Lack of sugar discrimination by human pol requires a single glycine residue

Lack o suga disc imina ion by human Pol m equi es a single glycine esidue Jose ÂF. Ruiz, Raquel Jua  ez, Miguel Ga cõ Âa-Dõ Âaz, Glo ia Te ados, Angel J. Piche , Se gio Gonza Âlez-Ba e a, An onio R. Fe na Ândez de Henes osa and Luis Blanco* Cen o de Biologõ Âa Molecula Se e o Ochoa (CSIC-UAM), Campus de la Uni e sidad Au o Ânoma de Mad id, Can oblanco, 28049 Mad id, Spain Recei ed Ap il 11, 2003; Re ised and Accep ed June 9, 2003 ABSTRACT DNA polyme ase mu (Pol m) is a no el amily X DNA polyme ase ha has been sugges ed o play a ole in mic o-homology media ed joining and epai o double s and b eaks. We show he e ha human Pol mis no able o disc imina e agains he 2¢-OH g oup o he suga moie y. I inse s NTPs wi h an e ®- ciency ha is <10- old lowe han ha o dNTPs, in sha p con as wi h he >1000- old disc imina ion cha ac e is ic o mos DNA-dependen DNA poly- me ases. The lack o suga disc imina ion by Pol m is demons a ed by i s abili y o add NTPs o bo h DNA and RNA p ime s ands, and o inse bo h deoxy- and ibonucleo ides on g owing nucleic acid chains. 3D-modelling o human Pol mbased on he a ailable Pol band TdT s uc u al in o ma ion allowed us o p edic candida e esidues in ol ed in suga disc imina ion. Thus, a single amino acid sub- s i u ion in which Gly433 esidue o Pol mwas mu a ed o he consensus y osine p esen in Pol b, p oduced a s ong inc ease in he disc imina ion agains ibonucleo ides. The unusual capaci y o inse bo h NTPs and dNTPs will be discussed in he con ex o he p edic ed oles o Pol min DNA epai . INTRODUCTION T adi ionally, polyme ases ha e been classi®ed in o se e al classes on he basis o hei subs a e speci®ci ies: usage o DNA and/o RNA as a empla e and dNTPs o NTPs as subs a es. The wo main classes, DNA and RNA poly- me ases, use DNA as empla e bu di e in hei abili y o inco po a e ei he dNTPs o NTPs on nucleic acid chains. Na u al DNA-dependen DNA polyme ases exhibi high suga disc imina ion, selec ing dNTPs o e NTPs by se e al o de s o magni ude (1±3). Based on exis ing e idence, he abili y o disc imina e be ween nucleo ide subs a es is a consequence o sub le s uc u al di e ences among DNA polyme ases. Thus, single amino acid subs i u ions in oduced in he ac i e si e o di e en DNA polyme ases had d ama ic e ec s on he deg ee o disc imina ion be ween he 2¢-H o -OH g oup o he suga moie y, p oducing mu an DNA polyme ases capable o inco po a ing NTPs e ®cien ly (2,4±7). A s iking excep ion o his ule is e minal deoxynucle- o idyl ans e ase (TdT), a DNA polyme ase which is no s ic ly DNA-dependen . TdT con ibu es o he di e si®ca ion o an igen ecep o s by adding non- empla ed nucleo ides o gene segmen junc ions o Ig and T-cell ecep o genes du ing V(D)J ecombina ion ( e iewed in 8). Soon a e i s isola ion, i was shown ha TdT could use bo h NTPs and dNTPs as nucleo ide polyme iza ion subs a es (9). These esul s ha e been ecen ly e-e alua ed, con® ming ha mu ine TdT, al hough ha ing a s ong p e e ence o deoxynucleo ide- e mina ed o e ibonucleo ide- e mina ed p ime s, is almos incapable o disc imina ing be ween NTPs and dNTPs (10). TdT belongs o he Pol X DNA polyme ase amily, e olu iona ily ela ed o a la ge g oup o nucleo idyl ans- e ases (11,12). Fou Pol X amily membe s a e exp essed in e eb a es: TdT (9) and Pol b(13), bo h s udied ea lie , and he ecen ly desc ibed Pol l(14) and Pol m(15). While Pol lis s uc u ally and unc ionally simila o Pol b,Polmis he closes ela i e o TdT, showing a 42% amino acid iden i y (16). Pol mis a ecen ly desc ibed DNA polyme ase wi h an in insic e minal ans e ase ac i i y, e o -p oneness du ing DNA-dependen DNA syn hesis, and speci®c exp ession pa e n in seconda y lymphoid o gans in mammals. Mo e- o e , Pol mhas he unusual abili y o p omo e ansien p ime - empla e misalignmen s, which mainly esul in ameshi (17) and base subs i u ion (J.F. Ruiz, K. Bebenek, T. Kunkel and L. Blanco, unpublished esul s) mu a ions. These da a, oge he wi h he ecen ly demons a ed in e ac ion wi h se e al componen s o he non-homologous end joining (NHEJ) machine y (18), sugges ha Pol mcould ha e an impo an ole in he end-joining pa hway o epai o DNA double s and b eaks. Mo eo e , he misalignmen capaci y o Pol mappea s o be c ucial o ca y ou bypass syn hesis o ce ain DNA lesions (19,20). We p o ide he e no el insigh s in o he biochemical cha ac e iza ion o human Pol m, demons a ing i s s iking abili y o inco po a e and elonga e NTPs o nucleic acid chains. This unusual capaci y mainly elies on a single glycine esidue ha subs i u es o a conse ed a oma ic esidue p esen in Pol band in mos membe s o he Pol X amily o DNA-dependen DNA polyme ases. *To whom co espondence should be add essed. Tel: +34 91 397 8493; Fax: +34 91 397 4799; Email: [email p o ec ed] Nucleic Acids Resea ch, 2003, Vol. 31, No. 15 4441±4449 DOI: 10.1093/na /gkg637 Nucleic Acids Resea ch, Vol. 31 No. 15 ãOx o d Uni e si y P ess 2003; all igh s ese ed MATERIALS AND METHODS Ma e ials Unlabelled ul apu e dNTPs and NTPs and [g-32P]ATP (3000 Ci/mmol) we e pu chased om Ame sham Pha macia Bio ech. Syn he ic DNA oligonucleo ides we e ob ained om In i ogen (P15-DNA, 5¢TCTGTGCAGGTTCTT 3¢; T32- DNA (A), 5¢TGAAGTCCCTCTCGACAAAGAACCTG- CACAGA 3¢; T32-DNA (C), 5¢TGAAGTCCCTCTCGAC- CAAGAACCTGCACAGA 3¢; T32-DNA (G), 5¢TGA- AGTCCCTCTCGACGAAGAACCTGCACAGA 3¢; T32- DNA (T), 