Chromosomal aberrations induced by double strand DNA breaks

Ch omosomal abe a ions induced by double s and DNA b eaks Tamas Va ga and Pe e D. Aplan Gene ics B anch, Cen e o Cance Resea ch, Na ional Cance Ins i u e, Na ional Ins i u es o Heal h, Be hesda, MD 20889, USA Abs ac I has been sugges ed ha in oduc ion o double-s and DNA b eaks in o mammalian ch omosomes can lead o g oss ch omosomal ea angemen s h ough imp ope DNA epai . To s udy his phenomenon, we employed a model sys em in which a double-s and DNA b eak (DSB) can be p oduced in human cells in i o a a p ede e mined loca ion. The ensuing ch omosomal changes lanking he b eakage si e can hen be cloned and cha ac e ized. In his sys em, he ecogni ion si e o he I-SceI endonuclease, whose 18 bp ecogni ion sequence is no no mally ound in he human genome, is placed be ween a s ong cons i u i e p omo e and he He pes simplex i us hymidine kinase (HSV- k) gene, which se es as a nega i e selec able ma ke . We ound ha he mos common mu a ion ollowing abe an DSB epai was an in e s i ial dele ion; hese dele ions ypically showed ea u es o non-homologous end joining (NHEJ), such as mic ohomologies and inse ions o di ec o in e ed epea sequences. We also de ec ed mo e complex ea angemen s, including la ge inse ions om adjacen o dis an genomic egions. The inse ion e en s ha in ol ed dis an genomic egions ypically ep esen ed ansc ibed sequences, and included bo h L1 LINE elemen s and sequences known o be in ol ed in genomic ea angemen s. This ype o abe an epai could po en ially lead o gene inac i a ion ia dele ion o coding o egula o y sequences, o p oduc ion o oncogenic usion genes ia inse ion o coding sequences. Keywo ds ch omosomal ea angemen ; double s and DNA b eak; non-homologous end joining; inse ion; I- SceI 1. In oduc ion DNA double s and b eaks (DSBs) appea in he genome o human cells on a egula basis [1]. These lesions can lead o cell dea h i le un epai ed, he e o e i is c ucial ha cells epai b oken ch omosomes. Double s and b eak epai can occu ia ei he homology-dependen o non-homologous mechanisms [2,3]. Two o ms o homology-di ec ed epai ha e been iden i ied. An in ac copy o he sequence can be copied in o he DSB om a sis e ch oma id o homologous ch omosome [4]; his epai p ocess is conse a i e. The e also exis s a non- conse a i e ype o homologous ecombina ion epai known as single-s and annealing (SSA), in which DSBs a e p ocessed o single-s anded ails, and epai occu s ia annealing o he single-s anded ail a a nea by di ec epea sequence [5]. The second majo o m o DSB epai is non-homologous end joining (NHEJ), which does no equi e he p esence o a homologous dono sequence. Sho segmen s (1–4 nucleo ides) o o e lapping nucleo ides (mic ohomologies) a e o en p esen a he NHEJ epai si e [6], Add ess co espondence o: D . Pe e Aplan, NCI/NIH/Gene ics B anch, Na ional Na al Medical Cen e , Building 8 Room 5101, 8901 Rock ille Pike, Be hesda, MD 20889-5105, Tel: 301-435-5005, FAX: 301-496-0047, Email: [email p o ec ed]. NIH Public Access Au ho Manusc ip DNA Repai (Ams ). Au ho manusc ip ; a ailable in PMC 2005 Sep embe 29. Published in inal edi ed o m as: DNA Repai (Ams ). 2005 Augus 15; 4(9): 1038–1046. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip sugges ing ha a sho DNA o e lap may be impo an o he eliga ion o b oken DNA ends. Since NHEJ is o en accompanied by dele ion o addi ion o nucleo ides, NHEJ po en ially comp omises gene ic in o ma ion [7,8]. Despi e he p opensi y o in oducing small gene ic changes upon epai , NHEJ is hough o ac as a ca e ake o genome, since cells de icien in NHEJ o en display gene ic ins abili y [9,10]. To explain he p esence o DNA end joining seen in sys ems lacking componen s o he NHEJ pa hway, an al e na i e, highly e o -p one NHEJ pa hway has been p oposed [11]. I has been sugges ed ha imp ope DSB epai can lead o g oss ch omosomal ea angemen s (GCRs), such as ansloca ions, dele ions and in e sions [12]. Indeed, many o he ch omosomal ansloca ion b eakpoin s seen in pa ien s wi h hema opoie ic malignancies display hallma ks o NHEJ, sugges