Topological constraints impair RNA polymerase II transcription and causes instability of plasmid-borne convergent genes

Topological cons ain s impai RNA polyme ase II ansc ip ion and causes ins abili y o plasmid-bo ne con e gen genes Ma ı ´a L. Ga cı ´a-Rubio and And e ´s Aguile a* Cen o Andaluz de Biologı ´a Molecula y Medicina Regene a i a CABIMER, Uni e sidad de Se illa-CSIC, A . Ame ´ ico Vespucio s/n 41092 Se illa, and Depa amen o de Gene ´ ica, Uni e sidad de Se illa, 41012 Se illa, Spain Recei ed June 30, 2011; Re ised Sep embe 8, 2011; Accep ed Sep embe 21, 2011 ABSTRACT Despi e he heo e ical bases o he associa ion o opoisome ases and supe coiling changes wi h ansc ip ion and eplica ion, ou knowledge o he impac o opological cons ain s on ansc ip ion and eplica ion is incomple e. Al hough mu a ion o opoisome ases a ec s exp ession and s abili y o he DNA egion i is no clea whe he he same is he case o RNAPII ansc ip ion and genome in eg- i y in o he egions. We de eloped new assays in which wo con e gen RNAPII-d i en genes a e ansc ibed simul aneously. Plasmid-based sys ems we e cons uc ed wi h and wi hou a ansc ip ion e mina o be ween he wo con e gen ansc ip- ion uni s, so ha he impac o ansc ip ion in e - e ence could also be e alua ed. Using hese assays we show ha Topos I and II play oles in RNAPII ansc ip ion in i o and educe he s abili y o RNAPII- ansc ibed genes in Saccha omyces ce e isiae. Supe coiling accumula ion in con e gen ansc ip ion uni s impai s RNAPII ansc ip ion in op1Ds ains, bu Topo II is also equi ed o e icien ansc ip ion independen o Topo I and o de ec able supe coiling accumula ion. Ou wo k shows ha opological cons ain s nega i ely a ec RNAPII ansc ip ion and gene ic in eg i y, and p o ides an assay o s udy gene egula ion by an- sc ip ion in e e ence. INTRODUCTION DNA in i s na u al o m in he cell is supe coiled. In euka yo es his s a e acili a es i s o ganiza ion in o ch o- ma in a ound nucleosomes. DNA opoisome ases con ol supe coiling and hei ac ion is equi ed in eplica ion and ansc ip ion as well as in ch omosome condensa ion and seg ega ion (1,2). The e a e wo ypes o DNA opoisom- e ases (Topo), I and II, which di e in hei ca aly ic way o ac ion, he fi s elying on a single-s and DNA clea age, he second on a double-s and b eak (DSB). Topoisome ases I and II a e he mos ep esen a i e o each ype in all o ganisms. They a e unc ionally in e - changeable in euka yo es, bu no in bac e ia. Du ing eplica ion DNA opoisome ases can ac a di - e en s ages. In Esche ichia coli econs i u ed eplica ion sys ems equi e DNA supe coiling o eplica ion ini i- a ion and he ac ion o DNA opoisome ases (3). Howe e , i is du ing elonga ion o eplica ion ha opo- isome ases a e expec ed o ha e a key ole. They a e equi ed o emo e he posi i e supe coiling ha a e gene- a ed by he con inuous opening o he empla e s ands. The in e wined daugh e DNA molecules esul ing a e eplica ion comple ion equi e he ac ion o ype II opo- isome ases o hei seg ega ion. This la e unc ion in ep- lica ion canno be pe o med by ype I opoisome ases bu only by ype II (4,5). In con as o eplica ion, ansc ip ion does no in ol e a con inuous sepa a ion o DNA s ands. Howe e , RNA polyme ase p og ession h ough he DNA empla e would lead o he accumula ion o local posi i e supe coiling ahead o he RNA polyme ase and nega i e supe coiling behind. The win-supe coiled-domain model was p oposed o explain emo al o his local supe coil- ing accumula ion by opoisome ases du ing ansc ip ion (6). E idence o ansc ip ional supe coiling associa ed wi h ansc ip ion has been p o ided in E. coli (6,7) and yeas (8,10). The elaxa ion o ansc ip ion-dependen supe coiling is gene ally conside ed o be he main