5¢TGAAGTCCCTCTCGACTAAGAACCTG- CACAGA 3¢; D16-DNA, 5¢GTCGAGAGGGACTTCA 3¢) and RNA oligonucleo ide om Geno ek (P15-RNA, 5¢ UCUGUGCAGGUUCUU 3¢). All he oligonucleo ides desc ibed abo e we e pu i®ed by elec opho esis on 8 M u ea, 20% polyac ylamide gels. T4 polynucleo ide kinase and T4 DNA ligase we e om New England Biolabs. TdT, Taq DNA polyme ase and P u DNA polyme ase we e om P omega. Cons uc ion and pu i®ca ion o wild- ype and mu an o ms o human Pol m Human Pol mcDNA was ob ained as p e iously desc ibed (15). Si e-di ec ed mu a ions we e in oduced in o a human Pol mo e exp ession plasmid (pRSETa-hPol m) by a PCR- based me hod (QuikChangeâSi e-Di ec ed Mu agenesis ki , S a agene) wi h he ollowing oligonucleo ides: 5¢ TGCTCGGTTTTACTGGCTCCAAGCT 3¢and i s e e se complemen a y oligonucleo ide o he W434F mu a ion; 5¢ CTTTCGCCCTGCTCGGTGGGACTG 3¢and i s e e se complemen a y oligonucleo ide o he W434R mu a ion; 5¢ TTCGCCCTGCTCTATTGGACTGGCTCC 3¢and i s e e se complemen a y oligonucleo ide o he G433Y mu a ion; 5¢ CTTTCGCCCTGCTCTATTTTACTGGCTCCA 3¢and i s e e se complemen a y oligonucleo ide o he G433Y/ W434F double mu a ion. Exp ession o Pol m a ian s was ca ied ou in he Esche ichia coli s ain BL21(DE3)pLysS unde s anda d condi ions. Pol mp o eins (ei he wild- ype o mu an ) we e pu i®ed o homogenei y as p e iously desc ibed (15). DNA polyme iza ion assays on de®ned DNA molecules Polyme ase ac i i y was e alua ed by using syn he ic double- s anded oligonucleo ides as subs a es. These subs a es we e p epa ed by annealing a 5¢-32P-end-labelled p ime (DNA o RNA) o di e en oligonucleo ides o gene a e open (P15/ T32) and gapped (P15/T32/D16) empla e/p ime subs a es. In polyme iza ion eac ions, he incuba ion mix u e con ained, in 12.5 ml, 50 mM T is±HCl, pH 7.5, 2 mM MgCl 2 o 1 mM MnCl 2 , 1 mM DTT, 4% glyce ol, 0.1 mg/ml BSA, di e en concen a ions o he indica ed dNTPs o NTPs, 4 nM o 5¢- labelled subs a e, and he indica ed amoun and concen a ion o ei he pu i®ed Pol m(wild- ype o mu an s) o he indica ed DNA polyme ase. A e incuba ion o 30 min a 30°C (condi ions shown o be linea bo h in ime and enzyme doses), eac ions we e s opped by adding gel loading bu e [95% ( / ) o mamide, 10 mM EDTA, 0.1% (w/ ) xylene cyanol and 0.1% (w/ ) B omophenol Blue]. Be o e loading he gel, samples we e dena u ed by hea ing a 95°C o 5 min. P oduc s we e esol ed and analysed by dena u ing 8 M u ea±20% PAGE and au o adiog aphy. Quan i a ion o au o adiog aphs was done by densi ome ic analysis o he band(s) co esponding o p ime ex ension p oduc s wi h he Bio-Rad Quan i y One so wa e. De e mina ion o Pol msuga selec i i y To analyse Pol msuga selec i i y, DNA oligonucleo ide P15 was hyb idized o he ou a ian s o he T32 empla e oligonucleo ide (T32(A), T32(C), T32(G) and T32(T), di e - ing in he ® s empla ing base) and o he D16 oligo- nucleo ide. Then 4 nM o each DNA-gapped subs a e was incuba ed wi h 100 nM Pol m, 2 mM MgCl 2 and he indica ed amoun s o each pai o nucleo ides (dATP + ATP; dCTP + CTP; dGTP + GTP; dTTP + UTP) in he p esence o he same incuba ion mix u e desc ibed abo e. A e incuba ion o 30 min a 30°C (condi ions shown o be linea bo h in ime and enzyme amoun ), eac ions we e s opped by adding gel loading bu e and p oduc s we e esol ed and analysed by dena u ing 8 M u ea±20% PAGE and au o adiog aphy. Calcula ion o he ela i e amoun o p ime s ex ended ei he wi h dNTP o wi h NTP was done by densi ome ic analysis o he band(s) co esponding o one- nucleo ide p ime ex ension p oduc s wi h he Bio-Rad Quan i y One so wa e. The suga selec i i y ac o (S) is he mean a io o he e ®ciencies o dNTP and NTP addi ions (% p ime ex ension wi h dNTP/% p ime ex ension wi h NTP) calcula ed o di e en nucleo ide concen a ions and h ee independen expe imen s. Amino acid sequence compa isons An ini ial alignmen o se e al membe s o he Pol X amily was done by using he MULTALIN p og am (h p://www. oulouse.in a. /mul alin.h ml). As a second s ep, he align- men ob ained was adjus ed manually and e®ned on he basis o he seconda y elemen s o a Pol b, as deduced om i s c ys al s uc u e (21). Th ee-dimensional s uc u e ex apola ions The s uc u al model o human Pol mwas au oma ically gene a ed using he p og am Swiss Model (h p://www. expasy.ch/swissmod/swiss-model.h ml) and he ecen ly ob ained h ee-dimensional s uc u e o mu ine TdT (22) as a empla e. TdT PDB coo dina es (1KEJ) we e ob ained om P o ein Da a Bank (h p://www. csb.o g/pdb/). Nex , his Pol mmodel was aligned wi h he s uc u e o DNA Pol b complexed wi h gapped DNA and ddCTP (PDB coo dina es 1BPY). 1BPY coo dina es we e selec ed as hey co espond o he ully closed con o ma ion o Pol b, shown o ha e he maximal s uc u al simila i y (o e lapping) o he c ys al s uc u e o TdT (22). Pu a i e amino acid esidues o bo h Pol m(modelled) and Pol bimplica ed in suga disc imina ion we e iden i®ed and compa ed by using he Swiss PDB Viewe p og am (23; h p://www.expasy.ch/spdb /). Single (G433Y, W434F, W434R) and double (G433Y/W434F) amino acid subs i u ions o be in oduced in Pol mwe e modelled by using he same p og am, and he side chain con o ma ions wi h he lowes ene gy we e selec ed. 4442 Nucleic Acids Resea ch, 2003, Vol. 31, No. 15 RESULTS Human Pol minco po a es and elonga es dNTPs and NTPs on a DNA p ime s and wi h a simila e ®ciency Al hough Pol mhas an in insic e minal ans e ase ac i i y, i s p e e ed subs a es a e empla