ing ha e oneous DNA epai may play a ole in malignan ans o ma ion [10,13,14]. Se e al model sys ems ha e been employed o in es iga e he link be ween DSB epai and GCRs. Following inse ion o a complemen a y pai o mu an neomycin phospho ans e ase genes (neo ) in o mouse ES cells [15,16], p oduc ion o a speci ic DSB wi hin he mu an neo genes led o econs uc ion o a unc ional neo gene ia homologous ecombina ion. Al hough his app oach allows one o de ec homologous ecombina ion, i does no iden i y epai p ocesses such as NHEJ, which do no p oduce a unc ional neo gene. Loss-o - unc ion epo e sys ems ha e also been used o gain in o ma ion abou mu a ions accompanying DSB epai in oden cells. In hese s udies, a DSB was in oduced ei he in o he coding egion o a He pes simplex i us hymidine kinase (Tk) gene [17] o an in on o he phospho ibosyl ans e ase (APRT) gene [3,18]. Nonconse a i e epai p ocesses ab oga ed he ac i i y o he epo e genes and enabled cells o su i e in selec ion media. In ano he expe imen al sys em [19], a DSB was in oduced in an in on o he endogenous Tk gene o a human lymphoblas oid cell line, and abe an DNA epai e en s we e eco e ed by selec ing o cells ha had los Tk ac i i y. Howe e , his assay a o s he eco e y o la ge scale changes, such as long dele ions, and does no p o ide a means o assessing he equency o sho dele ions. We epo he e a loss-o - unc ion epo e assay in which he ou come o he non-homologous epai o one single DSB can be s udied in human cells. This app oach p o ides an oppo uni y o analyze bo h he sho dele ions and he la ge scale changes ha accompany he epai o a single, induced DSB. 2. Resul s 2.1 Design o he model sys em We modi ied an exis ing expe imen al sys em (13, 14) o in oduce a single DSB and iden i y cells ha ha e sus ained mu a ions accompanying he epai p ocess. We gene a ed a plasmid (pEF1αTk) ha exp esses he He pes simplex i us hymidine kinase (HSV- k) gene unde he con ol o he EF1α p omo e ( ig. 1A), wi h he ecogni ion sequence o he es ic ion endonuclease I-SceI in e posed be ween he EF1α p omo e and he HSV- k gene. The I-SceI endonuclease, de i ed om he yeas Sacha omyces ce e isiae, has an 18 bp ecogni ion sequence which is no p esen in he human genome [20]. The pEF1αTk ec o can be in oduced in o mammalian cells, and hose cells ha ha e s ably in eg a ed he cons uc can be selec ed wi h G418. These cells will exp ess HSV- k and will he e o e be sensi i e o gancyclo i (GCV). We elec opo a ed he human monocy ic cell line U937 [21] wi h he pEF1αTk plasmid and isola ed clone F5, which con ains a single copy o he cons uc ( ig. 1). In e se PCR demons a ed ha he cons uc had been in eg a ed in o ch omosome 7 be ween nucleo ides Va ga and Aplan Page 2 DNA Repai (Ams ). Au ho manusc ip ; a ailable in PMC 2005 Sep embe 29. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip 112329914-112329932 (all ch omosomal coo dina es e e o he July 2003 eeze, based on Human Genome Build 34 om NCBI). The F5 cells we e hen ans ec ed wi h I-SceI exp ession ec o s. Fo some expe imen s, we used an episomal ec o (pCEP4-IsceI). As opposed o con en ional plasmid ec o s, in which only a ac ion o cells ha a e success ully ans ec ed will in eg a e he plasmid ec o and supply ongoing an ibio ic (eg., hyg omycin) esis ance, episomal ec o s can supply ongoing an ibio ic esis ance in he absence o in eg a ion. This allows selec ion o ela i ely la ge numbe s o unique, success ully ans ec ed clones, selec ed solely on he basis o exp ession o he an ibio ic esis ance ma ke . In o he expe imen s, we used p e iously desc ibed non- episomal I-SceI ec o s (pPGK3XnlsI-SceI and pCBASce). Since I-SceI clea es DNA be ween he EF1α p omo e and he HSV- k gene; his clea age migh esul in ch omosomal abe a ions due o an impe ec epai ha sepa a es he EF1α p omo e om HSV- k coding sequences. A combina ion o nega i e (GCV) and posi i e selec ion (G418) enabled us o eco e cells ha sus ained di e en o ms o genomic ea angemen s ( ig. 1B). 