unc ion o Topo I, which is conside ed o play a swi el ole in ansc ip ion (1,2). In euka yo es such as he yeas Saccha omyces ce e isiae, he ac ha op1Dmu an s a e iable sugges s ha Topo II can eplace he unc ion o Topo I; whe eas he opposi e is no he case, since Topo I *To whom co espondence should be add essed. Tel: +34 954 468 372; Fax: +34 954 461 664; Email: [email p o ec ed] 1050–1064 Nucleic Acids Resea ch, 2012, Vol. 40, No. 3 Published online 13 Oc obe 2011 doi:10.1093/na /gk 840 ßThe Au ho (s) 2011. Published by Ox o d Uni e si y P ess. This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion Non-Comme cial License (h p://c ea i ecommons.o g/licenses/ by-nc/3.0), which pe mi s un es ic ed non-comme cial 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. canno eplace Topo II a he end o eplica ion du ing seg ega ion. Despi e he heo e ical bases o he associa ion o opo- isome ases and supe coiling changes wi h ansc ip ion, expe imen al e idence o he impac o opoisome ases deficiency and opological cons ain s in ansc ip ion is incomple e. In bac e ia, he impo ance o Topo I has been well documen ed in DNA ansc ip ion (11) o an- sc ip ion om ope ons in ol ed in amino acid me abolism (12,13). In yeas , an e ec o opoisome ase ac i i y has been epo ed on ansc ip ion ini ia ion. This may be ela ed o he need o nega i ely supe coiled DNA a p omo e egions o acili a e ansc ip ion ini ia ion (14). The e ec o DNA opoisome ases on ansc ip ion elonga ion has been es ablished o DNA, in which he a e o RNAPI ansc ip ion is educed when bo h DNA opoisome ases a e inac i a ed in yeas (8,15). Ins ead, none o a poo e ec has been epo ed o mRNA an- sc ip ion as es ed in endogenous and lacZ epo e s used o GAL p omo e s in i o (10,16). In addi ion o he e ec o Topo mu a ions in cell g ow h, ansc ip ion and eplica ion, hey ha e an im- po an impac on genome in eg i y. S abili y o andem DNA epea s in yeas has been shown o be s ongly dependen on DNA opoisomeases I and II, whe eas his was no he case o di ec epea s ou side o he DNA egion (17). Hype - ecombina ion has been shown in plasmid-bo ne di ec - epea s in double op1 op2 mu an s (18). I has been p oposed ha his is mainly he esul o he ailu e o elax he nega i e a he han he posi i e supe coiling o DNA (18). Ne e heless, im- pai men o eplica ion o k p og ession by obs acles o DNA lesions can lead o eplica ion o k collapse ha in u n could igge ecombina ion as a mechanism o ep- lica ion es a (19). Thus, hype - ecombina ion o yeas opoisome ase mu an s could be linked o hei ole in eplica ion, bu he unc ional ele ance o Topo I in DNA ansc ip ion and i s la ge accumula ion in he nu- cleolus, makes plausible ha hype - ecombina ion could also be linked o ansc ip ion. Indeed, DNA genes con ain he RNAPI ansc ip ion-dependen HOT1 e- combina ion ho spo s (20,21), and one key unc ion o he Fob1 ba ie o DNA eplica ion is o a oid eplica- ion– ansc ip ion collisions ha could po en ially com- p omise genome in eg i y (22). In his sense, one in iguing class o hype - ecombina ion mu a ions in yeas is composed o hose o he THO complex. THO is a conse ed p o ein complex composed o s oichiome ic amoun s o Hp 1, Tho2, M 1 and Thp2, wi h a ole in ansc ip ion elonga ion, mRNP biogenesis and RNA expo (23). THO mu an s, as well as mu an s ela ed o mRNP biogenesis ac o s such as he THSC/TREX-2 complex, cause a s ong hype - ecombina ion be ween di ec epea s ha a e ansc ip ion-dependen . This pheno ype is linked o he co- ansc ip ional o ma ion o R-loops ha can impai he p og ession o eplica ion (24,25). In e es ingly, in E. coli mu an s lacking opoisome ase I, an excess o nega- i e supe coiling du ing ansc ip ion can lead o R-loops (26). Fu he mo e, THO