DNA resection in eukaryotes: deciding how to fix the break

DNA esec ion in euka yo es: deciding how o ix he b eak Pablo Hue as The Wellcome T us and Cance Resea ch UK Gu don Ins i u e, Uni e si y o Camb idge, Camb idge, UK Abs ac DNA double-s and b eaks a e epai ed by di e en mechanisms, including homologous ecombina ion and nonhomologous end-joining. DNA-end esec ion, he i s s ep in ecombina ion, is a key s ep ha con ibu es o he choice o DSB epai . Resec ion, an e olu iona ily conse ed p ocess ha gene a es single-s anded DNA, is linked o checkpoin ac i a ion and is c i ical o su i al. Failu e o egula e and execu e his p ocess esul s in de ec i e ecombina ion and can con ibu e o human disease. He e, I e iew ecen indings on he mechanisms o esec ion in euka yo es, om yeas o e eb a es, p o ide insigh s in o he egula o y s a egies ha con ol i , and highligh he consequences o bo h i s impai men and i s de egula ion. The epai o double-s and b eaks DNA is cons an ly challenged bo h by exogenous agen s such as mu agenic chemicals and adia ion and by endogenously a ising compounds such as eac i e oxygen species1. To minimize he impac o hese h ea s, cells ha e e ol ed a ious DNA epai mechanisms. DNA double-s and b eaks (DSBs) a e he mos cy o oxic o ms o DNA damage. Inaccu a e DSB epai leads o mu a ions and/o g oss ch omosomal ea angemen s (GCRs)1. Mo eo e , he con olled epai o p og ammed DSBs occu s du ing physiological p ocesses such as meiosis o he di e si ica ion o immunoglobulins. The e o e, inhe i ed de ec s in DSB epai genes cause emb yonic le hali y, s e ili y, de elopmen al diso de s, immune de iciencies, and p edisposi ion o neu odegene a i e diseases and cance . The e a e wo majo ways o epai ing DSBs1. Nonhomologous end-joining (NHEJ) liga es oge he he wo DNA ends wi h li le o no p ocessing2 (Fig. 1); i is highly e icien bu p one o gene a ing mu a ions a he si es o joining. Fu he mo e, because he e is no appa en mechanism o ensu e ha he wo ends being joined we e o iginally con iguous, NHEJ can yield GCRs such as in e sions and ansloca ions. The second DSB epai mechanism is a se o pa hways ha use an undamaged homologous DNA sequence as a empla e o accu a e epai , collec i ely known as homologous ecombina ion (HR)3 (Fig. 1). Al hough HR has been p ima ily s udied as a esponse o DSBs, i s p ima y unc ion is p obably o deal wi h s alled o collapsed eplica ion o ks1. HR has been ex ensi ely e iewed3. B ie ly, all HR subpa hways a e ini ia ed by a 5′–3′ deg ada ion o one s and a bo h sides o he b eak, gene a ing s e ches o single-s anded DNA (ssDNA) ha is hen coa ed by he ssDNA binding p o ein complex RPA— he so- called DNA-end esec ion. Th ee o he HR subpa hways use he ssDNA molecule o in ade Co espondence should be add essed o P.H. ([email p o ec ed]).. Published online a h p://www.na u e.com/nsmb/. Rep in s and pe missions in o ma ion is a ailable online a h p://npg.na u e.com/ ep in sandpe missions/. Eu ope PMC Funde s G oup Au ho Manusc ip Na S uc Mol Biol. Au ho manusc ip ; a ailable in PMC 2010 July 01. Published in inal edi ed o m as: Na S uc Mol Biol . 2010 Janua y ; 17(1): 11–16. doi:10.1038/nsmb.1710. Eu ope PMC Funde s Au ho Manusc ip s Eu ope PMC Funde s Au ho Manusc ip s a homologous DNA egion si ua ed elsewhe e in he genome (dono sequence), which is used as a empla e o DNA syn hesis. A e his he h ee mechanisms di e ge (Fig. 1)3. In double-s and-b eak epai (DSBR), he second end is cap u ed and ex ended and hen he newly syn hesized DNA is liga ed o he end o he esec ed s ands o o m wo c uci o m s uc u es known as Holliday junc ions, which can be esol ed by di e en mechanisms3. In b eak-induced eplica ion (BIR), a e one-end in asion, eplica ion simply p oceeds un il he end o he ch omosome. Syn hesis-dependen s and annealing (SDSA) can ollow ei he one-end o wo-end in asion e en s (one-ended in asion shown in Fig. 1); he pa ially eplica ed s ands eanneal and a e liga ed. The ou h subpa hway (single-s and annealing; SSA) is used only when wo homologous egions lank he DSB si e. In his case, he homologous egions a e exposed, and a e annealing and clea age o he DNA o e hang, he ends a e liga ed, esul ing in he dele ion o he in e ening egion. A mechanism ha sha es some gene ic equi emen s wi h bo h NHEJ and SSA—mic ohomology-media ed end-joining; MMEJ—has ecen ly been desc ibed as well (Fig. 1; o e iew see e . 