2.2 Exp ession o I-SceI in F5 cells F5 cells we e ans ec ed wi h he episomal exp ession ec o pCEP4-IsceI, which con ains a hyg omycin esis ance casse e (hyg o ) and exp esses he I-SceI enzyme unde he con ol o a CMV p omo e . Following ans ec ion, he cells we e g own in bulk, wi hou any selec ion o 5 days, and we e subsequen ly pla ed in o 96 well pla es a dilu ions o 3, 9, o 27 cells pe well. These pla es we e hen selec ed wi h hyg omycin alone ( o iden i y cells ha had been success ully ans ec ed) o hyg omycin and GCV. Al hough app oxima ely 2.5% o he ans ec ed cells we e hyg omycin esis an , only 0.4 % we e bo h hyg omycin and GCV esis an , indica ing ha app oxima ely 84 % o he cells (2.5 − 0.4 2.5 ) ha we e success ully ans ec ed wi h pCEP4-IsceI con inued o exp ess a unc ional HSV- k gene. This inding sugges ed ha minimal o no gene ic changes had occu ed in mos o he cells ha we e ans ec ed wi h he I-SceI exp ession ec o . Using a se ies o PCR assays ( ig. 1A), we ound ha hal o he 22 clones selec ed on he basis o hyg omycin esis ance alone ( e med 22-1, 2, 3 e c.) showed al e ed DNA sequences lanking he I-SceI clea age si e, indica ing ha he DNA had been clea ed and impe ec ly epai ed. Two clones had dele ions o EF1α o HSV- k, as e idenced by he inabili y o ampli y any o he segmen s indica ed in ig. 1A. Sequence analysis o PCR ampli ied agmen s demons a ed ha 4 clones displayed a sho dele ion (9, 7, 3 and 1 nucleo ides), 1 clone displayed a la ge dele ion (55 bp), and 4 clones had mo e complex ea angemen s due o DNA inse ions a he I-SceI si e. In hese 4 clones he nucleo ide sequence o he inse ed DNA segmen s ma ched sequences om ei he he ans ec ed episomal ec o (1), o om dis an egions o he genome (3) ( able. 1). As indica ed abo e, hal o he 22 clones ha had been success ully ans ec ed wi h he episomal I-SceI exp ession ec o con ained “ge mline” PCR agmen s and did no show clea e idence o sequence changes a he I-SceI si e upon sequence analysis. Howe e , he ch oma og am o se e al clones wi h “ge mline” PCR agmen s exhibi ed an inc ease in backg ound signal beginning p ecisely a he I-SceI si e (no shown), sugges ing ha hese ch oma og ams ep esen ed mixed clones, some o which had mu a ions a he I-SceI si e. These esul s sugges ha I-SceI media ed DSB occu ed in a leas hal o he cells ha we e success ully ans ec ed. 2.3 Imp o ed eco e y o cells wi h mo e ex ensi e ea angemen s As shown abo e, 50% o cells ans ec ed wi h he I-SceI exp ession ec o exhibi ed mu a ions a ound he clea age si e. Gi en ha malignan cells o en display GCRs, we wan ed o Va ga and Aplan Page 3 DNA Repai (Ams ). Au ho manusc ip ; a ailable in PMC 2005 Sep embe 29. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip de e mine i we could inc ease he eco e y o clones con aining GCRs by modi ying ou de ec ion scheme. We u ilized he nega i e selec ion p o ided by exp ession o HSV- k, and ans ec ed cells wi h pCBASce, a non-episomal I-SceI exp ession ec o . Following ans ec ion wi h he pCBASce ec o , app oxima ely 0.2 % o F5 cells we e GCV esis an , and had e ained G418 esis ance, indica ing ha GCV esis ance was no acqui ed h ough loss o he en i e in eg a ed ch omosome. We an icipa ed ha clones isola ed in his ashion migh ha e mo e ex ensi e mu a ions, as clones wi h sho (i.e., <20–30 bp) dele ions would con inue o exp ess Hs - k, and he e o e be elimina ed by GCV selec ion. A e 3 weeks o selec ion in GCV and G418, he ans ec ed cells we e single-cell cloned o ob ain pu e colonies; 58 GCV and G418 esis an colonies ( e med as 6-1,2,3 e c.) we e assayed o e idence o GCRs. 2.4 Sc eening s a egy o analyzing GCRs gene a ed by one DSB We used Sou he n blo hyb idiza ion o dis inguish clones which sus ained an in e s i ial dele ion om hose ha sus ained mo e complex ea angemen s, such as la ge inse ions o ansloca ions. To dis inguish be ween hese wo e en s, we hyb idized HindIII-diges ed genomic DNA om GCV /G418 clones wi h a ch omosome 7 speci ic p obe ha hyb idized ups eam o he EF1α in eg a ion si e ( e med p obe TV7), and a neo gene speci ic p obe ( ig. 1A). Since HindIII does no clea e wi hin he pEF1αTk cons uc , bo h p obes should iden i y iden ical