and THSC mu an s show syn- he ic g ow h de ec s in combina ion wi h op1 and op2 mu a ions (27). All his e idence aises he ques ion o whe he hype - ecombina ion in he DNA egion in opoisome ases mu an s could be linked o ansc ip ion. Thus, al hough opological cons ain s and opoisome - ase mu a ions ha e a clea impac in exp ession and s a- bili y o he DNA locus his is no ha clea o RNAPII ansc ip ion and genome in eg i y a egions o he s han DNA. To in es iga e his, new and highly sensi i e in i o assays we e de eloped based on plasmid-bo ne DNA sequences o inc easing leng hs ha a e ansc ibed o m s ong p omo e s wi h o wi hou con e gen an- sc ip ion coming om a second p omo e . We show ha bo h op1 and op2 mu a ions d as ically educe RNAPII ansc ip ion, a de ec ha is s eng hened as he leng h o he ansc ibed DNA sequence inc eases and by con e - gen ansc ip ion. Impo an ly, op2-1 mu a ion causes a significan inc ease in di ec - epea ecombina ion in a op1Dmu an backg ound; howe e , his hype - ecombina ion was p edominan ly independen o an- sc ip ion o he di ec - epea sys em i sel . The e o e, his s udy p o ides new e idence o he in i o ele ance o Topos I and II in RNAPII ansc ip ion and genome in eg i y. We p opose ha DNA opoisome ases has a non-in e changeable ole in he main enance o in eg i y o RNAPII genes p o iding e idence o he nega i e e ec o opological cons ain s in ansc ip ion and genome s abili y in DNA egions beyond he DNA. Al hough a simila s udy would need o be pe o med in ch omosome-bo ne cons uc s, we belie e ha hese esul s may also be alid o ch omosomal egions, p o ided ha local supe coiling dynamics ollow uni e sal ules. MATERIALS AND METHODS S ains and plasmids Yeas s ains and plasmids a e desc ibed in Supplemen a y Tables S1 and S2, espec i ely. Ch oma in immunop ecipi a ion analysis Fo ChIP expe imen s, s ains we e g own in syn he ic medium (SC) 2% glyce ol-2% lac a e o an OD 660 o 0.5 ( o GAL1). Fo GAL1 ChIPs he cul u e was spli in wo, and one hal was supplemen ed wi h 2% glucose ( ep essed ansc ip ion) and he o he wi h 2% galac ose (ac i a ed ansc ip ion). Fo P e -based sys ems, yeas s ains we e g own in syn he ic medium (SC), he cul u e was spli in wo, and one hal was supplemen ed wi h doxycycline. Samples we e hen aken a e 4 h o induc- ion and ChIP assays we e pe o med as desc ibed (28). Polyclonal an i-PolII (N-20) (San a C uz Bio echnology) and P o ein A–Sepha ose we e used o RNAPII immunop ecipi a ion. The GFX pu ifica ion sys em (Ame sham) was used o he las DNA pu ifica ion s ep. We used he PCR o he in e genic egion a posi ions 9716–9863 o ch omosome V as a nega i e con ol. Real- ime quan i a i e PCR and calcula ions o he ela- i e abundance o each DNA agmen was pe o med as desc ibed (29). Fo each expe imen , he DNA a ios in he di e en egions we e calcula ed om he DNA amoun o hese egions ela i e o he in e genic egion. Nucleic Acids Resea ch, 2012, Vol. 40, No. 3 1051 Median and S anda d De ia ion (SD) o h ee independ- en expe imen s a e shown. Nucleic acid isola ion Fo DNA isola ion, pelle s we e e-suspended o a final concen a ion o 1.5–2 10 9 /ml in NIB pH 7.2 (17% glyce ol, 50 mM MOPS, 150 mM KAc, 2 mM MgCl, 0.5 mM spe midine and 0.15 mM spe mine) wi h 3 mg/ml Zymoliase-20T and incuba ed a oom empe a u e o 25 min. Cellula suspensions we e dilu ed se en imes wi h chilled wa e and cen i uged o collec nuclei. Pelle s we e e-suspended in 720 ml o TE (50 mM T is–HCl, pH 8, 20 mM EDTA) and mixed wi h 80 mlo 10% SDS. A e phenol–chlo o o m–isoamylalcohol (25:24:1) ex ac ion and e hanol p ecipi a ion, DNA was finally e-suspend in 20–30 ml o TE. Chlo oquine gel elec opho esis Elec opho esis was pe o med a oom empe a u e on ho izon al slab