4). A key ea u e o HR-based epai , excep o SSA, is he p ese a ion o he gene ic ma e ial, as he dono sequence is usually he sis e ch oma id. Howe e , when he dono sequence used is no he sis e ch oma id bu ano he homologous egion, HR can yield GCRs such as dele ions, in e sions o loss o he e ozygosi y1. The choice be ween di e en DSBs epai pa hways is igh ly egula ed, and esec ion ep esen s a p ima y egula o y s ep. Resec ion is needed o MMEJ and all HR pa hways3,4, and esec ed DNA dec eases NHEJ e iciency, likely as a esul o poo binding o he NHEJ ac o Ku70–Ku80 o ssDNA5. Indeed, he balance be ween HR, MMEJ and NHEJ has been shown o be con olled by key DNA esec ion ac o s such as Sae2 ( e s. 6,7) and C IP8,9. Fu he mo e, o ma ion o RPA-coa ed ssDNA a e DNA-end esec ion is a c i ical in e media e o checkpoin ac i a ion10 and is key in he swi ch om he ATM-d i en o he ATR-con olled checkpoin 11. Consequen ly, DNA esec ion is a highly complex and egula ed p ocess. Mechanism o esec ion The co e esec ion machine y is conse ed in all kingdoms o li e (Table 1)3,9,12-21. An impo an componen is he M e11 complex, composed o M e11, Rad50, and a hi d p o ein known as X s2 in he budding yeas Saccha omyces ce e isiae and as Nbs1 in mos o he euka yo es6,22-27. M e11 is a nuclease ela ed o bac e ial SbcD, whe eas Rad50 is homologous o bac e ial SbcC. By con as , Nbs1/X s2 is less conse ed and is es ic ed o euka yo es. The C e minus o Nbs1/X s2 possess an in e ac ion mo i o ATM (in budding yeas , Tel1), a p o ein kinase ha con ols DNA damage–induced e en s28,29. The en i e M e11 complex ac s as a single unc ional uni , because loss o any o he h ee subuni s esul s in simila pheno ypes3: hype sensi i i y o DNA-damaging agen s, impai ed HR and de ec i e meiosis. In i o , he M e11 complex shows bo h endonuclease and exonuclease ac i i ies23. Howe e , budding yeas m e11 nuclease mu an s ha e a much milde pheno ype han cells lacking M e11, which ha e only pa ial de ec s in esec ion o endonuclease-induced DSBs24. This e lec s addi ional oles o he M e11 complex in checkpoin ac i a ion o main enance o ch omosome s uc u e3,22,25,26 bu also a gues agains he idea ha M e11 is he main nuclease o esec ion. Mo eo e , M e11 exonuclease ac i i y in i o ope a es in he 3′–5′ di ec ion, opposi e o he di ec ion o esec ion in i o . M e11 is a poo nuclease, bo h endo- and exo-, and hus is unlikely o be esponsible o gene a ing he ex ensi e ssDNA obse ed in i o 23,27. The poo in i o ac i i y o M e11 may e lec he lack o accesso y ac o s. One likely candida e is he budding yeas p o ein Sae2. sae2 dele ion phenocopies he nuclease- Hue as Page 2 Na S uc Mol Biol . Au ho manusc ip ; a ailable in PMC 2010 July 01. Eu ope PMC Funde s Au ho Manusc ip s Eu ope PMC Funde s Au ho Manusc ip s de ec i e m e11 mu an s and a speci ic amily o mu a ions in Rad50 called ad50S ( e . 30): ha is, sae2 Δ s ains a e comple ely de ec i e in p ocessing meio ic DSBs31-33 bu a e mildly sensi i e o DNA damaging agen s and impai DNA-end esec ion only pa ially34. Sae2 is an endonuclease ha coope a es wi h he M e11 complex in he p ocessing o a ious DNA s uc u es35. The cu en model p oposes ha he endonuclease ac i i ies o M e11 and/o Sae2 ini ia e esec ion (Fig. 2). This endonucleoly ic p ocessing will, heo e ically, elease small ssDNA oligonucleo ides. Such oligonucleo ides ha e been obse ed in he p ocessing o meio ic DSBs in yeas 32 and ha e been de ec ed in Xenopus lae is ex ac s36. DNA-end esec ion and HR a e ba ely a ec ed in he absence o Sae2 and no a all in m e11 nuclease–de ec i e mu an s24,34, sugges ing he exis ence o addi ional nucleases. One candida e is he 5′–3′ exonuclease Exo1, which is conse ed om yeas o humans19 and is essen ial o DNA-end p ocessing a uncapped elome es37. Like dele ion o sae2, exo1 dele ion esul s in only mild DNA damage sensi i i y and pa ial impai men o DNA- end esec ion38. sae2 exo1 and exo1 m e11 double mu an s show a syne gis ic dec ease in DNA-end esec ion and g ea e DNA-damage sensi i i y han he single mu an s38. O e exp ession o EXO1 pa ially escues he DNA sensi i i y pheno ype o m e11 mu an s39, sugges ing ha M e11 and Exo1 may unc ion in pa allel pa hways. Su p isingly, m e11 exo1 mu an s show esidual DNA-end esec ion, sugges ing ha a hi d pa hway also exis s39. In bac e ia, he mul i unc ional enzyme RecBCD, which ha bo s helicase and nuclease ac i i ies, does mos o he esec ion, bu in i s absence, he helicase RecQ ac s oge he wi h he nuclease RecJ o esec DNA ends12. Mu a ions in sgs1 , he budding yeas homolog o RecQ, in combina ion wi h exo1 dele ion comple ely abolish long- ange DNA-end esec ion, and only some minimal p ocessing close o he b eak can be de ec ed in such double mu an s1316,21. Residual p ocessing is dependen on Sae2 and M e11 ( e s. 13,16,21). As is he case o bac e ial RecQ, budding yeas Sgs1 wo ks in combina ion wi h a nuclease called Dna2 ( e . 