genomic DNA agmen s i he clone had sus ained an in e s i ial dele ion. Howe e , i a ansloca ion o o he GCR sepa a ed he 3′ and 5′ po ion o he in eg a ed cons uc , he wo p obes should hyb idize o dis inc genomic DNA agmen s. Fig. 2 shows ha mos o he clones sus ained an in e s i ial dele ion as e idenced by agmen s o iden ical sizes in he duplica e hyb idiza ions. In some cases, one hyb idiza ion signal was los , sugges ing ha a la ge dele ion, p esumably igge ed by I-SceI clea age be ween he hyb idiza ion a ge sequences, ex ended beyond he a ge sequence o he p obe. To in es iga e he mechanism o abe an epai ollowing I-SceI clea age, we analyzed he nucleo ide sequence lanking he I-SceI clea age si e in GCV esis an clones. We used a se ies o con en ional and in e se PCR eac ions o assess di e en egions o he in eg a ed pEF1αTk ec o ( ig. 1A). O he 58 clones isola ed, ou we e shown o be a mix u e o wo independen clones, ou clones we e ep esen ed wice, and one clone was ep esen ed h ee imes. Six clones showed no mu a ion o he egion su ounding he I-SceI si e. O he 50 independen clones which exhibi ed mu a ions a ound he DSB si e, 48 we e shown by Sou he n blo and/o PCR, o ha e sus ained an in e s i ial dele ion o DNA lanking he I- SceI si e. Two clones los he en i e EF1α p omo e and a leas 1 kb o adjacen ch omosome se en egion, bu e ained he neo gene downs eam o he I-SceI si e. To e i y ha he dele ions we e gene a ed as a esul o e oneous DSB epai , as well as iden i y any addi ional mu a ions (such as inse ions, in e sions, duplica ions) accompanying he dele ions, we de e mined he b eakpoin sequences o 38 independen mu an clones. In one case, due o a la ge inse ion o ec o sequences a he I-SceI si e, we did no de e mine bo h ends o he dele ed segmen . Al oge he , he nucleo ide sequence changes om 37 mu an clones we e comple ely cha ac e ized. As shown in ig. 3., I-SceI media ed clea age o he in eg a ed pEF1αTk cons uc occu s a posi ion 1352, be ween he 3′ end o he EF1α p omo e (1304) and he i s ATG codon o he HSV- k gene (1359). Fine analysis o he dele ions demons a ed ha he pEF1αTk cons uc sus ained asymme ic dele ions ollowing I-SceI clea age. Ups eam o he I-SceI clea age si e, owa ds he EF1α p omo e , he mean size o he dele ed segmen was 442 bp. Howe e , hese dele ions a e une enly dis ibu ed. A majo i y o he clones (19) sus ained small dele ions, sho e han 10 bp, whe eas nine clones had dele ions o 11–200 bp and nine clones had much longe dele ions ha a e aged 1752 bp. The leng h o he dele ed segmen s downs eam o he Va ga and Aplan Page 4 DNA Repai (Ams ). Au ho manusc ip ; a ailable in PMC 2005 Sep embe 29. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip I-SceI induced DSB, ex ending in o he HSV- k gene, showed a mo e uni o m dis ibu ion. In con as o he dele ions in ol ing EF1α, no dele ion was sho e han 100 bp o longe han 1700 bp. Six een clones showed sho egions o mic ohomology (1–6 bp) a he b eakpoin junc ion ( ig. 3), and a sho palind ome sequence (4–8 bp) was ound a he b eakpoin o 14 clones, nine o which also showed mic ohomologies. In addi ion o he dele ions desc ibed abo e, some clones con ained addi ional ea angemen s in he egion clea ed by I-SceI. The p edominan ype o addi ional e en was a nucleo ide inse ion; 18 clones con ained an inse ion in he gap c ea ed by he dele ion. These clones could be di ided in o se e al g oups acco ding o he cha ac e is ics o he inse ed DNA segmen . Six clones showed inse ions o only a ew nucleo ides, and i was no possible o de e mine whe he hese inse ions we e empla ed o no . A g oup o nine clones had sho inse ions o 11–24 bp. In hese clones, he co e o he inse ed nucleo ides was copied om ei he side o he b eakpoin junc ion, c ea ing ei he a di ec (8 clones) o in e ed (1 clone) epea wi h he lanking sequences ( ig. 3 ). Two clones con ained complex ea angemen s wi h inse ions de i ed om he pEF1αTK ec o . A single clone con ained I-SceI exp ession ec o sequences inse ed a he b eakpoin junc ion. 2.5 Modi ica ion o he expe imen al sys em Following cha ac e iza ion o