Non-canonical DNA structures in Double Strand Break repair

DOCTORAL THESIS Non-canonical DNA s uc u es in Double S and B eak epai MªRosa Cama illo Daza Non-canonical DNA s uc u es in Double S and B eak epai CABIMER Memo ia p esen ada po : MªRosa Cama illo Daza Pa a op a al g ado de Doc o a en Biología Molecula , Biomedicina e In es igación Clínica 2023 Di ec o /Tu o Pablo Hue as Sánchez Di ec o a Sonia Jimeno González Abs ac DNA double-s and b eaks (DSBs) pose a se ious h ea o genome s abili y, so hei p ope epai is c ucial o cell iabili y. To epai hem, he cell has e ol ed wo main al e na i e mechanisms, non-homologous end-joining (NHEJ) and homologous ecombina ion (HR). The co ec choice be ween hese wo pa hways is a key poin o ob ain a ai h ul es o a ion o he b oken DNA sequence. One o bes -known e en s egula ing his decision is he gene a ion o 3’ single-s anded DNA o e hangs in a p ocess known as DNA esec ion. Many di e en ac o s a e in ol ed in his p ocess such as cell s emness, ch oma in s a us o DNA con o ma ion. In his Thesis, we in es iga ed he ole o G-quad uplexes, a non- canonical DNA s uc u e, in DNA esec ion. Fi s , we s udied he impac o G-quad uplexes in his p ocess and showed ha DNA esec ion is impai ed in he p esence o hese s uc u es, bo h in i o and in i o. In addi ion, we demons a ed ha PIF1α helicase ac i i y unwinding G-quad uplexes p omo es esec ion o e hese s uc u es and ha his p ocess equi es o he ac o s such as BRCA1 and TOP2β. Fu he mo e, we ook ad an age o he key ole o ac o s modula ing DNA esec ion o egula e homologous ecombina ion e iciency o sea ch o new compounds ha could a ec CRISPR-Cas9 media ed homologous ecombina ion e iciency. Resumen Los co es que a ec an a las dos cadenas del ADN suponen una se ia amenaza a la es abilidad del genoma po lo que su epa ación es c ucial pa a la iabilidad celula . Pa a epa a los, la célula ha desa ollado p incipalmen e dos mecanismos, la unión de ex emos no homólogos (NHEJ) y la ecombinación homóloga (HR). La co ec a elección en e ambas u as es un pun o cla e pa a consegui es au an ielmen e la secuencia de ADN dañada. Uno de los e en os más conocidos que egulan es a decisión es la gene ación de ib as de ADN de cadena sencilla colgan es en el ex emo 3’ del co e de ADN en un p oceso conocido como esección del ADN. Exis en muchos ac o es in oluc ados en es e p oceso al y como el g ado de di e enciación de la célula, el es ado de la c oma ina o la con o mación del ADN. En es a Tesis, se ha in es igado el papel de los G-cuad uplexes, una es uc u a no canónica del ADN, en la esección del ADN. En p ime luga , se es udió el impac o de los G-cuad uplexes en es e p oceso y se ha mos ado que la esección del ADN se encuen a pe judicada en p esencia de es as es uc u as, an o in i o como in i o. Además, se ha demos ado que la ac i idad helicasa de PIF1α en la esolución de los G- cuad uplexes p omue e la esección a a és de es as es uc u as y que es e p oceso equie e de o os ac o es como BRCA1 y TOP2β. Asimismo, ap o echamos la impo ancia de la modulación de la esección del ADN po dis in os ac o es en la egulación la e iciencia de la ecombinación homóloga pa a busca new compues os que puedan in lui en la e iciencia de la ecombinación homóloga mediada po CRISPR-Cas9. Indexes Indexes 12  Tables Index  Table R1. Candida e compounds wi h a mean RPA oci o ma ion e iciency o e 2- old o unde 0.85- old ............................................................................................................... 85 Table R2. Selec ed candida es desc ip ion ........................................................................ 87 Table R3. Valida ed compounds selec ed o hei impac on CRSPR-Cas9 e iciency ... 92 Table M1. Cell lines used in his Thesis .......................................................................... 146 Table M2. Plasmids used in his Thesis .......................................................................... 146 Table M3. siRNAs used in his Thesis ............................................................................ 147 Table M4. P ime s used in his Thesis ............................................................................. 148 Table M5. P ima y an ibodies used in his Thesis ..................................................... 149 Table M6. Seconda y an ibodies used in his Thesis ....................................................... 150 Indexes 13  ABBREVIATIONS ROS Reac i e Oxida i e Species WRN We ne helicase UV Ul a iole ligh G4 G-quad uplex IR Ionizing Radia ion PQS Pu a i e Quad uplex Sequence SSB Single S and B eak ChIP-seq Ch oma in- Immunop ecipi a ion high- h oughpu DNA sequencing BER Base Excision Repai SF DNA helicases supe amilies NER Nucleo ide Excision Repai PIF1 Pe i e In eg a ion F equency 1 MMR Misma ch Repai BIR B eak-Induced Replica ion DSB Double S and B eak TOPOI Topoisome ase I DDR DNA Damage Response TOPOII Topoisome ase II DNA-PK DNA-dependen p o ein kinase TOP1 Topoisome ase IB-1 PIKKs Phosphoinosi ide 3-Kinase (PI3K)- ela ed Kinases amily TOP2 Topoisome ase IIA-2 IRIF Ionizing Radia ion Induced Foci TOP2cc TOP2 clea age complex NHEJ Non-Homologous End Joining CRISPR Clus e ed Regula ly In e spaced Palind omic Repea s HR Homologous Recombina ion Cas CRISPR-associa ed genes LIG4 Ligase IV c RNA CRISPR RNA C IP CTBP In e ac ing P o ein gRNA Guide RNA PARP1 Poly [ADP-Ribose] Polyme ase 1 ac RNA T ans-ac i a ing c RNA SSA Single-S and Annealing PAM P o ospace Adjacen Mo i e SSTR Single S anded Templa e Repai KO Knock-Ou Indels Inse ions and Dele ions KI Knock-IN CDKs Cyclin-Dependen K inases ssODN ssDNA oligonucleo ides ssDNA Single s anded DNA CAT-R Colo Assay T acing- Repai BRCA1 B eas Cance 1 BSA Bo ine Se um Albumin EXO1 Exonuclease 1 PBS-T PBS-Tween Indexes 14  BLM BLOOM helicase B dU 5-B omo-2 Deoxyu idine RPA Replica ion P o ein A SSR See-Saw Repo e   I. In oduc ion    In oduc ion  19 1. DNA damage and genome s abili y The gene ic in o ma ion con ained in he DNA is in cha ge o guiding all he p ocesses ha de e mine he beha iou o he cell and he cha ac e is ics o e e y gi en o ganism. Thus, i is o u mos impo ance he ai h ul p ese a ion o his in o ma ion and i s p ope ansmission o he p ogeny. Howe e , he e is a la ge body o e idence demons a ing he cons an exposu e o he genomic euka yo ic DNA o geno oxic s ess ha induces nume ous DNA insul s. I is es ima ed ha e e yday each human cell su e s oughly 105 spon aneous DNA lesions ha al e i s DNA sequence and s uc u e (Lam, 2022). This damage needs o be accu a ely epai ed o ensu e he s abili y o he genome and, o ha ma e , he cell has e ol ed mul iple impo an epai pa hways o sol e hese h ea s. The accumula ion o mu a ions in genes esponsible o sensing and epai ing DNA damage leads o lesion accumula ion, loss o gene ic in o ma ion, highe mu agenic a es and uncon olled gene exp ession. These cha ac e is ics a e conside ed hallma ks o cance de elopmen and is he eason why DNA damage is a key e en in neoplas ic ans o ma ion (Neg ini e al., 2010). The e a e mul iple sou ces o DNA lesions bu , mainly, hey can a ise endogenously du ing DNA ansac ions o be induced by exogenous agen s (Fig.I1). Endogenous damage can be caused by he e y p ocess o DNA eplica ion du ing cell di ision, o example, due o he inco ec nucleo ide addi ion ( he so-called misma ches), inse ions o dele ions du ing DNA polyme ase ac i i y (Budzowska and Kanaa , 2009). In addi ion, base alkyla ion, me hyla ion, eac i e ni ogen species and eac i e oxida i e species (ROS) can be gene a ed as by-p oduc s du ing he no mal unc ioning o he cell and in e ac wi h he DNA molecule al e ing i s composi ion (Cha e jee and Walke , 2017). Exogenous damage, on he o he hand, is induced when en i onmen al, physical, and chemical agen s al e he DNA. Examples o his ype o DNA lesions include alkyla ing and c osslinking agen s, ul a iole ligh (UV), ionizing i adia ion (IR), chemo he apeu ic agen s and o he en i onmen al s esses such as empe a u e, ciga e e smoke o con amina ion (Meh a and Habe , 2014). In oduc ion 20  Figu a I1Each o hese dis inc DNA h ea s impac s di e en ly on he DNA molecule and, hence, encompasses di e se consequences in he genome (Cha e jee and Walke , 2017; Hoeijmake s, 2001). Mos o he lesions ha al e he DNA sequence such as misma ches, abasic si es, bulky adduc s, py imidine dime s, co alen bounds be ween nucleo ides o single s and b eaks (SSB) usually a ec jus one o he DNA s ands and can be epai ed gene ally by BER (Base Excision Repai ), NER (Nucleo ide Excision Repai ) o MMR (Misma ch Repai ) (Fig.I1). Al hough he mechanism o epai in which he cell incu s o mend hese al e a ions is speci ic o each kind o damage, he pa hways ollowed when only one s and is dis u bed sha e some simila i ies and can be sa ely ca ied ou since he in o ma ion om he complemen a y s and is p esen o ac as a empla e. Basically, once he damage is ecognized by speci ic p o eins, he a ec ed base, nucleo ide o oligonucleo ide is excised and eplaced wi h he co ec sequence (Caldeco , 2022). A di e en si ua ion needs o be con on ed when he damage a ec s bo h s ands o he DNA since he e is no longe an undamaged empla e o copy he in o ma ion om. This is he Figu e I1. Types o DNA damage DNA damage can be induced by se e al sou ces (exogenous and/o endogenous) ha gene a e di e en kinds o lesions in he DNA sequence al e ing ei he one o bo h DNA s ands. Each al e a ion equi es a speci ic mechanism o be epai ed. Modi ied om (Hoeijmake s, 2001). Ionizing adia ion Oxygen adicals Alkyla ing agen s Spon aneous eac ions Ul a- iole ligh DNA c osslinke s Replica ion e o s Ionizing adia ion Geno oxic agen s Meiosis U acil Abasic si e 8-Oxoguanine Al e ed base Single-s and b eak Bulky adduc In as and c osslink (6-4) pho op oduc Ciclobu ane py imidine dime A-G misma ch T-C misma ch Inse ion Dele ion In e s and c osslink Double-s and b eak (DSB) Base Excision Repai (BER) Nucleo ide Excision Repai (NER) MIsma ch Repai (MMR) Non-Homolohous End-Joining (NHEJ) Homologous Recombina ion (HR) DNA damaging sou ces DNA damage ypes DNA epai mechanisms In oduc ion  21 scena io o double-s and b eaks (DSB), which implies a mo e complex and coo dina ed pa hway o mul iple p o eins and cellula e en s o accu a ely es o e he sequence (Ceccaldi e al., 2016). 1.1 DNA Double S and B eaks (DSBs) DNA double s and b eaks a e conside ed one o he mos cy o oxic and di icul o epai lesion on he DNA. Only one o hese DSBs is su icien o induce cell dea h o e minal cell cycle a es . In all cases, he gene a ion o a DSB implies he b eak o bo h s ands o he double DNA helix h ough he clea age o he phosphodies e linkages o he backbone. Howe e , DSBs di e widely in hei s uc u e and in he mechanisms o hei gene a ion. While es ic ion endonucleases gi e ise o blun o s agge ed clean ends, clas ogens, such as adiomime ic chemicals, display a much mo e complex gene a ion o DSBs, chemically modi ying he s ands e mini (Bignold, 2009). Among he ac o s inducing DSBs, IR is a well-known one. IR a ge s DNA in wo di e en ways. Fi s , i can di ec ly a ack he DNA backbone inducing, mos ly, DSBs. In his p ocess IR migh also gene a e wa e adiolysis which can esul in he o ma ion o eac i e oxida i e species (ROS) (San i asi and Xia, 2014). These adicals can hen a ack mac omolecules in he cell, such as DNA, o ming single s and b eaks (SSBs) ha can u n in o DSBs a e DNA eplica ion. An icance chemo e apeu ic agen s a e also conside ed an impo an sou ce o DSBs. Some examples a e DNA-alkyla ing agen s, c oss-linking and adiomime ic compounds (Chen and S ubbe, 2005; Wy obek e al., 2005). Finally, unusual ch oma ic s uc u es such as G-quad uplexes, seconda y hai pin loops and R-loops a e ano he impo an sou ce o genomic ins abili y ha leads o DSBs gene a ion due o he s alling o he eplica ion o k p og ession and he inc eased exposu e o he DNA sequence o damaging agen s (Aguile a and Ga cía-Muse, 2013; Lopes e al., 2011; Ma ia e al., 2020; Rondón and Aguile a, 2019). Howe e , DSBs a e also necessa y o some impo an cell biology p ocesses (Os e and Aqeilan, 2020). Fi s , his kind o damage is a g ea sou ce o gene ic a iabili y due o he inse ions, dele ions and la ge ch omosomal ea angemen s ha i s un ai h ul epai migh gene a e (Ho e al., 2018). In addi ion, p og ammed DSBs a e also induced du ing meiosis o allow gene c oss o e and exchange o achie e an accu a e ch omosome In oduc ion 28  esec ion o unde go SSA has u ned his pa hway in o one o he mos common ead-ou o DNA esec ion. SSA equi es longe esec ed DNA and longe homology sequences han al -EJ and he annealing o he wo complemen a y sequences is esol ed by diges ing he single s anded ails and, he e o e, leading o loss o gene ic in o ma ion (Bha ga a e al., 2016). Recen ly, since he upswing o genome edi ing, ano he RAD51-independen DSB epai mechanism has been desc ibed. This pa hway uses a ssDNA empla e o epai a si e- speci ic DSB and is known as single s anded empla e epai (SSTR) (Gallaghe and Habe , 2021). The e is no much in o ma ion abou his p ocess and mos o i comes om s udies using Cas9 in human cells. 1.4 DSB epai pa hway choice E e y ime he cell aces a DSB, he DDR mus ac i a e a cascade o p ocesses o co ec ly epai he damage. The choice o he pa hway ac i a ed in esponse o he DSB is hea ily egula ed and de e mines ge ing ei he a ai h ul epai o an al e ed DNA sequence. This decision is in luenced by se e al ac o s ha he DDR conside s o choose he co ec epai mechanism. NHEJ is he as es and p edominan pa hway o epai DSBs in human cells. As al eady men ioned, NHEJ is no limi ed o cell cycle phases and, he e o e, can be employed whene e is needed. Howe e , i is conside ed o be an e o -p one pa hway when he ends o he b eaks need some p ocessing (Bu ma e al., 2006). Fo example, when he DSB gene a ed p esen s p o ein adduc s, sho o e hangs o al e ed chemical s uc u es hey need o be emo ed p io o liga ion o he ends. Then, he gaps gene a ed a e he liga ion a e illed by NHEJ associa ed polyme ases. In hese p ocesses, some inse ions and dele ions (indels) migh be gene a ed o igina ing and un ai h ul epai (Rodge s and McVey, 2016). On he o he hand, HR equi es longe pe iods o ime o epai he b eak and is limi ed o S/G2 phases o cell cycle, bu he ou come is usually mo e accu a e han NHEJ. Ne e heless, he accu acy o he epai by HR depends on he use o he co ec empla e o es o e he damaged DNA sequence. I he sis e ch oma id o he DNA sequence is p esen du ing he epai , i s use as a empla e is p omo ed and he chances o ha ing a co ec epai a e highe (Ciccia and Elledge, 2010; Heye e al., 2010). Since he sis e ch oma id is only a ailable when he DNA has been duplica ed in he G2/S phases, i In oduc ion  29 is clea ly explained why HR is es ic ed o hese phases. In he case ha his empla e is no p esen , o he HR is misadjus ed, he e a e o he subs a es ha can be used ins ead bu , usually, wi h mo e dele e ious ou comes. Fo example, he u iliza ion o he homologous ch omosome leads o non-allelic ecombina ion and, he e o e, loss o he e ozygosi y. In addi ion, non-iden ical epe i i e sequences a e ano he sou ce o homology o HR ha migh gene a e an inco ec epai o he damage due o indels (Aguile a and Gómez- González, 2008; Rodge s and McVey, 2016). Taking in o accoun all he de imen al consequences o he cell in case o unde going he w ong epai pa hway, he e a e many ac o s egula ing he choice o he co ec one. Fi s and p e iously men ioned, cell cycle is c i ical in se ing he igh pa hway. Cyclin-dependen kinases (CDKs) ac i a ed h oughou he cell cycle ha e a key ole in igge ing one pa hway o he o he (B andsma and an Gen , 2012; Dasika e al., 2000; Scully e al., 2019). Indeed, when a DSB is induced and G2/S phases a e ini ia ed, CDKs phospho yla es C IP p omo ing DNA esec ion. By con as , du ing G1 phase, p o easome deg ades C IP skewing he epai balance owa ds NHEJ (B andsma and an Gen , 2012). Also, he cell ype acing he DSB in luences he choice. P oli e a i e cells, such as s em cells, p omo e HR o epai DSBs (Checa-Rod íguez e al., 2020; Mujoo e al., 2017; Swi e al., 2021) in opposi ion o pos -mi o ic o mo e di e en ia ed cells which usually applies NHEJ. O he cell ea u es in luencing DNA epai pa hway choice a e epigene ic changes (Fe nandez e al., 2021; Min e al., 2022), complexi y o he b eak (B andsma and an Gen , 2012), ch oma in s uc u e (Fe nandez e al., 2021) and e en he nuclea posi ion o he damaged DNA (DNA in he pe iphe y is p one o epai DSB by NHEJ o e HR) (Lemaî e e al., 2014). The de egula ion o any o he elemen s desc ibed migh esul in genome ins abili y and disease (Hosoya and Miyagawa, 2014; Majidinia and Youse i, 2017). Tha is why DSB epai pa hway choice needs o be a ine- uned p ocess. Despi e all he ac o s al eady known o a ec his p ocess, he mos decisi e s ep guiding his esolu ion is DNA esec ion ac i a ion since i s inhibi ion by KU70/80 binding o 53BP1, o example, guides he epai h ough NHEJ and inhibi s HR (Syming on and Gau ie , 2011). In oduc ion 30  1.5 DNA esec ion DNA esec ion consis s o he nucleoly ic deg ada ion o he 5’- e mina ed b oken ends in he 5’-3’ di ec ion, lea ing 3’ ssDNA o e hangs o up o 1 kb (Fig.I4). This p ocess ini ia es and commi s DSB epai o he HR pa hway. Once esec ion is pe o med, ssDNA is apidly coa ed by Replica ion P o ein A (RPA) and homology-s and sea ch and in asion by RAD51 is ini ia ed. Al hough he e a e mo e s eps downs eam RAD51 s and-in asion o accomplish HR e icien ly, hey a e no needed o be desc ibed in his Thesis and an ex ensi e e iew on he ma e can be ound elsewhe e (K ejci e al., 2012; San Filippo e al., 2008; Sun e al., 2020; W igh e al., 2018). DNA esec ion can be di ided in wo s eps: sho - ange esec ion and long- ange esec ion (Fig. I4). Fi s , DNA esec ion is ini ia ed by DNA ecogni ion by MRN complex which is kep inac i e un il MRE11 ac i i y is licensed by C IP (Takeda e al., 2007). In he MRN complex, MRE11 is he co e componen and exhibi s di e en ac i i ies, including 3’-5’ exonuclease and endonuclease ac i i y, needed o pe o m he i s sho - ange esec ion s ep (Ga cia e al., 2011; Regina o and Cejka, 2020). Pa adoxically, 3’ssDNA gene a ion du ing esec ion is desc ibed o ake place in he 5’-3’ di ec ion, hus he esec ing ac i i y o MRE11 is p oposed o equi e a wo-s eps bidi ec ional mechanism and he associa ion wi h C IP (Canna o