21). Al hough Dna2 has bo h helicase and lap-endonuclease ac i i y40, only he nuclease ac i i y is equi ed o DNA-end esec ion21. The ollowing model o DNA-end esec ion has been p oposed in S. ce e isiae 13,16,21 (Fig. 2). Fi s , he M e11 complex and Sae2 a e esponsible o he ini ial p ocessing h ough hei endonucleoly ic ac i i ies. The esul ing pa ially esec ed DNA is u he p ocessed by he ac ion ei he o Exo1 o o Sgs1 and Dna2. In he absence o Sgs1 and Exo1, he ac i i ies o M e11 and Sae2 a e esponsible o sho - ange p ocessing (Fig. 2). Al hough Sae2 and M e11 a e comple ely essen ial o esec ion du ing meio ic ecombina ion, hei unc ions can be bypassed du ing mi o ic ecombina ion24,34. This di e ence is p obably due o he speci ic na u e o meio ic DSBs, which equi es M e11 and Sae2 o emo e he co alen ly bound nuclease, Spo11, ha c ea es he b eaks32,41. The na u e o his bypass is unknown, bu i p obably in ol es Exo1 and Sgs1, as sae2 exo1 sgs1 mu an s a e unable o esec DNA and o e exp ession o Exo1 pa ially escues m e11 mu an s16,21,39 (Fig. 2). Resec ion in e eb a es Fo a long ime, he only componen o he DNA-end esec ion machine y known in highe euka yo es was he M e11 complex11,22,26. Recen ly, unc ional coun e pa s o Sae2 ha e been ound in se e al o ganisms9,15,17,18,20 (Table 1). Human C IP, as well as ission yeas Schizosaccha omyces pombe C p1, physically in e ac wi h he M e11 complex and ha e a majo ole in ssDNA o ma ion a he si e o DSBs9,15,18, bu i is s ill unknown whe he hey unc ion as endonucleases like Sae2. In i o , human C IP oge he wi h M e11 showed an inc eased nuclease ac i i y compa ed wi h M e11 alone18, bu whe he his ac i i y elies on M e11, C IP o bo h emains o be es ablished. C IP down egula ion Hue as Page 3 Na S uc Mol Biol . Au ho manusc ip ; a ailable in PMC 2010 July 01. Eu ope PMC Funde s Au ho Manusc ip s Eu ope PMC Funde s Au ho Manusc ip s comple ely abolishes ssDNA o ma ion, as measu ed om RPA ocus o ma ion8,9,13,18, in con as wi h S. ce e isiae sae2 ( e s. 6,34,38) o S. pombe c p1 ( e . 15). Whe he his is due o di e ences in he echniques used o e lec s a ue impossibili y o esec ion in he absence o C IP is s ill an open ques ion. Al hough in i o M e11, Rad50 and C IP a e su icien o ca alyze he nuclease ac i i y18, hey equi e addi ional ac o s in i o . Apa om Nbs1, which is necessa y o ec ui men o he M e11 complex o si es o b eaks41, p ope DNA esec ion in e eb a es equi es he ac ion o speci ic ac o s such as he umo supp esso BRCA1 ( e s. 9,42). BRCA1 is an ubiqui in ligase ha physically in e ac s wi h and polyubiqui ina es C IP43. In e ac ion o C IP and BRCA1 is con olled by phospho yla ion and is essen ial o C IP ec ui men o si es o DNA damage43 and p ope DNA esec ion9,42. Howe e , he ole o BRCA1-media ed ubiqui ina ion in DNA-end esec ion emains o be de e mined. Despi e he s ong e ec o C IP down egula ion, bo h he Exo1 and Sgs1 esec ion pa hways a e unc ional in highe euka yo es13. Al hough in humans he e is only one o holog o Exo1, he e a e i e homologs o RecQ and Sgs1 (Table 1), and a leas one o hese (BLM) is in ol ed in DNA esec ion13. As in yeas , he BLM pa hway appea s o be pa allel and independen o Exo1, as he simul aneous down egula ion o Exo1 and BLM se e ely impai ed ssDNA o ma ion13. Howe e , in i o BLM in e ac s wi h Exo1 and s imula es i s ac i i y44, a guing ha BLM and Exo1 migh unc ion in he same pa hway. Fu u e wo k will be equi ed o cla i y hese disc epancies be ween he in i o and in i o da a. The ole o e eb a e Dna2 is also unclea . Human Dna2 is an endonuclease45 (Table 1), and al hough i is p ima ily loca ed in he mi ochond ia46, i is also p esen in he nucleus46. Xenopus Dna2 possesses he majo ac i i y esponsible o 5′–3′ DNA p ocessing in ex ac s47. The helicase-nuclease andem o DNA esec ion seems is a gene al heme o DNA-end p ocessing machine y6. In addi ion o Sgs1 and Dna2, RecBCD and RecQJ, in he a chean Py ococcus u iosus SbcCD-media ed esec ion is s imula ed by he ac ion o he He A- Nu A helicase-nuclease pai 48. The e o e, i is possible ha in u u e o he helicases will be ound o be in ol ed in esec ion. S ong candida es a e he addi ional membe s o he RECQ amily (Table 1). In ac , human RECQ5 has been shown o be ec ui ed o si es o DNA damage by he M e11 complex and has been epo ed o inhibi he 3′–5′ nuclease ac i i y o M e11 ( e . 