he clones desc ibed abo e, we modi ied he expe imen al sys em in an e o o iden i y GCV esis an clones ha exhibi ed GCRs. We used di e en exp ession ec o s (pCEP4-IsceI s. pCBASce), a ied he ans ec ion condi ions (sho e pe iod o bulk cul u e) and, in some expe imen s omi ed G418 om he selec ion medium. We eco e ed 76 addi ional clones ( e med 5-1, 2 3, e c. and 29-1, 2, 3, e c.) om hese expe imen s, and de ec ed p ima ily in e s i ial dele ions simila o hose desc ibed abo e. We again ound unequal dele ions a ound he I-SceI clea age si e, he p esence o sho di ec epea sequences, and a e inse ion e en s. Fi e clones exhibi ed la ge inse ions ha ma ched gene segmen s om dis an ch omosomal egions; wo addi ional clones sus ained complex ea angemen s in ol ing DNA inse ions om ch omosome 7 egions adjacen o he pEF1αTk in eg a ion si e ( able 1). The Sou he n blo hyb idiza ion pa e n o clone 5–22 could no be explained by a simple dele ion ( ig. 2) and was consis en wi h a ch omosomal ansloca ion, as unique HindIII agmen s hyb idized o p obes Neo and TV7. PCR eac ions e ealed ha his clone had e ained he EF1α p omo e , and sus ained a sho dele ion o HSV- k sequences. In e se PCR eac ions showed ha EF1α p omo e sequences immedia ely adjacen o he I-SceI si e had been joined o a sequence o a leas 982 bp de i ed om ch omosome 15 α sa elli e cen ome ic epea sequences, sugges ing ha his clone had sus ained a ch omosomal ansloca ion. O he clones ha we e comple ely cha ac e ized, 56% had in e s i ial dele ions, 24 % had dele ions accompanied by small (<30 bp) inse ions, and 20% dele ions accompanied by la ge (>30 bp) inse ions. 3. Discussion Impe ec epai o DSBs is hough o play an impo an ole in gene a ing GCRs associa ed wi h malignan ans o ma ion. To examine how DSBs migh lead o GCRs, we in es iga ed he ou come o DSB epai in he human monocy ic cell line U937. To ob ain a de ailed analysis o epai e en s ollowing a single, speci ic DSB, we induced DSBs wi h he I-SceI endonuclease. The simples mechanism o epai o an I-SceI induced DSB is a pe ec eliga ion e en , which would ec ea e an I-SceI ecogni ion sequence ha would emain suscep ible o I-SceI clea age. Howe e , i he clea ed ecogni ion sequence we e epai ed impe ec ly, no u he I-SceI clea age would be possible, since he I-SceI ecogni ion si e Va ga and Aplan Page 5 DNA Repai (Ams ). Au ho manusc ip ; a ailable in PMC 2005 Sep embe 29. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip would ha e been des oyed. We ound ha hal (11/22) o he clones success ully ans ec ed wi h an episomal I-SceI exp ession ec o showed clea e idence o I-SceI clea age, sugges ing ha I-SceI media ed clea age o genomic a ge s was easonably e icien . To en ich o clones which may ha e sus ained mo e ex ensi e mu a ions (such as in e sions, inse ions, ansloca ions, o la ge dele ions) ollowing epai o an I-SceI-induced DSB, we made use o he nega i e selec ion p o ided by he HSV- k gene. Simila o p io epo s ha used oden cells[3,18], we ound ha he mos common mu a ion in he human U937 sub- clones ha su i ed he GCV selec ion was an in e s i ial dele ion, o en accompanied by addi ional e en s (inse ions, in e sions, duplica ions). Many o he clones showed hallma ks o NHEJ, such as inse ion o di ec o in e ed epea s o mic ohomologies a he b eakpoin s. We eco e ed a o al o ou een clones ha con ained la ge segmen s o o eign DNA inse ed a he I-SceI si e, and a single clone (clone 5–22) wi h a ea angemen consis en wi h a ch omosomal ansloca ion. Fou clones con ained sequences om he I-SceI exp ession ec o s, simila o p e ious epo s [22], and wo clones con ained ch omosome se en sequences de i ed om egions nea he pEF1αTK in eg a ion si e. O in e es , eigh clones con ained DNA inse ions ha ma ched sequences om dis an egions o he genome ( able 1); se en o he eigh inse ion e en s in ol ed known ansc ibed egions, sugges ing wo possibili ies as o he sou ce o hese sequences. The inse ed segmen s could ha e been de i ed om double-s anded DNA agmen s, o DNA/RNA he e o-duplexes, as sugges ed o he ec o cap u e e en s. In his scena