e al., 2019; Ga cia e al., 2011). MRN-C IP complex i s makes a nick in he p one- o-be esec ed s and se e al using MRE11 endonuclease ac i i y and hen, he 3’-5’ exonuclease diges he DNA owa ds he DSB end p oducing 3’ long ssDNA ails (Ga cia e al., 2011) (Fig. I4). C IP is conside ed a c i ical addi ional co ac o o he MRN complex and, upon i s phospho yla ion by CDKs in a cell-cycle dependen manne , i egula es he endonucleoly ic ac i i y o MRE11 (Hue as and Jackson, 2009; You and Bailis, 2010; Zd a ko ić e al., 2021). In addi ion, in humans, C IP phospho yla ion by ATM is equi ed o elici DNA esec ion (Makha ash ili and Paull, 2015; You and Bailis, 2010). The complex o igina ed by he binding o C IP o MRE11 is conside ed o be he minimal equi emen o DNA esec ion wi h C IP being he bes - known egula o o his p ocess and, he e o e, o DSB epai pa hway choice. Apa om C IP, he e a e addi ional accesso y ac o s in ol ed in his i s sho - ange esec ion such as he umou supp esso B eas Cance 1 (BRCA1). BRCA1 ole in DNA esec ion is no In oduc ion  31 e y unde s ood bu i is hough o help ec ui ing C IP o DSBs by i s in e ac ion wi h bo h C IP and he MRN complex (Densham and Mo is, 2019; Rosen and Walle, 2013) and, he e o e, p omo ing DNA esec ion. Mo eo e , BRCA1 has also been desc ibed o accele a e DNA esec ion p ocessi i y by binding MRN-C IP complex (C uz-Ga cía e al., 2014). Finally, i has been p oposed ha BRCA1 is equi ed o o e come he an i- esec ion ac i i y o he Shieldin complex (De e al., 2018). BRCA1also plays a ole in se e al s eps o DNA epai and cell cycle checkpoin s ac i a ion, so i is impo an o main ain genomic in eg i y. Mu a ed e sions o his gene a e p esen in high p e alence in he edi a y b eas and o a ian cance (Jiang and G eenbe g, 2015; Rosen and Walle, 2013). Despi e he impo ance o BRCA1, no all cells lacking a unc ional o m o his p o ein unde go apop osis. The ini ial sho esec ion by MRN-C IP is ollowed by a mo e ex ensi e esec ion o he 3’ ends in wha is known as long- ange esec ion (Fig. I4). This second s ep is ca alysed by wo al e na i e pa hways in human cells depending on he exonuclease used o esec he DNA. One o he wo pa ially edundan mechanism is he 5’−3’ exonuclease 1 (EXO1) while he o he mechanism in ol es DNA2 nuclease wo king in conce wi h BLOOM helicase (BLM) o We ne synd ome p o ein (WRN) (Hue as, 2010; Nimonka e al., 2011). In bo h cases, he esul is he gene a ion o ~1Kb long ssDNA acks ha will be co e ed by RPA and will acili a e he homology sea ch needed o homologous ecombina ion (San Filippo e al., 2008). This quick coa ing o he DNA ibe by RPA makes i he golden ead-ou o esec ion and i s quan i ica ion allows he s udy o DNA esec ion e iciency (Rade schall e al., 1999). The gene a ion o his ex ensi e ssDNA ails emo e chemical adduc s om he 5’ ends and displace KU70/80 he e odime s bound o he DNA, inhibi ing NHEJ (Hue as, 2010). Being DNA esec ion such a key poin in he balance be ween HR and NHEJ, he e has been ex ensi e s udy o he ac o a ec ing and egula ing DNA p ocessing like he p esence o DNA seconda y s uc u es o he p esence o p o ein adduc s co alen ly bound o he DNA. In oduc ion 32  Figu e I4. DNA end esec ion DNA esec ion is igge ed by he ec ui men o C IP and BRCA1 by he senso p o ein comple x MRN o DSBs. This p ocess consis s o wo s eps, a sho - ange and a long- ange DNA esec ion. Sho - ange esec ion is ini ia ed by he endonucleoly ic clea age o he 5’-ended DNA s and by MRE11. Then, he exonuclease 3’-5’ ac i i y o he MRN complex c ea es sho 3’ o e hangs o ssDNA ha a e u he esec ed in he long- ange esec ion by ei he DNA2 and BLM (le ) o EXO1 ( igh ) in a 5’-3’ pola i y. The 3’ssDNA o e hangs a e coa ed by RPA and acili a e he HR pa hway. Al e na i ely, he i s sho esec ion can be used o unde go he al e na i e end joining (al -EJ) pa hway (dashed a ow). DSB 5’ 3’ 5 ’ 3 ’ C IP C IP BRCA1 BRCA1 BRCA1 BRCA1 C IP C IP MRN MRN MRN MRN MRN MRN DNA2 EXO1 RPA RPA Sho - ange DNA esec ion L ong- ange DNA esec ion Al e na i e end-joining In oduc ion  33 2. DNA non-canonical seconda y s uc u es in DNA biology DNA is usually depic ed in i s canonical o m, a igh -handed B-DNA con igu a ion o wo an i-pa allel linea s ands ha wis a ound he same axis gi ing ise o a double helix wi h a mino g oo e and a majo g oo e (Fig.I5). This con o ma ion is he mos Figu e I5. Schema ic ep esen a ion o non-canonical DNA s uc u es On op le , Canonical B- o m duplex DNA s uc u e. On op igh , ep esen a ion o an R-loop s uc u e o med by an RNA-DNA hyb id and he consequen displacemen o he complemen a y DNA s and. On he bo om le , he ep esen a ion o a pu a i e quad uplex sequence (PQS) and he consequen G-quad uplex (G4) can be obse ed. On he bo om igh , di e en con o ma ions o G4s. In amolecula (gene a ed on one s and o he DNA) o in e molecula (gene a ed by se e al s ands o DNA). In each case, hey can be pa allel o an i- p a allel depending on he o ien a ion o he s ands. In he in e molecula box, a ep esen a ion o an an i-pa allel dime ic ( wo di e en s ands) and a pa allel e ame ic ( ou di e en s ands) s uc u es a e ep esen ed. CANONICAL DNA B- o m Mino g oo e Majo g oo e ssDNA RNA ssDNA R-loops An i-pa allel Pa allel In amolecula In e molecula An i-pa allel Pa allel G-quad uplex G-qua e GG GG GG GG PQS (Dime ic) (Te ame ic) In oduc ion 34  common o m in which DNA can be ound in he cell, howe e , du ing di e en cell p ocesses DNA can adop o he s ha a e conside ed non canonical. The di e si y o he di e en con o ma ions o DNA is a ibu ed o he di e en bonds and backbone con igu a ions ha he double helix can adop . Thei p esence is en iched in a eas wi h a highe a e o epe i i e sequences such as di ec epea s o sho -handed epea s (Bansal e al., 2022). Fo example, Z-DNA equi es an al e na ing pu ine-py imidine/GC ich sequences and c uci o ms equi e in e ed epea s o o m. The p esence o hese non-canonical DNA s uc u es is qui e abundan in he human genome and has been associa ed o DNA damage, endogenous genome ins abili y and disease (Pa lo a e al., 2021; Puge e al., 2019; Rinaldi e al., 2021; Rondón and Aguile a, 2019). In e es ingly, many ho spo s o b eakage si es and egula o s o cance -associa ed genes a e loca ed nea o non-B DNA s uc u es. One example is he p omo e o he c- MYC gene ha is abundan in guanine ich sequences ha o igina e some o hese s uc u es (Siddiqui-Jain e al., 2002). Indeed, some diseases, such as Hun ing on’s disease o agile X synd ome, a e caused by inucleo ide epea s and expansion o epe i i e sequences ha weaken he genome and inc ease gene ins abili y (Kh is ich and Mi kin, 2020; Möncke- Buchne e al., 2002). The mos s iking ac abou non-canonical DNA s uc u es is ha , despi e being conside ed an impo an sou ce o genome ins abili y, hey a e s ongly conse ed ac oss e olu ion sugges ing an addi ional ole o hem in cell biology. In ac , in he las yea s, he e has been an upsu ge o many s udies desc ibing he e ec s o non-canonical DNA s uc u es in DNA me abolism. Appa en ly, hese s uc u es egula e many p ocesses c i ical o cell biology no only by ac ing as physical ba ie s o he p ocessi i y o DNA machine ies bu also by ope a ing as sca olds o ec ui p o eins whe e needed (A ab e al., 2019; Boque-Sas e e al., 2015; De Magis e al., 2020). Among he wide a ie y o non-canonical con o ma ions o DNA, R-loops and G- quad uplexes (G4s) a e he mos s udied ones in ela ion wi h DNA me abolism. This is due o he b oad scena ios whe e hei p esence in luences he ac i i y o he di e en p o eins in ol ed. R-loops a e h ee-s anded non canonical s uc u es o med by an RNA-DNA hyb id plus a displaced ssDNA (Fig.I5). They a e expec ed o occu na u ally as a esul o In oduc ion  35 nega i e supe coiling du ing physiological cell p ocedu es, such as eplica ion and ansc ip ion. The p og ession o he ansc ip ion bubble and he syn hesis o he DNA lagging s and p omo es i s o ma ion in cis spanning leng hs om 0.1 o 2 kb (Bade and Bushell, 2020; Malig e al., 2020; Massé e al., 1997; Roy and Liebe , 2009). Recen ly, he e has been a lo o con o e sy ega ding he ole o R-loops in he cell biology. Fi s , R-loops a e desc ibed as induce s o DNA damage and genome ins abili y since hei occu ences co ela e wi h blocking o he ansc ip ion bubble and DNA eplica ion and cell s ess (C is ini e al., 2019; Hue as and Aguile a, 2003; Lang e al., 2017). Indeed, hey a e eme ging as po en ial sou ces o se e al diseases, such as cance and neu odegene a ion (Cama illo e al., 2021; Chen e al., 2004; C ow e al., 2006; Ga cía- Muse and Aguile a, 2019) Howe e , he physiological p esence o hese DNA s uc u es also unc ion as egula o s o class swi ch ecombina ion o immunoglobulins (Yu e al., 2003), elome e main enance (Tan e al., 2020), gene egula ion (G unseich e al., 2018; Sun e al., 2013) o in double-s and b eaks (DSBs) epai (S. Liu e al., 2021). The u ning poin be ween hese wo opposi e aces o R-loops elies in he e iciency o hei egula ion. In ac , hei u no e is igh ly con olled by di e en e olu iona y conse ed pa hways. Among he ac o s egula ing R-loops le els, RNA unwinding o deg ading enzymes a e he mos known ones as i is he case o SETX o RNase H (Ce i elli and C ouch, 2009; Cohen e al., 2018). Ne e heless, o he RNA p ocessing ac o in ol ed in splicing, opoisome ases and di e en DNA epai pa hways a e also in cha ge o esol ing hese s uc u es (Ha chi e al., 2015; Li e al., 2007; Massé e al., 1997). The deep s udy o R-loops has unco e ed hei coupled o ma ion wi h ano he ele an non-canonical DNA seconda y s uc u e, he G4s. Indeed, he mapping o R-loops in he genome shows an o e lapping loca ion wi h G4 (Kuzne so e al., 2018). As well as R-loops, G4s ha e dual oles in genome biology ac ing bo h as damage induce s and egula o s o DNA me abolism. In oduc ion 36  2.1 G-quad uplexes (G4s) s uc u es, mapping and cha ac e iza ion G4s a e guanine ich sequences assembled by Hoogs een hyd ogen bonding in o squa ed plana a angemen s, known as G-qua e s, ha s ack on op o each o he gene a ing hese ou -s anded helical s uc u es (Fig.I5). A la ge a ie y o DNA (DNA G4) and RNA (RNA G4) sequences ha can assembly in o G4s ha e been desc ibed. Basically, all G4s con ain a leas ou islands o a leas wo guanines sepa a ed by andom nucleo ides (loops) (Spiegel e al., 2020). Depending on he o ien a ion o he na u e o he s ands con o ming he G4, di e en ypes can be obse ed (Spiegel e al., 2020). In i s place, depending on he na u e o he nucleic acid adop ing he G4 s uc u e, he e a e DNA G4s and RNA G4s. In addi ion, depending on he numbe o s ands in ol ed in hei o ma ion, we can di e en ia e in e molecula (mul ime ic) G4s and in amolecula (monome ic) G4s (Fig. I5). In e molecula s uc u es can be dime ic, ime ic o e ame ic depending on he numbe o s ands in ol ed and hese s ands can be all DNA, RNA o a mix o RNA and DNA s ands. In compa ison, in amolecula G4s a e conside ed mo e s able and wi h as e con o ma ion dynamics han in e molecula G4s. Finally, he nucleo ide s and/s can old in o pa allel (all ou G-s ands in he same di ec ion) o an ipa allel (all adjacen G-s ands an ipa allel o each o he ) con o ma ion (F asson e al., 2022; Kim e al., 2003; Reshe niko e al., 2010; Yang, 2019) (Fig. I5). The con o ma ion ha a sequence wi h po en ial o o m a G4 (Pu a i e Quad uplex Sequences, PQS) can adop is di icul o p edic and usually in ol es some expe imen al s uc u e de e mina ion. G4s we e i s de e mined in i o using a G4-speci ic an ibody o isualise G4s in elome es (Scha i zel e al., 2001), bu hey also occu na u ally in i o. Ce ainly, se e al s udies ha e mapped oughly 700000 G4s alongside he whole genome wi h special emphasis in de e mined egions such as p omo e s, splicing si es, elome es and 5’ un ansla ed egions (Chambe s e al., 2015; Huppe and Balasub amanian, 2007, 2005; Mu a and Balasub amanian, 2014). Ce ainly, he p esence o his s uc u es in such key loca ions has inc eased he e o s o ully cha ac e ise hem. To do so, se e al s a egies ha e eme ged. Among hem, he mos ex ended one is he use o he BG4 an ibody ha binds a wide a ie y o G4 s uc u es wi h high a ini y (Bi i e al., 2013). This an ibody has been employed in many s udies o map endogenous G4 s uc u es along he genome and In oduc ion  37 o isualize hem unde he mic oscope. Al hough he BG4 an ibody seems o ecognize wi h high selec i i y di e en G4s s uc u es, i emains elusi e i his an ibody can bind a single G4 o i , on he con a y, i needs a high densi y o de ec hem. In addi ion, he numbe o G4s de ec ed by BG4 was lowe han he expec ed numbe p edic ed by nex - gene a ion sequencing (Bi i e al., 2013; Chambe s e al., 2015). Howe e , his could be explained by he dynamic and ansien olding and un olding o hese s uc u es in i o. O he me hods o G4s de ec ion include Ch oma in-Immunop ecipi a ion high- h oughpu DNA sequencing (ChIP-seq) using an ibodies agains p o eins known o bind hese s uc u es (G ay e al., 2014) o p o eins ecognizing he DNA damage induced by G4s ligands (Rod iguez e al., 2012; Xu e al., 2017). As expec ed, all hese app oaches showed an en ichmen o G4s in he a o emen ioned places, such as p omo e s and elome es. The p esence o G4s in hese key places suppo s hei egula o y ole in gene exp ession and poin hem ou as a no el s a egy o disease ea men . 2.2 Physiological oles o G4 s uc u es Despi e many epo s associa ing G4s o DNA damage and genome ins abili y, hei high conse a ion ac oss e olu ion in mammalian species suppo s hei ole as egula o y elemen s. Indeed, se e al s udies ha e ou lined he many DNA ansac ions in which hese non-canonical DNA s uc u es a e in ol ed. One o he i s p ocesses associa ed o a physiological ole o G4s was elome e biology. Telome es a e en iched in G- ich andem epea s wi h high po en ial o o m G4s. The p esence o hese s uc u es in hese a eas o ch omosomes ha e a dual ole in elome e in eg i y bo h p o ec ing elome es om deg ada ion by ac ing as a cap o elome ic DNA and egula ing elome e syn hesis by he elome ase hanks o he cell cycle dependen u no e egula ion o elome ic G4s (Ju anek and Paeschke, 2012; Smi h e al., 2011). In ac , elome ase ac i i y is inhibi ed by G4s s abilizing compounds leading o elome e sho ening (Sun e al., 1997), al hough elome ase ac i i y is no inhibi ed by e e y G4 ligand since i is able o p ocess DNA h ough pa allel G4s, bu no an ipa allel ones (Moye e al., 2015; Paudel e al., 2020; Zhang e al., 2010). In oduc ion 44  Apa om he a o emen ioned G4-d i en genome ins abili y mechanism, he mos in es iga ed pa hway o G4 induced DNA damage is DNA eplica ion. These non- canonical DNA s uc u es cons i u e a h ea o eplica ion machine y in e e ing wi h he eplica ion o k p og ession o bo h he lagging and leading s ands (Le ne and Sale, 2019) and he unc ion o DNA polyme ases (Es ep e al., 2019). Pa icula ly, s alling o he eplica ion o k due o an un esol ed G4 induces DSBs, one o he majo p omo e s o mu agenesis when i is no p ope ly ixed. Besides, eplica ion is he main mechanism o DSBs gene a ion by G4s since he SSB p oduced can be con e ed in o a DSB in he second ound o eplica ion. 2.5.1 G4 s abilising compounds and cance ea men E idence o he nega i e G4 impac on DNA in eg i y is alida ed by he dele e ious e ec s o he ac ion o small-molecule G4 ligands, which s abilize hese s uc u es and induce DNA damage by eplica ion s ess a PQS (Piazza e al., 2010; Rod iguez e al., 2012). Fu he mo e, he use o G4 ligands induce an inc ease in cell cy o oxici y in cells whe e HR- ac o s, such as BRCA1/2 and EXO1, a e impai ed (S oik e al., 2020; Xu e al., 2017; Zimme e al., 2016). To da e, a moun ing numbe o G4 in e ac ing compounds ha e been de eloped and mos o hem can be ound in he G4 Ligands Da abase 2.1 (h p://www.g4ldb.com/). The common cha ac e is ics sha ed by mos o hese compounds is he p esence o an a oma ic co e and he selec i i y be ween G4 and duplexes opologies. Some o he bes -known G4 ligands include py idos a in, BRACO-19 and TMPyP4 (Ko san is e al., 2020; Rod iguez e al., 2008). The inding o G4s in elome es and p omo e s o oncogenes has led much a en ion o he use o G4 ligands as he apeu ic compounds. This gene al concep was suppo ed by he obse a ion ha cance cells can be addic ed o hype ac i a ed oncogenes and/o augmen ed ac i i y o he elome ase p e en ing senescence. In consequence, speci ic down egula ion o oncogenes exp ession o elome ase inhibi ion by G4s s abiliza ion could inhibi cell p oli e a ion and cause cell dea h esul ing in an icance ac i i y. Despi e he p omising a ge ha G4 ligands o e , no many o hese compounds ha e en e ed phase II ials. This is explained by he low selec i i y ha G4 in e ac ing molecules shows owa ds a ce ain G4 con o ma ion (Zu o e al., 2018) complica ing hei use as an speci ic an i- In oduc ion  45 ca ce ea men . Ne e heless, e o s a e being made o design new compounds wi hou hese d awbacks and, a he same ime, new s udies showing he e icacy o hei use as no el s a egies o cance ea men keep a ising (Ama o e al., 2020; Gowan e al., 2002; Hu e al., 2020). 