49). Regula ion o esec ion DNA-end esec ion has a majo ole in egula ing he balance be ween HR and NHEJ4,6,8,9 and is a key modula o o checkpoin ac i a ion10. The e o e, i is highly egula ed and esponds o many di e en cellula signals. An o e iew o he mul iple laye s o egula ion o DNA-end esec ion is shown in Figu e 3 and explained in mo e de ail below. DNA-end esec ion du ing he cell cycle HR is a highly accu a e epai p ocess when he sis e ch oma id is eadily a ailable and held in close p oximi y a e DNA eplica ion ei he in S o G2. The e o e, DNA-end esec ion and HR a e almos comple ely con ined o S and G2 ( e s. 11,50,51). Al hough om now on we will dis inguish me ely_be ween he G1 (li le o no esec ion) and S/G2 (high esec ion) phases, esec ion occu s as e in S han in G2 ( e . 52). The mechanism unde lying his di e ence emains unclea . One emp ing idea is ha he DNA eplica ion machine y i sel can ec ui he esec ion machine y. Acco dingly, C IP has been shown o be ec ui ed o ac i e eplica ion si es ia an in e ac ion wi h he eplica ion ac o PCNA53. As a esul o his di e ence be ween G2- and S-phase esec ion, i is also di icul o Hue as Page 4 Na S uc Mol Biol . Au ho manusc ip ; a ailable in PMC 2010 July 01. Eu ope PMC Funde s Au ho Manusc ip s Eu ope PMC Funde s Au ho Manusc ip s compa e esul s ob ained wi h cycling e sus G2-a es ed cells, and om his poin on we no e when a es ed cells we e used. DNA esec ion akes place only when cyclin-dependen kinases (CDKs), mas e egula o s o cell cycle p og ession, a e ac i e (S/G2)6,11,50,51. So a Rad9 and Sae2 ha e been implica ed in he CDK-dependen egula ion o DNA esec ion in S. ce e isiae . Dele ion o he checkpoin p o ein Rad9 inc eases DNA-end esec ion e en when CDKs a e no ac i e (G1)54. Rad9, a la ge ch oma in-binding p o ein, could pose a physical obs acle o p ocessi e DNA esec ion. Indeed, ad9 Δ mu an s esec as e and u he han wild ype, sugges ing ha CDK-media ed phospho yla ion o ei he Rad9 i sel o an unknown subs a e can diminish his physical block54. Rad9, and i s o hologs (Table 1) 53BP1 (highe euka yo es) and C b2 ( ission yeas ), unde go mul iple CDK-dependen phospho yla ions55,56, bu i is unknown whe he hese modi ica ions a ec on esec ion. Sae2 is di ec ly phospho yla ed by CDK a Se 267 ( e . 6). Impai men o his phospho yla ion leads o a educ ion in DNA-end esec ion, a delay o HR, an inc ease in NHEJ and an inc ease in DNA-damage sensi i i y6. Mo e in e es ingly, sae2-S267E mu an s, which mimic cons i u i e phospho yla ion, esec in he absence o CDK ac i i y and, as a consequence, ha e as e HR and dec eased NHEJ6. The sae2-S267E s ain is no sensi i e o DSBs ha a ise du ing DNA eplica ion, bu i shows enhanced hype sensi i i y when DSBs appea in G1 ( e . 6). Al hough sae2-S267E mu an s esec in he absence o CDK, such esec ion is limi ed o a ew kilobases lanking he b eak6, sugges ing a lack o ac i a ion o he Exo1 and/o Sgs1 pa hways14,17,22. Sae2 and C IP sha e only a small s e ch o sequence homology, bu his sho egion includes Sae2 Se 267 and i s equi alen C IP Th 847 ( e s. 6,8,9,17,18,20). Mo eo e , C IP Th 847 phospho yla ion con ols DNA-end esec ion in human cells much as Sae2 Se 267 phospho yla ion does in budding yeas . Impai men o his phospho yla ion, as well as cons i u i e phospho yla ion, leads o he appea ance o GCRs due o an imbalance be ween NHEJ and HR8. Phospho yla ion o chicken C IP a he equi alen esidue has simila unc ions9. As all he homologs o Sae2 and C IP excep S. pombe C p1 sha e his small egion o homology6,8,9,15,17,18,20, i is emp ing o specula e ha a simila mechanism egula es DNA esec ion in mos euka yo es. In ac , S. pombe C p1, al hough lacking a esidue homologous o Se 267, is con olled du ing he cell cycle bo h ansc ip ionally and by CDK phospho yla ion15,57. Addi ional laye s o egula ion by CDKs con ol C IP unc ion. C IP p o ein le els a e minimal in G1 and inc ease in S/G2 ( e . 