gels (0.7% aga ose) con aining TPE bu e (50 mM T is–phospha e [pH 7.2], 1 mM EDTA and 25 mM phospho ic acid). Chlo oquine (Sigma) was added o he mel ed aga ose om a 10 mg/ml s ock p io o cas ing he gel and was p esen in he elec opho - esis bu e . The bu e was e-ci cula ed con inuously du ing elec opho esis. The concen a ion o chlo oquine was 4mg/ml. Each lane had 20 mg o o al nucleic acids. The gels we e un a 40 V o 48 h. Miscellaneous Yeas me hods, a- 32 P-labeled DNA p obes, no he n and Sou he n analyses we e pe o med acco ding o s anda d p ocedu es. RNA was isola ed om mid-log phase cells. Recombina ion equencies we e ob ained by fluc u- a ion es s as he median alue o six independen colonies isola ed om SC pla es. The final equency gi en o each s ain and condi ion is he mean and SD o h ee o ou median alues as desc ibed (30). Recombinan s Leu+we e selec ed in SC- Leu pla es. RESULTS Con e gen ansc ip ion in sys ems wi h o wi hou e mina o sequences Fo a fine assessmen o he possible e ec o opological cons ain s in ansc ip ion, a plasmid-based sys em was cons uc ed in which a 4-kb sequence was ansc ibed om wo con e gen and di e en ially egula ed p o- mo e s loca ed on each side. Fo his cons uc ion, we used a 2-kb lacZ agmen unde he GAL1 p omo e o a 2-kb lys2 agmen unde he quime ic e p omo e so ha ansc ip ion d i en om he wo p omo e s was con e gen (Figu e 1). Two e sions o his sys em we e cons uc ed in monocopy CEN-based plasmids: one wi h a 0.4-kb CYC1 e mina o be ween he lys2 and lacZ sequences (pTGlylaTER) and he o he wi hou ansc ip- ion e mina o (pTGlyla). The CYC1 e mina o sequence con ains wo con e ging e mina o s so ha i ac s in bo h o ien a ions. As expec ed, whe eas in he pTGlyla sys em wi hou e mina o he 4-kb lys2-lacZ mRNA was p oduced when cells we e cul i a ed ei he wi h 2% gal- ac ose (GAL1 p omo e ac i a ed) o wi hou doxycycline ( e p omo e ac i a ed), bu no in 2% glucose + doxy- cycline (GAL1 and e p omo e s ep essed) (Figu e 1A). To confi m ha he band de ec ed in 2% galac ose was d i en om he GAL1 p omo e and he band de ec ed in he absence o doxycycline was d i en om he e p omo e no he n analysis was pe o med in which hyb idiza ion was made wi h specific ssDNA p obes. As expec ed, he 4-kb mRNA band was specific o a DNA s and. I was de ec ed by he Wa son-chain-specific (W) p obe in he absence o doxycycline (P e ac i e), and wi h he C ick-chain-specific (C) p obe in galac ose (P GAL1 ac i e), acco ding o he p omo e om which each an- sc ip was ansc ibed (Figu e 1B). Ins ead, in he pTGlylaTER sys em ei he he 2-kb lacZ o he 2-kb lys2 mRNAs we e de ec ed in galac ose ega dless o he p esence o doxycycline, o in doxycycline ega dless o he ca bon sou ce, espec i ely (Figu e 1C and D). Impo an ly, in he pTGlyla sys em, he le el o 4-kb lys2-lacZ ansc ip s was lowe when bo h he P GAL1 and P e p omo e s we e simul aneously ac i e han when only one was ac i e, as obse ed wi h ei he he lys2 o he lacZ p obe (Figu e 1A and B). This esul is consis en wi h ansc ip ion in e e ence ei he a he p omo e o an- sc ip ion s alling p o ided by RNAPII collisions. Since he 4-kb ansc ip d i en om ei he he e o GAL1 p omo e s eaches he downs eam con e ging p omo e , he la e may unde go ansc ip ion in e e ence. When he ansc ip s we e analyzed wi h specific ssDNA p obes i was ound ha a e galac ose ac i a ion he P e -d i en 4-kb ansc ip , as de ec ed wi h he W p obe, was educed, whe eas he P GAL1 -d i en ansc ip was below de ec ion le els (Figu e 1B). This indica es ha since he P e -d i en RNAPII eaches P GAL1 , i causes p omo e in e e ence impeding P GAL1 ac i a ion, e en hough an- sc ip ional s alling canno be excluded ei he . Simila ly, P GAL1 -d i en ansc ip ion