e-p ime s uc u es (15). Thus, in he p esence o a DNA empla e/p ime and he ou dNTPs as subs a es, Pol mca alyses e ®cien polyme iza ion o dNTPs on o he DNA p ime s and (Fig. 1a). Mo eo e , his polyme iza ion occu s in a empla e-di ec ed manne , as sugges ed by he change in he elonga ion pa e n obse ed by p o iding subse s o he ou dNTPs (Fig. 1a, lanes 1 o 4). Howe e , al hough mos Pol minse ions a e empla e- di ec ed, i equen ly ca alyses misinco po a ion o dNTPs (Fig. 1a, lane 2), p e e en ially due o ansien empla e misalignmen s (17). We checked he possibili y ha NTPs could be used as subs a es by Pol m. As shown in Figu e 1b, human Pol m e ®cien ly inco po a es NTPs on a DNA p ime s and, as judged by he p opo ion o ex ended p ime s ands (compa e Fig. 1a and b). Howe e , elonga ion o ibonucleo ides o wi h ibonucleo ides is less e ®cien , as he pa e n obse ed is sho e han ha ob ained wi h iden ical concen a ions o dNTPs (compa e Fig. 1a and b, lanes 1 o 6). In ac , a a low ibonucleo ide concen a ion (1 mM), he inse ion/elonga ion pa e n ob ained in he p esence o he ou NTPs is e y simila o ha ob ained wi h GTP alone ( he ® s NTP o be inse ed acco ding o he empla e). This limi ed pa e n sugges s ha ibonucleo ide inco po a ion by Pol mmigh be comp omised by he empla e sequence con ex o by a speci®c ibonucleo ide p e e ence (see below). Howe e , his pa e n could be u he elonga ed ei he by inc easing he NTP concen a ion in he eac ion (Fig. 1b, lanes 5 and 6), o by using Mn2+ ins ead o Mg2+ as me al co ac o in he eac ion (lane 7). Inse ions o NTPs occu ed mainly as empla e-di ec ed e en s, e en in he p esence o Mn2+ ions, and esul ed om empla e misalignmen s ( esul s no shown). Pol mis able o polyme ize nucleo ides on RNA p ime s ands The s ong accumula ion o he +1 p oduc shown in Figu e 1b (lanes 1±4) could e¯ec a lowe e ®ciency o ecognize a p ime - e minus ha ing a 2¢-OH g oup. To cla i y his poin , we di ec ly es ed whe he Pol mhas he capaci y o use a p e- exis ing RNA p ime o inco po a e ibo- o deoxynucleo ides. Fo a mo e di ec compa ison wi h he esul s shown in Figu e 1, we used a p ime s and ha ing he same nucleo ide sequence, bu made o NTPs, hyb idized o he same DNA empla e sequence. As shown in Figu e 2, Pol mis able o polyme ize bo h dNTPs (Fig. 2a) and NTPs (Fig. 2b) on a RNA p ime hyb idized o a DNA empla e s and. By p o iding dGTP alone, elonga ion is mainly es ic ed o he inse ion o wo esidues (Fig. 2a), as expec ed o a ai h ul eac ion. Howe e , addi ion o di e en combina ions o dNTPs allowed u he ex ension, gene a ing p oduc s longe han expec ed, consis en wi h Pol m's p opensi y o inse nucleo ides in a a ie y o misalignmen -media ed e en s. As shown in Figu e 2, he e is a signi®can amoun o unex ended RNA p ime s, highe han ha o he unex ended DNA p ime molecules shown in Figu e 1a. This was due o a 3- old lesse hyb idiza ion e ®ciency o he RNA p ime s o he empla e DNA, as de e mined by polyac ylamide gel elec opho esis a low ionic s eng h ( esul s no shown). On he o he hand, he nucleo ide ex ension pa e n o hese RNA and DNA p ime s is e y simila . Again, he elonga ion is mo e e ®cien wi h dNTPs han wi h NTPs subs a es (Fig. 2a and b, Figu e 1. Pol me ®cien ly polyme izes ibonucleo ides on a DNA p ime . The assay was ca ied ou using 4 nM o a 5¢-32P-labelled 15me DNA p i- me (dP) hyb idized o a DNA empla e oligonucleo ide (a scheme is shown), as desc ibed in Ma e ials and Me hods. Polyme iza ion was assayed in he p esence o 2 mM MgCl 2 , 100 nM pu i®ed human Pol m, and he indica ed concen a ions o ei he dNTPs (a) o NTPs (b). Whe e indica ed, MnCl 2 (1 mM) was used ins ead o MgCl 2 . A e incuba ion o 30 min a 30°C, ex ension o he 5¢-labelled oligonucleo ide was analysed by 8 M u ea±20% PAGE and au o adiog aphy. Rele an ex ension p oduc s a e indica ed. Figu e 2. Pol me ®cien ly polyme izes bo h deoxy- and ibonucleo ides on RNA p ime s. The assay was ca ied ou using 4 nM o a 5¢-32P-labelled 15me RNA p ime ( P) hyb idized o a DNA empla e oligonucleo ide (a scheme is shown), as desc ibed in Ma e ials and Me hods. Polyme iza ion was assayed in he p esence o 2 mM MgCl 2 , 100 nM pu i®ed human Pol m, and he indica ed concen a ions o ei he dNTPs (a) o NTPs (b). Whe e indica ed, MnCl 2 (1 mM) was used ins ead o MgCl 2 . A e incuba- ion o 30 min a 30°C, ex ension o he 5¢-labelled oligonucleo ide was analysed by 8 M u ea±20% PAGE and au o adiog aphy. Rele an ex ension p oduc s a e indica ed. Nucleic Acids Resea ch, 2003, Vol. 31, No. 15 4443 espec i ely). RNA p ime elonga ion could be s ongly imp o ed by using Mn2+ ins ead o Mg2+ as me al co ac o in he eac ion (Fig. 2b, lane 6). Acco ding o hese esul s we can conclude ha Pol mis also able o use a DNA/RNA empla e/p ime subs a e o inco po a e and elonga e bo h ibo- and deoxynucleo ides. Ribonucleo ide inse ion/ex ension by o he Pol X amily membe s Pol mbelongs o he Pol X amily o DNA polyme ases, oge he wi h TdT, Pol b,Polland he mos dis an ly ela ed Pol s( e iewed in 24). While Pol lis mo e ela ed o Pol b, Pol mis closely ela ed o TdT, an enzyme whose abili y o inse NTPs has been p e iously desc ibed (10). Unlike Pol b, Pol l, Pol mand TdT ha e a BRCT domain a hei N- e minus. Howe e , al hough all o hem could sha e a simila h ee-dimensional s uc u e, hei biochemical ac i - i ies appea o be qui e di e se. The e o e, we compa ed he abili y o DNA polyme