io, RNA molecules may ha e se ed as a sou ce o DNA “pa ches” ia e e se ansc ip ion, as seen in endonuclease-independen LINE L1 e o ansposi ions [23–25]. A second possibili y is ha dis an genomic egions se ed as a empla e o hese sequences. These segmen s could ha e been copied in o he gap ini ia ed by I-SceI clea age o , al e na i ely, could ha e been excised om he genome and inse ed a he b eakpoin . Two o hese eigh inse ions we e likely he esul o a h ee-way liga ion, since he inse ed DNA segmen s we e de i ed om wo dis inc egions o he genome. In bo h ins ances, he e we e no sequence mo i s sha ed be ween he wo inse ed sequences. Addi ionally, i is in e es ing o no e ha a leas one inse ed DNA segmen (D4Z4) is associa ed wi h a known in- i o genomic ins abili y phenomenon. This sequence is ound in andem epea on 4q35.2, a egion ha is hough o ha e been de i ed om an in e ch omosomal exchange wi h he nea ly iden ical sub elome ic 10q26 [26], and o en pa icipa es in in ach omosomal ecombina ions ha lead o epea con ac ions [27]. Al hough he mu a ions in his s udy ook place wi hin an a i icial DNA cons uc , his ype o aul y epai could easily lead o inac i a ion o umo supp esso genes o he p oduc ion o oncogenic usion genes. Fo ins ance, a sho in e s i ial dele ion o he BRCA1 gene, which shows nucleo ide sequence changes simila o hose desc ibed he e, leads o inac i a ion o he gene [28], and inse ions o AF5Q31 sequences wi hin he MLL locus can lead o oncogenic MLL-AF5Q31 usions [29]. The expe imen al sys em desc ibed he e p o ides a pla o m use ul o elabo a ing he mechanisms by which epai o DSBs by NHEJ may lead o ine and g oss ch omosomal abe a ions. 4. Ma e ials and Me hods 4.1 Vec o cons uc ion pEF1α Tk was gene a ed by diges ing pEF1αNUPHOX (a pcDNA3 based ec o con aining a human EF1α p omo e segmen ) wi h BamHI and No I o emo e he NUPHOX inse and o p o ide a ec o wi h an EF1α p omo e . The HSV- k gene was ampli ied om he pTkNeo ec o using p ime s Tk e e se (AATGCGGCCGCAGTTAGCCTCCCCCATCTC) and TkFwBamHI (AAGGGATCCTAGGGATAACAGGGTAATGGCTTCGTACCCCTGCCAT), wi h he Va ga and Aplan Page 6 DNA Repai (Ams ). Au ho manusc ip ; a ailable in PMC 2005 Sep embe 29. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip la e p ime con aining an I-SceI ecogni ion si e immedia ely 5′ o he HSV- k gene. The PCR p oduc was diges ed wi h No I and BamHI, pu i ied, and liga ed in o he EF1α ec o backbone. The pCEP4-IsceI plasmid was gene a ed by cloning he 3Xnls-I-SceI agmen om he pPGK3XnlsI-SceI ec o [30], in o he HindIII (blun ed) and BamHI si es o he episomal exp ession ec o pCEP4 (In i ogen, Ca lsbad, CA). 4.2 T ans ec ions The U937 human monocy ic leukemia cell line (main ained in RPMI 1640, supplemen ed wi h 10% e al cal se um, 2 mM L-glu amine, 100 U/ml penicillin and 100 μgml S ep omycin) was elec opo a ed wi h ScaI linea ized pEF1αTk using he Gene Pulse II elec opo a ion sys em (Bio-Rad Labo a o ies, Inc., He cules, CA), wi h pulses o 0.4 kV and 975 μF, and s able G418R clones we e isola ed. T ans ec ions wi h ec o s pCEP4-IsceI, pPGK3XnlsI-SceI o pCBASce [4]we e pe o med using DMRIE-C eagen (In i ogen, Ca lsbad, CA). . On he ou h o i h day ollowing ans ec ion, cells we e expanded in bulk, o we e pla ed in o 96 o 24 well pla es a 3 cells/ well, 9 cells/well and 27 cells/well, o a 250 cells/well, 750 cells/well, 2250 cells/well and 6750 cells/well densi y, espec i ely. Selec ion agen s GCV ( om In i oGen, San Diego, CA, CAS n°: 82410-32-0), G418 ( om In i ogen, Ca lsbad, CA, CAS n°: 1405-41-0) and Hyg omycin B ( om In i ogen, Ca lsbad, CA, CAS n°: 31282-04-9) we e added o he wells. The ollowing selec ion condi ions we e applied: 40 μ M GCV, o 40 μM GCV and 400 μg/ ml G418 in ans ec ions done wi h pCBASce; 200 μg/ml Hyg omycin, 200 μg/ml Hyg omycin and 20 μM GCV, o 40 μM GCV in ans ec ions done wi h pCEP4-IsceI. 