2.6 G4s and Topoisome ases Non-canonical DNA s uc u es, such as G4s and R-loops, usually a ise as consequence o nega i e supe coiling behind he ansc ip ion machine y, pa icula ly in a eas wi h a high ansc ip ion a e. To ace hese changes in he opology and o elax he supe coiled DNA a eas, cells use opoisome ases (Yada e al., 2014). Topoisome ases cons i u e a la ge amily o ubiqui ous nuclea enzymes ha in oduce ansien b eaks in he DNA backbone o main ain DNA opology main enance du ing cellula ansac ions. In euka yo ic cells, Topoisome ases can be classi ied in o wo g oups, opoisome ase I (TOPOI) and opoisome ase II (TOPOII), each o hem being u he di ided in he subg oups A o B. Each g oup is cha ac e ised by di e en s uc u al o ganiza ion, enzyma ic ca alysis modali ies and biological unc ions. TOPOI sub amily ac s as monome s ha clea es ansien ly only one s and o he DNA o elax he DNA supe coiling, ollowed by passage o he in ac s and h ough he b eak ( ype IA) o by con olled o a ion o he helix a ound he b eak ( ype IB). Con a y, Type II opoisome ases a e homodime s ha co alen ly a ach o 5’ ends and gene a e a ansien DSB o ca y ou he passage o ano he duplex DNA. Du ing he clea age, a co alen bond is o med be ween he p o ein and he DNA gene a ing ca aly ic in e media es known as opoisome ases clea age complexes (TOPOcc). Once he opology h ea is esol ed, he DNA b eak is e-liga ed. Due o hei eac ion mechanism, TOPOI can media e winding and unwinding ac i i ies bu is no able o esol e kno s and angles while TOPO II can modula e bo h ac i i ies (Champoux, 2001). Human cells possess six di e en opoisome ases wi h some edundan ac i i ies in ansc ip ion, eplica ion and DNA epai : TOP1 and TOP1m ( ype IB), TOP2α and TOP2β ( wo iso o ms o ype IIA) and TOP3α and TOP3β ( ype IA) (Pommie e al., 2016). In oduc ion 46  Despi e TOP2 unc ion being c i ical o he co ec p og essing o DNA ansac ion, i s mechanism o ac ion gene a es po en ially dele e ious DSBs i no p ope ly e-liga ed o i TOP2 ac i i y is impai ed o blocked. Indeed, opoisome ase inhibi o s a e desc ibed o induce DNA damage and genome ins abili y and, in consequence, some o hem a e alida ed as an i-cance ea men s (Ni iss, 2009). To unde s and he mechanism o ac ion o TOP2 inhibi o s, i should be explained he ca aly ic cycle o TOP2 (Fig.I7). The cycle s a s when TOP2 binds o an in ac segmen , he G (ga e) segmen , and hen e ains ano he DNA egion, he T ( anspo ed) segmen . Then, TOP2 expe iences a con o ma ional change using ATP om an open o a closed clamp o m. In he p esence o Mg2+, each TOP2 monome s ike a DNA phosphodies e bond ou base apa on opposi e s ands o he G duplex and becomes co alen ly linked o he 5’ end o he b oken DNA Figu e I7. Topoisome ase 2 (TOP2) ca aly ic cycle ATP Mg 2+ TOP2cc ADP+Pi Me ba one G segmen T segmen In oduc ion  47 Figu e I7. Topoisome ase 2 (TOP2) ca aly ic cycle The TOP2 homodime binds wo duplex DN agmen s. Fi s , he G segmen ( ed) en e s he clamp which is in he open con o ma ion, ollowed by he in e ac ion wi h a second DNA segmen (T segmen , black). When he T segmen en e s TOP2 complex, ATP p o ides he ene gy equi ed o close he clamp, clea e he G segmen and, in p esence o Mg+2, co alen ly ge linked o each s and o he G segmen ia a 5′-phospho- y osyl bond. o ming he TOP2 clea age complex (TOP2cc). Then, he T segmen passes h ough clea ed G agmen o he cen al ca i y and ge s eleased ollowed by he elega ion o he G segmen . The hyd olysis o he emaining ATP and he dissocia ion o wo ADP molecules inally eleases he G-segmen and p epa es he TOP2 complex o an addi ional ca aly ic cycle. TOP2 can be inhibi ed o poisoned a se e al di e en poin s in he enzyme eac ion cycle. Fo example, me ba one inhibi s TOP2 cycle be o e DNA is clea ed. gene a ing a clea age complex in e media e (TOP2cc). Once he T segmen is ans e ed h ough he induced b eak, he T segmen is eleased and he b oken ends o he G segmen a e esealed. The cycle inishes wi h he elease o he G segmen , ha is eleased upon ATP hyd olysis ha con e s he complex back o i s open clamp o m, lea ing he DNA p oduc chemically in ac om he ini ial backbone (Deweese and Oshe o , 2009). Taking TOP2 ca aly ic cycle in o accoun , TOP2 inhibi o s a ge di e en s eps h ough his cycle and, hus, can be classi ied in o wo g oups: TOP2 poisons and TOP2 ca aly ic inhibi o s. On one hand, TOP2 poisons blocks he ca aly ic cycle a e DNA is clea ed bu p io o DNA elega ion, leading o inc eased le els o DNA clea age and, hus, DNA damage. In he o he hand, ca aly ic inhibi o s a ge TOP2 wi hou s abilizing TOP2cc o inc easing DNA clea age. In his g oup, Me ba one is pa icula ly in e es ing since his agen is able o block TOP2 cu ing ac i i y wi hou a ec ing p o ein-DNA binding (Fo une and Oshe o , 1998; Ni iss, 2009; Pas o e al., 2012). Al hough he connec ion be ween opoisome ases and G4s has mos ly been limi ed o he pa icipa ion o he i s ones in G4 esolu ion, ecen s udies show ha hey migh also be in ol ed in G4 media ed genomic ins abili y. In ac , TOP2-induced DSBs seem o colocalise wi h DNA seconda y s uc u es and human TOPOI is able o bind G4 s uc u es and e en pa icipa e in hei o ma ion (A imondo e al., 2000; Szlach a e al., 2020). In e es ingly, no el da a demons a es ha he cy o oxici y o some G4 ligands, such as py idos a in, is media ed by TOP2 apping (Bossae e al., 2021; B uno e al., 2020; Oli ie i e al., 2020). Al hough hese expe imen s show ha TOP2 p omo es cell-killing ac i i y by some G4 s abilising agen s, his con ibu ion canno be gene alised o all o hem In oduc ion 48  since, o example, PhenDC3 media ed ins abili y is no p omo ed by TOP2 (Oli ie i e al., 2020). Oli ie i e al., 2020 indica ed as well, ha , despi e he beha iou o py idos a in being di e en o e oposide (a TOP2 poison) he ou come ob ained was he same (TOP2cc le els inc eased). This in e wining be ween TOP2 ligands and G4 in e ac ing compounds should be u he add essed since his inding is no an isola ed coincidence and o he compounds ha e also showed a high ac i i y as TOP2 inhibi o s and G-quad uplex binde s (Palluo o e al., 2016). 3. Clus e ed Regula ly In e spaced Palind omic Repea s (CRISPR)–Cas9 The abili y o modi ying he DNA sequence o ou desi e a a p ecise loca ion is o immense ele ance o he molecula biology, bio echnology and e en medicine and is known as genome edi ing. Genome edi ing has de eloped and imp o ed imp essi ely o e he yea s om he disco e y o es ic ion enzymes o clone a DNA sequence in o a plasmid o he use o Clus e ed Regula ly In e spaced Palind omic Repea s (CRISPR)–Cas sys em o accu a ely modi y a single nucleo ide in he genomic DNA. Du ing his e olu ion, many echniques ha e been explo ed such as C e- ecombinases, meganucleases, ZINC- inge nucleases o TALEN, nicely e iewed in (Khalil, 2020; Pe ez-Pine a e al., 2012; U no e al., 2010; Zhang e al., 2019). Howe e , he bigges ad ance in his ield was made wi h he disco e y o CRISPR-Cas sys em. CRISPR-Cas sys em o gene edi ing de i es om he disco e y o CRISPR in P oka yo es. I is based in hei adap i e immune sys em ha p e en s phage in ec ion by s o ing small sequences o i al DNA, called space s, in bac e ial hos ch omosomes. These i al DNA agmen s om di e en in ec ions a e sepa a ed by epe i i e nucleo ide sequences called di ec epea s. These CRISPR elemen s we e adjacen o mul iple well- conse ed genes known as CRISPR-associa ed (Cas) genes ha encode nucleases. A e a i al challenge, bac e ia in eg a e new space s de i ed om he in ec ing phage genomic sequence in o i s genome. These space s a e impo an o he sys em because, upon a ein ec ion, hey ge ansc ibed in o some sho guiding CRISPR RNA (c RNA) ha lead In oduc ion  49 he a ge ing by Cas o he i al genome, hus, p o ec ing he bac e ia agains he phage (Jo e e al., 2012). The cons an su i al s uggle o p oka yo es o con on he en i onmen al challenges has p omo ed he e olu ion o di e en CRISPR-Cas sys ems. The sys em conside ed as classical mainly comp ises six ypes o CRISPR-Cas ha can be classi ied in o wo classes based on he s uc u e o he Cas genes. Class I includes ype I, II and IV sys ems and uses se e al Cas p o eins, whe eas Class II ( ype II, V and VI) possesses only one Cas p o ein wi h mul iple domains, making i easie o enginee (Maka o a e al., 2018). Due o i s simplici y, he mos widely used in gene ic enginee ing is he ype II CRISPR- Cas9 om S ep ococcus pyogenes. Cas9 p o ein is an endonuclease wi h wo domains, Ru c and HNH. The HNH domain clea es single DNA s ands complemen a y o he sgRNA, while he Ru c domain clea es he opposi e s and (Fig.I8) (H yho owicz e al., 2017). The basic componen s o he ypical CRISPR-Cas9 sys em consis o a Cas9 nuclease and a single guide RNA (gRNA). The gRNA is o med by he syn he ic usion o he c RNA and he ans-ac i a ing c RNA ( ac RNA). To ecognise i s a ge sequence, c RNA encodes a 20 bases ecogni ion sequence ha binds speci ically he a ge si e wi h he help o he ac RNA. In ype II CRISPR-Cas9, his a ge si e needs o be placed nex o a p o ospace adjacen mo i e (PAM) ha can be ei he NGG o NAG, wi h N being ei he A, T, G, o C. Thus, he gRNA binds by i s 5’end he Cas9 p o ein o ming he sRNA/Cas9 complex and, by i s 3’ end, i binds he a ge si e, guiding he complex. Finally, upon a ge ecogni ion, he gRNA/Cas9 complex clea es he DNA 3-4 nucleo ides downs eam he PAM sequence inducing a DSB (Fig.I8) (H yho owicz e al., 2017). In oduc ion 50  3.1 Genome edi ing by CRISPR-Cas9 The si e-speci ic DNA DSBs gene a ed by Cas9 induce he cellula DDR o epai he b eak which can be exploi ed o enginee he genome. As al eady explained, DSBs can be epai ed ei he by NHEJ o HR, bo h o which can be used o ob ain he desi ed genome edi ing ou come (Fig.I8). In he absence o a p ope empla e, he DNA ends o he DSB a e e-liga ed by he e o -p one NHEJ pa hway. As a consequence, he b eak can be epai ed ai h ully econs i u ing he sequence o p oduce a di e se a ay o mu a ions, like indels. The gene a ion o hese indels in he coding sequence o a gene can p oduce ameshi mu a ions Figu e I8. Clus e ed Regula ly In e spaced Palind omic Repea s (CRISPR)-Cas9 sys em A guide RNA (gRNA) o med by he c RNA and he ac RNA is he esponsible o ecognising he genomic sequence which p ecedes he p o ospace adjacen mo i e (PAM). This ecogni ion guides he Cas9 nuclease o he a ge sequence which gene a es a DSB 3-4 nucleo ides ups eam o he PAM sequence by he wo Cas9 nuclease domains (Ru C and HNH) and, he e o e, ac i a ing he DSB epai machine y. In he absence o a homologous epai empla e, he b eak is epai by N HEJ which can esul in indels dis up ing he a ge sequence. Al e na i ely, he epai o he DSB by HR o SSTR can be p omo ed by he p o ision o a homologous dono empla e. c RNA ac RNA DSB NHEJ HR/SSTR Indels Indels Dono C as9-Ru C Cas9-HNH 5’ 5’ 5’ 3’ 3’ 3’ PAM NNNNNNNNNNNNNNNNN NNNNGG NNNNNNNNNNNNNNNN NNN NNNNNNNNNNNNNNNN NNNNNCC N In oduc ion  51 o p ema u e s op codons ha leads o loss o he unc ion o he gene and, hus, gene a e a knock-ou (KO). Al e na i ely, he DSB can be epai ed by HR in oducing speci ic nucleo ide changes ha span om a single nucleo ide a ia ion o he inse ion o la ge DNA agmen s (Ran e al., 2013b). To p omo e Cas9-DSB epai by HR a empla e is used which can be endogenous ( esul ing in an unedi ed sequence) o exogenous. In gene al, he e a e wo majo ypes o exogenous empla es o be used in he CRISPR-Cas9 sys em depending on he desi ed edi ion o make: a con en ional dsDNA a ge ing cons uc s, such as a plasmid o a PCR p oduc , wi h homology a ms lanking he inse ion sequence, o ssDNA oligonucleo ides (ssODNs). dsDNA dono empla e epai occu s mainly h ough HR while ssODN epai is media ed by SSTR, simila o SSA (Gallaghe and Habe , 2021). Al hough hese epai empla es ha e some di e ences, bo h mus con ain he desi ed edi as well as addi ional homologous sequence (homology a ms) immedia ely ups eam and downs eam o he a ge sequence. While ssODNs p o ide a much mo e e ec i e and simple me hod o making small edi s in he genome, such as he in oduc ion o single- nucleo ide mu a ions. dsDNA cons uc s o e he possibili y o in oducing la ge DNA agmen s (a ag o a luo opho e) hence gene a ing a knock-in (KI). Ne e heless, he DNA sequence p o ided in he empla e can in oduce a mu a ion as well ha leads o he KO o he gene (Renaud e al., 2016; Xue and G eene, 2021). 3.2 CRISPR-Cas9 applica ions Apa om gene a ing KO and KI in genes, CRISPR-Cas lexibili y as a gene manipula ion ool ha e been exploi ed o widen he epe oi e o hei applica ion. This is possible hanks o he abili y o Cas enzymes o bind a ge DNA independen ly o hei abili y o clea e i . The e o e, he modi ica ion o he clea age ac i i y o Cas9 ende s a new ool o modi y gene exp ession o sequence wi hou inducing DNA damage. The poin mu a ions D10A and H840A o he endonuclease domains o he Cas9, Ru c and HNH espec i ely, esul in a ca aly ically nuclease dead Cas9 (dCas9) molecule ha canno clea e a ge DNA bu can bind o i . Then, his dCas9 can be used o ansc ip ional ac i a o s (CRISPRa) o ep esso s (CRISPRi) o bind gene p omo e s and egula e hei exp ession (Gilbe e al., 2013; Maede e al., 2013; Qi e al., 2013), o epigene ic egula o s o c ea e p og ammable epigenome egula o s (B ocken e al., 2018) In oduc ion 52  o luo escen ags o isualise he a ge ed genomic loci (Ye e al., 2017). Since dCas9 does no modi y pe manen ly he DNA sequence, his modula ion is e e sible. This c ea es a po en ial ool o in es iga e essen ial genes (Pe e s e al., 2016; Singh e al., 2016) o e en egula e biosyn he ic me abolic pa hways in bac e ia (Gao e al., 2018; Kim e al., 2017). In some o ganisms, he induc ion o Cas9-DSB is e y oxic hus complica ing he manipula ion o hei genome. To o e come his se back, dCas9 can be used by e he ing i o DNA deaminases (Komo e al., 2016; Rees e al., 2017) o edi an speci ic loci wi hou damaging he genome. Ano he s a egy used in hese si ua ions includes he use o a pa ially inac i e Cas9 in which jus one o he endonucleases domains is inac i e. This Cas9 only o ms a non-le hal SSB hence is conside ed a nickase (nCas9) (S andage-Beie e al., 2015). The simplici y o gRNA design and p oduc ion, as well as he abili y o a ge almos any genomic locus, has poin ed ou CRISPR as he pe ec sys em o la ge-scale genome-wide gene ic sc eenings (Shi e al., 2015; Zhong e al., 2015). Cu en ly, he mos popula me hod o conduc ing genome-wide sc eens using CRISPR in ol es he use o pooled len i i al CRISPR lib a ies whe e each len i i al ans e ec o s encodes an indi idual gRNA a ge ing a single gene (McDade e al., 2016). CRISPR-Cas9 ha e also aised in e es o i s po en ial use o co ec gene ic diseases. In cance esea ch, he applica ions o CRISPR-Cas9 a e ex ensi e and ange om he de elopmen o in i o and in i o cance models and he sc eening o no el oncogenic mu a ions and umou supp esso s o cance gene he apy. The e icien and s aigh o wa d gene-edi ing capaci y o CRISPR-Cas9 accele a es he p ocess o ob aining a ep esen a i e cance model by mu a ing he gene in ol ed in he unde -s udy cance (Kal enbache e al., 2022; H. Wang e al., 2013). In cance ea men , CRISPR-Cas9 can be used o a ge speci ic sequences in he cance genome o d i e mu a ions associa ed wi h cance p og ession. Doing so, only selec i e apop osis o umou cells would be induced, educing ad e se e ec s on no mal issue cells, which a e usually associa ed o o he clinical cance ea men s (Ma inez-Lage e al., 2020). In addi ion, CAR-T cells, used in cance immuno he apy, could be gene ically modi ied by CRISPR-Cas9 o inc ease hei e ec i eness (Razeghian e al., 2021). In oduc ion  53 Ne e heless, CRISPR-Cas9 applica ion in medicine should no be limi ed o cance . Indeed, o he gene ic he edi a y disease a e in spo o be ea ed by CRISPR-Cas9. Cys ic ib osis (G aham and Ha , 2021) o Duchenne’s muscula dys ophy (Nelson e al., 2016; Zhang e al., 2022) a e examples o hem. Co ec ing he gene mu a ion ha o igins he disease could es o e he al e ed gene ic code. In ec ious diseases like HIV (He sko i z e al., 2021) o COVID-19 (Abbo e al., 2020) a e ano he ecen a ge o CRISPR-Cas. Fo ins ance, esea che s ha e de eloped a me hod o selec i ely a ge and des oy he gene ic ma e ial in oduced by he i us in he cell, p e en ing in ec ion. 3.3 Enhancing CRISPR-Cas9 echnology The moun ing numbe o s udies employing CRISPR-Cas9 o such an ex ended numbe o applica ions is p oo enough o all he u u e pe spec i es his gene ic enginee ing ool may p o ide. CRISPR-Cas9 is cheap, easy o use, cus omise and p oduce, has a wide a ge ed sequence ange and a obus ac i i y in euka yo es. Al hough CRISPR- Cas9 echnology has cons an ly alida ed i s signi ican ad an ages, i is undeniable ha he e a e some majo d awbacks in i s use ha limi i s po en ial. The less conce ning weakness o CRISPR-Cas9 migh be he equi emen o a speci ic PAM sequence nex o he a ge DNA. While he PAM sequence has a c i ical ole in CRISPR-Cas in bac e ia immune sys em o a oid sel - ecogni ion, i cons ains he selec ion o he ideal DNA a ge . To deal wi h his limi a ion, o he Cas9 o hologs and a ian s a e being op imised o p o ide a wide ca alogue o PAM sequence ha adap be e o he needs o he edi ing (Collias and Beisel, 2021). One o he majo p oblems in CRISPR-Cas9 echnology is nuclease clea age a unin ended genomic si es, caused by he nuclease ole a ing misma ches be ween gRNA and “o - a ge ” DNA (Fu e al., 2013). These induced DSBs in unp og ammed a eas gene a es o - a ge e ec s ha , among o he consequences, impede CRISPR-Cas9 use in clinical applica ion. Since his disad an age has e y ele an de imen al consequences o CRISPR-Cas9 de elopmen as he ideal gene ic ool, he e a e many esea che s ocusing on smoo hing hese o - a ge e ec s ou . One s a egy o imp o ing Cas9 speci ici y is o induce wo colocalizing SSBs, one in each opposi e s and, ins ead o di ec ly gene a ing a     III. Resul s Resul s-Chap e I  63  Chap e I: DNA esec ion o e G4-p one sequences by PIF1α DNA esec ion is a complex p ocess which in ol es a highly coo dina ed esponse o di e en ac o s. The esea ch o he cen al machine y in ol ed in his p ocess has been he cen al opic o many s udies, so i is well cha ac e ised. Howe e , li le is known abou he addi ional ac o s necessa y o o e come DNA s uc u es ha migh impede DNA esec ion. Sc eening o helicases in ol ed in DNA esec ion As in any o he DNA ansac ion, an impo an ac o in DNA esec ion a e helicases. Helicases, by de ini ion, unwind DNA so we hypo hesised ha , in a con ex o a ba ie due o an abno mal DNA con o ma ion, hey migh be a key ac o o p omo e DNA esec ion. To ha e a be e knowledge o helicases ha may be in ol ed in his key s ep o HR, we e iewed he e ec o all human helicases in a p e ious genome-wide sc eening pe o med in he lab (López-Saa ed a e al., 2016). In such s udy, mo e han 300 p o eins ha a ec he balance be ween HR and NHEJ we e desc ibed using he SeeSaw Repo e 2.0 (SSR 2.0). This epo e analyses he choice be ween HR and NHEJ a e y ea ly s ages being pa icula ly sensi i e o changes in DNA esec ion eloci y and/o p ocessi i y. Wi h such epo e , he ou come o he epai o a single, inducible DSB, can be isualized by he appea ance o g een (cells ha ha e epai ed by NHEJ) o ed ( hose ha ha e comple ed HR, speci ically SSA) cells (Fig. R1A). No helicases desc ibed o ha e a ole in DNA esec ion we e in his candida e sho lis , so he da a was e-analyzed by lowe ing he h eshold o posi i e candida es (Fig. R1B). In ha e ision, PIF1α deple ion, among o he helicases, was ound o aise NHEJ o e HR. Resul s-Chap e I 64  Figu e R1. DNA helicases e ec in DNA epai balance A) Schema ic ep esen a ion o he See Saw Repo e (SSR) 2.0. An I-SceI DSB is induced and epai ed ei he by NHEJ, leading o GFP econs