seems o educe he e ficiency o P e -d i en ansc ip ion, al hough o a lesse ex en . To assay whe he he no he n esul s could be in e - p e ed in e ms o RNAPII p og ession, a ch oma in– immunop ecipi a ion (ChIP) analysis was pe o med wi h an i-Rpb1 N- e an ibodies in bo h he pTGlylaTER and pTGlyla sys ems unde all possible condi ions (wi h ei he none, only one o bo h p omo e s ac i a ed). As can be seen in Figu e 2, RNAPII ec ui men was no obse ed in ei he o he egions analyzed when bo h p omo e s we e ep essed in ei he sys em, as expec ed. In he pTGlylaTER sys em RNAPII was de ec ed in lys2 when only P e was ac i e and in bo h lys2 and lacZ when only P GAL1 was ac i e (Figu e 2A). The la e was consis en wi h ansc ip ion e mina ion ead- h ough obse ed o ansc ip ion ini ia ed a he s ong P GAL1 p omo e , as obse ed by no he n (Figu e 1). In he pTGlyla sys em, RNAPII ec ui men was obse ed a bo h he lys2 and lacZ egions when ei he P e o P GAL1 we e he only ac i e p omo e s, whe eas ec ui men o one p omo e was only de ec ed when ansc ip ion was d i en om he opposi e p omo e (Figu e 2B), implying ha such polyme ases co - esponded o elonga ing RNAPIIs. When bo h p omo e s 1052 Nucleic Acids Resea ch, 2012, Vol. 40, No. 3 we e ac i e, he lys2 and lacZ signals we e de ec ed, as expec ed, in he pTGlylaTER sys em, whe eas only he lys2 signal was de ec ed in pTGlyla. This confi ms ha he no he n analysis used can be in e p e ed in e ms o RNAPII capaci y o ansc ibe he di e en egions analyzed. T ansc ip ion elonga ion is sensi i e o Topos I and II inac i a ion Nex , he wo cons uc s we e used o analyze ansc ip- ion in he p esence o doxycycline, so ha he P e p omo e was inac i a ed and ansc ip ion was only d i en om P GAL1 . Unde hese condi ions kine ics expe imen s we e pe o med o P GAL1 -d i en ansc ip ion wi h 2% galac ose o de e mine he e ficiency o ansc ip- ion o he 4-kb lacZ-lys2 ansc ip in pTGlyla and he 2-kb LacZ ansc ip in pTGlylaTER. As can be seen in Figu e 3, ansc ip ion o LacZ was clea ly educed in op1Dmu an s, mo e educed in op2-1 mu an s and u - he in he double op1D op2-1 mu an s. T ansc ip ion o he 4-kb lys2-lacZ was d ama ically dec eased in all single and double op mu an s. This esul is consis en wi h a de ec in ansc ip ion elonga ion so ha he longe he RNA is ansc ibed he s onge he e ec . The double mu an showed he s onges e ec and a simila d ama ic dec ease in he accumula ion o bo h o he 2- and 4-kb ansc ip s. Di e en s udies ha e e ealed ha yeas mu a ions im- pai ing ansc ip ion elonga ion causes a educ ion in mRNA accumula ion ha in mos cases inc eases wi h he leng h o he ansc ip uni analyzed (23). Using plasmid-bo ne P GAL1 ::PHO5 and P GAL1 ::lacZ cons uc s con aining he 1.2-kb and 3.1-kb yeas PHO5 and bac e - ial lacZ ORFs, espec i ely, unde he same GAL1 p omo e , he nega i e e ec o opoisome ase mu a ions was confi med in ansc ip ion o long DNA sequences. The single and double Topo mu an s ha e a significan e ec on he e ficiency o ansc ip ion o he PHO5 ORF, he educ ion in he accumula ion o he lacZ mRNA being s onge in he single op2-1 and in he double hp1D op2-1 (Supplemen a y Figu e S1). Gi en ha ansc ip ion in bo h cons uc s is d i en om he same GAL1 p omo e , his esul excludes he possibili y Figu e 1. T ansc ip ion analyses o he lys2-CYC -lacZ and lys2-lacZ usion cons uc s. (A) No he n analyses o lys2-lacZ-con aining mRNAs d i en om he e and GAL1 p omo e s. The 3-kb XbaI-EcoRV lacZ and a 589-bp 25 S DNA in e nal agmen s ob ained by PCR ( RNA) we e used as DNA p obes. (B) No he n analyses o lys2-lacZ. The oligos C and W we e used as single-s anded DNA p obes. (C), (D) No he n analyses o pTGlylaT-1 and pTGlylaT-5 plasmids. A ows e e s o he mRNA