42). Mo eo e , CDK-dependen phospho yla ion o C IP a Se 327 p omo es i s in e ac ion wi h BRCA1 in S/G2 ( e . 58), which is essen ial o C IP ec ui men o si es o DSBs and C IP-media ed DNA-end esec ion9,42,58. How CDK-dependen phospho yla ions o Se 327 and Th 847 collabo a e o egula e DNA esec ion and HR is no clea , bu bo h seem o be essen ial. Se 327 phospho yla ion, C IP- BRCA1 in e ac ion and C IP ec ui men o si es o damage a e no a ec ed by Th 847 phospho yla ion8, and mu an s ha mimic cons i u i e Th 847 phospho yla ion canno supp ess he lack o BRCA1 ( e . 9). On he con a y, a CTIP-T847E phosphomime ic mu an is able o esec DSB in G1 o a ce ain ex en , e en in he absence o an in e ac ion wi h BRCA1. One possible model is ha CDK-dependen phospho yla ion a C IP Th 847 is equi ed o ac i a e he DNA- esec ion machine y, bu BRCA1 is equi ed o e icien ly a ge C IP o si es o DSB in G2. Al hough phospho yla ion o Sae2 Se 267 o C IP Th 847 e lec s a conse ed mechanism o ac i a ion o DNA-end esec ion, he a ge ing o Sae2, C p1 o C IP o si es o DNA damage has di e ged h oughou e olu ion; ec ui men o C IP equi es BRCA1 ( e . 42), C p1 equi es Nbs1 ( e . 57) and Sae2 is ec ui ed by i sel o si es o DSBs59. Despi e he conse a ion be ween Sae2 Se 267 and C IP Th 847 in he Hue as Page 5 Na S uc Mol Biol . Au ho manusc ip ; a ailable in PMC 2010 July 01. Eu ope PMC Funde s Au ho Manusc ip s Eu ope PMC Funde s Au ho Manusc ip s licensing o DNA esec ion, li le is known abou he molecula mechanism unde lying his ac i a ion. One emp ing idea is ha phospho yla ion a hese esidues s imula es he nuclease ac i i y o Sae2 ( e . 35), bu such ac i i y has ye o be p o en o C IP. Ano he possibili y is ha such modi ica ions a ec DNA-end esec ion by ei he helping he ec ui men o posi i e ac o s o blocking he ac ion o nega i e modula o s. In addi ion o he CDK-dependen phospho yla ions o Sae2, C IP and Rad9, i is clea ha o he cell cycle– egula ed phospho yla ions o pos - ansla ional modi ica ions o he esec ion machine y a e equi ed o ully ac i a e long- ange DNA esec ion. DNA-end esec ion and he checkpoin The na u e o he b eaks also egula es DNA-end esec ion. Low numbe s o endonuclease- gene a ed DSBs a e no esec ed in G1 ( e s. 6,50-52,60). Howe e , simila numbe s o DSBs p oduced by γ-i adia ion esul in limi ed DNA esec ion, su icien o p omo e RPA ocus o ma ion in yeas 59,61. γ-i adia ion–induced esec ion, simila o ha in sae2-S267E mu an s, co e s only a ew kilobases and is p obably due o he ac ion o M e11 and Sae2 a he han o Exo1 o o Sgs1 and Dna2. I has been p oposed ha cells dis inguish hese adia ion-induced b eaks as ‘ agged’ ends, as opposed o he ‘clean’ endonuclease-induced b eaks, and ac i a e p ocessing ac i i ies o ‘clean’ hem61. In addi ion o he ype, he numbe o b eaks also plays a ole in he ac i a ion o DNA esec ion52. In S/G2, he mo e DNA b eaks gene a ed, he as e he esec ion akes place52. In G1, induc ion o up o h ee endonuclease-induced b eaks esul s in no esec ion, bu ou b eaks a e su icien o ac i a e DNA-end esec ion52. Simila o wha occu s wi h ei he “ agged” b eaks o in sae2-S267E mu an s, esec ion ac i a ed in G1 by mul iple b eaks is limi ed o he p oximi y o he end52. How he numbe o ype o b eaks modula e he esponse is no unde s ood, bu i is p obably ela ed o checkpoin ac i a ion, a p ocess in ima ely connec ed o DNA esec ion. The M e11 complex is equi ed, independen ly o i s esec ion ac i i y, o he ec ui men 28 and ac i a ion25 o he apical checkpoin kinase ATM (Tel1 in budding yeas ). The esec ion machine y is also a downs eam subs a e o his checkpoin kinase. Tel1 phospho yla es Sae2 and ATM phsopho ya es C IP in esponse o DNA damage, and hese phospho yla ions a e essen ial o esec ion18,62. As discussed p e iously, he checkpoin p o ein Rad9 blocks DNA esec ion54. Thus, i is possible ha ac i a ion o Rad9 by checkpoin kinases also acili a es esec ion54. Once esec ion is ac i a ed, i c ea es a posi i e eedback loop ha ampli ies he signal. The p oduc ion o sho , M e11- gene a ed oligonucleo ides con ibu es o u he ac i a ion o ATM in Xenopus36 . In addi ion, DNA-end esec ion gene a es ssDNA, which ac i a es ano he checkpoin kinase, ATR10,18,38 (Mec1 in budding yeas ), p o iding a po en ial mechanism o ATM-media ed ATR ac i a ion11,38. Mec1 phospho yla es Sae2 a he same si es as Tel1 ( e . 