i u ion, o by HR using RFP agmen s, hus gene a ing a unc ional RFP gene. B) O e iew o he e ec o he deple ion o di e en DNA helicases in NHEJ/HR balance ob ained om López-Saa ed a e al. (2016). Helicase genes o which deple ion p oduces an unbalance owa d inc eased HR a e ma ked in ed. Genes ha a ou s ecombina ion (NHEJ inc eased when hey we e down egula ed) a e ma ked in g een. The plo ep esen s he mean and SD o h ee independen expe imen s. A B BACH1 BLM CHD2 CHD7 ERCC2 ERCC3 HELLS INO80 MCM2 MCM3 MCM4 MCM5 MCM6 MCM7 MCM8 MCM9 MCM10 NCL PEO1 PIF1 RECQL RECQL4 RECQL5 RTEL1 RUVBL1 WRN 0.5 1 2 4 Inc eased NHEJ Inc eased HR NHEJ/HR a io ela i e o con ol siLuc RWDD2A SeeSaw Repo e 2.0 I-SceI BFP in ec ion Resul s-Chap e I  65  PIF1α is equi ed o DNA esec ion So, PIF1α deple ion seemed o a ou NHEJ o e HR. As s a ed be o e, he balance be ween bo h epai pa hways is mos ly egula ed by DNA end esec ion, he ini ial s ep in HR. Indeed, as he SSR is speci ically designed o de ec ac o s a ec ing DNA esec ion, we in es iga ed he ole o PIF1α in HR h ough i s in ol emen in ha p ocess. As explained in he in oduc ion, i p oduces long single-s anded DNA (ssDNA) ails ha a e immedia ely coa ed by RPA p o ein complex o p o ec ion. The accumula ion o RPA a he si es o he b eak p oduces oci ha can be isualised by immuno luo escence and he e o e can be used as a eadou o DNA esec ion in mammalian cells. Figu e R2. PIF1α is in ol ed in DNA esec ion  * % RPA posi i e cells siNT siNT siPIF1α siPIF1α PIF1α HSP70 Ladde siNT siC IP siPIF1α siNT siPIF1α 0 50 100 G1 S/G2 % Cells A D B C RPA γH2AX DAPI Resul s-Chap e I 66  Figu e R2. PIF1α is in ol ed in DNA esec ion A) Rep esen a i e Wes e n-blo image o siRNA-media ed deple ion o PIF1α. P o ein ex ac s we e ob ained om U2OS cells 48h a e siRNA ans ec ion. siNT ep esen s a con ol, Non-Ta ge , siRNA. HSP70 was used as a loading con ol. B) DNA esec ion p o iciency measu ed as he pe cen age o RPA posi i e cells upon DSBs induc ion by ionizing adia ion (10 Gy). The a e age and SD o h ee independen expe imen s a e shown. Signi icance was de e mined by - es S uden . (*P < 0.05). C) Cell cycle dis ibu ion o U2OS cells 48h a e ans ec ion wi h he indica ed siRNA. The mean and SD o h ee independen expe imen s a e plo ed. D) Median DNA esec ion leng h ob ained by SMART a e ionizing adia ion (10 Gy) ep esen ed as he mean and SD o h ee di e en expe imen s (le ) and a ep esen a i e luo escence mic oscopy image o he indi idual single s anded DNA ibe s ob ained ( igh ). Signi icance was de e mined by - es S uden . (*P < 0.05, **P < 0.01). In o de o es he e ec o PIF1α in DNA esec ion, we deple ed his p o ein by siRNA (Fig. R2A) and analysed i s impac on RPA oci o ma ion in esponse o a high dose o IR (10 Gy). The damage‐induced phospho yla ion o H2AX (γH2AX) was used as a ma ke o DSBs. In ag eemen wi h a ole a ou ing HR o e NHEJ, PIF1α siRNAs ende ed a de ec in RPA oci o ma ion (Fig. R2B). DNA esec ion and, in consequence, HR a e limi ed o G2/S phases o cell cycle so any change in he pe cen age o cells going h ough each one o hese phases ha e an impac in he DNA epai pa hway unde aken. As we obse ed in Figu e R2C, he RPA oci impai men associa ed o PIF1α deple ion did no ely on any e ec on he cell cycle since he knockdown o PIF1α did no a ec cell cycle dis ibu ion. Despi e RPA oci o ma ion being he gold s anda d eadou o esec ion, i is a low- esolu ion echnique ha does no p o ide in o ma ion abou he leng h o speed o he esec ion p ocess. To o e pass his de iciency, an al e na i e me hod named Single Molecule Analysis o Resec ion T acks (SMART) (C uz-Ga cía e al., 2014) was used o assess DNA esec ion in indi idual DNA ibe s. In e es ingly, no only he numbe o b eaks esec ed was educed upon PIF1α deple ion, bu he a e age leng h o ssDNA o med du ing esec ion was educed in a simila ex en o ha caused by he deple ion o he bona ide esec ion ac o C IP, as can be app ecia ed in he g aph and ep esen a i e in Figu e R2D. This sugges s ha PIF1α migh play a ele an ole in DNA end esec ion. Resul s-Chap e I  67  PIF1α is ec ui ed o DSBs In o de o u he s udy his po en ial ole in DNA epai , we decided o check i PIF1α was ec ui ed o DSBs. To do so, we i s s udied he abili y o GFP-PIF1α o o m oci a e i adia ion, a common ea u e o p o eins in ol ed in DNA epai . This analysis showed ha , once damage was induced, he numbe o GFP-PIF1α oci aised, hus sugges ing he ec ui men o PIF1α o DSBs (Fig. R3A). To u he s udy his ec ui men Figu e R3. PIF1α is ec ui ed o DNA-damage induced oci Cells wi h GFP-PIF1α oci a he a ay (%) Time a e DOX addi ion ** * ** * * Cells wi h γH2AX oci a he a ay (%) 0h 14 h 20 h 0 20 40 60 Time a e DOX addi ion α -IR +IR Numbe o GFP-PIF1α oci A B CD Resul s-Chap e I 68  Figu e R3. PIF1α is ec ui ed o DNA-damage induced oci A) U2OS s ably exp essing GFP-PIF1α we e used o s udy PIF1α oci o ma ion. Quan i ica ion o he median numbe o oci pe cell a e mock ea men o 10 Gy o ionizing adia ion (le ) and ep esen a i e images ( igh ) o GFP-PIF1α oci ( ed), γH2AX (g ey) and DAPI (blue) a e shown. B) Schema ic ep esen a ion o he expe imen al sys em used o measu e p o ein ec ui men o DSBs (le ). In U2OS19p igh 13, I-SceI is exp essed upon doxycycline addi ion o he medium and hen he endonuclease p oduces a single DSB. Binding o a Che y-lacI cons uc o he lacO andem epea s loca ed nex o he I-SceI clea age si e allows i s isualiza ion. Rep esen a i e luo escence mic oscopy images o he lacO a ay in ed, GFP-PIF1α in g een and γH2AX in g ey a e shown ( igh ). C) Quan i ica ion o he pe cen age o colocaliza ion o he lacO a ay wi h γH2AX in U2OS19p igh 13 cells exp essing GFP-PIF1α a e DSB induc ion upon Dox ea men . D) GFP- PIF1α ec ui men o DNA damage sco ed as he pe cen age o colocaliza ion o he a ay wi h GFP- PIF1α in cells ea ed as in C. S a is ical signi icance as in Fig. R2D. o PIF1α o damaged ch oma in we ook ad an age o a gene ic ool in which a single DSB can be c ea ed by he endonuclease I-SceI a e doxycycline induc ion (Lemaî e e al., 2014). This I-SceI a ge si e is lanked by 256 epea s o he lac ope a o (lacO) ha can be isualized by he accumula ion o Che y-lacI (Fig. R3B). The e o e, a e i s induc ion wi h doxycycline, we c ea ed a single DSB ha we can be isualize as a ed spo whe e p o eins in ol ed in i s epai would be ec ui ed. DNA damage indica o s, o example, γH2AX, will also co-localize wi h he Che y-lacI. In o de o analyse he ec ui men o PIF1α o DSBs, we measu ed colocaliza ion o GFP-PIF1α wi h he lacO a ay. Fi s , we con i med DSB induc ion a e doxycycline ea men by he de ec ion o γH2AX a he lacO a ay (Fig. R3C). Al hough he e was some backg ound binding o PIF1α be o e doxycycline addi ion, likely due o he s uc u e c ea ed by lacO epea s, PIF1α showed an inc eased colocaliza ion wi h he a ay a e DSB induc ion (Fig. R3D). These esul s ein o ce he hypo hesis ha PIF1α is ec ui ed o DSBs. G-quad uplexes con aining sequences ac as oadblocks o DNA esec ion Among he di e en ac i i ies o PIF1α in DNA ansac ions, i s ole in he esolu ion o a ypical DNA species seemed o pa icula in e es o PIF1α equi emen in DNA end p ocessing. One o hese noncanonical s uc u es a e G-quad uplexes (G4s). As hese G4s ac as a ba ie o DNA eplica ion machine y (Sa kies e al., 2010), we hough G4s could also a ec DNA esec ion. To es his hypo hesis, we used a G4 s abilize , py idos a in (Rod iguez e al., 2008) ha is desc ibed o induce DNA damage (Rod iguez e al., 2012). This small molecule, on i s own, p oduced a de ec in DNA esec ion a e Resul s-Chap e I  69  ionizing i adia ion o a simila ex en o he one p oduced by PIF1α deple ion (Fig. R4A) and no linked o changes in cell cycle dis ibu ion (Fig. R4B). Fu he mo e, py idos a in ea men was epis a ic o e PIF1 deple ion, indica ing a gene ic connec ion be ween he helicase and he G4 dissolu ion. Figu e R4. G4s s abiliza ion impai s DNA esec ion B dU G4 Me ge % RPA posi i e cells * * * ** * ** * B A C D E -Py idos a in -Py idos a in +Py idos a in +Py idos a in siNTsiPIF1α RPA γH2AX DAPI Resul s-Chap e I 76  Figu e R8. PIF1α pa icipa es in bo h sho and long- ange esec ion  A C B EFG D siNT siPIF1α * * * %MRE11 RNA le els ela i e o con ol %DNA2 RNA le els ela i e o con ol %EXO1 RNA le els ela i e o con ol %Cells Resul s-Chap e I  77  Figu e R8. PIF1α pa icipa es in bo h sho and long- ange esec ion A) DNA esec ion p o iciency measu ed as he pe cen age o RPA posi i e cells in cells co- ans ec ed wi h he indica ed siRNAs. O he de ails as in Fig. R2B. B) Cell cycle dis ibu ion o U2OS cells 48h a e ans ec ion wi h he indica ed siRNA. The mean and SD o h ee independen expe imen s a e plo ed. C) Same as in A, bu ela i ised o cell cycle dis ibu ion. D) mRNA le el exp ession o PIF1α in U2OS cells deple ed o he indica ed p o eins when compa ed o a con ol condi ion. Cells ans ec ed wi h he co esponding siRNA we e ha es ed 48 hou s upon ans ec ion. mRNA was ex ac ed om cell pelle s and e e sed ansc ibed in o cDNA. cDNA abundance was measu ed by qPCR wi h di e en p ime pai s. The mean and s anda d e o o h ee independen expe imen s is shown in he g aph. E, F and G) Same as in D, bu o he indica ed p o eins. G-quad uplex con aining sequences in luences DNA esec ion in i o In an a emp o gain a nea e knowledge o he impac o G4s s uc u e in esec ion, we mo ed o in i o DNA esec ion assays (Pin o e al., 2018). Fo ha , we c ea ed a dsDNA subs a e bea ing wo G4s mo i s om he C-MYC p omo e in he p one- o-be esec ed 5’- 3’ s and. In addi ion, wo lacO casse es we e inco po a ed in he 3’ end o he plasmid o inhibi DNA esec ion om ha end when lacI is added o he eac ion (Fig. R9A, le ). Nex , we s udied he esec ion e iciency in his linea ised dsDNA subs a e compa ed o a con ol wi hou he G4 sequences (Fig. R9A, igh ) in he p esence o di e en human p o eins in ol ed in DNA esec ion. Fi s , we did a kine ics wi h DNA2 nuclease in combina ion wi h BLM helicase o analyse long- ange DNA esec ion. The all in he in ensi y o he subs a e ha emains unp ocessed is conside ed he pe cen age o subs a e ha has been p ocessed (% o subs a e u iliza ion), o he wise said, esec ed. Con a y o he expec a ions, he p esence o he G4s s uc u es in he subs a e p omo ed subs a e u iliza ion compa ed o he con ol in he i s 30 min o he eac ion (Fig. R9B). A simila esul was obse ed when WRN helicase was used ins ead o BLM, al hough in his case he inc ease was p esen in all he imepoin s selec ed (Fig. R9C). In con as , EXO1 esec ion o he subs a es seemed unal e ed by he p esence o he G4s, p obably due o a sa u a ion o he eac ion by he high concen a ion o EXO1 p o ein used (Fig. R10). Gi en he close p oximi y o he G4 s uc u es o he simula ed DSB, one plausible possibili y o explain he inc eased DNA deg ada ion obse ed in he G4s-bea ing subs a es could be due o a change in he mel ing condi ions o he dsDNA and, in consequence, an easie access o he esec ion machine y o he DNA s and. To es his possibili y, we Resul s-Chap e I 78  Figu e R9. G4s p esence a ou s DNA esec ion by DNA2-BLM and DNA2-WRN in dsDNA subs a es A B C 30 nM hDNA2 20 nM hBLM -- -+ - + - + + + + + + + -- -+ - + - + + + + + + + 0 15 30 60 15 30 60 0 15 30 60 15 30 60 Time (min) pUC19x2LacO pUC19x2LacOx2G4 non empla e hRPA 576 nM (300%) Unp ocessed dsDNA ssDNA Resec ion p oduc s Unp ocessed dsDNA ssDNA Resec ion p oduc s LACO LACO LACOLACO 5’ 5’3’ 3’ LACI LACI Resec ion Radioac i e label Con ol DNA subs a e2xG4s DNA subs a e LACO LACO LACOLACO 5’ 5’3’ 3’ LACI LACI Resec ion Radioac i e label Resul s-Chap e I  79  Figu e R9. G4s p esence a ou s DNA esec ion by DNA2-BLM and DNA2-WRN in dsDNA subs a es A) Schema ic ep esen a ion o he dsDNA subs a es used in he DNA esec ion in i o assay. In he le , dsDNA subs a e bea ing 2 G4s sequences in andem 44 bp om he 5’end o he linea ized pUC19x2lacO plasmid and adioac i ely labelled wi h [α-32P]-dCTP in he 3’ end o he op s and. In he igh , con ol subs a e wi h he same cha ac e is ics as in he le , bu wi hou he G4 sequences. B) Quan i ica ion o subs a e u iliza ion kine ics by esec ion o dsDNA subs a es bea ing o no wo G4 sequences by DNA2 (30 nM) and BLM (20 nM) (le ). A ep esen a i e aga ose gel (1%) o he esec ion assay is shown ( igh ). The mean and SD o h ee independen expe imen s a e shown. Signi icance was de e mined by - es S uden . (*P < 0.05) C) Same as in B, bu wi h DNA2 and WRN (50 nM). boiled he dsDNA o ob ain a labelled ssDNA wi h he G4 mo i s and wi hou hem (Fig. R11A), whe e no unwinding o he double helix would be necessa y. Nex , we epea ed he esec ion assay wi h only DNA2, since no helicase is equi ed. This expe imen showed a dec eased subs a e u iliza ion when he G4s we e p esen (Fig. R11B), u he suppo ing ou p e ious in i o indings ha G4s inhibi DNA esec ion. Howe e , u he expe imen s wi h dsDNA subs a es in which he G4s a e loca ed u he inside he DNA will be equi ed o cla i y comple ely his poin . Figu e R10. DNA esec ion e icienc y b y EXO1 is no a ec ed b y G4 s uc u es in dsDNA subs a es Quan i ica ion o subs a e u iliza ion by esec ion o dsDNA subs a es bea ing o no wo G4 sequences by EXO1 (le ). A ep esen a i e aga ose gel (1%) o he esec ion assay is shown ( igh ). The mean and SD o h ee independen expe imen s a e shown. Unp ocessed dsDN A ssDNA Resec ion p oduc s hEXO1 (nM) 0 1 10 40 pUC19x2LacO pUC19x2LacOx2G4 non empla e hRPA 576 nM (300%) 0 1 10 40 Resul s-Chap e I 80  Figu e R11. G4 s uc u es impai DNA esec ion by DNA2 in ssDNA subs a es A) Schema ic ep esen a ion o he ssDNA subs a es used in he DNA esec ion in i o assay. On he le , ssDNA subs a e bea ing 2 G4s sequences in andem 44 bp om he 5’end o he linea ize d pUC19x2lacO plasmid and adioac i ely labelled wi h [α-32P]-dCTP in he 3’ end. On he igh , con ol subs a e wi h same cha ac e is ics as in he le , bu wi hou he G4 sequences. B) Quan i ica ion o subs a e u iliza ion by esec ion o ssDNA subs a es bea ing o no wo G4 sequences by DNA2 (le ). A ep esen a i e aga ose gel (1%) o he esec ion assay is shown ( igh ). The mean and SD o h ee independen expe imen s a e shown. Signi icance was de e mined by - es S uden . ( *P < 0.05 ) LACO LACO 5’ 3’ Resec ion Con ol DNA subs a e2xG4 DNA subs a e A B LACO LACO 5’ 3’ Resec ion Radioac i e label Radioac i e label hDNA2 (nM)0153060 pUC19x2LacO pUC19x2LacOx2G4 non empla e hRPA 576 nM (300%) 0153060 dsDNA ssDNA Resec ion p oduc s Resul s-Chap e I  81  FAN1 nuclease coope a es wi h PIF1α in G4-associa ed DSBs In DSB epai , i is equen he o ma ion o a helicase/nuclease complex o p ocess he damaged DNA. Being PIF1α a helicase wi h a possible ole in DSB epai , we wonde ed which nuclease could be he associa ed wi h PIF1α ac i i y. To do so, we checked FAN1, a known nuclease wi h a s ablished ole in DNA in e s and c oss-link epai by HR and in ol ed in G4s-induced o k collapse esolu ion (Po o e al., 2017), al hough i s implica ion in DNA esec ion had no been de ined. To s udy his possible ole, we analysed he capaci y o FAN1 down egula ed cells o o m RPA oci a e DNA damage. In acco dance wi h ou expec a ions, we obse ed a all in he pe cen age o cells able o ec ui RPA and, in e es ingly, his e ec was epis a ic o e PIF1α deple ion (Fig. R12A) and independen o cell cycle (Fig. R12B). Since FAN1 deple ion caused a de ec in DNA esec ion ini ia ion, measu ed by RPA oci o ma ion, we wonde ed i i would also a ec he esec ion leng h o he DNA ibe s. Con a y o ou expec a ions, he SMART assay o FAN1 deple ed cells showed a ise in he leng h o he esec ed ibe s ha was abolished when G4s we e s abilized by py idos a in (Fig. R12C). This inc ease was also app ecia ed when we measu ed he pe cen age o ibe s con aining, a leas , one G4 s uc u e (Fig. R12D). Howe e , in his case, he enhancemen obse ed compensa ed he d op a e py idos a in ea men . Nex , we wan ed o check i PIF1α o e exp ession could escue he pheno ype ob ained wi h FAN1 deple ion. Using RPA oci o ma ion as a eadou , we obse ed a lessening a e py idos a in addi ion bo h in he con ol and in FAN1 deple ed cells and his educ ion disappea ed when PIF1α was o e exp essed (Fig. R12E). FAN1 down egula ion was checked by qPCR (Fig. R12F). These indings suppo a ole o FAN1 in DNA esec ion ha is hampe ed by G4s and s imula ed by PIF1α. Resul s-Chap e I 82  Figu e R12. FAN1 is in ol ed in DNA esec ion o e G4 sequences A) DNA esec ion p o iciency measu ed as he pe cen age o RPA posi i e cells as in Fig. R11A. B) Cell cycle dis ibu ion o U2OS cells 48h a e ans ec ion wi h he indica ed siRNA. The mean and SD o h ee independen expe imen s a e plo ed. C) Median DNA esec ion leng h ob ained by SMART a e ionizing adia ion (10 Gy) in cells deple ed o FAN1 o o a con ol sequence (siNT) and ea ed ei he wi h H2O o py idos a in (10 µM) o 2h. O he de ails as in Fig. R2D. D) Quan i ica ion o he pe cen age o ssDNA ibe s con aining a leas one G4, o he in o ma ion as in C. E) Same as in A, bu in U2OS cells s ably exp essing GFP, GFP-PIF1α o GFP-PIF1α E307Q and ea ed ei he wi h H2O o py idos a in (10 µM) o 2h. F) mRNA le el exp ession o FAN1 in U2OS cells deple ed o ei he FAN1 o a Non-Ta ge sequence when compa ed o a con ol condi ion. O he de ails as in Fig. R11D. A CB EF D %RNA le els ela i e o con ol %Cells GFP-PIF1αGFP *** Resul s-Chap e II  83  Chap e II: Sea ching new ac o s a ec ing CRISPR-Cas9 gene-edi ing e iciency As we ha e al eady s a ed in he in oduc ion, CRISPR-Cas9 has been s ablished as an e icien ool o genome edi ing. Despi e i s mul iple ad an ages, one impo an se - back o his echnique is he low a es o KI e iciency ob ained. Tha is he eason why many s udies ha e ocused on inc easing hese a es by di e en means. To pa icipa e in his sea ch, we ha e ca ied ou a sc eening o look o new d ugs ha migh ha e an impac in he DNA epai pa hway choice. A d ug sc eening o compounds a ec ing DNA esec ion e iciency Since he gene a ion o a KI sequence is media ed by HR, we hypo hesized he d ugs ha would p omo e DNA esec ion, and he e o e, HR, would also inc ease CRISPR- media ed KI e iciency. Fi s , we op imized a sys em based on he de ec ion o RPA oci o ma ion in U2OS cells seeded in 96-well pla es. These cells we e p e- ea ed wi h h ee di e en d ug lib a ies ha con ained ei he an i-cance , an i-diabe ic o na u al compounds and we e exposed o 10 Gy i adia ion a e wa ds. Then, an immuno luo escence was pe o med on hese pla es and de ec ed using a high- h oughpu