being p oduced om he specified p omo e . No a ow indica es ha he e is no ansc ip ion om ha p omo e . All expe imen s a e made in he absence o doxycycline, so ha he e p omo e is pe manen ly ac i e. A o al o 2% glucose (Glu) o galac ose (Gal) a e used o ei he ep ess o ac i a e he GAL1 p omo e . O he de ails as desc ibed in ‘Ma e ials and Me hods’ sec ion. Nucleic Acids Resea ch, 2012, Vol. 40, No. 3 1053 ha he majo ansc ip ion de ec s obse ed in op mu an s occu a ansc ip ion ini ia ion, bu a elong- a ion. In addi ion, consis en wi h he idea ha euka yo ic Topos I and II can pa ially eplace each o he , we showed ha o e exp ession o Top1 can pa ially supp ess he lacZ mRNA accumula ion de ec o op2-1 mu an s (Supplemen a y Figu e S2). Con e gen ansc ip ion is impai ed unde Topos I and II inac i a ion as a unc ion o supe coiling accumula ion Nex , we assayed whe he o no con e gen ansc ip ion a ec ed ansc ip ion e ficiency. This analysis o con e - gen ansc ip ion was made in he absence o doxycycline o allow con inuous P e -d i en ansc ip ion. Time-cou se kine ics was pe o med o he P GAL1 -d i en ansc ip s a e galac ose addi ion. As can be seen in Figu e 4, in he pTGlylaTER sys em wi h he ansc ip ion e mina o he o e all le els o he P e -d i en lys2 ansc ip did no show a significan change a e ac i a ion o he con e g- ing P GAL1 . A 200 min o P GAL1 ac i a ion he o e all le els o lys2 ansc ip s we e simila o he le els be o e ac i a ion. A in e media e imes some di e ences be ween wild- ype and mu an s ains we e e iden , bu hese we e mino . This may be a esponse o cell adap a ion o new condi ions. Impo an ly, a 200 min o ac i a ion he lys2 ansc ip le els we e he same in wild- ype and mu an s ains, wi h a sligh dec ease in op1D op2-1 mu an s. Ins ead, ansc ip ion om he GAL1 p omo e , as de e mined by accumula ion o he 2-kb lacZ ansc ip , was clea ly a ec ed in op1Dand op2-1 mu an s (38–40% o wild- ype le els) and much mo e in he double op1D op2-1 mu an s (10% o wild- ype le els) (Figu e 4A). The clea impai men in ansc ip ion o he 2-kb lacZ an- sc ip caused by opoisome ase mu a ions as compa ed o he poo o no e ec on he 2-kb lys2 ansc ip (bo h ansc ip s a e o he same size bu a e d i en om p omo e s o di e en s eng hs), indica es ha he highe he ansc ip ion a e he highe he equi emen o opoisome ases. In he pTGlyla sys em he e is li le change in he o al amoun o 4-kb ansc ip s (Figu e 4B) du ing he kine ics pe o med. As shown abo e, hese esul s indica e ha pTGlyla, which con ains no e mina o sequence be ween he con e ging p omo e s, unde goes p omo e in e e - ence caused a P GAL1 by he RNAPII coming om P e . Howe e , despi e he in e e ence phenomenon, an- sc ip s we e mode a ely low in single op1 and op2 mu an s (60–65% o wild- ype le els a e 200 min ac i a- ion) and lowe in double op1 op2 mu an s (38–45%) as compa ed o wild- ype le els. This is consis en wi h he impo ance o opoisome ase ac i i ies in con e ging an- sc ip ion. I is wo h no ing he di e en opoisome ase equi emen s o ansc ip ion o he 4-kb lys2-lacZ and 2-kb lys2 agmen s in he pTGlyla and pTGlylaTER sys ems (compa e Figu es 4A. up and D, bo om). This is due o he longe leng h o he lacZ-lys2 agmen , con- sis en wi h he conclusion ha he longe he ansc ip he highe he dependency o ansc ip ion on opoisom- e ase ac i i y (Figu e 1). Al oge he , he esul s indica e ha opoisome ases I and II a e equi ed o e ficien RNAPII ansc ip ion elonga ion. To assay whe he he ansc ip ion de ec caused by op mu a ions was linked o he incapaci y o esol e Figu e 2. Dis ibu ion o RNAPII along he lys2-lacZ ansc ip ion uni s o plasmids pTGlyla and pTGlylaTER. ChIP analyses in he wild- ype s ain (HRN1-4A) ca ying he plasmids pTGlylaTER (A) o TGlyla (B). Schemes o he gene and he PCR-amplified agmen s a e shown. DNA a ios in egions 1–4 we e calcula ed om he amoun s ob ained o hese egions ela i e o he amoun s o he in e genic egion. ChIPs we e pe o med om h ee independen cul u es, and quan i a i e PCRs we e epea ed h ee imes o each cul u e. SDs a e indica ed as e o ba s. 1054 Nucleic Acids Resea ch, 2012, Vol. 40, No. 3 supe coiling, a he han o a specific unc ion o any o he opoisome ases a specific a ge DNA si es, supe coil- ing o he plasmids was analyzed ca ying bo h he pTGlylaTER and pTGlyla sys ems unde condi ions o ansc ip ion ac i a ion o bo h p omo e s. As can be seen in Figu e 5, a la ge p opo ion o supe coiled DNA is accumula ed in op mu an s, in pa icula in op1 and op1 op2 when bo h p omo e s a e ac i e, bu no when hey we e inac i e. Expe imen s we e pe o med simila ly o p e ious epo s using chlo oquine concen- a ions in which all opoisome s show nega i e supe - coiling (9). Impo an ly, accumula ion o he supe coiled plasmid band is clea ly s onge in pTGlyla e sus pTGlylaTER, consis en wi h he conclusion ha he longe he leng h o he ansc ip ion uni , he highe he change in supe coiling p oduced by ansc ip ion and he s onge he equi emen o opoisome ases. E en hough on a heo e ical basis he impac o ansc ip- ion on DNA supe coiling is local, TOP1 inac i a ion has a global impac on plasmid supe coiling as has been p e- iously epo ed (9). As p e iously shown (9) accumula- ion o global supe coiling was no e iden in op2-1 mu an s. E ficien RNAPII-d i en con e gen ansc ip ion equi es Topos I and II So a , all ou p e ious expe imen s we e pe o med a 30C, a semi-pe missi e empe a u e o allow g ow h o Figu e 3. T ansc ip ion analyses o lys2-CYC -lacZ and lys2-lacZ usion ansc ip d i en o m he GAL1 p omo e . (A) No he n analyses o lys2-CYC -lacZ con aining mRNAs d i en om he GAL1 p omo e in wild- ype (HRN1-4A), op1D(TOHR-11A), op2-1 (TORH-6A) and op1D op2-1 (TOHR-13C) s ains. Mid-log phase plasmid- ans o med cells we e dilu ed in 3% glyce ol–2%lac a e syn he ic comple e (SC)-T p medium plus doxycycline and dilu ed in o iden ical esh media o an OD 600 o 0.4 and incuba ed o 16 h. Galac ose was hen added and samples we e aken o no he n analyses a di e en imes. RNA le els in a bi a y uni s we e ob ained by quan ifica ion o signals in ensi ies in a FUJI FLA 5000 and no malized wi h espec o RNA le els o each sample. Wild- ype mRNA le els we e aken as 100%. (B) No he n analyses o lys-lacZ-con aining mRNAs d i en om he GAL1 p omo e using he same s ains as in A. Nucleic Acids Resea ch, 2012, Vol. 40, No. 3 1055 Figu e 4. No he n analyses o con e gen ansc ip ion in he lys2-CYC -lacZ and lys2-lacZ usion cons uc s. (A) No he n analyses o lys2-CYC -lacZ mRNAs d i en om he e and GAL1 p omo e s in wild- ype (HRN1-4A), op1D(TOHR-11A), op2-1 (TORH-6A) and op1D op2-1 (TOHR-13C) s ains. The 3-Kb XbaI-EcoRV lacZ in e nal agmen we e used as DNA p obe. (B) No he n analyses o lys2-lacZ mRNAs d i en om he P e and P GAL1 p omo e s. O he de ails as desc ibed o Figu e 3. 1056 Nucleic Acids Resea ch, 2012, Vol. 40, No. 3 op2-1 he mosensi i e mu an s. Ye , i was possible o see a syne gis ic e ec on double op1 op2 mu an s sugges ing ha Topo II has a ole in ansc ip ion in i o. This conclusion was confi med by assaying he e ec on comple e Top II inac i a ion. These expe imen s we e epea ed on con e gen ansc ip ion in bo h he pTGlyla and pTGlylaTER sys ems a e shi ing cells o he es ic i e empe a u e o 37C. As can be seen in Figu e 6A, in he pTGlylaTER sys em in which bo h con- e gen p omo e s a e ac i a ed, whe eas lys2 and lacZ ansc ip le els show li le e ec on op1Dand op2-1 single mu an s, which we e d ama ically educed