ±110. [10] Conn KJ, Rich CB, Jensen DE, Fon anilla MR, Bashi MM, Rosenbloom J, and Fos e JA ( 1996 ) . Insulin- like g ow h ac o - I egula es ansc ip ion o he elas in gene h ough a pu a i e e inoblas oma con ol elemen Ð a ole o Sp3 ac ing as a ep esso o elas in gene ansc ip ion. J Biol Chem 271, 28853± 28860. [11] Tsao DA, Wu CW, and Lin YS ( 1998 ) . Molecula cloning o bo ine elome ase RNA. Gene 221, 51± 58. [12] Faniello MC, Be ilacqua MA, Condo elli G, de C omb ugghe B, Mai y SN, A edimen o VE, Cimino F, and Cos anzo F ( 1999 ) . The B subuni o he CAAT- binding ac o NFY binds he cen al segmen o he Co- ac i a o p300. J Biol Chem 274, 7623 ± 7626. [13] Man o ani R, Li XY, Pessa a U, Hoo an Huisjduijnen R, Benois C, and Ma his D (1994). Dominan nega i e analogs o NF -YA. J Biol Chem 269, 20340± 20346. [14] Lania L, Majello B, and De Luca P ( 1997 ) . T ansc ip ional egula ion by he Sp amily p o eins. In J Biochem Cell Biol 29, 1313 ± 1323. [15] Selle s WR, and Kaelin WG (1996) . RB [ co ec ed ] as a modula o o ansc ip ion [ published e a um appea s in Biochim Biophys Ac a 1996 Dec 9;1288 ( 3 ) :E - 1]. Biochim Biophys Ac a 1288,M1-5. [16] Selle s WR, No i ch BG, Miyake S, Hei h A, O e son GA, Kaye FJ, Lassa AB, and Kaelin WGJ ( 1998 ) . S able binding o E2F is no equi ed o he e inoblas oma p o ein o ac i a e ansc ip ion, p omo e di e en ia ion, and supp ess umo cell g ow h. Genes De 12, 95 ±106. [17] Blasco MA, Lee HW, Hande MP, Sampe E, Lansdo p PM, DePinho RA, and G eide CW (1997). Telome e sho ening and umo o ma ion by mouse cells lacking elome ase RNA [see commen s ] . Cell 91, 25±34. [18] Mai y SN, and de C omb ugghe B ( 1998 ) . Role o he CCAAT - binding p o ein CBF / NF -Y in ansc ip ion. T ends Biochem Sci 23, 174± 178. [19] Good LF, and Chen KY (1996 ) . Cell cycle - and age - dependen ansc ip ional egula ion o human hymidine kinase gene: he ole o NF- Y in he CBP / k binding complex. Biol Signals 5, 163±169. [20] Gu Z, Kun z - Simon G, Rommelae e J, and Co nelis J (1999 ) . Oncogenic ans o ma ion - dependen exp ession o a ansc ip ion ac o NF-Y subuni . Mol Ca cinog 24, 294 ± 299. [21] Pang JH, Good LF, and Chen KY ( 1996 ) . The age -dependen binding o CBP/ k, a CCAAT binding p o ein, is de egula ed in ans o med and immo alized mammalian cells bu absen in p ema u e aging cells. Exp Ge on ol 31, 97 ±109. [22] Mo a MC, Ca e i G, Bada acco GF, and Man o ani R ( 1999 ) . In e ac ions o he CCAAT - binding ime NF -Y wi h nucleosomes. J Biol Chem 274, 1326± 1333. [23] Jin S, and Sco o KW ( 1998 ) . T ansc ip ional egula ion o he MDR1 gene by his one ace yl ans e ase and deace ylase is media ed by NF- Y. Mol Cell Biol 18, 4377 ± 4384. [24] Cu ie RA ( 1998 ) . NF - Y is associa ed wi h he his one ace yl ans- e ases GCN5 and P / CAF. J Biol Chem 273, 1430 ± 1434. [25] Rei h W, Sieg is CA, Du and B, Ba as E, and Mach B ( 1994 ) . Func ion o majo his ocompa ibili y complex class- ii p omo e s equi es coope a i e binding be ween ac o s R x and N - Y. P oc Na l Acad Sci USA 91, 554 ± 558. [26] W igh KL, Moo e TL, Vilen BJ, B own AM, and Ting JPY ( 1995 ) . Majo