mic oscope. A schema ic layou o he p ocess is ep esen ed in Figu e R13A. To ca y ou he expe imen , se e al con ols we e included: i s , he ehicle o each d ug (DMSO o wa e we e used o no malize he esul s), second, as a posi i e con ol, he neddyla ion inhibi o d ug (MLN2449) ha we ha e p e iously demons a ed ha inc eases DNA esec ion (Jimeno e al., 2015) and inally, as an addi ional posi i e con ol we also included an siRNA deple ing he bona ide DNA esec ion ac o C IP. Each expe imen was epea ed independen ly a leas wice. Using his sys em wi h he 552 d ugs es ed, we ob ained se e al candida es (Table R1) wi h an e ec in DNA epai balance. We de ec ed 61 d ugs ha inc eased DNA end esec ion (g een ellipse, see Fig. R13B) whe eas 62 dec eased DNA esec ion ( ed ellipse, Resul s-Chap e II 84  see Fig. R13B). The selec ed candida es ( hose wi h RPA oci o ma ion e iciency o e 1.5 o below 0.85 o he le els ob ained wi h he mock ea men in a leas 2 epea s) a e lis ed in Table R2. The nume ic esul s ob ained o hose candida es and o he con ols indica ed abo e a e shown in Figu e R14. Figu e R13. D ug-based sc eening o compounds a ec ing RPA o ma ion e iciency A) Schema ic ep esen a ion o he wo k low o he sc eening. P e iously seeded cells we e ea ed o one hou wi h each one o he componen s o he indica ed d ug lib a ies, i adia ed wi h 10 Gy and incuba ed o an addi ional hou . A e pe o ming an immuno luo escence o RPA, images we e acqui ed and analysed. The compounds ha skewed he balance o e 2- old o unde 0.85- old in his assay we e selec ed as posi i e candida es o u he p og ess in he alida ion o he sc eening and be, inally, es ed as enhance s o CRISPR-Cas9 media ed KI e iciency. B) Rep esen a ion o he esul s o he sc eening. The mean o he numbe o RPA oci pe cell o , a leas , wo independen expe imen s ela i ised o hei con ol is ep esen ed. A g een ellipse ci cles he candida es wi h a p ossible enhancing e ec on DNA esec ion. On he con a y, a ed ellipse ci cles he ones appa en ly impai ing DNA esec ion. B U2OS cells D ug ea men RPA oci immuno luo escence D ugs 1h 1h 10 Gy IR An i-cance lib a y Na u al p oduc s lib a y An i-diabe ic lib a y Posi i es candida es Impac on CRISPR-Cas9 KI e iciency? A 8,5 7,5 8 7 6,5 6 5,5 5 4,5 4 3,5 3 2,5 2 1,5 1 0,5 0,2 Numbe o RPA oci ela i e o con ol D ugs p omo ing DNA esec ion D ugs inhibi ing DNA esec ion Resul s-Chap e II  85  Table R1. Candida e compounds wi h a mean RPA oci o ma ion e iciency o e 2- old o unde 0.85- old COMPOUND REP1 REP2 REP3 MEAN Oba oclax mesyla e (GX15-070) 13.0125564 2.94957302 7.9810647 IWP-2 8.23432425 4.17502986 6.20467706 BAY 11-7082 (BAY 11-7821) 3.82323012 4.46919164 4.14621088 Dexame hasone ace a e 1.9498026 5.77044005 3.86012132 PCI-32765 (Ib u inib) 2.49655107 5.18867904 3.84261506 GF109203X 2.24778224 5.23458376 3.741183 CNX-2006 4.21524092 3.09523808 3.6552395 To in 2 2.06928353 4.86069152 3.46498752 GW4064 3.80631321 2.73721605 3.27176463 Dacomi inib (PF299804, PF- 00299804) 3.15494416 3.25471659 3.20483038 So as au in (AEB071) 3.76488391 2.57679034 3.17083712 AMG 900 1.97620712 4.32255819 3.14938265 VS-5584 (SB2343) 3.81607298 2.42160183 3.11883741 Dalce apib (JTT-705) 3.1424422 3.05643471 3.09943846 Si inol 2.92936847 2.93022639 2.92979743 Cabozan inib mala e (XL184) 3.4683126 2.17929125 2.82380192 GSK2606414 3.20502611 2.2761858 2.74060595 PF-543 1.88168507 2.98427654 2.4329808 Pazopanib 1.2114414 3.64121678 2.42632909 Romidepsin (FK228, Depsipep ide) 2.14778303 2.54803132 2.34790717 TPCA-1 1.34172541 3.23327599 2.2875007 INK 128 (MLN0128) 1.51916431 3.04111025 2.28013728 GSK923295 2.04637396 2.48121909 2.26379653 (S)-10-Hyd oxycamp o hecin 1.81296816 2.61470275 2.21383546 AZ628 2.23723596 2.18266307 2.20994951 CHIR-99021 (CT99021) HCl 2.29811273 2.06721981 2.18266627 PFK15 3.10956441 1.22641502 2.16798971 Anag elide HCl 2.09856588 2.20284128 2.15070358 Ch ysophanic acid (Ch ysophanol) 2.01230822 2.26439668 2.13835245 Z-VAD-FMK 1.39150183 2.77557096 2.0835364 PH-797804 1.64296655 2.51755736 2.08026195 A bu in 2.65405405 1.33363513 2.18675722 2.0581488 Geldanamycin 1.376905 2.72147338 2.04918919 Lisinop il 1.0374315 1.57242102 3.43803056 2.01596103 Ho denine 0.95537822 2.1593602 2.87266553 1.99580132 De o olimus (Rida o olimus) 1.8135221 2.15314182 1.98333196 GW3965 HCl 1.21749876 2.66652345 1.94201111 Mycophenolic (Mycophenola e) 2.04737667 1.80782649 1.92760158 LY2603618 (IC-83) 1.78603927 1.87758293 1.8318111 Teniposide (Vumon) 2.0505313 1.53989941 1.79521535 DAPT (GSI-IX) 1.77945572 1.79313286 1.78629429 Sinomenine 0.89837838 1.85785423 2.57555178 1.77726146 E oposide (VP-16) 0.55059795 2.9586515 1.75462472 IOX1 1.38800497 2.07733487 1.73266992 LDE225 (NVP-LDE225, E ismodegib) 1.20804726 2.24840419 1.72822572 Yohimbine HCl 2.16586824 0.84552163 2.07385399 1.69508129 Bosen an 1.41787661 1.96155843 1.68971752 (+)-Usniacin 1.68216216 0.87369851 2.41595925 1.65727331 Silyma in 0.57297297 1.82797646 2.51782683 1.63959209 Apocynin 1.37249807 1.8993725 1.63593528 Mi oxan one HCl 1.79566353 1.44233315 1.61899834 Hema oxylin 1.08564125 1.6734282 2.08488964 1.61465303 Tie2 kinase inhibi o 1.61510669 1.59897191 1.6070393 Te ahyd opapa e ine HCl 1.25837838 1.70574921 1.83531409 1.59981389 Piogli azone (Ac os) 1.894971 1.25733294 1.57615197 Silibinin 1.12864865 1.79538253 1.71816638 1.54739919 Nalidixic acid 0.38594595 1.61158895 2.63327674 1.54360388 Topo ecan HCl 0.08614865 2.97206864 1.52910865 Clad ibine 0.87884802 2.17700289 1.52792546 Os hole 0.51675676 1.78451788 2.25806452 1.51977972 Pue a in 0.7772973 1.89859665 1.85229202 1.50939532 Resul s-Chap e II 92  au ophagy and apop osis (Liu e al., 2013). (-)-Epigalloca echin galla e (PubChem CID 65064) is a phenolic an ioxidan ( ound in g een and black ea) ha seems o impai DNA esec ion in ou sc eening. I displays an ineoplas ic e ec s by inducing apop osis and cell cycle a es , inhibi ion o me as asis and umou cell g ow h (B aicu e al., 2013; Han e al., 2011; Zan e al., 2019). Ida ubicin HCl (PubChem CID 636362) is an an h acycline om he an i-cance lib a y ha educes DNA esec ion e iciency. I is conside ed a TOP2 poison, hus, impai ing DNA eplica ion and ansc ip ion (Bigioni e al., 1994; Da sch and Giesele , 2007). I ino ecan HCl ihyd a e (PubChem CID 60837) is an an i-cance compound ha educes RPA oci o ma ion e iciency as well. I is he hyd ochlo ide sal o camp o hecin, a cy o oxic opoisome ase I inhibi o ha s abilizes he clea able complex be ween TOP1 and DNA, esul ing in DNA b eaks ha inhibi DNA eplica ion and igge apop o ic cell dea h. I s cy o oxic ac i i y is conside ed S-phase speci ic since i needs ongoing DNA syn hesis (Thomas and Pommie , 2019). Me ba one (PubChem CID 4990817) is a TOP2 inhibi o om he an i-cance lib a y wi h DNA esec ion inhibi ing ac i i y. I is a TOP2 ca aly ic inhibi o ha nei he in e cala e DNA no s abilize TOP2cc. I is known o p olong cell cycle p og ession, e a ding S phase and a es ing G2 phase (Fo une and Oshe o , 1998; Pas o e al., 2012). Table R3. Valida ed compounds selec ed o hei impac on CRSPR-Cas9 e iciency COMPOUND LIBRARY Ho denine Na u al p oduc s Lisinop il An i-diabe ic Dimesna An i-cance AZ628 An i-cance GF109203X An i-cance IWP-2 An i-cance I ino ecan HCl T ihyd a e An i-cance Ida ubicin HCl An i-cance Me ba one An i-cance To in2 An i-cance (-)-Epigalloca echin galla e An i-cance Resul s-Chap e II  93  Analysis o induced DNA damage signalling in he selec ed candida es Since some o he inal candida e compounds selec ed we e DNA damage induce s, we s udied he in ensi y o he DNA damage ma ke γH2AX o ule ou ha he e ec ob ained in he RPA oci o ma ion assay was due o a change in he DNA damage signalling (Fig. R17). As obse ed, om he candida es inc easing DNA esec ion e iciency, only IWP-2 induced a signi ican up ise in he in ensi y o he DNA damage le els de ec ed. Rega ding he compounds ha impai ed DNA esec ion, To in2 caused a d ama ic all in he le els o γH2AX. Being To in2 an inhibi o o he DDR ini ial signalling s ep by blocking ATM/ATR/DNA-PK ac i i y, his esul was expec ed. Su p isingly, we de ec ed a big d op as well wi h he ea men o TOP1 inhibi o Ida ubicin HCl. Since his candida e induces SSBs ha can po en ially u n in o DSB a e DNA eplica ion, his esul came unexpec edly. In addi ion, we es ed cell iabili y a e he ea men wi h he selec ed candida e compounds using an MTT assay. Al hough wi h he condi ions s ablished a sligh inc ease could be app ecia ed in some o he ea men s, none o hem ende ed a signi ican change in cell iabili y. (Fig. R18). Figu e R17. DNA damage signallin g le els induced b y he posi i e candida es om he sc eening In he le , γH2AX in ensi y le els in U2OS cells upon DSBs induc ion by ionizing adia ion (10 Gy) and d ug ea men . O he de ails as in Fig. R2B. In he igh , ep esen a i e images o he candida es om he le showing a signi ican change compa ed o he con ol. γ DMSO γH2AX DAPI Ida ubicin HCl IWP-2 To in2 Resul s-Chap e II 94  Analysis o he e ec o DNA esec ion-modula ing d ugs in CRISPR-Cas9 e iciency The main aim o pe o ming his d ug sc eening was looking o new compounds ha could enhance he low e iciency o CRISPR-media ed KI e iciency. To assess ha , we used he HEK-293 CAT-R cell line. A mo e de ailed explana ion o his assay can be ound in he Ma e ials and Me hods sec ion. B ie ly, he epai o a Cas9-induced DSB in a GFP gene linked o a mChe y sequence ende s di e en ou comes depending on he epai pa hway unde aken (Fig. R19A). In case o he epai pa hway inducing indels, we would ge a loss o he GFP luo escence and, he e o e, a KO. In addi ion, hype - esec ion o he b eak could lead as well o he loss o luo esce o GFP and mChe y, also gene a ing a KO. On he con a y, i he DSB is epai ed using he ssODN dono empla e added by SSTR, he epai would lead o he loss o GFP, bu no mChe y, and he acquisi ion o BFP luo escence, hus p o iding a measu e o KI e iciency. Finally, e o - ee epai o un- ans ec ed cells would show bo h a GFP and mChe y posi i e p o ile. As con ols o he expe imen , DMSO, H2O and MLN2449 we e included. As expec ed, ea men wi h he neddyla ion inhibi o MLN2449, which is known o a ou DNA esec ion and, in consequence, HR, p oduced a signi ican inc ease in KI e iciency, as measu ed by he inc ease in he pe cen age o BFP posi i e cells. This esul alida es Figu e R18. Posi i e candida es selec ed do no indue cell cy o oxici y A) MTT assay showing he iabili y o U2OS cells ea ed wi h 1 µM o he posi i e candida es om he sc eening o 2h, no malized o hei ehicle (DMSO o wa e ) and exp essed as pe cen age. S a is ical da a as in Fig. R2B. Resul s-Chap e II  95  Figu e R19. D ug e ec on CRISPR-Cas9 e iciency measu ed by he CAT-R epo e A) Schema ic ep esen a ion o he CAT-R epo e showing he di e en possible DNA epai ou comes a e a Cas9-media ed DSB. The CRISPR/Cas9 a ge si e is indica ed a he eGFP locus. I he DSB is esol ed h ough epai leading o small InDels gene a ion, ameshi mu a ions will p oduce a KO o eGFP, and only he mChe y will be exp essed; i he b eak is epai ed wi h la ge dele ions bo h mChe y and eGFP sequences will be los and no luo escence will be obse ed. Fo simplici y, double nega i e popula ion a e e e ed as la ge dele ions, bu hey migh include addi ional classes o ea angemen s as well. Finally, i he DSB is epai ed by SSTR using an exogenous ssODN bea ing he nucleo ide change needed o u n eGFP in o BFP, BFP posi i e cells will be p oduced. B) Quan i ica ion o he low cy ome y analysis o HEK293 CAT-R cell lines 72 h pos - ans ec ion wi h he syn he ic gRNA and d ug addi ion. In he box and whiske plo s, cen elines ma k he medians, do s indica e each eplica e om he expe imen s, box limi s indica e he 25 h and 75 h pe cen iles, and whiske s ex end o minimum and maximum, showing all poin s. The ed line indica es he le el o he mock ea ed con ol. Da a a e de i ed om, a leas , h ee independen expe imen s. S a is ical signi icance was de e mined wi h an ANOVA es . *p<0.05, **p<0.01, ***p<0.001. B A mChe y P2A eGFP mChe y P2A eGFP mChe y BFP mChe y P2A eGFP mChe y P2A eGFP mChe y P2A eGFP mChe y P2A P2A eGFP DSB DSB epai 5’ 3’ ssODN dono empla e Small InDels La ge Dele ions Un ans ec ed/ E o - ee Single S and Templa e Repai (SSTR) < 50-bp indel > 250-bp indel Cas9 a ge * * *** *** *** *** *** * Small Indels La ge dele ions SSTR E o - ee/Un ans ec ed % Popula ion ela i e o con ol Resul s-Chap e II 96  he assay pe o med. Mos o he candida es, namely me ba one, IWP-2, i ino ecan HCl ihyd a e, ho denine, AZ628, (-)-epigalloca echin galla e, lisinop il o bisindolylmaleimide I, did no p oduce any change in any o he epai p oduc s o he a iabili y was oo high o conside i ep oducible (Fig. R19B). On he con a y, a shi in, a leas , one o he luo escen cell popula ions ob ained was obse ed when cells we e ea ed wi h dimesna, o in2 and ida ubicim HCl. In he case o dimesna, he e is a non-signi ican d op in he KI e iciency and KO e iciency by he gene a ion o indels. This is in acco dance wi h he dec eased DNA esec ion e iciency obse ed in Figu e R17. Ne e heless, i signi ican ly ei he p omo es he e o - ee epai o he DSB o hinde s cell ans ec ion o he gARN and he ssODN. Su p isingly, o in2 p omo ed BFP KI e iciency despi e hampe ing DNA esec ion in ou p e ious expe imen s, p obably e lec ing addi ional e ec s on epai h ough ATM/ATR independen o esec ion. In addi ion, i induced a all in he pe cen age o cells unde going la ge dele ions due o he DSB epai . Finally, Ida ubicin HCl dec eased BFP posi i e and only Che y posi i e cell popula ions, consis en wi h i s obse ed ole dec easing RPA oci o ma ion e iciency. Fu he mo e, he e was a aise in he induc ion o la ge dele ions and e o - ee/un ans ec ed cells. To u he alida e he in luence o he compounds in CRISPR-Cas9 KI e iciency and o clea some con adic ions be ween he e ec in he RPA oci o ma ion assay and he KI e iciency measu ed by he CAT-R sys em, we employed ano he assay o uniquely measu e KI e iciency. This ime a plasmid ca ying bo h he Cas9 and gRNA we e elec opo a ed in o U2OS cells wi h a GFP-bea ing plasmid wi h homology a ms complemen a y o he RAG1 gene, whe e he DSB would be induced. Then, he measu emen o he di e en pe cen ages o GFP posi i e cells would ep esen he e iciency o CRISPR-Cas9 media ed KI (Fig. 20A). Once mo e, MLN2449 was used as a posi i e con ol o he pe o ming o he assay and, as an icipa ed, inc eased KI e iciency was obse ed. Con a y o ou expec a ions, he e was no ep oducibili y in he esul s ob ained wi h he candida e compounds in his assay compa ed o he ones o CAT-R epo e (Fig. 20B). He e, lisinop il is he only candida e which inc eased KI e iciency while i was no e idenced as a possible in luence in he CAT-R sys em. Simila ly, AZ628 and bisindolylmaleimide I also seemed o a ec he in eg a ion e iciency o GFP bu in he opposi e di ec ion o he p e ious obse a ion, lowe ing he e iciency. The same e ec was Resul s-Chap e II  97  ob ained wi h o in2. I is wo h no icing ha ida ubicin HCl could no be assessed wi h his assay since he concen a ion used in p e ious expe imen s was oo cy o oxic o he new se ings applied. So, we conclude ha he selec ed d ugs migh a ec DSB epai and KI e iciency in i o, bu he a iabili y be ween eplicas and epo e s was oo high o ende conclusi e esul s. This migh be due o he pleio opic e ec s o hese d ugs. So, we easoned ha u he e inemen was equi ed o isola e d ugs wi h po en ial. Me ba one inhibi s DNA esec ion One s iking obse a ion om he g oup o candida es ob ained om he sc eening is ha is en iched in opoisome ase inhibi o s, speci ically, i ino ecan HCl ihyd a e (TOP1 inhibi o ), ida ubicin HCl (TOP2 inhibi o ) and me ba one (TOP2 inhibi o ). The inhibi ion o opoisome ase is o g ea in e es o cance ea men because i impedes he epai o he b eak induced and, hus, p omo es DNA damage which, ul ima ely, would lead o cell Figu e R20. D ug e ec on CRISPR-Cas9 media ed KI e iciency A) Schema ic ep esen a ion o he assay used o assess KI e iciency by CRISPR-Cas9. A dono plasmid ca ying he GFP sequence lanked by homology a ms o he RAG1 gene was co- elec opo a ed in o U2OS cells wi h ano he plasmid bea ing he Cas9 nuclease and he gRNA a ge ing RAG1 gene. Then, cells we e seeded and ea ed wi h he candida e d ugs. 72h pos ans ec ion, he pe cen age o GFP posi i e cells was measu ed by low cy ome y. B) Quan i ica ion o he KI e iciency o GFP gene ep esen ed by he pe cen age o GFP posi i e cells measu ed by low cy ome y. O he de ails as in Fig. R2B. AB *** *** * * * RAG1 GFP + D ug Resul s-Chap e II 98  dea h. Many d ugs used as chemo he apeu ics wo k inhibi ing DSB epai pa hways, ei he NHEJ o HR, since hey a e usually mu a ed in many cance cells (Che niko a e al., 2012; Luo e al., 2022; Pa e son-Fo in e al., 2022; Xie e al., 2022). Conside ing he opoisome ase inhibi o s ound in ou d ug sc eening impai ed DNA esec ion (Fig. R16), he ini ial s ep o HR, we decided o s udy he e ec o opoisome ase inhibi o s in his p ocess in mo e de ail. Since I ino ecan HCl T ihyd a e ga e he opposi e esul in he sc eening compa ed o he alida ion, his candida e was ejec ed. Tu ning o he TOP2 inhibi o s ida ubicin HCl and me ba one, bo h d ugs block TOP2 ac i i y using di e en mechanisms. While ida ubicin HCl is conside ed a TOP2 poisons because i p e en s TOP2 u no e a e inducing he DSB (Bigioni e al., 1994), me ba one is a ca aly ic inhibi o o TOP2I and i blocks TOP2 ac i i y be o e DNA clea age is gene a ed (Ni iss, 2009). So, in b ie , ida ubicin block TOP2 ac i i y and c ea es DSBs, whe eas me ba one inhibi s he enzyme wi hou b eaking he DNA. Gi en ha he gene a ion o DSBs a he same ime o TOP2 apping could shade he eal ole o TOPOII in DNA esec ion, we decided o ocus on me ba one as a DNA esec ion inhibi o . Me ba one e ec in DNA esec ion is ela ed o PIF1 helicase. As discussed in he in oduc ion, opoisome ases ha e ecen ly eme ged as egula o s o he homeos asis and unc ion o G4s in human cells. In addi ion, a gene ic CRISPR/Cas9 based mul i-sc een demons a ed ha he cy o oxici y caused by py idos a in, a G4-s abilising d ug, is media ed by TOP2 apping (Oli ie i e al., 2020). Since in he p e ious Resul s chap e o his Thesis we showed ha py idos a in inhibi ed DNA esec ion in he same ashion as PIF1α deple ion, we wonde ed i he e was a connec ion be ween he e ec o he TOP2 inhibi o me ba one in DNA esec ion and he ole o G4s in his p ocess. Thus, we decided o combine me ba one wi h he deple ion o PIF1α helicase. As p edic ed, ou esul s poin o an epis a ic e ec in he RPA oci o ma ion pheno ype when bo h me ba one addi ion and PIF1α deple ion a e combined (Fig. R21A). F om his da a we can conclude ha PIF1α and TOP2 a e wo king in he same gene ic pa hway. Since bo h enzymes show a ole in G4 esolu ion (By d and Raney, 2017; Yada e al., 2014), we wonde ed i his was he common pa hway in which hey we e collabo a ing o media e Resul s-Chap e II  99  DNA esec ion. Indeed, he combined ea men o py idos a in and me ba one was epis a ic o e he ea men wi h jus one o hem and in all he cases he o e exp ession o PIF1α supp essed he DNA esec ion de ec caused by hose ea men s (Fig. R21B). To u he suppo his gene ic in e ac ion, we pe o med a co-IP o es i TOP2 and PIF1α we e physically in e ac ing. Using a GFP-PIF1α usion p o ein we we e able o eco e TOP2β Figu e R21. Me ba one e ec in DNA esec ion is ela ed o PIF1α helicase ac i i y A) Cells p e- ea ed wi h me ba one o DMSO as a con ol and ans ec ed ei he wi h an siRNA agains PIF1α o wi h a con ol siRNA (siNT) we e i adia ed and collec ed a e 1 h and he p esence o RPA oci was isualized by immuno luo escence. O he de ails as in Fig. R2B. B) Same as in A, bu in cells ans ec ed ei he wi h a plasmid ca ying he GFP-PIF1α usion o wi h a con ol plasmi d and ea ed o no wi h py idos a in. O he de ails as in Fig. R2B. C) An i-GFP Dynabeads we e used o immunop ecipi a e GFP-PIF1α and TOP2β om p o ein ex ac s o U2OS cells exp essing GFP- PIF1α o GFP. In he le , a ep esen a i e image o he expe imen is shown. The a e age and SD en ichmen o wo independen expe imen s a e shown in he igh . Resul s-Chap e II 100  in ag eemen wi h a physical in e ac ion be ween bo h ac o s (Fig. R21C). All in all, we conclude ha TOP2β ac i i y was equi ed o o e come some opological cons ain ha also equi es PIF1α.      