4.3 Con en ional and in e se PCR The PCR eac ions indica ed in ig. 1A we e ca ied ou using 100 ng o genomic DNA as empla e in he p esence o 0.2 μM p ime s wi h ei he Taka a LA Taq (PCR #2, 3) (Taka a Mi us Bio, Madison, WI) o wi h PCR Supe mix ( eac ions 1, 4 and 5) (In i ogen, Ca lsbad, CA), acco ding o he manu a u e s’ ecommenda ions. Fo in e se PCR eac ions, 1 μg o genomic DNA was diges ed wi h ei he C oI, HindIII o AluI in 100 μl. The samples we e ex ac ed wi h phenol-chlo o o m, e hanol p ecipi a ed, and sel -liga ed in 250 μl using T4 Ligase (P omega, Madison, WI). F agmen s con aining EF1α sequences we e ampli ied in Supe mix (In i ogen) using p ime s designed o a nes ed PCR. De ails o p ime sequences and ampli ica ion p o ocols a e a ailable upon eques . 4.4 Sou he n Blo analysis Duplica e samples con aining 10 μg o genomic DNA om GCV esis an clones we e diges ed wi h HindIII, size- ac ioned on 0.8% aga ose gels, and ans e ed o a ni ocellulose memb ane and hyb idized o a ch omosome 7 speci ic agmen (p obe TV7; nuc 112330867-112331229) o a NcoI-SmaI G418 agmen isola ed om pPRC/CMV (In i ogen, Ca lsbad, CA). The agmen s we e labeled wi h 32P using Ready-To-Go DNA labelling beads (Ame sham Biosciences) and hyb idized as p e iously desc ibed [31]. Final washing condi ions we e 0.1% SDS/0.1x SSC a 52°C o bo h p obes. 4.5 Sequence Analysis Nucleo ide sequences we e de e mined using an Applied Biosys ems 3730 and compa ed o he human genome assembly (Uni e si y o San a C uz July 2003 eeze, based on NCBI Human Genome Build 34) [32]. Va ga and Aplan Page 7 DNA Repai (Ams ). Au ho manusc ip ; a ailable in PMC 2005 Sep embe 29. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip Acknowledgemen s The au ho s hank Ma ia Jasin o p o iding he I-SceI exp ession ec o , pCBASce and G eg Donoho o p o iding he I-SceI exp ession ec o , pPGK3XnlsI-SceI. We also hank Ilan Ki sch and Michael Kuehl o help ul and s imula ing discussions. Re e ences 1. Vilenchik MM, Knudson AG. Endogenous DNA double-s and b eaks: p oduc ion, ideli y o epai , and induc ion o cance . P oc Na l Acad Sci U S A 2003;100:12871–12876. [PubMed: 14566050] 2. Liang F, Han M, Romanienko PJ, Jasin M. Homology-di ec ed epai is a majo double-s and b eak epai pa hway in mammalian cells. P oc Na l Acad Sci U S A 1998;95:5172–5177. [PubMed: 9560248] 3. Sa gen RG, B enneman MA, Wilson JH. Repai o si e-speci ic double-s and b eaks in a mammalian ch omosome by homologous and illegi ima e ecombina ion. Mol Cell Biol 1997;17:267–277. [PubMed: 8972207] 4. Richa dson C, Moynahan ME, Jasin M. Double-s and b eak epai by in e ch omosomal ecombina ion: supp ession o ch omosomal ansloca ions. Genes De 1998;12:3831–3842. [PubMed: 9869637] 5. Lin FL, Spe le K, S e nbe g N. In e molecula ecombina ion be ween DNAs in oduced in o mouse L cells is media ed by a nonconse a i e pa hway ha leads o c osso e p oduc s. Mol Cell Biol 1990;10:103–112. [PubMed: 2294396] 6. Ro h DB, Wilson JH. Nonhomologous ecombina ion in mammalian cells: ole o sho sequence homologies in he joining eac ion. Mol Cell Biol 1986;6:4295–4304. [PubMed: 3025650] 7. Liebe MR, Ma Y, Pannicke U, Schwa z K. Mechanism and egula ion o human non-homologous DNA end-joining. Na Re Mol Cell Biol 2003;4:712–720. [PubMed: 14506474] 8. Gu H, Fo s e I, Rajewsky K. Sequence homologies, N sequence inse ion and JH gene u iliza ion in VHDJH joining: implica ions o he joining mechanism and he on ogene ic iming o Ly1 B cell and B-CLL p ogeni o gene a ion. Embo J 1990;9:2133–2140. [PubMed: 2113468] 9. Smi h J, Riballo E, Kysela B, Baldey on C, Manolis K, Masson C, Liebe MR, Papadopoulo D, Jeggo P. Impac o DNA ligase IV on he ideli y o end joining in human cells. Nucleic Acids Res 2003;31:2157–2167. [PubMed: 12682366] 10. Fe guson DO, Al FW. DNA double s and b eak epai and ch omosomal ansloca ion: lessons om animal models. Oncogene 2001;20:5572–5579. [PubMed: 11607810] 11. Lee GS, Neidi ch MB, Salus SS, Ro h DB. RAG p o eins shephe d double-s and b eaks o a speci ic pa hway, supp essing e o -p one epai , bu RAG nicking ini ia es homologous ecombina ion. Cell 2004;117:171–184. [PubMed: 15084256] 12. Mo gan WF, Co co an J, Ha mann A, Kaplan MI, Limoli CL, Ponnaiya B. DNA double-s and b eaks, ch omosomal ea angemen s, and genomic ins abili y. Mu a