his ocompa ibili y complex class ii - associa ed in a ian chain gene-exp ession is up - egula ed by coope a i e in e ac ions o Sp1 and N - Y. J Biol Chem 270, 20978 ± 20986. [27] Dooley KA, Millinde S, and Osbo ne TF ( 1998 ) . S e ol egula ion o 3- hyd oxy- 3 - me hylglu a yl - coenzyme A syn hase gene h ough a di ec in e ac ion be ween s e ol egula o y elemen binding p o ein and he ime ic CCAAT - binding ac o nuclea ac o Y. J Biol Chem 273, 1349±1356. [28] De Luca P, Majello B, and Lania L ( 1996 ) . Sp3 ep esses ansc ip ion when e he ed o p omo e DNA o a ge ed o p omo e p oximal RNA. J Biol Chem 271, 8533 ± 8536. [29] Kyo S, Takaku a M, Tai a T, Kanaya T, I oh H, Yu sudo M, A iga H, and Inoue M ( 2000 ) . Spl coope a es wi h c - Myc o ac i a e ansc ip ion o he human elome ase e e se ansc ip ase gene (hTERT ) . Nucleic Acids Res 28, 669±677. [30] Kim SJ, Wagne S, Liu F, O'Reilly MA, Robbins PD, and G een MR ( 1992 ) . Re inoblas oma gene p oduc ac i a es exp ession o he human TGF - be a 2 gene h ough ansc ip ion ac o ATF - 2. Na u e 358, 331 ± 334. [31] MuÈlle H, Lukas J, Schneide A, Wa hoe P, Ba ek J, Eile s M, and S auss M (1994) . Cyclin D1 exp ession is egula ed by he e inoblas oma p o ein. P oc Na l Acad Sci USA 91, 2945 ± 2949. [32] Yamabe Y, Shimamo o A, Go o M, Yoko a J, Sugawa a M, and Fu uichi Y ( 1998 ) . Sp1 - media ed ansc ip ion o he We ne helicase gene is modula ed by Rb and p53. Mol Cell Biol 18, 6191 ± 6200. [33] He wig S, and S auss M ( 1997 ) . The e inoblas oma p o ein: a mas e egula o o cell cycle, di e en ia ion and apop osis. Eu J Biochem 246, 581± 601. [34] Pa k K, Choe J, Osi chin NE, Temple on DJ, Robbins PD, and Kim SJ ( 1994 ) . The human e inoblas oma suscep ibili y gene p omo e is 538 hTERC P omo e Regula ion Zhao e al. Neoplasia .Vol. 2, No. 6, 2000 posi i ely au o egula ed by i s own p oduc . J Biol Chem 269, 6083 ± 6088. [35] Kaelin WG ( 1999 ) . Func ions o he e inoblas oma p o ein. Bioessays 21, 950 ± 958. [36] Ud adia AJ, Roge s KT, and Ho owi z JM ( 1992 ) . A common se o nuclea ac o s bind o p omo e elemen s egula ed by he e inoblas- oma p o ein. Cell G ow h Di e 3, 597 ±608. [37] Chen L, Nishinaka T, Kwan K, Ki abayashi I, Yokoyama K, Fu Y, G unwald S, and Chiu R (1994) . The e inoblas oma gene p oduc RB s imula es Sp1 - media ed ansc ip ion by libe a ing Sp1 om a nega i e egula o . Mol Cell Biol 14, 4380 ± 4389. [38] Sode AI, Hoa e SF, Mui S, Going JJ, Pa kinson EK, and Kei h WN ( 1997 ) . Ampli ica ion, inc eased dosage and in si u exp ession o he elome ase RNA gene in human cance . Oncogene 14, 1013 ± 1021. [39] Blasco MA, Funk W, Villepon eau B, and G eide CW ( 1995 ). Func ional- cha ac e iza ion and de elopmen al egula ion o mouse elome ase Rna. Science 269, 1267 ± 1270. [40] P owse KR, and G eide CW ( 1995 ) . De elopmen al and issue - speci ic egula ion o mouse elome ase and elome e leng h. P oc Na l Acad Sci USA 92, 4818 ± 4822. [41] Chen JL, Blasco MA, and G eide CW ( 2000 ) . Seconda y s uc u e o e eb a e elome ase RNA. Cell 100, 503± 514. Neoplasia .Vol. 2, No. 6, 2000 hTERC P omo e Regula ion Zhao e al. 539