ases o he Pol X amily o inco po a e NTPs on DNA p ime s in ou expe imen al condi ions. As shown in Figu e 3, Pol b,Polland Pol mbeha ed as p o®cien DNA-dependen DNA polyme ases, polyme izing dNTPs on a DNA empla e/p ime subs a e. TdT, as expec ed, was less e ®cien o polyme ize dNTPs on his kind o subs a e, since i displayed i s op imal ac i i y on non- empla ed single s anded DNA molecules (no shown; 25). On he o he hand, he abili y o inco po a e NTPs on hese DNA empla e/ p ime molecules, which can be clea ly seen wi h bo h TdT and Pol m, was null in Pol land e y low in Pol b(Fig. 3). The same esul s we e ob ained unde di e en eac ion condi ions and o he DNA subs a es as gapped DNA (da a no shown). Mo eo e , as judged by he ex ension pa e n shown in Figu e 3, ibonucleo ide inco po a ion by Pol mwas empla e- di ec ed, while TdT polyme ized NTPs in a empla e- independen manne . These esul s sugges ha he abili y o e ®cien ly use NTPs as subs a es o elonga e DNA p ime s appea s o be speci®c o TdT-like DNA polyme ases (TdT, Pol m), while o he membe s o he Pol X amily, such as Pol band Pol ldo no ha e his capaci y. Pol mpolyme ize ibo- and deoxynucleo ides dis ibu i ely on empla e/p ime subs a es DNA polyme ases om he Pol X amily display a s ic ly dis ibu i e polyme iza ion mode, i.e. hey dissocia e om a empla e-p ime molecule a e each nucleo ide addi ion cycle. This beha iou has been shown o Pol b(26), Pol l (27) and TdT (10), bu i has no been epo ed ye in he case o Pol m. In his pape , we assessed Pol m's p ocessi i y when polyme izing ei he dNTPs o NTPs on DNA empla e/p ime subs a es by analysing he chain leng h dis ibu ion a a ious enzyme/DNA subs a e a ios. As shown in Figu e 4, he leng h o he p oduc s syn hesized by Pol mwhen polyme iz- ing ei he dNTPs o NTPs dec eased wi h he enzyme/ subs a e a io (Fig. 4a and b, le panels). In bo h cases, he limi leng h is 1, in ag eemen wi h a ully dis ibu i e polyme iza ion pa e n. We also analysed a possible e ec o he me al co ac o used in he polyme iza ion eac ion on his low Pol mp ocessi i y. As shown in Figu e 4a and b ( igh panels), he dis ibu i e beha iou o Pol mis also main ained in he p esence o Mn2+ ac i a ing ions. Figu e 3. Compa a i e analysis o NTP inse ion by Pol X DNA poly- me ases. The assay was ca ied ou using 4 nM o a 5¢-labelled 15me DNA p ime (dP) hyb idized o a DNA empla e oligonucleo ide (see Ma e ials and Me hods). Polyme iza ion was assayed in he p esence o 2 mM MgCl 2 , 1mM o ei he dNTPs o NTPs, and 100 nM o ei he Pol b, Pol l, TdT o Pol m. A e incuba ion o 30 min a 30°C, ex ension o he 5¢-labelled oligonucleo ide was analysed by 8 M u ea±20% PAGE and au o adiog aphy. Figu e 4. Dis ibu i e polyme iza ion o bo h dNTPs and NTPs by human Pol m. Reac ions we e pe o med by using 4 nM o a 5¢-labelled 15me DNA p ime hyb idized o a 32me empla e oligonucleo ide, as desc ibed in Ma e ials and Me hods. Polyme iza ion eac ions we e ca ied ou in he p esence o ei he 2 mM MgCl 2 o 1 mM MnCl 2 ,10mM dNTPs, and he indica ed amoun s ( om 100 o 0.1 nM) o Pol m. A e incuba ion o 30 min a 30°C, ex ension o he 5¢-labelled oligonucleo ide was analysed by 8 M u ea±20% PAGE and au o adiog aphy. 4444 Nucleic Acids Resea ch, 2003, Vol. 31, No. 15 The e o e, we can conclude ha Pol mpolyme izes bo h NTPs and dNTPs on a DNA p ime in a dis ibu i e manne , independen ly o he me al ac i a o p esen (Mg2+ o Mn2+). Suga disc imina ion by Pol mdepends on each pa icula nucleo ide base pai To analyse he disc imina ion a io o each NTP/dNTP pai by Pol m, we designed a se o 1-nucleo ide gapped DNA subs a es wi h each o he ou (A, C, G o T) bases as empla es. Since Pol mis no able o ca y ou s and displacemen (ou unpublished da a) such DNA subs a es will es ic polyme iza ion o a single nucleo ide inco po - a ion, and elimina e he addi ional complexi y esul ing om he equency o p omiscuous inse ion e en s ca alysed by Pol m. Thus, we expe imen ally de e mined he alue o he Pol msuga selec i i y ac o (S) o each deoxy/ ibonucle- o ide pai (dATP/ ATP, dCTP/ CTP, dGTP/ GTP and dTTP/ UTP) using a compe i ion assay (10) whe e bo h suga a ian s ( ibose and deoxy ibose) o each nucleo ide a e simul aneously p o ided. Since he NTP and dNTP ha e di e en molecula weigh s, he +1 ex ended p ime s can be easily sepa a ed by gel elec opho esis and hen quan i®ed. The S ac o is gi en by he a io be ween he amoun o p ime s ex ended wi h ei he dNTPs o NTPs (% p ime ex ension wi h dNTP/% p ime ex ension wi h NTP). Figu e 5 shows a ep esen a i e assay o an S ac o de e mina ion, and Table 1 summa izes he alues ob ained o all ou dNTP/ NTP pai s. The S alues ange be ween 1.34 o dGTP o e GTP and 11.04 o dATP o e ATP. These alues a e compa able o hose ob ained o TdT in un empla ed eac ions (10), and se e al o de s o magni ude smalle han he suga selec i i y ac o s epo ed o o he DNA polyme ases (1±3). In e es ingly, he selec i i y ac o s o Pol m a y depending on he na u e o he base. The lack o disc imina ion be ween ibo- and deoxynucleo ides is mo e p onounced o Gs and Cs han o Ts and As (Table 1). Ribonucleo ides inhibi DNA polyme iza ion by Pol m As has been p e iously shown, cells ha e a ibo/deoxy ibo- nucleo ide pool imbalance, wi h an excess o NTPs o e dNTPs (28). Thus, once es ablished ha Pol mis able o inco po a e NTPs on bo h DNA and RNA p ime s, we analysed he inco po a ion o bo h dNTPs and NTPs on DNA