62), hus u he hype ac i a ing Sae2. Addi ionally, C IP also con ols he ec ui men o he human PCNA-like DNA-damage senso , he 9-1-1 complex o si es o ionizing adia ion-induced DSBs63. In budding yeas he hype phospho yla ion o he majo downs eam checkpoin kinase Rad53 canno be de ec ed on agged IR-induced b eaks in G1. Howe e , low le els o checkpoin ac i a ion, as measu ed by deg ada ion o Sml1, can be obse ed61. This limi ed checkpoin ac i a ion can igge limi ed esec ion, o example ia Mec1 o Tel1-dependen hype phospho yla ion o Sae2. In con as , Rad53 hype phospho yla ion is eadily obse ed when ou endonuclease-induced b eaks a e p oduced in G1, in which case DNA-end esec ion is also obse ed52. The eason o his di e ence emains a mys e y. One possibili y is ha a h eshold o ssDNA mus be su passed in o de o ully ac i a e Mec1 and cause Rad53 hype phospho yla ion. The limi ed esec ion o ou HO-induced b eaks, when combined oge he , may ully s imula e Rad53 in a way ha one o wo agged ends canno . A simila h eshold mechanism has been p oposed o checkpoin ac i a ion by Hue as Page 6 Na S uc Mol Biol . Au ho manusc ip ; a ailable in PMC 2010 July 01. Eu ope PMC Funde s Au ho Manusc ip s Eu ope PMC Funde s Au ho Manusc ip s s alled eplica ion o ks, in which mul iple uncoupled o ks oge he p o ide enough ssDNA o ac i a e he checkpoin 64. Despi e all ha , ully p ocessi e esec ion is only ob ained in S/G2 when CDKs a e ac i e6,8,52,61. To add an addi ional laye o complexi y, i has been shown ha he checkpoin machine y can nega i ely egula e esec ion o uncapped elome es. This is achie ed by he phospho yla ion, and consequen inhibi ion, o Exo1 ( e . 65). Howe e , i is unknown whe he his nega i e eedback loop ac s on DSBs o whe he i is speci ic o elome es. DNA esec ion and NHEJ NHEJ and DNA-end esec ion machine ies compe e in i o o he same subs a es. NHEJ is gene ally ini ia ed by he binding o he b eak o he he e odime Ku70–Ku80 ( e . 2), which se es as a sca old o o he p o eins ha con ibu e o he end-joining eac ion. Ku dime s ha e a high a ini y o DSBs, bu hey bind poo ly o ssDNA5 such as ha gene a ed by DNA-end esec ion. The e o e, esec ion educes he abili y o Ku o bind, and consequen ly, lack o M e11, Rad50, X s2 o Sae2 lead o inc eased amoun s o Ku bound o DSBs66. By con as , in he absence o NHEJ p o eins such as Ku o ligase IV, an inc ease in DNA esec ion and in he amoun o M e11 bound o he b eak is obse ed52,60,66. Cells lacking Ku a e able o esec in G1 close o he b eak, in ag eemen wi h an M e11- and Sae2-media ed esec ion52,60. This limi ed esec ion o a single b eak is enough o g an ull checkpoin ac i a ion as measu ed by Rad53 phospho yla ion60. Also, in S/G2 cells, esec ion is as e in he absence o Ku, and o e exp ession o Ku70–Ku80 educes DNA-end esec ion in G2 cells52,60. Biological ele ance o esec ion He e, I ha e discussed how DNA-end esec ion plays a key ole in he epai o DSBs and con ols he balance be ween HR and NHEJ. This is especially ele an because ailu e o epai DSB is associa ed wi h human diseases, including cance . No only he lack o epai , bu also he use o an inapp op ia e DSB epai pa hway, can be a sou ce o GCRs and he appea ance o po en ially dele e ious mu a ions1. Acco dingly, comple e loss o any o he majo playe s in DNA-end esec ion, such as M e11, Rad50, Nbs1 o C IP, leads o emb yonic le hali y in mice67-69 and inc eased DNA-damage hype sensi i i y in yeas and mammalian cells6,8,9,18,24,67,68. In addi ion, poin mu a ions in MRE11, NBS1 and RAD50 esul in inhe i ed synd omes ha a e associa ed wi h inc eased genomic ins abili y and cance p edisposi ion68,70. CTIP mu a ions ha e also been de ec ed in se e al cance s71,72. Mo eo e , haploid insu iciency o C ip in mice also p edisposes o cance 69. Addi ionally, hype ac i e sae2-S267E and CTIP-T847E esul in inc eased sensi i i y o ionizing adia ion due o a dec ease in NHEJ e iciency and an inc ease in GCRs esul ing om abe an HR6,8,9. Along hese lines, o e exp ession o C IP can be de ec ed in se e al b eas cance s72. Bloom synd ome, caused by mu a ions in BLM , is associa ed wi h genomic