in he double op1D op2-1 mu an s (20% o wild- ype le els a e 2 h a 37C). In he pTGlyla sys em, whe eas in wild- ype and, o a lesse ex en , in he op1Dmu an , he 4-kb ansc ip le els inc eased wi h ime a 37C con- sis en wi h a as e me abolism a his empe a u e, hey did no change in op2-1 mu an s and d opped d ama ic- ally in op1D op2-1 mu an s (Figu e 6B). These esul s we e confi med by analyzing he e ec o op2-1 a di e - en empe a u es, om 30C o33 C in which i could be seen a d ama ic dec ease in he accumula ion o lacZ mRNA in he op1Dbackg ound as he empe a u e inc eased (Supplemen a y Figu e S2). Al oge he , he esul s indica e ha Topo II plays a c i ical ole in RNAPII ansc ip ion in i o, which indeed is quan i a- i ely mo e impo an han ha o Topo I. op1Dand op2-1 con e a gene al hype - ecombina ion pheno ype I is known ha op1Dmu an s con e a s ong hype - ecombina ion a he DNA locus (17). To es whe he his lack o e ec was gene al o any DNA egion and o di e en ypes o ecombina ion e en s and whe he i was ela ed o ansc ip ion, he e ec o op mu a ions on ecombina ion was analyzed in di e en and highly sensi i e plasmid-bo ne ecombina ion sys ems. Fi s we assayed ecombina ion in he pTINV sys em based on wo in e ed epea s o he LEU2 gene, one copy o which is a 50-end unca ion (leu2D50) and he o he a 21-bp inse ion mu a ion unde he con ol o he egula ed e p omo e (P e ::leu2-HO ) (31). This sys em allows de ec ion o gene con e sion e en s o he leu2-HO allele wi h o wi hou c osso e s as a unc ion o an- sc ip ion, which leads o Leu+ ecombinan s ha a e sco ed on SC-leu pla es (Figu e 7A). As can be seen in Figu e 7A, whe eas Leu+ ecombina ion le els in op1D and op2-1 mu an s we e simila o he wild- ype bo h wi h o wi hou ansc ip ion, ecombina ion was inc eased 26- and 37- old wi hou o wi h ansc ip ion. Nex , we es ed he e ec on ecombina ion in he pGL-PHO5 sys em ca ying wo 0.6-kb leu2 di ec epea s unde he con ol o he GAL1 p omo e and he PHO5 ORF loca ed be ween he epea s (32).This sys em is used o s udy de- le ions occu ing by ecombina ion be ween he wo leu2 Figu e 5. Analysis o he e ec o op1Dand op2-1 on DNA supe coiling. (A) Chlo oquine gel elec opho esis analysis o plasmid pTGlylaTER in wild- ype (HRN1-4A), op1D(TOHR-11A), op2-1 (TORH-6A) and op1D op2-1 (TOHR-13C). S ains we e cul u ed unde non- ansc ip ion condi ions in glucose plus doxycycline (Glu +dox) and unde ansc ip ion condi ions in galac ose wi hou doxycycline (Gal) a 30C. Elec opho esis was ca ied ou in he p esence o 4 mg/ml chlo oquine, and hyb idiza ion was pe o med wi h a labelled LYS2 DNA p obe. Sou he n e eals bands ep esen ing opoisome s di e ing in linking numbe by s eps o one. Wi h he chlo oquine concen a ion used, all opo- isome s a e nega i ely supe coiled, wi h hose o inc easing nega i e supe helici y mig a ing as e in he gel. The band ma ked as SC consis s o he mos nega i ely supe coiled species ha a e no esol ed by he concen a ion o chlo oquine used he e. (B) Chlo oquine gel elec opho esis analysis o plasmid pTGlyla. Nucleic Acids Resea ch, 2012, Vol. 40, No. 3 1057 Figu e 6. E ec o op1 and op2 mu a ions on ansc ip ion o lys-CYC -lacZ and lys-lacZ usion cons uc s a 37C. (A) No he n analyses o lys-CYC -lacZ-con aining mRNAs d i en om he e and GAL1 p omo e s in wild- ype (HRN1-4A), op1D(TOHR-11A), op2-1 (TORH-6A) and op1D op2-1 (TOHR-13C) s ains a 37C. Mid-log phase plasmid- ans o med cells we e dilu ed o an OD 600 o 0.4 syn he ic comple e (SGal)-T p medium a 26C and shi ed o 37C, samples we e aken o no he n analyses a di e en imes. (B) No he n analyses o lys-lacZ mRNAs d i en om he P e and P GAL1 p omo e s. O he de ails as desc ibed in Figu e 3. 1058 Nucleic Acids Resea ch, 2012, Vol. 40, No. 3