s ands in i o unde condi ions ha ep oduce his pool imbalance. We ca ied ou polyme iza ion eac ions in which each indi idual NTP was added a a 10- old excess o e he ou dNTPs. As shown in Figu e 6, he e ec o he NTP a ied depending on he na u e o he base. Thus, while he addi ion o py imidines ( UTP o CTP) had no appa en e ec on he Pol melonga ion pa e n (Fig. 6, lanes 3 and 4, espec i ely), he addi ion o pu ines ( ATP and mainly GTP) inhibi ed elonga ion (lanes 2 and 5, espec i ely). The s onge inhibi ion by GTP could be explained ei he by he ac ha he ® s wo consecu i e empla ing bases a e dC, combined wi h he low disc imina ion o GTP o e dGTP (Table 1), o by a poo e ex ension o GMP- e mina ed p ime s by Pol m. To clea ly demons a e ha NTPs can be inco po a ed in o DNA by Pol mac i i y, we ca ied ou in i o polyme iza ion eac ions in which we added h ee dNTPs and he ou h was eplaced by he NTP coun e pa in a 10- old excess, con® ming ha , unde hese condi ions, Pol mac i i y is able o inco po a e NTPs in o DNA polyme iza ion p oduc s (da a no shown). Gly433 o Pol mis essen ial o allow inse ion o NTPs As p e iously desc ibed, he e a e h ee Pol b esidues ha in e ac by an de Waals con ac s wi h he suga moie y (C2¢ and C3¢ca bons) o he incoming nucleo ide: Ty 271, Phe272 and Gly274, loca ed a he end o a-helix M in he humb subdomain, and p edic ed o pa icipa e in nucleo ide selec - i i y o DNA o e RNA (21). As shown in Figu e 7, a coun e pa o Pol bGly274 is in a ian ly p esen in all membe s o he Pol X amily, including TdT and Pol m. In e es ingly, Ty 271 and Phe272 esidues o Pol ba e subs i u ed by in a ian Gly and T p esidues, espec i ely, in TdT and Pol m om di e en species, and in Pol IV om Schizosaccha omyces pombe. The e o e, acco ding o his, i seemed e y likely ha esidues Gly433 and T p434 o Pol m could be esponsible o he lack o suga disc imina ion be ween NTPs and dNTPs. To es his p edic ion, we mu a ed he amino acid esidues G433 and W434 o Pol mby si e-di ec ed mu agenesis (see Figu e 5. Suga disc imina ion by Pol mdepends on he na u e o he base. Single nucleo ide gap-®lling assays we e pe o med using he ou 1 n - gapped DNA subs a es (a scheme is shown) di e ing in he empla ing base (X). Reac ions we e pe o med by using 4 nM o each 5¢-32P-labelled 1 n -gapped DNA subs a e, 100 nM Pol m, 2 mM MgCl 2 , and he indica ed concen a ion o each o he ou pai s o deoxy- and ibonucleo ide. A e 30 min a 30°C, eac ions we e subjec ed o 8 M u ea±20% PAGE, and he deoxy/ ibonucleo ide +1 ex ended p ime s we e de ec ed by au o adiog aphy. Table 1. Suga selec i i y ac o s o he inse ion o deoxy/ ibonucleo ide pai s by Pol m Templa e Nucleo ide mix WT W434F W434R G433Y dT dA/ A 11.04 60.70a17.09 60.59 17.87 62.65 >30 dG dC/ C 2.70 60.67 3.41 60.74 7.54 61.65 >40 dC dG/ G 1.34 60.26 1.77 60.20 7.67 60.91 >25 dA dT/ U 8.64 61.18 5.93 61.42 7.65 61.55 >30 dNbdN/ N 5.93 60.70 7.05 60.73 10.18 61.69 >31 aFigu es indica e he mean alues o independen expe imen s and hei s anda d de ia ion. bThe suga selec i i y ac o s ob ained o each deoxy/ ibonucleo ide pai we e used o calcula e he mean alue o he ou bases (N). Nucleic Acids Resea ch, 2003, Vol. 31, No. 15 4445 Ma e ials and Me hods), and checked how e ®cien ly he mu an p o eins disc imina e be ween dNTPs and NTPs. Single mu a ions selec ed we e G433 o Ty and W434 o Phe, o es o e indi idual consensus esidues o Pol b-like enzymes. Two addi ional mu an s we e ob ained: W434 o A g, o e alua e he e ec o a non-conse a i e change a his posi ion, and he double mu an G433Y/W434F, o ully es o e he Pol bconsensus. In e es ingly, he W434F mu an p o ein had a polyme iz- a ion ac i i y simila o ha o he wild- ype p o ein, as assessed by dNTP inco po a ion, and was s ill able o polyme ize ibonucleo ides (Fig. 8). As shown in Table 1, suga selec i i y ac o s o mu an W434F we e simila o hose ob ained wi h he wild- ype p o ein, wi h a maximal di e ence o 1.5- old o dA/ A selec i i y. In ag eemen wi h he conse a ion o an a oma ic esidue in mos Pol X polyme ases, he ca aly ic ac i i y o mu an W434R dec eased abou 5- old compa ed o he wild- ype p o ein. This e ec appea s o be highe o NTP han dNTP polyme iza ion, as shown in Figu e 8. Howe e , as shown in Table 1, he change in disc imina ion be ween ibo- and deoxy ibonucleo ides by he W434R mu an p o ein seems o be dependen on he na u e o he base, as selec i i y ac o s o C and G we e he mos a ec ed (abou 3- and 6- old, espec i ely) by he mu a ion (Table 1). No change in suga disc imina ion was obse ed o he dA/ A and dT/ U pai s. S ikingly, he abili y o inco po a e NTPs d as ically d opped in he G433Y mu an p o ein, as shown in Figu e 8. The S alues ob ained o his mu an (Table 1) e¯ec a la ge gain o suga disc imina ion o all dNTP/ NTP pai s. On he o he hand, as shown in Figu e 8, his mu a ion esul s in a dec ease in he dNTP polyme iza ion e ®ciency in ela ion o wild- ype Pol m, which sugges s ha his amino acid esidue is also impo an o he op imal ca aly ic ac i i y o he enzyme. The double mu an p o ein G433Y/W434F had a e y low, almos null, polyme iza ion ac i i y wi h bo h dNTPs and NTPs (Fig. 8), indica ing he essen ial ole o hese amino acid esidues o an op imal polyme iza ion ac i i y o Pol m, in addi ion o hei ole in suga disc imina ion. DISCUSSION Pol mis a no el nucleo idyl ans e ase belonging o he Pol X amily. Despi e i s p e iously desc ibed