ins abili y and cance p edisposi ion73. Al hough i is di icul o co ela e hese gene ic synd omes wi h esec ion de ec s, i is emp ing o specula e ha abe an esec ion is a leas pa ially esponsible o he inc eased genomic ins abili y and cance p edisposi ion obse ed in indi iduals wi h such condi ions. The e o e, he unde s anding o DNA esec ion egula ion bea s g ea impo ance o he unde s anding o cance de elopmen . In addi ion, many cance he apies a e based on he idea ha DSBs a e ex emely po en p omo e s o cell dea h, especially in cance cells ha di ide apidly and a e usually de ec i e in some DSB epai pa hway. In ac , a ge ing DNA epai mechanisms o inc ease he le hali y o endogenous damage has p o en a success ul way o selec i ely kill cance cells74. The de elopmen o new he apeu ic s a egies ha a ge he esec ion Hue as Page 7 Na S uc Mol Biol . Au ho manusc ip ; a ailable in PMC 2010 July 01. Eu ope PMC Funde s Au ho Manusc ip s Eu ope PMC Funde s Au ho Manusc ip s machine y, h ough ei he inhibi ion o spu ious ac i a ion, could inc ease he e ec i eness o con en ional cance ea men s. Acknowledgmen s I would like o apologize o all au ho s whose wo k could no be ci ed due o space limi a ions. I am g a e ul o all he membe s o S. Jackson’s labo a o y in Camb idge, UK, o help ul discussions and especially o S. Jackson, A. Kaidi, J. Ha igan, K. Mille and R. Belo se ko skaya o hei help ul sugges ions and commen s on he manusc ip . I would also like o hank BBSRC and Cance Resea ch UK o unding my wo k. Re e ences 1. Aguile a A, Gómez-González B. Genome ins abili y: a mechanis ic iew o i s causes and consequences. Na . Re . Gene . 2008; 9:204–217. [PubMed: 18227811] 2. Liebe MR. The mechanism o human nonhomologous DNA end joining. J. Biol. Chem. 2008; 283:1–5. [PubMed: 17999957] 3. K ogh BO, Syming on LS. Recombina ion p o eins in yeas . Annu. Re . Gene . 2004; 38:233–271. [PubMed: 15568977] 4. McVey M, Lee SE. MMEJ epai o double-s and b eaks (di ec o ’s cu ): dele ed sequences and al e na i e endings. T ends Gene . 2008; 24:529–538. [PubMed: 18809224] 5. Dynan WS, Yoo S. In e ac ion o Ku p o ein and DNA-dependen p o ein kinase ca aly ic subuni wi h nucleic acids. Nucleic Acids Res. 1998; 26:1551–1559. [PubMed: 9512523] 6. Hue as P, Co es-Ledesma F, Sa o i AA, Aguile a A, Jackson SP. CDK a ge s Sae2 o con ol DNA-end esec ion and homologous ecombina ion. Na u e. 2008; 455:689–692. [PubMed: 18716619] 7. Lee K, Zhang Y, Lee S. Saccha omyces ce e isiae ATM o hologue supp esses b eak-induced ch omosome ansloca ions. Na u e. 2008; 454:543–546. [PubMed: 18650924] 8. Hue as P, Jackson SP. Human C IP media es cell cycle con ol o DNA end esec ion and double s and b eak epai . J. Biol. Chem. 2009; 284:9558–9565. [PubMed: 19202191] 9. Yun MH, Hiom K. C IP-BRCA1 modula es he choice o DNA double-s and-b eak epai pa hway h oughou he cell cycle. Na u e. 2009; 459:460–463. [PubMed: 19357644] 10. Ha ison JC, Habe JE. Su i ing he b eakup: he DNA damage checkpoin . Annu. Re . Gene . 2006; 40:209–235. [PubMed: 16805667] 11. Jazaye i A, e al. ATM- and cell cycle-dependen egula ion o ATR in esponse o DNA double- s and b eaks. Na . Cell Biol. 2006; 8:37–45. [PubMed: 16327781] 12. Amundsen SK, Smi h GR. In e changeable pa s o he Esche ichia coli ecombina ion machine y. Cell. 2003; 112:741–744. [PubMed: 12654241] 13. G a el S, Chapman JR, Magill C, Jackson SP. DNA helicases Sgs1 and BLM p omo e DNA double-s and b eak esec ion. Genes De . 2008; 22:2767–2772. [PubMed: 18923075] 14. Khakha RR, Cobb J, Bje gbaek L, Hickson I, Gasse SM. RecQ helicases: mul iple oles in genome main enance. T ends Cell Biol. 2003; 13:493–501. [PubMed: 12946629] 15. Limbo O, e al. C p1 is a cell-cycle- egula ed p o ein ha unc ions wi h M e11 complex o con ol double-s and b eak epai by homologous ecombina ion. Mol. Cell. 2007; 28:134–146. [PubMed: 17936710] 16. Mimi ou EP, Syming on LS. Sae2, Exo1 and Sgs1 collabo a e in DNA double-s and b eak p ocessing. Na u e. 2008; 455:770–774. [PubMed: 18806779] 17. Penkne A, e al. A conse ed unc ion o a Caeno habdi is elegans Com1/Sae2/C IP p o ein homolog in meio ic ecombina ion. EMBO J. 2007; 26:5071–5082. [PubMed: 18007596] 18. Sa o i AA, e al. Human C IP p omo es DNA end esec ion. Na u e. 2007; 450:509–514. [PubMed: 17965729] 19. T an PT, E deniz N, Syming on LS, Liskay RM. EXO1-A mul i- asking euka yo ic nuclease. DNA Repai (Ams .). 