Res 1998;404:125–128. [PubMed: 9729329] 13. Zucman-Rossi J, Legoix P, Vic o JM, Lopez B, Thomas G. Ch omosome ansloca ion based on illegi ima e ecombina ion in human umo s. P oc Na l Acad Sci U S A 1998;95:11786–11791. [PubMed: 9751743] 14. Ellio B, Jasin M. Double-s and b eaks and ansloca ions in cance . Cell Mol Li e Sci 2002;59:373– 385. [PubMed: 11915950] 15. Richa dson C, Jasin M. F equen ch omosomal ansloca ions induced by DNA double-s and b eaks. Na u e 2000;405:697–700. [PubMed: 10864328] 16. Richa dson C, Jasin M. Coupled homologous and nonhomologous epai o a double-s and b eak p ese es genomic in eg i y in mammalian cells. Mol Cell Biol 2000;20:9068–9075. [PubMed: 11074004] 17. Lukacso ich T, Yang D, Waldman AS. Repai o a speci ic double-s and b eak gene a ed wi hin a mammalian ch omosome by yeas endonuclease I-SceI. Nucleic Acids Res 1994;22:5649–5657. [PubMed: 7838718] 18. Phillips JW, Mo gan WF. Illegi ima e ecombina ion induced by DNA double-s and b eaks in a mammalian ch omosome. Mol Cell Biol 1994;14:5794–5803. [PubMed: 8065314] Va ga and Aplan Page 8 DNA Repai (Ams ). Au ho manusc ip ; a ailable in PMC 2005 Sep embe 29. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip 19. Honma M, Izumi M, Saku aba M, Tadoko o S, Sakamo o H, Wang W, Ya agai F, Hayashi M. Dele ion, ea angemen , and gene con e sion; gene ic consequences o ch omosomal double-s and b eaks in human cells. En i on Mol Mu agen 2003;42:288–298. [PubMed: 14673874] 20. Jasin M. Gene ic manipula ion o genomes wi h a e-cu ing endonucleases. T ends Gene 1996;12:224–228. [PubMed: 8928227] 21. Sunds om C, Nilsson K. Es ablishmen and cha ac e iza ion o a human his iocy ic lymphoma cell line (U-937). In J Cance 1976;17:565–577. [PubMed: 178611] 22. Lin Y, Waldman AS. Cap u e o DNA sequences a double-s and b eaks in mammalian ch omosomes. Gene ics 2001;158:1665–1674. [PubMed: 11514454] 23. Syme DE, Connelly C, Szak ST, Capu o EM, Cos GJ, Pa migiani G, Boeke JD. Human l1 e o ansposi ion is associa ed wi h gene ic ins abili y in i o. Cell 2002;110:327–338. [PubMed: 12176320] 24. Mo ish TA, Gilbe N, Mye s JS, Vincen BJ, S ama o TD, Taccioli GE, Ba ze MA, Mo an JV. DNA epai media ed by endonuclease-independen LINE-1 e o ansposi ion. Na Gene 2002;31:159–165. [PubMed: 12006980] 25. Eickbush TH. Repai by e o ansposi ion. Na Gene 2002;31:126–127. [PubMed: 12006979] 26. an Geel M, Dickson MC, Beck AF, Bolland DJ, F an s RR, an de Maa el SM, de Jong PJ, Hewi JE. Genomic analysis o human ch omosome 10q and 4q elome es sugges s a common o igin. Genomics 2002;79:210–217. [PubMed: 11829491] 27. Lemme s RJ, Van O e eld PG, Sandkuijl LA, V ieling H, Padbe g GW, F an s RR, an de Maa el SM. Mechanism and iming o mi o ic ea angemen s in he sub elome ic D4Z4 epea in ol ed in acioscapulohume al muscula dys ophy. Am J Hum Gene 2004;75:44–53. [PubMed: 15154112] 28. Ha douin A, Baumann J, Roussel G, Quillien V, Dugas C, Be he P. A new mu a ion in he BRCA1 gene (g.5196–5201del6, 5195–5202ins12), a 6 bp dele ion eplaced by he duplica ion o a 12 bp adjacen ups eam in onic sequence. Hum Mu a 2001;17:154. [PubMed: 11180603] 29. De eney R, Che insky DS, Jani-Sai SN, G ossi M, Aplan PD. Inse ion o MLL sequences in o ch omosome band 5q31 esul s in an MLL-AF5Q31 usion and is a a e bu ecu en abno mali y associa ed wi h in an leukemia. Genes Ch omosomes Cance 2003;37:326–331. [PubMed: 12759932] 30. Donoho G, Jasin M, Be g P. Analysis o gene a ge ing and in ach omosomal homologous ecombina ion s imula ed by genomic double-s and b eaks in mouse emb yonic s em cells. Mol Cell Biol 1998;18:4070–4078. [PubMed: 9632791] 31. Aplan PD, Che insky DS, S anulla M, Bu hans WC. Si e-speci ic DNA clea age wi hin he MLL b eakpoin clus e egion induced by opoisome ase II inhibi o s. Blood 1996;87:2649–2658. [PubMed: 8639880] 32. Ken WJ. BLAT-- he BLAST-like alignmen ool. Genome Res 2002;12:656–664. [PubMed: 11932250] Va ga and Aplan Page 9 DNA Repai (Ams ). Au ho manusc ip ; a ailable in PMC 2005 Sep embe 29. NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip NIH-PA Au ho Manusc ip