Discussion  108  unc ions o PIF1 in ei he he nucleus o he mi ochond ia (Chisholm e al., 2012; Li e al., 2021) and has been associa ed wi h an inc eased isk o b eas cance (Chisholm e al., 2012). Ano he DNA esec ion ela ed p o ein wi h a high p e alence in b eas cance when mu a ed is BRCA1 (Jiang and G eenbe g, 2015; Rosen and Walle, 2013). BRCA1 is a key p o ein in DNA esec ion and i is known o in e ac wi h o he essen ial p o eins in ol ed in his p ocess o ec ui and ac i a e hem (Jiang and G eenbe g, 2015; Yu e al., 2006). In e es ingly, BRCA1 deple ion has been associa ed wi h sensi i i y o py idos a in (Zimme e al., 2016) hus, suppo ing a possible connec ion be ween his p o ein and G4s. We specula ed ha his connec ion could be h ough he ac i i y o PIF1α, since hey we e bo h pa icipa ing in DNA esec ion. In ag eemen wi h his sha ed ole o BRCA1 and PIF1α in G4 unwinding, an inc ease in he immunop ecipi a ion e iciency o BRCA1 an PIF1α was de ec ed a e he py idos a in ea men (Fig. R6A). Indeed, his in e ac ion is di ec as he use o BRCA1 ecombinan p o ein also e icien ly pulled down i s ecombinan PIF1α pa ne in an in i o binding assay (R6C). In addi ion, u he expe imen s o con i m he ela ionship be ween hese wo p o eins show ha hey also in e ac gene ically (Fig. R7A). As deduced om he s udy o PIF1α oci o ma ion, BRCA1 deple ion seems o impai PIF1α e-localiza ion a e IR, hus sugges ing he equi emen o BRCA1 o ec ui his helicase o DNA damage. This is in ag eemen wi h a ecen e iew ha showed ha BRCA1 haploinsu iciency migh lead o de ec i e G4 esolu ion (G oelly e al., 2022; Kim and Hwang, 2022) likely because PIF1α is no longe ec ui ed o hese s uc u es. In acco dance wi h he ec ui men o PIF1α oci o py idos a in a ge si es obse ed in Rod iguez e al., 2012, he py idos a in ea men wi h no o he sou ce o DNA damage mildly inc eased he numbe o PIF1α oci. Howe e , con a y o ou expec a ions, in py idos a in ea ed cells, PIF1α ec ui men o DNA damage did no inc ease a e i adia ion (Fig. R7A). In ac , he e ec seemed o be he con a y. The eason o his a he con adic o y esul is no ye comple ely clea , bu a likely explana ion could be he e-localiza ion o PIF1α oci a e i adia ion. Ini ially, PIF1α migh colocalize wi h G4s (Rod iguez e al., 2012) bu , once we induce new DNA damage oci wi h i adia ion, PIF1α ha e o e-localize o IR-induced oci. This is acco dance wi h he signi ican aise in PIF1α oci de ec ed in ou IR expe imen s (Fig. R7A). Howe e , in he p esence o py idos a in, PIF1α migh be mo e igh ly associa ed wi h s abilized G4s, hampe ing his e-localiza ion. Discussion 109   I is possible ha he ime ame o ou expe imen was no enough o he shi om s able G4s o he o ma ion o new IR-induced PIF1α oci. This would esul in a diminu ion o he numbe o oci obse ed. Fu he expe imen s wi h longe eco e ies a e IR ea men will clea his possibili y. Up o his poin , all he e idence con i ms a ole o PIF1α helicase in DNA esec ion. This DNA esec ion needed o unde go HR epai is no always ini ia ed om b oken DNA ends conside ed chemically “clean”. Indeed, DSB-DNA-ends may p esen p o eins co alen ly a ached o hem, chemical modi ica ions o seconda y s uc u es ha migh impede he ini ia ion o he esec ion. Depending on he ela i e posi ion o hese oadblocks o he b oken DNA ends, hey may in e e e wi h ei he he ini ial sho - ange esec ion o he ollowing long- ange esec ion. Since G4s a e conside ed one o hese oadblocks, hei p esence alongside he p one- o-be esec ed DNA sequence obs uc he esec ion. The equi emen o hei esolu ion o p oceed wi h he esec ion sa is ac o ily explains he pa icipa ion o PIF1α in bo h s eps o he p ocess as obse ed by he epis a ic e ec o he double deple ion o PIF1α and ac o s in ol ed in ei he long o sho - ange esec ion (Fig. R8). An impo an conside a ion when s udying DNA esec ion is ha i is egula ed by cell cycle p og ession wi h many o he ac o s in ol ed in his p ocess being only exp essed du ing S/G2 phases (Falck e al., 2012; Hue as e al., 2008; Tomima su e al., 2014). Thus, he signi ican cell cycle modi ica ions p oduced by he deple ion o MRE11 and DNA2 had o be aken in o accoun . The e o e, we no malized he DNA esec ion e iciency wi h cell cycle p o iles measu ed by he DNA con en , conside ing he ela i e changes in he subpopula ion o S/G2 cells compa ed o he con ol si ua ion. Gi en ha no signi ican change in he esul s was no iced a e he no maliza ion, we concluded ha he pheno ype obse ed is no a ibu ed o any change in cell cycle dis ibu ion (Fig. R8C). In S.ce e isiae, M e11 is indispensable o p ocess “di y” ends and ini ia e esec ion (Cejka, 2015). Howe e , in ou expe imen s pe o med in human cells, MRE11 down egula ion did no show such an essen ial ole o DNA esec ion ini ia ion (Fig. R8C) sugges ing he collabo a ion wi h accesso y ac o s o o e come hese impedimen s. Following his a ionale, ano he explana ion o he signi ican di e ence ound in MRE11- PIF1α deple ed cells could be he conside a ion o PIF1α as one o hese accesso y ac o s con ibu ing o he p ocessing o he ends by he emo ing o G4 s uc u es. Discussion  110  The p esence o G4s s uc u es o PQS in he DNA sequence has long been p oposed o ac as ba ie s o many DNA ansac ions (see In oduc ion 2.2). Indeed i is well-known ha hey in e e e wi h he p og ession o he eplica ion machine y du ing DNA syn hesis (Kaguni and Clay on, 1982; Kama h-Loeb e al., 2001). Su p isingly, he da a ob ained om ou in i o expe imen s poin ed o he opposi e di ec ion. In ac , he in oduc ion o wo G4- o ming sequences in he esec ed s and imp o ed he DNA deg ada ion e iciency o he DNA by bo h DNA2/BLM (Fig. R9B) and DNA2/WRN andem (Fig. R9C). Rema kably, DNA esec ion by EXO1 was no p omo ed by he p esence o he G4s compa ed o he con ol sequence (Fig. R10). This could be explained by he sa u a ion o he eac ion wi h high concen a ions o EXO1 since he use o lowe le els o his p o ein did sugges a s imula ion o he p ocess by G4s. The disc epancies obse ed wi h ou ini ial hypo hesis o G4s being a oadblock o esec ion can be a ibu ed o a as e access o he esec ion p o eins o he DNA due o he easie mel ing o he DNA duplex induced by he p esence o he G4s close o he DNA ends. Indeed, o he non- canonical DNA s uc u es, such as DNA hai pins, seems o in luence bo h he a e and e iciency o DNA hyb idiza ion by lowe ing he ee ene gy equi ed o hei dissocia ion (Sch eck e al., 2015). In line wi h his idea, no imp o emen was obse ed wi h DNA2/BLM a e 60 min o eac ion mos likely because i was ime enough o DNA2/BLM o unwind he DNA and equal he le el o DNA deg ada ion p oduced by he G4-bea ing subs a e (Fig. R9B). Fu he suppo ing his heo y, when we emo ed he equi emen o DNA duplex unwinding p io o DNA deg ada ion by using a ssDNA subs a e, i inally ende ed he expec ed de ec in DNA esec ion when G4s a e p esen (Fig. R11B). I is no ewo hy o men ion ha he o ma ion o he G4 s uc u e was no checked in hese expe imen s. The e o e, ano he easonable explana ion o he p omo ion o DNA esec ion by G4s could be a ibu ed o he p esence o he PQS bu no he s uc u e i sel , ha ing his a di e en e ec in he deg ada ion e iciency. Addi ional expe imen s con i ming he s a e o he po en ially G4- o ming sequences and posi ioning hese sequences u he om he DNA ends o a oid he in luence on he opening o he duplex should be ca ied ou o con i m his hypo hesis. As men ioned in he In oduc ion, R-loops we e desc ibed as ano he sou ce o genome ins abili y when no p ope ly egula ed, simila ly o G4s. In ac , R-loops o ma ion Discussion 111   is a ou ed by he p esence o a G- ich sequence in he coding s and. Thus, he o ma ion o an R-loop would gene a e a displaced G- ich s and wi h he po en ial o assemble a G4 s uc u e. The concu ence o R-loops and G4s c ea es a no el s uc u e ha combines bo h o hem and is called G-loop (Duque e e al., 2004). The cu en a ionale indica es ha he R-loop induced o ma ion o a G4 s abilises in e u n he R-loops s uc u e c ea ing a e y he modynamically esis an s uc u e (Lim and Hohng, 2020). In acco dance o his da a, bo h non-canonical DNA s uc u es ha e pa allel genome dis ibu ions (Kuzne so e al., 2018) and e en sha e simila cell unc ions. This e idence suppo s a syne gis ic ac i i y be ween hese wo s uc u es in hei biological ac i i y. In addi ion, s abiliza ion o G4s using ligands inc eases R-loop o ma ion in a simila ashion o G4 o ma ion and endo se R-loop-media ed eplica ion s ess (Chappidi e al., 2020; De Magis e al., 2019; Ko san is e al., 2020). One o he key oles o G4s is ansc ip ion egula ion and, ecen ly, he di e en ou comes o his egula ion by G4s ha e been ela ed o he o ma ion o G-loops (Lee e al., 2020). In e es ingly, he esul ob ained di e ed depending on which s and he G4 was loca ed. Thus, i he G4 is p esen in he empla e s and, i inhibi s ansc ip ion limi ing he mRNA p oduc ion whe eas, i loca ed in he coding s and, i induces R-loop o ma ion in he empla e s and and, in consequence, p omo es ansc ip ion. In line wi h his collabo a i e ole o G4s and R-loops in ansc ip ion, i would be in e es ing o in es iga e i he obse ed e ec in DNA esec ion by G4s is dependen on he s and in which hey a e loca ed and i i is also media ed by R-loop o ma ion. Indeed, an expe imen pe o med in Cama illo e al., 2021 sheds some ligh in his ega d. Appa en ly, he o e exp ession o RNase H1, an R-loop emo ing ac o , seems o be enough o o e come he DNA esec ion de ec p oduced by PIF1α deple ion. Al hough ou in es iga ions in his a ea a e s ill ongoing, his esul sugges s ha R-loop esolu ion is enough o eco e he de ec a ibu ed o G4 p esence and, hus, suppo a combined ac i i y o G4s and R-loops in DNA esec ion. Indeed, PIF1α is no only in ol ed in G4 esolu ion bu also in R-loops emo al (Zhou e al., 2014). Thus, any o hese s uc u es, o he combina ion o bo h, can be he ele an s uc u e ha PIF1α mus unwind o an e icien DNA esec ion. Discussion  112  Apa om PIF1α ole in DNA esec ion unwinding complica ed subs a es ha migh in e e e wi h his p ocess, his helicase migh be u he in ol e in he DDR. This is suppo ed by P ados-Ca ajal e al., 2020 ha unco e s he ole o a a ian o PIF1α ob ained by al e na i e splicing. This a ian lacks he helicase ac i i y and i is no equi e o he ea ly e en s o DNA epai bu i s equi ed o p omo e su i al upon exposu e o damaging agen s. Ano he ema kable cha ac e is ic o DNA esec ion is he conse ed coo dina ed ac i i y o a helicase wi h a nuclease o ini ia e HR. In Esche ichia coli, he RecQ helicase and he RecJ exonuclease coope a e in DSB esec ion (Amundsen and Smi h, 2003) and, he human equi alen s, BLM helicase and DNA2 also need o ac coupled o p ocess DNA ends (S u zenegge e al., 2014). The co ela ion o nucleases wi h hei speci ic helicase is di icul o de e mine because he e a e many candida e p o eins ha could be eligible as he igh pa ne . Since we demons a ed ha PIF1α is a po en ial impo an helicase in DNA esec ion, i s associa ion wi h a nuclease o e icien ly exe i s ole in DNA esec ion o e G4s p one s uc u es was conside ed logical. In a simila ashion o PIF1α, he nuclease FAN1 seemed o be in ol ed in HR and G4 esolu ion (MacKay e al., 2010; Po o e al., 2017). FAN1 is a p o ein wi h endo and exonuclease ac i i y impo an o he HR s ep o in e -s and-c osslink epai . Al hough p e ious s udies showed ha FAN1 deple ion led o a delayed disappea ance o RAD51 oci, hey did no obse e a ole o his nuclease in DNA esec ion (MacKay e al., 2010). In con adic ion wi h hese indings, we de ec ed a de ec in DNA esec ion, measu ed by RPA oci o ma ion, a e FAN1 down egula ion (Fig. R12A). Fu he mo e, his deple ion seemed o be epis a ic wi h PIF1α deple ion since he double-down egula ion o bo h p o eins ende ed he same de ec le el as he single deple ion o hem (Fig. R12A). This disc epancy could be explained by he di e ences in he expe imen condi ions used, such as he sou ce o DNA damage (cispla in e sus IR) and he eco e y ime a e he damage induc ion (1h s 24 h). Indeed, om his da a, i could be in e ed ha he e is a delay in esec ion ini ia ion and RPA emo al om ssDNA when FAN1 is deple ed. Tha would be he eason why he e a e less RPA posi i e cells a e 1 h bu , 24h la e , he pe cen age o RPA oci is highe . In addi ion, al hough he analysis o RPA oci o ma ion is conside ed he golden- ead ou o DNA esec ion, i is a low esolu ion echnique ha gi es in o ma ion abou he numbe o DSBs ha ini ia e Discussion 113   esec ion bu no abou he ex ension o his esec ion (C uz-Ga cía e al., 2014). S ikingly, using he SMART echnique, which o e comes his h owback, we obse ed a ise in esec ed ibe s leng h a e FAN1 deple ion (Fig. R12C). A simila pheno ype is obse ed in Domingo-P im e al., 2019 when EXOSC10 is deple ed, inhibi ion o RPA oci o ma ion and u he ex ension o ssDNA esec ed ibe s. A possible explana ion migh be ha FAN1 is pa icipa ing in bo h b anches o DNA esec ion. In he sho - ange esec ion, i may ac as a sca old o ec ui p o eins du ing his i s s ep bu hen, i is needed o s op he esec ion ex ension. The e o e, i s emo al om he DNA sequence p omo es he ex ension o he esec ed ibe s. Mo e da a elucida ing he ole o FAN1 in DNA esec ion should be collec ed om expe imen s widening he se ings used. Maybe he compa ison o di e en sou ces o DNA damage and di e en ime poin s could cla i y his ma e . Despi e ou esul s no being solid enough o s ablish he ole o FAN1 in DNA esec ion, hey do indica e ce ain ela ionship be ween PIF1α and FAN1 in his p ocess. Apa om hei co-down egula ion being epis a ic, as al eady men ioned, PIF1α o e exp ession seems o eco e he esec ion ini ia ion de iciency obse ed in cells lacking FAN1 (Fig. R12E). Al hough hese esul s a e p elimina y, hey migh help o u he elucida e how he cell deal wi h G4 s uc u es du ing DNA esec ion. 2. New sc eening o new ac o s a ec ing DNA esec ion Gi en he ele ance o CRISPR-Cas9 in genome edi ing, di e en app oaches ha e been applied o cope wi h he low-e iciency o CRISPR-Cas9 d i en HR. This, nex o i s high o - a ge e ec equency, s ongly a ec s i s conside a ion as he ideal gene-edi ing ool. Indeed, his ine iciency hampe s he gene a ion o KI, limi ing he modi ica ions ha can be in oduced in he genome. As al eady discussed in he In oduc ion, some o he s a egies de eloped o o e come his limi a ion include modi ying CRISPR-Cas9 componen s, such as Cas9 a ian s, gRNA modi ica ions o imp o ed dono s, o con olling he exp ession o ce ain p o eins ha migh edi ec he epai owa ds HR (Ghan a e al., 2021; Hu e al., 2018; Méndez-Mancilla e al., 2022). Ne e heless, despi e hese a emp s g ea ly con ibu ing o CRISPR-Cas9 de elopmen , hey migh be unp ac ical o a common Discussion  114  use in he lab since hey equi e an addi ional adjus men o he o iginal sys em. Tha is why he iden i ica ion o small-molecule compounds ha enhance CRISPR-Cas9 e iciency is a p omising app oach o easily and e e sibly a ou HR. Among he s udied compounds o imp o e CRISPR-Cas9 HR, we can ind a wide a ie y o a ge s. They can ei he modi y he ch oma in s a us o make he DNA ends mo e accessible, a ec cell cycle p og ession o a es cells in G2/S phases, when HR machine y is a ailable, o inhibi NHEJ ac o s (Lin e al., 2014; B. Liu e al., 2020; Ma uyama e al., 2015; Wiene e al., 2020). In gene al, he main aim o hese compounds is o dis up he balance be ween he wo main DSB epai pa hways owa ds HR. In line wi h his idea, ou sc eening o ac o s ha could be la e used o al e CRISPR-Cas9 e iciency was based on he sea ch o compounds ha could in luence DNA esec ion. In addi ion, he sc eening was pe o med using h ee di e en lib a ies: an i-diabe ic, na u al and an i-cance compounds. The majo i y o he chemo he apeu ics agen s a e based on he ac ha one o he hallma ks o cance is he exace ba ed gene ic ins abili y ha inc ease he mu a ion a e. As a consequence, he loss-o - unc ion mu a ions in key p o eins o cell biology, such as he DNA epai machine y, would obliga e he cell o ind an al e na i e o he los pa hway. Thus, a ge ing hese al e na i e epai pa hways using DSB epai p o eins inhibi o s is becoming an in e es ing s a egy o selec i ely kill cance cells (Kaelin, 2005). Tha is why i did no come as a su p ise he ac ha mos o he candida es ob ained in he sc eening came om he an i-cance lib a y (Fig. R16). Some e y-well known chemo he apeu ic agen s a e TOPO inhibi o s ha induce DNA damage and block DNA epai (He ene e al., 2018; Pommie e al., 2022; Sun e e al., 2010). Unexpec edly, h ee o ou candida es we e