2004; 3:1549–1559. [PubMed: 15474417] 20. Uanschou C, e al. A no el plan gene essen ial o meiosis is ela ed o he human C IP and he yeas COM1/SAE2 gene. EMBO J. 2007; 26:5061–5070. [PubMed: 18007598] Hue as Page 8 Na S uc Mol Biol . Au ho manusc ip ; a ailable in PMC 2010 July 01. Eu ope PMC Funde s Au ho Manusc ip s Eu ope PMC Funde s Au ho Manusc ip s 21. Zhu Z, Chung WH, Shim E, Lee S, I a G. Sgs1 helicase and wo nucleases Dna2 and Exo1 esec DNA double-s and b eak ends. Cell. 2008; 134:981–994. [PubMed: 18805091] 22. Williams RS, Williams JS, Taine JA. M e11-Rad50-Nbs1 is a keys one complex connec ing DNA epai machine y, double-s and b eak signaling, and he ch oma in empla e. Biochem. Cell Biol. 2007; 85:509–520. [PubMed: 17713585] 23. Paull TT, Gelle M. The 3′ o 5′ exonuclease ac i i y o M e 11 acili a es epai o DNA double- s and b eaks. Mol. Cell. 1998; 1:969–979. [PubMed: 9651580] 24. B essan DA, Oli a es HA, Nelms BE, Pe ini JH. Al e a ion o N- e minal phosphoes e ase signa u e mo i s inac i a es Saccha omyces ce e isiae M e11. Gene ics. 1998; 150:591–600. [PubMed: 9755192] 25. La in MF. ATM and he M e11 complex combine o ecognize and signal DNA double-s and b eaks. Oncogene. 2007; 26:7749–7758. [PubMed: 18066087] 26. Buis J, e al. M e11 nuclease ac i i y has essen ial oles in DNA epai and genomic s abili y dis inc om ATM ac i a ion. Cell. 2008; 135:85–96. [PubMed: 18854157] 27. T ujillo KM, Yuan SS, Lee EY, Sung P. Nuclease ac i i ies in a complex o human ecombina ion and DNA epai ac o s Rad50, M e11, and p95. J. Biol. Chem. 1998; 273:21447–21450. [PubMed: 9705271] 28. Falck J, Coa es JA, Jackson SP. Conse ed modes o ec ui men o ATM, ATR and DNA-PKcs o si es o DNA damage. Na u e. 2005; 434:605–611. [PubMed: 15758953] 29. Nakada D, Ma sumo o K, Sugimo o K. ATM- ela ed Tel1 associa es wi h double-s and b eaks h ough an X s2-dependen mechanism. Genes De . 2003; 17:1957–1962. [PubMed: 12923051] 30. Usui T, Pe ini JH, Mo ales M. ad50S alleles o he M e11 complex: ques ions answe ed and ques ions aised. Exp. Cell Res. 2006; 312:2694–2699. [PubMed: 16857186] 31. McKee AH, Kleckne N. A gene al me hod o iden i ying ecessi e diploid-speci ic mu a ions in Saccha omyces ce e isiae , i s applica ion o he isola ion o mu an s blocked a in e media e s ages o meio ic p ophase and cha ac e iza ion o a new gene SAE2 . Gene ics. 1997; 146:797–816. [PubMed: 9215888] 32. Neale MJ, Pan J, Keeney S. Endonucleoly ic p ocessing o co alen p o ein-linked DNA double- s and b eaks. Na u e. 2005; 436:1053–1057. [PubMed: 16107854] 33. P inz S, Amon A, Klein F. Isola ion o COM1 , a new gene equi ed o comple e meio ic double- s and b eak-induced ecombina ion in Saccha omyces ce e isiae . Gene ics. 1997; 146:781–795. [PubMed: 9215887] 34. Cle ici M, Man ie o D, Lucchini G, Longhese MP. The Saccha omyces ce e isiae Sae2 p o ein p omo es esec ion and b idging o double s and b eak ends. J. Biol. Chem. 2005; 280:38631– 38638. [PubMed: 16162495] 35. Lengs eld BM, Ra ay AJ, Bhaska a V, Ghi lando R, Paull TT. Sae2 is an endonuclease ha p ocesses hai pin DNA coope a i ely wi h he M e11/Rad50/X s2 complex. Mol. Cell. 2007; 28:638–651. [PubMed: 18042458] 36. Jazaye i A, Bales ini A, Ga ne E, Habe JE, Cos anzo V. M e11-Rad50-Nbs1-dependen p ocessing o DNA b eaks gene a es oligonucleo ides ha s imula e ATM ac i i y. EMBO J. 2008; 27:1953–1962. [PubMed: 18596698] 37. Ma ingele L, Lydall D. Telome ase- and ecombina ion-independen immo aliza ion o budding yeas . Genes De . 2004; 18:2663–2675. [PubMed: 15489288] 38. Man ie o D, Cle ici M, Lucchini G, Longhese MP. Dual ole o Saccha omyces ce e isiae Tel1 in he checkpoin esponse o double-s and b eaks. EMBO Rep. 2007; 8:380–387. [PubMed: 17347674] 39. Mo eau S, Mo gan EA, Syming on LS. O e lapping unc ions o he Saccha omyces ce e isiae M e11, Exo1 and Rad27 nucleases in DNA me abolism. Gene ics. 2001; 159:1423–1433. [PubMed: 11779786] 40. Bae SH, e al. Dna2 o Saccha omyces ce e isiae possesses a single-s anded DNA-speci ic endonuclease ac i i y ha is able o ac on double-s anded DNA in he p esence o ATP. J. Biol. Chem. 1998; 273:26880–26890. [PubMed: 9756935] Hue as Page 9 Na S uc Mol Biol . Au ho manusc ip ; a ailable in PMC 2010 July 01. Eu ope PMC Funde s Au ho Manusc ip s Eu ope PMC Funde s Au ho Manusc ip s