e minal ans e ase ac i i y (15), Pol mshould be conside ed a DNA-dependen DNA polyme ase e sa ile in i s use o subs a es. In ac , Pol mwas o iginally desc ibed as an ex emely un ai h ul DNA polyme ase (15), and i was la e shown ha i s in®deli y is ela ed o i s ex ao dina y abili y o accep misaligned empla e-p ime s uc u es as a subs a e (17). As we ha e shown he e, an addi ional ea u e ha e¯ec s he e sa ili y o subs a e usage by Pol mis i s abili y o inse ibonucleo ides o he 3¢-hyd oxyl p ime ends o ei he DNA empla e/p ime Figu e 6. Inse ion o NTPs inhibi DNA polyme iza ion by Pol m. P ime ex ension assays we e pe o med using 4 nM 5¢-32P-labelled DNA p ime / empla e, 100 nM Pol m, 2 mM MgCl 2 , and he indica ed concen a ion o ei he dNTPs and NTPs. Di e en combina ions o nucleo ides we e used o e alua e he in e e ence o NTPs on dNTP polyme iza ion. A e 30 min a 30°C, polyme iza ion p oduc s we e subjec ed o 8 M u ea±20% PAGE and au o adiog aphy. Figu e 7. Mul iple alignmen o he amino acid egion likely in ol ed in suga disc imina ion in he Pol X amily. Numbe s be ween slashes indica e he amino acid posi ion ela i e o he N- e minus o each polyme ase. Acco ding o Pol bs uc u al da a (21), his egion connec s subdomains palm and humb, and con ains one o he h ee aspa a es ac ing as me al ligands (black do abo e he sequences). In a ian esidues among Pol X amily membe s (29) a e indica ed in whi e le e s o e a black backg ound. The wo amino acid esidues mos likely implica ed in suga disc imina ion a e indica ed in bold, inside a box. Abb e ia ions used a e: Hom.sa., Homo sapiens, Mus mu., Mus musculus, Bos a., Bos au us, Gal.ga., Gallus gallus, Xen.la., Xenopus lae is, Mon.do., Monodelphis domes ica, Amb.me., Ambys oma mexicanum, Onc.my., Onco hynchus mykiss, Sch.po., Schizosaccha omyces pombe, Ra .no., Ra us no e gicus, Sac.ce., Saccha omyces ce e isiae, C i. a., C i hidia ascicula a, Bac.su., Bacillus sub ilis, Me . h., Me hano he mobac e he mau o ophicus, Aqu.ae., Aqui ex aeolicus and The.aq., The mus aqua icus. 4446 Nucleic Acids Resea ch, 2003, Vol. 31, No. 15 o DNA-gapped molecules, in sha p con as wi h mos DNA polyme ases. Mo eo e , he suga selec i i y ac o s measu ed in single nucleo ide addi ion assays (be ween 1 and 11, o dG/ G and dA/ A, espec i ely; see Table 1), indica e ha Pol m e ®cien ly inco po a es bo h ibo- and deoxy ibonucleo ides. This is a e y a e beha iou o a DNA syn hesizing enzyme, since mos DNA polyme ases s ongly a ou he inse ion o deoxynucleo ides o a oid he inciden al inse ion o ibo- nucleo ides in o he DNA. In ac , as shown he e, o he Pol X enzymes as Pol bo Pol l, s ongly disc imina e agains ibonucleo ides. Excep ionally, TdT ( he closes coun e pa o Pol m) is known o ha e a simila ibonucleo ide inse ion capaci y (10). A e ou wo k was comple ed, Nick McElhinny and Ramsden epo ed a se ies o esul s e y simila o hose desc ibed he e, demons a ing ha Pol mis a DNA-dependen DNA and RNA polyme ase (30). Fo se e al DNA polyme ases, suga disc imina ion has been shown o depend on a single esidue in mo i A ha ac s as a s e ic ba ie o he 2¢-hyd oxyl o an incoming ibonucleo ide (1,5). Howe e , c ys allog aphic analysis o Pol bsugges ed a di e en mechanism, in which exclusion o he ibose moie y is ca ied ou by a h ee- esidue pep ide backbone (Ty 271±Phe272±Th 273) loca ed in a-helix M o he humb subdomain (21). Mo e ecen ly, he c ys al s uc u e o mu ine TdT has been esol ed (22). Acco ding o his s uc u e, he suga moie y o he incoming nucleo ide is loca ed in he neighbou hood o a numbe o a-helix M esidues (Fig. 7): T p450 on one side, and he cis-pep ide bond be ween Gly452 and Se 453 on he o he side. Howe e , none o hese esidues a e close enough o he 2¢posi ion o he suga moie y o an incoming nucleo ide, hus accoun ing o he abili y o TdT o e ®cien ly inco po a e ibonucleo ides. The ex ensi e amino acid sequence simila i y be ween Pol mand TdT allowed us o gene a e a model o Pol m's h ee-dimensional s uc u e based on he c ys al s uc u e o mu ine TdT (22). This model indica es a simila o ganiza ion o he ac i e si e o bo h enzymes. I sugges s ha he Pol m esidues closes o he 2¢posi ion o he ibose o an incoming nucleo ide would be Gly433 and T p434, loca ed in a-helix M (Fig. 9, op le ), whose coun e pa s in Pol ba e Ty 271 and Phe272 (Fig. 9, bo om igh ). Ou esul s demons a e ha , while ibonucleo ide inse ion is no g ea ly a ec ed when T p434 is subs i u ed, he small size o he Gly433 side chain is essen ial o allow he polyme iza ion o ibonucleo ides by Pol m. In ac , when Gly433 is changed o a esidue wi h a bulkie side chain (Ty ) as ha p esen in Pol b(Fig. 9, bo om le ), a d ama ic inc ease in he suga selec i i y ac o s be ween deoxy- and ibonucleo ides is obse ed. Thus, he size o he side chain p esen a his posi ion p obably egula es ibonucleo ide inse ion in Pol X amily enzymes. Based on his mu a ional analysis and he amino acid sequence alignmen shown in Figu e 7, a simila e ®ciency o NTP inse ion, as ha shown by TdT and Pol m, is also p edic ed o Pol IV om S.pombe. The highly e ®cien inse ion o NTPs by bo h TdT and Pol mis ex emely in iguing aking in o accoun ha inco po a- ion o ibonucleo ides in he DNA c ea es a lesion ha has o be epai ed a a isk o ec ui ing a ious e o - ee and e o - p one DNA epai sys ems. Howe e , he biological conse- quences could di e as TdT syn hesis is mos ly empla