TOPO inhibi o s ha impai ed DNA esec ion. The mechanism o ac ion o hese d ugs in hinde ing DNA esec ion is la e u he discussed. Con a y o ou expec a ions o a g oup o compounds wi h a DNA-damage inducing ac i i y, he ea men wi h any o hese TOPO inhibi o s candida es did no inc ease γH2AX in ensi y. In ac , he use o ida ubicin HCl e en ende ed a educed in ensi y o his his one a ie y , in ag eemen wi h he obse ed his one e ic ion induced by an h acyclines, such as doxo ubicin, bu no by o he TOPO inhibi o s, such as e oposide (Pang e al., 2013). Discussion 115   Fo he es o he candida es, we could no ind a common pa hway ha would ela e hei ac i i y wi h hei ole in DNA esec ion. Despi e each o hem a ec ing di e en mechanisms, a gene al explana ion o hei e ec in DNA epai was he inhibi ion o a pa hway in ol ed in HR. Fo example, IWP-2 inhibi s Wn signalling pa hway. I has been desc ibed ha his signalling up egula ed LIG4 exp ession which is equi ed o NHEJ (Chappidi e al., 2020). The e o e, i s inhibi ion could be p omo ing esec ion ini ia ion due o he absence o a p ope NHEJ epai . Ano he example is he inhibi ion o mTOR signalling by To in2 which seemed o impai DNA esec ion e iciency. Indeed, i has been p e iously shown ha o in2 ea men a ec ed C IP oci o ma ion and g ea ly delayed γH2AX oci o ma ion and dissolu ion, in ag eemen wi h ou esul s (Udayakuma e al., 2016) (Fig. R17). Reg e ably, no all ou candida es p oduced he same e ec as desc ibed in he bibliog aphy. This is he case o AZ628, a BRAF inhibi o . Appa en ly, abolishing BRAF educed ATM and Chk1 ac i a ion which would lead o an ine icien DNA epai (Colome e al., 2019). Howe e , we obse ed he con a y e ec as an inc ease in RPA oci o ma ion was de ec ed. Simila ly, (-)-Epigalloca echin galla e was expec ed o a ou DNA esec ion since, in yeas , was desc ibed o inc ease he exp ession o HR genes (Chong e al., 2019). Con a y o his esea ch, we ob ained a dec ease in DNA esec ion e iciency in cells p e- ea ed wi h his compound. Fo he es o compounds, he p ocess in which hey pa icipa e ha e no been p e iously linked o an e ec in DNA epai , such as Ho denine, Bisindolylmaleimide I o he an i-diabe ic ac i i y o Lisinop il. Rega ding dimesna, i s ole in DNA esec ion could be a ibu ed o i s pa icipa ion in ac olein me abolism. Ac olein is an en i onmen al con aminan also p oduced endogenously by lipid pe oxida ion (Chung e al., 1996). This me aboli e induces DNA damage and causes an inhibi o y e ec on DNA epai (Sa ka , 2019; Tang e al., 2011). The e o e, he al e a ion o i s me abolism egula ion could in luence he DNA epai landscape, as obse ed by dimesna ea men . Fu he analysis will be equi ed o de ine he molecula mechanisms and biological meaning behind hese ac o s in DNA esec ion. Discussion  116  Rega dless o he mechanism by which ou d ug candida es in luence DNA esec ion, hei e ec in CRISPR-Cas9 e iciency was ou eal in e es . The use o a colou - based sys em o ace he epai ou comes a e he induc ion o a a ge ed DSB by CRISPR- Cas9 (Roidos e al., 2020) was employed o assess he in luence in his gene edi ing ool. Un o una ely, al hough he use o MLN2449 as a con ol did p oduce he expec ed e ec in CRISPR-Cas9 epai ou comes (inc eased KI e iciency), he esul s o he es o he compounds ob ained by his assay we e no in line wi h he e ec s obse ed in he RPA oci o ma ion assay (Fig. R19B). Indeed, only dimesna, o in2 and ida ubicin HCl ga e a signi ican pheno ype in his epo e . Bo h dimesna and ida ubicin HCl inc eased he popula ion o cells which conse ed GFP and mChe y exp ession. In o he wo ds, hey inc eased e o - ee epai o impai ed he ans ec ion o he cells wi h he gRNA. Rega ding he o he pa ame e s, o in2 and ida ubicin HCl ende s opposi e esul s despi e p oducing he same impai ed DNA esec ion pheno ype in he RPA oci o ma ion assay. To explain his disc epancy, we should ge a deepe unde s anding o he CAT-R epo e . La ge dele ions could be conside ed a p oduc o DNA esec ion o o he la ge ea angemen s such as ansloca ions, in e sions o la ge inse ions (Kosicki e al., 2018). This di e en causes leading o he loss o bo h GFP and mChe y signal canno be di e en ia ed wi h his epo e and hus, may explain why we ob ain opposi e e ec s. To in2 could be blocking DNA esec ion bu no in luencing o he kind o la ge dele ions and, in consequence, educe he la ge-dele ion epai ou come. On he con a y, despi e impai ing he gene a ion o esec ed acks, he ea men wi h ida ubicin HCl could be inc easing he numbe o ansloca ions o la ge in e sions e en s leading o a ne up ise o la ge dele ions e en s. Indeed, he use o TOP2 inhibi o s ha e been linked wi h inc eased ch omosomal ansloca ions and he TOP2-induced DSBs a e also conside ed o be a sou ce o genomic ea angemen s (Gómez-He e os e al., 2017; S anulla e al., 1997). Fu he suppo ing his heo y, we obse e a dec ease in he KI e iciency o he ssODN a e he ea men wi h ida ubicin HCl. This KI e iciency would ep esen he e iciency o SSTR, a ype o DNA- esec ion dependen epai mechanism ha is independen o RAD51 (Gallaghe and Habe , 2021). This kind o epai is based in SSA and, he e o e, is e y sensi i e o DNA esec ion. Changes in DNA esec ion e iciency would hen be mani es ed in changes in SSTR, as he changes obse ed in ida ubicin HCl ea men . Finally, we also Discussion 117   obse e an inc ease in SSTR a e o in2 ea men . This disag ees wi h ou p e ious esul s and u he da a collec ion is equi ed o disce n he cause o his ise. Gi en ha he ocus o he s udy was on assaying he e ec o some compounds in CRISPR-Cas9 media ed KI e iciency and he use o CAT-R epo e p oduced some con using esul s, we alida ed he esul s wi h ano he epo e sys em. This ime he DNA sequence empla e was p o ided by a plasmid, so he epai pa hway measu ed was no longe SSTR bu he classical HR mechanism. Once mo e, he use o MLN2449 con i med he co ec pe o mance o he expe imen since he use o his neddyla ion inhibi o inc eased he HR e iciency (Fig. R20B), as expec ed (Jimeno e al., 2015). In addi ion, in acco dance wi h he esul s ob ained by he RPA oci o ma ion assay, lisinop il sligh ly inc eased CRISPR-Cas9 media ed KI e iciency, al hough he ele ance o his inc ease is no e y no o ious. Simila ly o p e ious esul s, o in2 ea men dec eased HR e iciency which ag ees wi h i s ole inhibi ing DDR signalling (Liu e al., 2013). Su p isingly, AZ628 and Bisindolylmaleimide I dec eased he in eg a ion e iciency o he empla e in opposi ion o wha we expec ed om he inc ease in DNA esec ion e iciency obse ed a e he ea men wi h hese d ugs. Since he d ug ea men ime is longe in his expe imen compa ed o he RPA o ma ion assay, some a ia ions in he pheno ypes we e somehow expec ed and op imisa ion o he se ings, such as lowe ing he concen a ion in compensa ion o he longe pe iod o incuba ion wi h he compounds, should be pe o med. He e we s ablished a p o ocol o sea ch o new small molecule compounds ha could in luence he e iciency o CRISPR-Cas9 media ed genome edi ing. Al hough he se ings s ablished wo ked well in p edic ing he e ec in his p ocess by he con ols, hey we e no able o ende a clea ep oducible esul ega ding he candida es selec ed. The selec ion o a di e en subse o candida es om he sc eening o u he elucida ing he wo king condi ions o hese compounds migh gi e a mo e success ul esul in his aim.    VI. Ma e ials and me hods Ma e ials and me hods  127 1. CELL CULTURE PROCEDURES 1.1 Cell lines, g ow h media and condi ions The main cell lines used in his hesis include he human os eosa coma (U2OS), human emb yonic kidney (HEK293) and se e al s able cell lines c ea ed om U2OS using di e en cons uc s (Table M1). HEK293 and U2OS cells we e cul u ed in high-glucose Dulbecco’s Modi ied Eagle Medium (DMEM; Sigma, D6546), supplemen ed wi h 10% e al bo ine se um (FBS; Sigma, F7524), 2 mM L-glu amine (Gibco, 25030024), 100 U/ml penicillin and 100 μg/ml s ep omycin (Gibco, 15140122). S able cell lines de i ed om he abo e-men ioned ones we e cul u ed in he same condi ions bu adding ei he 0.5 mg/ml G418 (Sigma, A1720) o 1 μg/ml pu omycin (Sigma, P8833) o he medium, depending on he selec ion ma ke o he inse ed plasmid. Fo U2OS19p igh 13 cells, which con ain a doxycycline-inducible sys em (Table M2), DMEM wi hou phenol ed (Sigma, D1145) supplemen ed wi h 10% e acycline- ee FBS (Biowes , S181T) was used. HEPA class 100 incuba o s (The mo) we e used o main ained cell lines a 37ºC and 5% CO2. T ypsin-EDTA solu ion (Sigma, T4049) o accu ase (In i ogen, 00-4555-56) we e used o de ach he cells when equi ed. To seed speci ic numbe s o cells, an au oma ic cell coun e was used (Z2 Coul e Coun e , Beckman Coul e ). Fo long- e m p ese a ion o he cells, hey we e ha es ed and pelle ed by cen i uga ion a 500 g o 3 minu es. Cells we e hen esuspended in eezing solu ion (10% dime hyl sul oxide [DMSO] in FBS), aliquo ed in ubes and main ained a -80ºC o a leas 24 hou s. Then, hey we e ans e ed o liquid ni ogen-con aining anks o long- e m s o age. Ma e ials and me hods 128  1.2 siRNA ans ec ion Fo ansien p o ein knockdown, siRNA ans ec ion was pe o med using Lipo ec amine RNAiMAX ans ec ion eagen (The mo Fishe , 13778500) and di e en pla e o ma s, depending on he equi emen s o each speci ic expe imen . B ie ly, cells we e seeded and g own o 24 hou s. The day o ans ec ion, medium was eplaced by esh DMEM wi hou an ibio ics and cells we e incuba ed wi h a mix o siRNA and RNAiMAX dilu ed in Op i-MEM, acco ding o manu ac u e ’s ins uc ions. Cells we e hen incuba ed a 37ºC o 6 hou s be o e eplacing he media wi h esh comple e DMEM o minimise cell dea h. All siRNA-media ed knockdowns we e alida ed 48 hou s a e ans ec ion by wes e n blo o quan i a i e RT-PCR. The lis o all siRNAs used in his hesis can be ound on Table M3. 1.3 Plasmid DNA ans ec ion (Fugene) FuGENE HD T ans ec ion Reagen (P omega, E2311) was used o in oduce exp ession ec o s (Table M2) and p oduce s able cell lines. Fo 60 mm pla es, 300,000 cells we e seeded he day be o e ans ec ion. Then, 2 μg o plasmid DNA was mixed wi h FuGENE in a 3:1 FuGENE:DNA a io and dilu ed in Op i-MEM (Gibco, 11058- 021). The mix was incuba ed o 15 min a oom empe a u e and added d opwise o he pla e wi h gen le ocking. 1.4 DSB induc ion by ionizing adia ion To induce double s and b eaks in he cells, an i adia o de ice emi ing gamma ays (BIOBEAM GM 8000, Gamma-Se ice Medical GmbH) was used. The damage was gene a ed by exposu e o he samples o an encapsula ed adioac i e sou ce (Cs-137). Cells we e i adia ed in pla es con aining cul u e medium, using a dose o 10 Gy. The i adia o is adjus ed o he dose a e so he co e su ace does no exceed 3 μS /h. Ma e ials and me hods  129 2. MOLECULAR BIOLOGY PROCEDURES 2.1 Nucleic acids manipula ions 2.1.1 Plasmid DNA ampli ica ion Plasmid DNA was ampli ied by ans o ming compe en DH5α Esche ichia coli cells (Model O ganism Se ice, CABIMER) using he hea shock p o ocol. B ie ly, 0.1 ml o compe en bac e ia we e mixed wi h plasmid DNA and incuba ed on ice o 30 min. Cells we e hen hea shocked by placing hem a 42ºC o 35 seconds and incuba ed on ice o 5 min. Then, 1 ml o LB b o h (Fo medium, LB-B o h Lennox) was added, and he ans o med cells we e incuba ed a 37ºC o 30 min. Cul u es we e hen ha es ed by cen i uga ion a 6,000 pm o 2 min and pla ed in LB aga pla es supplemen ed wi h 100 μg/ml ampicillin (Sigma, 9518) o 25 μg/ml kanamycin (Sigma, K4000) depending on he an ibio ic esis ance casse e o he ans o med plasmid. Plasmid DNA was pu i ied om one single colony using Pu eYield Plasmid Maxip ep (P omega, A2393), ZymoPURE II Plasmid Maxip ep Ki (Zymo esea ch, D4203) o Midip ep (P omega, A2492) Sys ems, ollowing he manu ac u e ’s ins uc ions. DNA concen a ion was quan i ied by measu ing 260 nm abso bance using a NanoD op ND-1000 spec opho ome e . 2.1.2 DNA diges ion wi h es ic ion enzymes Fo DNA cloning and plasmid checking, es ic ion endonucleases om Taka a o New England Biolabs we e used acco ding o manu ac u e 's ins uc ions. DNA agmen s we e esol ed in aga ose gels o non-dena u ing polyac ylamide gels as desc ibed in he ollowing sec ions. 2.2 Cloning s a egies To p oduce he pUC19-2xLacO-2G4-non empla e plasmid, wo agmen s bea ing he c-myc p omo e po en ial G-quad uplex sequence we e cloned in andem in he pUC19_2xLacO plasmid (Table M2). Fi s , he G4 sequence was ampli ied om a p e ious plasmid om he lab (SA-GFP-G4_N uI, Table M2) by PCR using oligos con aining ei he Ma e ials and me hods 130  he HincII and Ps I si es as ups eam and downs eam p ime s, espec i ely, o he Ps I and EcoRI si es. Then, bo h he pUC19 plasmid and he ampli ied G4 sequences we e diges ed wi h HincII and Ps I. The p oduc s o he es ic ion eac ion we e liga ed by incuba ion wi h T4 DNA ligase (Taka a, 2011A) a 15ºC o e nigh . To in oduce he second G4 sequence in he plasmid con aining al eady one G4, he same s a egy was ollowed bu using he Ps I and EcoRI si es ins ead. 2.2.1 DNA elec opho esis in aga ose gels DNA elec opho esis was pe o med on gels con aining a a iable pe cen age o aga ose (P onadisa, 8010.22) depending on he size o he bands o di e en ia e, and RedSa e (In on Bio echnology, 21141) o E hidium B omide o DNA s aining, bo h dilu ed in 1x TAE bu e (40 mM T is-HCl pH7.6, 20 mM ace ic acid, and 1 mM EDTA). Loading bu e (Taka a) was added o DNA samples p io o loading in he gel, and 1 kb DNA ladde (gTPbio, GTPBM0002) was used o size es ima ion o he bands. S ained DNA agmen s we e isualized using an ul a iole ansillumina o (Bio-Rad) and analysed by Quan i y One so wa e. 2.2.2 DNA elec opho esis in non-dena u ing polyac ylamide gel DNA elec opho esis o esol e agmen s sho e han 200 bp was pe o med on 12% 29:1 ac ylamide:bis-ac ylamide gels unde non-dena u ing condi ions. Loading bu e (1 mL 6X T ackI Cyan/O ange Loading Bu e ) was added o DNA samples p io o loading in he gel, and 50 bp DNA ladde (The mo ishe , 10416014) was used o size es ima ion o he bands. Fo DNA s aining, he gel was incuba ed o 15 min in b omodeoxyu idine (B dU) dilu ed in 1X TE bu e (10mM T is-HCl con aining 1mM EDTA), washed wice o 10 min wi h dH2O and isualized using an ul a iole ansillumina o (Bio-Rad) and analysed by Quan i y One so wa e. 2.2.3 Si e-di ec ed mu agenesis Di ec ed changes in he DNA sequence o plasmids we e pe o med using QuickChange Ligh ning Si e-Di ec ed Mu agenesis ki s (Agilen Technologies, 210518) acco ding o manu ac u e 's ins uc ions. B ie ly, mu agenesis was pe o med by PCR using Ma e ials and me hods  131 designed p ime s con aining he desi ed mu a ion (Table M4). A e he PCR, pa en al DNA was deg aded by DpnI nuclease diges ion. Finally, mu a ion-con aining syn hesized DNA was ans o med in o compe en bac e ia and ampli ied. Candida es we e sen o DNA sequencing. 2.2.4 RNA ex ac ion RNA ex ac s we e ob ained om cells using NZY To al RNA Isola ion ki (NZY ech, MB13402) acco ding o manu ac u e 's ins uc ions. RNA concen a ion was quan i ied by measu ing 260 nm abso bance using a NanoD op ND-1000 spec opho ome e . To emo e ace amoun s o DNA, 1 μg RNA was ea ed wi h RQ1 RNase-F ee DNase (P omega, M6101) acco ding o manu ac u e 's ins uc ions. 2.2.5 Re e se ansc ip ion To syn hesize complemen a y DNA (cDNA), 1 μg RNA was subjec ed o e e se ansc ip ion eac ion using Maxima H Minus Fi s S and cDNA Syn hesis ki (The mo Scien i ic, K1652) acco ding o manu ac u e 's ins uc ions. 2.2.6 Quan i a i e PCR (qPCR) Quan i a i e PCR om cDNA was pe o med in iplica e o check siRNA- media ed knockdown o se e al p o eins. Fo his, iTaq Uni e sal SYBR G een Supe mix (Bio-Rad, 172-5124) was used ollowing manu ac u e ’s ins uc ions. DNA p ime s used o qPCR a e lis ed in Table M4. qPCR was pe o med in an Applied Biosys em 7500 FAST Real-Time PCR sys em. The compa a i e h eshold cycle (C ) me hod was used o de e mine ela i e ansc ip s le els (Bulle in 5279, Real-Time PCR Applica ions Guide, Bio-Rad), using he exp ession o he housekeeping gene β-ac in as in e nal con ol. Exp ession le els ela i e o β-ac in we e de e mined wi h he o mula 2-ΔΔC (Li ak and Schmi gen, 2001). 2.2.7 PCR PCR was pe o med using he indica ed p ime s in each expe imen (Table M4) and he Veloci y polyme ase (Bioline, BIO-21098) ollowing manu ac u e ’s ins uc ions. PCR Ma e ials and me hods 132  p oduc s we e hen pu i ied using NucleoSpin Gel and PCR Clean-up ki (Mache ey-Nagel, 740609.50). 2.3 P o ein analysis 2.3.1 P o ein ex ac ion unde dena u ing condi ions P o ein ex ac ion was ca ied ou by lysing he cells in Laemmli bu e 2x (125 mM T is-HCl pH 6.8, 4% SDS, and 20% glyce ol) using a plas ic cell sc ape (Sa s ed , 83.1830). Al e na i ely, cells we e ha es ed wi h ypsin, insed wi h PBS and esuspended in Laemmli 2X bu e . To educe sample iscosi y, p o ein ex ac s we e passed h ough a sy inge wi h a 0.5x16 mm needle (BD Plas ipak, 303175) a leas 10 imes. P o ein concen a ion was de e mined by he measu emen o abso bance a 280 nm using a NanoD op ND-1000 spec opho ome e . 2.3.2 Sodium dodecyl sulpha e polyac ylamide gel elec opho esis (SDS-PAGE) P o ein samples we e esol ed by size in 29:1 ac ylamide:bis-ac ylamide gels p epa ed a di e en concen a ions acco ding o he molecula weigh o he p o eins o analyse. SDS-PAGE was pe o med acco ding o p e iously desc ibed me hod (Laemmli, 1970). Samples we e dilu ed o ob ain simila p o ein concen a ion and hen SDS-PAGE loading sample bu e 4x (250 mM T is-HCl pH6.8, 8% SDS, 40% glyce ol, 20% β- me cap oe hanol (Sigma, M6250) and b omophenol blue) was added o a inal concen a ion o 1x. Samples we e incuba ed a 100ºC o 5 min and hen loaded in o he gels. P e-s ained p o ein ladde (gTPbio, GTPBM003) was also loaded as a molecula weigh ma ke . Elec opho esis was pe o med in a Mini-PROTEAN Te a Cell (Bio-Rad) wi h unning bu e (25 mM T is-HCl pH 8.3, 190 mM glycine, and 0.1% SDS) a 100-150 V. 2.3.3 Wes e n blo analysis A e elec opho esis, p o eins we e we - ans e ed using Mini T ans-Blo sys em (Bio-Rad) o 2-3 h a 400 mA in ans e bu e (25 mM T is-HCl pH 8.3, 190 mM glycine, 20% me hanol and 0.1% SDS) in o PVDF memb anes (Immobilon-FL; Millipo e, IPFL00010) p e iously ac i a ed in me hanol o 1 min and equilib a ed in ans e bu e . Ma e ials and me hods  133 Comme cial Odyssey Blocking Bu e (LI-COR Biosciences, 927-40000) was used o blocking he memb ane o a leas 1 hou a oom empe a u e o educe backg ound signal wi h he an ibodies, and hen, memb anes we e incuba ed o e nigh a 4ºC wi h he app op ia e p ima y an ibodies (Table M5) dilu ed in Odyssey Blocking Bu e con aining 0.1% Tween-20. Then, memb anes we e washed h ee imes o 5 minu es wi h 0.1% Tween-20 in TBS, ollowed by 1 hou incuba ion a oom empe a u e p o ec ed om ambien ligh wi h he co esponding IRDye seconda y an ibodies (Table M6) dilu ed in blocking bu e con aining 0.1% Tween-20. A e wa ds, memb anes we e washed again h ee imes wi h 0.1% Tween-20 in TBS and d ied be o e scanning. Image acquisi ion was pe o med in Odyssey CLx Imaging Sys em (LI-COR Biosciences) a wo in a ed wa eleng hs (700 and 800 nm) o di e en ial imaging o bo h an i- abbi and an i-mouse seconda y an ibodies a he same ime. ImageS udio .2.1 so wa e was used o scanning and analysis o he images. 