e- independen , while he ibonucleo ide inse ion capaci y desc ibed he e o Pol mis di ec ed by he empla e. Fu he mo e, whe eas TdT ac i i y is es ic ed o a e y speci®c s age o lymphocy e di e en ia ion, Pol mseems o be Figu e 8. Lack o suga disc imina ion by Pol m esides in Gly433. Single nucleo ide gap-®lling assays we e pe o med using 4 nM o a 5¢-32P-labelled 1 n -gapped DNA subs a e ha ing a dC as empla ing base (a scheme is shown), 2 mM MgCl 2 , he indica ed concen a ion o bo h dGTP and GTP, and 100 nM o ei he he wild- ype Pol m(WT), o each o he si e-di ec ed mu an p o eins W434F, W434R, G433Y and G433/W434F. A e 30 min a 30°C, he a io o deoxy- e sus ibonucleo ide-ex ended p ime s was analysed by 8 M u ea±20% PAGE and au o adiog aphy. Figu e 9. S uc u al basis o he lack o suga disc imina ion by Pol m. Gene a ion o a s uc u al model o Pol mbased on TdT s uc u e (22), and compa ison wi h ha o a Pol b e na y complex (21) we e pe o med as indica ed in Ma e ials and Me hods. The ®gu e shows he spa ial posi ion o he wild- ype Pol mamino acid esidues (Gly433, T p434 and Th 435) ha a e closes o he suga moie y o an incoming nucleo ide (le - op panel). Si e di ec ed subs i u ion o ei he Gly433 o Ty (G433Y; le -bo om panel) o T p434 o Phe (W434F; igh - op panel) was modelled. A s uc- u al compa ison o he Pol bcoun e pa esidues (Ty 271, Phe272 and Th 273) is also shown ( igh -bo om panel). Figu e was c ea ed using he Swiss PDB Viewe p og am (23; h p://www.expasy.ch/spdb /), and ende ed wi h POV Ray (www.po ay.o g). Nucleic Acids Resea ch, 2003, Vol. 31, No. 15 4447 exp essed in mos cell ypes, which poin s o he po en ial ele ance o NTP inse ion in o DNA by Pol min a/some widely occu ing p ocess. As shown he e, Pol me ®cien ly inse s NTPs in o DNA, bu polyme iza ion e ®ciency dec eases a e he inco po a ion o a limi ed numbe o NMPs. When bo h NTPs and dNTPs a e p esen , Pol msyn hesizes mixed p oduc s. Howe e , since Pol mex ends ibonucleo ide- e mina ed p ime s wi h a lowe ca aly ic a e han deoxy ibonucleo ide- e mina ed ones, mixed p oduc s a e sho e in leng h. A simila e ec has been desc ibed ecen ly o mu ine TdT (10), leading o he p oposal ha NTP inse ion could se e o limi non- empla ed nucleo ide addi ions (N-addi ions) by TdT o he junc ions o ea anging V, D and J gene segmen s. The e minal ans e ase ac i i y o Pol m, oge he wi h i s lack o suga disc imina ion o nucleo ides sugges ha Pol mcould ha e a ole in V(D)J ecombina ion, pe haps complemen ing TdT addi ions a V(D)J ecombina ional join s. In e es ingly, second V(D)J ecombina ion wa es ( ecep o edi ing) ha e been desc ibed o occu in a subse o ge minal cen e B lymphocy es (31), jus he place whe e highe Pol mle els ha e been ound (15,16). Recen da a demons a ed ha Pol m o ms a complex wi h he end-joining ac o s Ku and XRCC4-Ligase IV in a manne e y simila o ha o TdT (18,32). Mo eo e , his complex acili a es he abili y o XRCC4-Lig IV and Ku o join ends wi h pa ially complemen a y o e hangs in i o, suppo ing he p oposal o a pa icipa ion o Pol min he epai o DNA double s and b eaks (16±18). I is no ob ious how inse ion o NTPs by Pol mcould ha e a ele an ole in NHEJ and pe haps also in lesion bypass o DNA damage, bu a emp ing specula ion is ha inse ion o NTPs could s abilize empla e misalignmen in e media es du ing hese p ocesses. An in e es ing p oposal, ecen ly made by Nick McElhinny and Ramsden (30), is ha Pol mmay ha e e ol ed NTP polyme iza ion ac i i y o be e ec i e when nuclea dNTP pools a e lowes , since NTP pools emain high h oughou he cell cycle and in non-cycling cells. I is wo h no ing ha he unusual capaci y o Pol m o accep misaligned empla e-p ime molecules as a subs a e, a con enien ea u e o mic ohomology-media ed NHEJ, is no a ec ed in he G433Y mu an (ou unpublished da a), sugges ing ha he abili y o inse ibonucleo ides is no a side-e ec (i.e. as a consequence o a less igh ened ac i e si e) and he e o e is likely o ha e a biological meaning. The iden i®ca ion o he c i ical esidue in ol ed in he lack o suga disc imina ion by human Pol mwill guide u u e s udies o demons a e he ele ance o NTP inco po a ion o he biological unc ion o his enzyme. ACKNOWLEDGEMENTS We wish o hank Kasia Bebenek (NIEHS, NC, USA) o c i ical eading o he manusc ip . This wo k was suppo ed by Minis e io de Ciencia y Tecnologõ Âa G an BMC2000±1138, and by an ins i u ional g an om Fundacio ÂnRamo Ân A eces. J.F.R, R.J., M.G.-D. and A.J.P. we e ecipien s o a ellowship om he Minis e io de Educacio Ân y Ciencia. S.G.-B. was pos doc o al ellow om he Comunidad de Mad id and A.R.F.d.H. was a ellow om he Ramo Ân y Cajal Resea ch P og am. REFERENCES 1. Joyce,C.M. (1997) Choosing he igh suga : How polyme ases selec a nucleo ide subs a e. P oc. Na l Acad. Sci. USA,94, 1619±1622. 2. As a ke,M., Ng,K., G indley,N.D.F. and Joyce,C.M. (1998) A single side chain p e en s Esche ichia coli DNA polyme ase I (Klenow agmen ) om inco po a ing ibonucleo ides. P oc. Na l Acad. Sci. USA,95, 3402± 3407. 3. Cases-Gonza Âlez,C.E., Gu ie  ez-Ri as,M. and Mene Ândez-A ias,L. (2000) Coupling ibose selec ion o ®deli y o DNA syn hesis. The ole o Ty - 115 o human immunode®ciency i us ype 1 e e se ansc ip ase. J. Biol. Chem.,275, 19759±19767. 4. Gao,G., O lo a,M., Geo giadis,M.M., Hend ickson,W.A. and Go ,S.P. (1997) Con e ing RNA polyme ase ac i i y o a DNA polyme ase: a single esidue in e e se ansc ip ase con ols subs a e selec ion. P oc. Na l Acad. Sci. USA,94, 407±411. 