2.3.4 P o ein A immunop ecipi a ion To immunop ecipi a e endogenous BRCA1, U2OS cells we e ha es ed in lysis bu e (50 mM T is-HCl, pH 7.4, 100 mM NaCl, 1 mM EDTA, 0.2% T i on X-100, 1X p o ease inhibi o s (Roche), 1X phospha ase inhibi o cock ail 1 (Sigma)) and incuba ed wi h 100 U/ml Benzonase (VWR, 70746-4) 30 min on ice o deg ade DNA. 1 mg o p o ein ex ac s was hen p eclea ed wi h 50 uL o washed magne ic p o ein A Dynabeads (No ex, 10002D) unde gen le agi a ion a 4ºC o 1h. P eclea ed samples we e hen incuba ed wi h 1 μg an i-BRCA1 an ibody (Table M5) o an equi alen amoun o mouse IgG (Sigma, I8140) as nega i e a con ol o 30 min a 4ºC. The emaining dynabeads we e added a e wa ds o he mix u e o p o eins and an ibodies o be incuba ed oge he o e nigh a 4ºC wi h gen le agi a ion. Beads we e hen washed h ee imes wi h lysis bu e , and he p ecipi a e was elu ed in Laemmli bu e and esol ed in SDS-PAGE as desc ibed in he p e ious sec ions. 2.3.5 GFP immunop ecipi a ion (IP) U2OS cells exp essing GFP o GFP-PIF1 we e subjec ed o p o ein ex ac ion unde na i e condi ions by sc apping wi h an immunop ecipi a ion lysis bu e (10 mM T is-HCl Ma e ials and me hods 140  manu ac u e ’s p o ocol in Va ioskan Flash (The mo Elec on co po a ion) and no malized wi h con ol. 3.4 Fluo escen p o ein analysis by low cy ome y 3.4.1 CAT-R epo e To s udy he e ec o he compounds selec ed om he d ug sc eening on CRISPR- media ed knock-ou (KO) and knock-in (KI) e iciency, he HEK293 cell line bea ing he CAT-R (Colo Assay T acing-Repai ) epo e (Roidos e al., 2020) was used. 3.4.1.1 Desc ip ion o he CAT-R epo e The CAT-R epo e (Fig. M2) consis s o wo coding sequences o he luo escen p o eins mChe y and euka yo ic g een luo escen p o ein (eGFP) linked wi h a sel - clea ing P2A pep ide. This epo e was hen in eg a ed a a single genomic locus in he HEK293 cell line o exp ess doxycycline inducible Cas9. Basically, a DSB is induced by CRISPR-Cas9 using a guide RNA (gRNA, Table M4) a ge ing he eGFP locus loca ed 355 bp downs eam o he P2A pep ide. The dis inc pa hways used o epai his DSB can gene a e h ee di e en popula ions ha di e s om each o he in hei luo escen signals: 1) Small InDels epai -de i ed ameshi mu a ions lead o loss o GFP signal (mChe y+/GFP-). In mos cases, his small InDels a e expec ed o be due he ac ion o c-NHEJ o al -EJ. The pe cen age o cells wi h his ou come p oduc could be conside ed as he equency o KO e iciency. 2) Dele ions o o he ea angemen s la ge han ~420 base pai s (bps) lead o loss o bo h mChe y and eGFP signal (mChe y-/GFP-). The mChe y sequence is 420 bp away om he cu ing si e so he loss o i s signals by la ge dele ions a e supposed o be p oduc s o DNA esec ion. 3) GFP-BFP con e sion by single-s and empla e epai (SSTR) leads o gain o Blue Fluo escen P o ein (BFP) signal by losing eGFP signal (mChe y+/BFP+/GFP-). eGFP can be con e ed o a BFP by a single amino acid change, allowing he measu e o SSTR by p o iding a single-s anded oligo-deoxynucleo ide (ssODN, Table M4) empla e oge he wi h he gRNA Ma e ials and me hods  141 a ge ing eGFP. The pe cen age o cells wi h his ou come p oduc could be conside ed as he equency o KI e iciency. 4) Un ans ec ed cells/e o - ee epai (ei he by NHEJ o HR) lea es bo h mChe y and eGFP sequences in ac (mChe y+/GFP+). The exac e en s ha lead o he mChe y+/GFP+ popula ion canno be esol ed wi h his epo e . 3.4.1.2 The CAT-R epo e in i o Cells bea ing a single copy in eg a ion o he CAT-R sys em we e used o analyse he in luence o a ba e y o d ugs in he CRISPR-media ed KI and KO e iciency. To ca y ou he assay, 20000 cells/well we e pla ed in U-bo om 96-well pla es and he cul u e media was supplemen ed wi h 1 μg/ml Doxycycline (Sigma, D9891) o induce Cas9 exp ession. One day a e seeding, cells we e 60-80% con luen o ans ec ion. The Al -RR CRISPR- Cas9 c RNA and ac RNA (IDT, 1072532) we e used o o m he guide RNA complex (gRNA, Table M4) a ge ing he eGFP sequence. To o m he gRNA, each RNA oligo (Al - RR CRISPR-Cas9 c RNA, ac RNA) was esuspended in nuclease- ee Duplex bu e (IDT, 11-01-03-01) o a inal concen a ion o 100 μM. Then, he wo RNA oligos we e Figu e M2. Schema ic ep esen a ion o he CAT-R (Colo Assay T acing-Repai ) epo e In he CAT-R sys em, he di e en epai mechanisms o he CRISPR-induced DSB de i es in di e en epai ou comes ha can be aced by he luo escen signal o he cell popula ion. Small InDels lead o loss o GFP signal (mChe y+/GFP-). Dele ions o o he la ge ea angemen s lead o loss o bo h mChe y and eGFP signal (mChe y-/GFP-). GFP o BFP con e sion by single-s and empla e epai (SSTR) leads o gain o Blue Fluo escen P o ein (BFP) signal by losing eGFP signal (mChe y+/BFP+/GFP-). Finally, un ans ec ed cells o e o - ee epai lea es bo h mChe y and eGFP sequences in ac (mChe y+/GFP+). mChe y P2A eGFP mChe y P2A eGFP mChe y BFP mChe y P2A eGFP mChe y P2A eGFP mChe y P2A eGFP mChe y P2A P2A eGFP DSB DSB epai 5’ 3’ ssODN dono empla e Small InDels La ge Dele ions Un ans ec ed/ E o - ee Single S and Templa e Repai (SSTR) < 50-bp indel > 250-bp indel Cas9 a ge Ma e ials and me hods 142  mixed in equimola concen a ions o c ea e a inal duplex concen a ion o 3 μM in nuclease- ee Duplex bu e . Finally, gRNA complex was hea ed a 95 °C o 5 min and hen allowed o cool o oom empe a u e (20-25 °C). We hen ans ec ed he gRNA a ge ing eGFP oge he wi h he single-s anded oligodeoxynucleo ides (ssODNs, Table M4) as a empla e o GFP-BFP con e sion. The ssODN was p e iously dilu ed in nuclease- ee Duplex bu e o a inal concen a ion o 100 µM. To ans ec he cells Lipo ec amine RNAiMAX ans ec ion eagen (The mo Fishe , 13778500) was used acco ding o he use manual. The gRNA complex was combined wi h Lipo ec amine RNAiMAX in a a io o 2:1 in Op i-MEM Medium o a inal concen a ion o 30 nM o he gRNA and 9 nM o ssODN. A he same ime o he ans ec ion wi h CRISPR gRNA, cells we e incuba ed wi h 1 µM o he compounds selec ed (see Table R3) o h ee days o s udy hei e ec on CRISPR e iciency. Then hey we e analysed in a high- h oughpu FACS LSR Fo essa analyze (BD Biosciences). These compounds we e s o ed in −20 °C as 1 mM s ocks in DMSO un il needed. 3.4.2 GFP KI e iciency To speci ically assess CRISPR-Cas9 media ed KI e iciency, U2OS cells we e elec opo a ed wi h CRISPR/Cas9-RAG1 and GFP dono -RAG1 plasmids using he Gene Pulse Xcell elec opo a ion sys em (BioRad). Fi s , 1.5x106 U2OS cells we e ha es ed, washed wi h PBS and esuspended in 100 µL o elec opo a ion bu e (2 mM HEPES pH 7.2, 250 mM D-Mani ol, 1 mM MgCl2 and 15 mM K2HPO4). Then, 2.5 µg GFP dono - RAG1 and 0.6 µg CRISPR/Cas9-RAG1 we e added o he mix o cells and elec opo a ed (squa e-wa eleng h, 140 V, 6 pulses, 1.5 ms pulse leng h, 1.5 s in e al leng h) using 0.2 mm cu e es (The mo Scien i ic, 10227570). Finally, app oxima ely 280000 cells we e seeded in 12-well pla es and ea ed wi h 1 µM o he posi i e candida es om he sc eening o MLN2449, DMSO o H2O as con ols. 72h a e seeding cells we e ha es ed and he pe cen age o GFP posi i e cells was de ec ed using he FACS LSR Fo essa analyze (BD Biosciences). Ma e ials and me hods  143 3.5 RPA/γH2AX oci immuno luo escence U2OS cells ei he down egula ed o o e exp essing di e en p o eins o ea ed wi h a a ie y o compounds, as indica ed in each speci ic case, we e seeded on co e slips (Labolan, 20012). Cells g owing on co e slips we e i adia ed (10 Gy) o mock ea ed, incuba ed o 1 hou o allow oci o ma ion, washed once wi h cold PBS and collec ed. Co e slips we e ea ed wi h p e-ex ac ion bu e (25 mM T is-HCl pH 7.5, 50 mM NaCl, 1 mM EDTA, 3 mM MgCl2, 300 mM suc ose and 0.2% T i on X-100) o 5 min on ice o emo e he nucleoplasmic and cy oplasmic p o eins, lea ing behind he ch oma in-bound and ma ix-associa ed p o eins. Cells we e hen washed once wi h cold PBS, ixed wi h 4% pa a o maldehyde (w/ ) in PBS o 15 minu es on ice, washed h ee imes wi h PBS and incuba ed o 1 h wi h blocking bu e (5% FBS in PBS). A e wa ds, cells we e co-s ained wi h he app op ia e p ima y an ibodies (Table M5) in blocking bu e o 2 h a oom empe a u e o a 4ºC o e nigh , washed h ee imes wi h PBS and hen co-immunos ained wi h he app op ia e seconda y an ibodies (Table M6) in blocking bu e o 1 hou a oom empe a u e in he da k. Co e slips we e washed again wi h PBS, dehyd a ed in o inc easing concen a ions o e hanol, d ied and moun ed on o glass slides using Vec ashield moun ing medium con aining 4',6-diamidino-2-phenylindole (DAPI) (Vec o Labo a o ies, H-1200). Samples we e isualized using Leica AF6000 luo escence mic oscope and a 63x objec i e. In each expe imen , a leas 200 cells we e analysed. 3.6 D ug sc eening o a ec DNA end esec ion In o de o sea ch o ac o (s) a ec ing end esec ion, a sc eening o h ee di e en compounds lib a ies was pe o med. The h ee lib a ies used we e:  Na u al P oduc Sc eening Lib a y (Selleckchem, L1400): 133 inhibi o s  An i-cance Compound Lib a y (Selleckchem, L3000): 386 bioac i e compounds  An idiabe ic Compound Lib a y (Selleckchem, L2900): 33 inhibi o s P io o s a he immuno luo escence expe imen s, each lib a y was manually dilu ed ei he in DMSO o Milli-Q wa e o a inal s ock o 20 µM in 96-well pla es ha we e kep a -20ºC un il use. To pe o m he sc eening, 6000 U2OS cells we e pla ed in 96- Ma e ials and me hods 144  well pla es (6005550, Pe kinElme ). One day a e seeding, con ol wells we e ans ec ed wi h C IP o con ol siRNA in medium wi hou glu amine. This medium was changed a e 6 h. 72 h a e seeding, cells we e ea ed ei he wi h he indica ed d ugs (1 µM inal concen a ion) om h ee p e iously men ioned Selleckchem lib a ies o wi h he co esponding con ols DMSO, H2O o MLN2449. One hou a e d ug ea men , cells we e i adia ed wi h 10 Gy and kep a 37ºC o 1h. The p o ocol ollowed o pe o m he immuno luo escence expe imen is desc ibed in he p e ious sec ion (see sec ion 3.5). B ie ly, pla es we e washed once wi h PBS ollowed by ea men wi h p e-ex ac ion bu e (25 mM T is-HCl, pH 7.5, 50 mM NaCl, 1 mM EDTA, 3 mM MgCl2, 300 mM suc ose and 0.2% T i on X-100) o 5 min on ice. Then, cells we e ixed wi h 4% pa a o maldehyde (San a C uz) o 15 min on ice. A e wo washes wi h PBS, cells we e blocked wi h blocking solu ion (5% FBS in PBS) o 1h a RT. P ima y an ibodies agains RPA and γH2AX (Table M5) we e incuba ed o 2h a RT, washed wi h PBS and hen incuba ed wi h seconda y an ibodies (Table M6) o 1h a RT in he da k. To isualize cell nuclei, cells we e washed wi h PBS, s ained wi h Hoechs 33258 (2 µg/mL, 861405, Sigma) o 10 min and washed wi h PBS again. The pla es we e sealed wi h pa a ilm and kep in he da k a 4ºC un il image acquisi ion. To isualize he cells, pla es we e imaged using an ImageXp ess Mic o (Molecula De ices) a 40X magni ica ion, and blue, g een and Texas Red il e s we e used o de ec nuclei, γH2AX and RPA signals espec i ely. Images we e analysed, and he numbe o RPA oci pe cell was quan i ied au oma ically wi h he Image Me aExp ess So wa e (Molecula De ices). Fo his, a leas 200 cells we e sco ed pe well. To acili a e he compa ison be ween expe imen s, he median o he numbe o RPA oci pe cell was no malized wi h each DMSO o H2O con ol. Condi ions ha skewed he balance owa ds inc eased HR epai esul ed in a old inc ease abo e 1. In con as , a ne dec ease o his a io ( o example, alues below 1) ep esen ed an imbalance owa ds NHEJ. Ma e ials and me hods  145 3.7 PIF1α oci a I-SceI DSBs To s udy he ec ui men o PIF1 o DSB, U2OS19p igh 13 cells (Lemaî e e al., 2012) ba ing an I-SceI clea age si e is lanked by andem epea s o he lac ope a o DNA sequences (lacO) we e used. These cells we e co- ans ec ed wi h a Che y-lacI and GFP- PIF1α plasmids in a 10mm pla es con aining 2 co e slips and ea ed o no wi h doxycycline hycla e (Sigma, D9891) a a inal concen a ion o 1 μg/ml o 14 h o 20 h, as indica ed. Then, cells we e washed and ixed wi h 4% pa a o maldehyde (w/ ) in PBS o 15 min, washed again wice wi h PBS and pe meabilized wi h 0.5% T i on X-100 (Sigma Ald ich 9036-19-5) in PBS o 15 min a oom empe a u e. Cells we e hen washed wi h PBS and blocked o 1h a oom empe a u e using 3% BSA (Sigma Ald ich, A4503) dilu ed in PBS wi h 0.1% Tween 20 (Sigma Ald ich). A e wa ds, co e slips we e incuba ed o 1 hou a oom empe a u e wi h p ima y an ibodies agains GFP and γH2AX (Table M5) p epa ed in blocking solu ion, washed h ee imes wi h 0.1% Tween-20 in PBS, incuba ed o 1 h a oom empe a u e wi h he app op ia e seconda y an ibodies (Table M6) p epa ed in blocking solu ion and washed h ee imes wi h PBS. Co e slips we e moun ed using Vec ashield moun ing medium (Vec o Labo a o ies) con aining DAPI. To isualize and acqui e he images, a LEICA con ocal mic oscope TCS SP5 was used wi h a HCX PL APO lambda blue 63X/ 1.4 OIL objec i e. 4. STATISTICAL ANALYSIS Unless o he wise speci ied in he igu e legend, s a is ical signi icance was de e mined wi h a -S uden es using he PRISM so wa e (G aphpad So wa e Inc.). S a is ically signi ican di e ences we e labelled wi h one, wo o h ee as e isks i P < 0.05, P < 0.01 o P < 0.001, espec i ely. Ma e ials and me hods 146  5. TABLES Table M1. Cell lines used in his Thesis CELL LINE SELECTION SOURCE/REFERENCE U2OS No ATCC HTB-96 U2OS-GFP G418 This Thesis U2OS-GFPPIF1α G418 Kind gi om S ephen P. Jackson (Rod iguez e al., 2012) U2OS19p igh 13 G418, pu omycin Kind gi om E i Sou oglou (Lemaî e e al., 2014) HEK293-CATR G418 Kind gi om Balca R. Ma din (Roidos e al., 2020) Table M2. Plasmids used in his Thesis PLASMID DESCRIPTION SELECTION MARKER SOURCE/REFERENCE pEGFP-C1 Exp ession o EGFP Kanamycin/G418 Clon ech (6084-1) GFP-PIF1α Exp ession o GFP used o hPIF1α Kanamycin/G418 Kind gi om S ephen P. Jackson (Rod iguez e al., 2012) GFP-PIF1α E307Q Exp ession o GFP used o hPIF1α con aining E307Q mu a ion Kanamycin/G418 This Thesis pUC19 2xLacO Vec o con aining wo lacO sequences o assay DNA esec ion Ampicilin Kind gi om Pe Cejka pUC19 2xLacO 2xG4 NT Vec o con aining wo lacO sequences and wo G4 sequences clones in andem in he non- empla e s and o assay DNA esec ion Ampicilin This Thesis Ma e ials and me hods  147 Con inua ion o Table M2. Plasmids used in his Thesis PLASMID DESCRIPTION SELECTION MARKER SOURCE/REFERENC E SA-GFP G4 N uI Vec o ha bou ing he SA-GFP epo e wi h a G4 sequence cloned nex o he GFP sequence o analyse he in luence o G4 in SSA e iciency Ampicilin/Pu om ycin Gene a ed in Pablo Hue as’ lab Che y-LacI Exp ession o Che y-lacI Kanamycin/G418 Kind gi om E i Sou oglou CRISPR/Cas9- RAG1 Exp ession o Cas9 and gRNA a ge ing RAG1 gene Ampicilin Kind gi om Ron Jachimowicz and Toni Ca homen GFP Dono - RAG1 Dono o he GFP sequence o be in eg a ed in RAG1 gene Ampicilin Kind gi om Ron Jachimowicz and Toni Ca homen Table M3. siRNAs used in his Thesis siRNA SENSE SEQUENCE (5’-3’) REFERENCE SUPLIER Non- a ge (NT) pool UGGUUUACAUGUCGACUAA D-001810-10 Dha macon C IP GCUAAAACAGGAACGAAUC López- Saa ed a e al., 2016) Sigma PIF1 3’UTR GGAUGUUCUCAGGUUGUAUUUAUTT This Thesis Sigma MRE11 GAAAGGCUCUAUCGAAUGU D-009271-02- 0005 Dha macon EXO1 GAAGUUUCGUUACAUGUGU D-013120-01- 0005 Dha macon DNA2 GCUAAACCGUGAAGCAAGA D-026431-01- 0005 Dha macon FAN1 GUAAGGCUCUUUCAACGUA This Thesis Sigma BRCA1 CAGCUACCUUCCAUCAUA This Thesis Sigma Ma e ials and me hods 148  Table M4. P ime s used in his Thesis PRIMER NAME SEQUENCE (5’-3’) USE PIF1α F/ CCCTGGATTGTGTGGAGATT R/ ACTCCAGACTGAGGCTCCTG RT-qPCR o PIF1α PIF1α F/CCTATGTGGCCCTTTCTCG R/GGTTTGGGTCCATGTTCTCC RT-qPCR o PIF1α Mu E307Q Fw CGGTTGGTCATTGACCAGATC TCAATGGTGGAG Mu a ion o E307Q o PIF1α by QuickChange Mul isi e ki Mu E307Q R CTCCACCATTGAGATCTGGTC AATGACCAACCG Mu a ion o E307Q o PIF1α by QuickChange Mul isi e ki GFPPIF1α Pu exp ess Fw GCGAATTAATACGACTCACT ATAGGGCTTAAGTATAAGGA GGAAAAAATATGGTGAGCAA GGGCGAGGAGC TGTTC In i o p oduc ion o GFP-PIF1α p o ein GFPPIF1α Pu exp ess R AAACCCCTCCGTTTAGAGAG GGGTTATGCTAGTTTCAGAGG ATTGGGTCCATGTTCTC In i o p oduc ion o GFP-PIF1α p o ein FAN Fw GATGAGCAGGAGAAGGGAAT TG RT-qPCR o FAN1 FAN R ACGGAATCTGAATTGTGGAC AC RT-qPCR o FAN1 EXO1 Fw CTGAAGTGTTTGTGCCTGAC RT-qPCR o EXO1 EXO1 R CCACAACTGCACCAC RT-qPCR o EXO1 DNA2 Fw GGAGAAGAGTGGCAGTT RT-qPCR o DNA2 DNA2 R TCTGTCACCTGCCATTAG RT-qPCR o DNA2 MRE11 Fw RT-qPCR o MRE11 MRE11 R RT-qPCR o MRE11 F1/F2 Ps I G4 Fw TAAGCACTGCAGTAATGGGA ACCCGGGAGGGG Cloning o he c-Myc G4 sequence in pUC19 2xLacO F1/F2 Ps I G4 R TGCTTACTGCAGTTATTCGCG ACCCCACCT Cloning o he c-Myc G4 sequence in pUC19_2xLacO F1 HincII G4 Fw TAAGCAGTTGACGTAATGGG AACCCGGGAGGGG Cloning o he c-Myc G4 sequence in pUC19 2xLacO F1 HincII G4 R TGCTTAGTCAACTTATTCGCG ACCCCACCTT Cloning o he c-Myc G4 sequence in pUC19_2xLacO Ma e ials and me hods  149 Con inua ion o Table M4. P ime s used in his Thesis PRIMER NAME SEQUENCE (5’-3’) USE F2/F3 EcoRI G4 R TGCTTAGAATTCTTATTCGCG ACCCCACCT Cloning o he c-Myc G4 sequence in pUC19 2xLacO gRNA GFP CTCGTGACCACCCTGACCTAC GG c RNA sequence a ge ing GFP gene ssODN BFP CCTGAAGTTCATCTGCACCAC CGGCAAGCTGCCCGTGCCCT GGCCCACCCTCGTGACCACC CTGAGCCACGGGGTGCAGTG CTTCAGCCGCTACCCCGACCA CATGAAGCAGCACGACTTCT TCAAGTCCGCCATGCC Sequence complemen a y o GFP o con e GFP sequence o BFP sequence Table M5. P ima y an ibodies used in his Thesis ANTIBODY SPECIES REFERENCE APPLICATION (DILUTION) RPA32 Mouse Abcam (ab2175) IF (1:500) ϒH2AX Rabbi Cell Signaling (2577L) IF (1:1000) GFP Rabbi San a C uz (sc-8334) IF (1:500), WB (1:1000) B dU Mouse Sigma Ald ich (RPN202) IF (1:100) PIF1 Mouse San a C uz (sc-48377 IF (1:500), WB (1:500) B ca1 Mouse San a C uz (sc-6954) IF (1:500), WB (1:500) G4 Goa Absolu e An ibody (Ab00174-24.1) IF (1:100) An i-His Mouse GE Heal hca e (27-4710-01) WB (1:500) TOP2β (H-286) Mouse San a C uz (sc-13059) WB (1:500) αTubulin Mouse Sigma (T9026) WB (1:50000)