5. Bonnin,A., La Âza o,J.M., Blanco,L. and Salas,M. (1999) A single y osine p e en s inse ion o ibonucleo ides in he euka yo ic- ype phi29 DNA polyme ase. J. Mol. Biol.,290, 241±251. 6. Pa el,P.H. and Loeb,L.A. (2000) Mul iple amino acid subs i u ions allow DNA polyme ases o syn hesize RNA. J. Biol. Chem.,275, 40266± 40272. 7. Ogawa,M., Tosaka,A., I o,Y., Yoshida,S. and Suzuki,M. (2001) Enhanced ibonucleo ide inco po a ion by an O-helix mu an o The mus aqua icus DNA polyme ase I. Mu a . Res.,485, 197±207. 8. Bassing,C.H., Swa ,W. and Al ,F.W. (2002) The mechanism and egula ion o ch omosomal V(D)J ecombina ion. Cell,109, S45±S55. 9. Ka o,K.I., Goncal es,J.M., Hou s,G.E. and Bollum,F.J. (1967) Deoxynucleo ide-polyme izing enzymes o cal hymus gland. II. P ope ies o he e minal deoxynucleo idyl ans e ase. J. Biol. Chem., 242, 2780±2789. 10. Boule Â,J.B., Rougeon,F. and Papanicolaou,C. (2001) Te minal deoxynucleo idyl ans e ase indisc imina ely inco po a es ibonucleo ides and deoxy ibonucleo ides. J. Biol. Chem.,276, 31388± 31393. 11. Holm,L. and Sande ,C. (1995) DNA polyme ase bbelongs o an ancien nucleo idyl ans e ase supe amily. T ends Biochem. Sci.,20, 345±347. 12. A a ind,L. and Koonin,E.V. (1999) DNA polyme ase b-like nucleo idyl ans e ase supe amily: iden i®ca ion o h ee new amilies, classi®ca ion and e olu iona y his o y. Nucleic Acids Res.,27, 1609± 1618. 13. SenGup a,D.N., Zmudzka,B.Z., Kuma ,P., Cobianchi,F., Skow onski,J. and Wilson,S.H. (1986) Biochem. Biophys. Res. Commun. 136, 341±347. 14. Ga cõ Âa-Dõ Âaz,M., Domõ Ânguez,O., Lo Âpez-Fe na Ândez,L.A., Laõ Ân de Le a,T., Sanõ Âge ,M.L., Ruiz,J.F., Pa  aga,M., Ga cõ Âa-O iz,M.J., Ki chho ,T., del Mazo,J., Be nad,A. and Blanco,L. (2001) DNA polyme ase lambda (Pol l), a no el euka yo ic DNA polyme ase wi h a po en ial ole in meiosis. J. Mol. Biol.,301, 851±867. 15. Domõ Ânguez,O., Ruiz,J.F., Laõ Ân de Le a,T., Ga cõ Âa-Dõ Âaz,M., Gonza Âlez,M.A., Ki chho ,T., Ma õ Ânez,A.C., Be nad,A. and Blanco,L. (2000) DNA polyme ase mu (Pol m), homologous o TdT, could ac as a DNA mu a o in euka yo ic cells. EMBO J.,19, 1731±1742. 16. Ruiz,J.F., Domõ Ânguez,O., Laõ Ân de Le a,T., Ga cõ Âa-Dõ Âaz,M., Be nad,A. and Blanco,L. (2001) DNA polyme ase mu, a candida e hype mu ase? Phil. T ans. R. Soc. Lond. B,356, 99±109. 17. Zhang,Y., Wu,X., Yuan,F., Xie,Z. and Wang,Z. (2001) DNA polyme ase m.Mol. Cell. Biol.,21, 7995±8006. 18. Mahajan,K.N., Nick McElhinny,S.A., Mi chell,B.S. and Ramsden,D.A. (2002) Associa ion o DNA polyme ase m(pol m) wi h ku and ligase IV: ole o pol min end-joining double-s and b eak epai . Mol. Cell. Biol., 22, 5194±5202. 19. Zhang,Y., Wu,X., Guo,D., Rechkobli ,O., Taylo ,J.S., Geacin o ,N.E. and Wang,Z. (2002) Lesion bypass ac i i ies o human DNA polyme ase m.J. Biol. Chem.,277, 44582±44587. 20. Ha ene ,J.M., Nick McElhinny,S.A., Basse ,E., Gauge ,M., Ramsden,D.A. and Chaney,S.G. (2003) T anslesion syn hesis pas pla inum DNA adduc s by human DNA polyme ase m.Biochemis y,42, 1777±1788. 21. Pelle ie ,H., Sawaya,M.R., Kuma ,A., Wilson,S.H. and K au ,J. (1994) S uc u es o e na y complexes o a DNA polyme ase b, a DNA empla e-p ime , and ddCTP. Science,264, 1891±1903. 22. Dela ue,M., Boule Â,J.M., Lesca ,J., Expe -BezancËon,N., Jou dan,N., Sukuma ,N., Rougeon,F. and Papanicolaou,C. (2002) C ys al s uc u e o 4448 Nucleic Acids Resea ch, 2003, Vol. 31, No. 15 a empla e-independen DNA polyme ase: mu ine e minal deoxynucleo idyl ans e ase. EMBO J.,21, 427±439. 23. Guex,N. and Pei sch,M.C. (1997) SWISS-MODEL and he Swiss-Pdb Viewe : an en i onmen o compa a i e p o ein modeling. Elec opho esis,18, 2714±2723. 24. Bu ge s,P.M.J., Koonin,E.V., B u o d,E., Blanco,L., Bu is,K.C., Ch is man,M.F., Copeland,W.C., F iedbe g,E.C., Hanaoka,F., Hinkle,D.C. e al. (2001) Euka yo ic DNA polyme ases: p oposal o a e ised nomencla u e. J. Biol. Chem.,276, 43487±43490. 25. Bollum,F.J. (1962) Oligodeoxy ibonucleo ide-p imed eac ions ca alyzed by cal hymus polyme ase. J. Biol. Chem.,237, 1945±1949. 26. Kunkel,T.A. (1985) The mu a ional speci®ci y o DNA polyme ases- alpha and gamma du ing in i o DNA syn hesis. J. Biol. Chem.,260, 12866±12874. 27. Ga cõ Âa-Dõ Âaz,M., Bebenek,K., Saba iegos,R., Domõ Ânguez,O., Rod õ Âguez,J., Ki chho ,T., Ga cõ Âa-Palome o,E., Piche ,A.J., Jua  ez,R., Ruiz,J.F., Kunkel,T.A. and Blanco,L. (2002) DNA polyme ase l, a no el DNA epai enzyme in human cells. J. Biol. Chem.,277, 13184±13191. 28. Meu h,M. (1984) The gene ic consequences o nucleo ide p ecu so pool imbalance in mammalian cells. Mu a . Res.,126, 107±112. 29. Oli e os,M., Ya Ân Äez,R.J., Salas,M.L., Vin Äuela,E. and Blanco,L. (1997) Cha ac e iza ion o an A ican swine e e i us 20-kDa DNA polyme ase in ol ed in DNA epai . J. Biol. Chem.,272, 30899±30910. 30. Nick McElhinny,S.A. and Ramsden,D.A. (2003) Polyme ase mu is a DNA-di ec ed DNA/RNA polyme ase. Mol. Cell. Biol.,23, 2309±2315. 31. Han,S., Dillon,S.R., Zheng,B., Shimoda,M., Schlissel,M.S. and Kelsoe,G. (1997) V(D)J ecombinase ac i i y in a subse o ge minal cen e B lymphocy es. Science,278, 301±305. 32. Mahajan,K.N., Gangi-Pe e son,L., So sche ,D.H., Wang,J., Ga hy,K.N., Mahajan,N.P., Ree es,W.H. and Mi chell,B.S. (1999) Associa ion o e minal deoxynucleo idyl ans e ase wi h Ku. P oc. Na l Acad. Sci. USA,96, 13926±13931. Nucleic Acids Resea ch, 2003, Vol. 31, No. 15 4449