Analysis of the DNA replication competence of the xrs-5 ...1994; Smider et al., 1994; Taccioli et al., 1994). Ku (reviewed in Tuteja and Tuteja, 2000) is a heterodimeric DNA-binding

IntroductionThe radiosensitive mutant cell line, xrs-5, a derivative of theChinese hamster ovary (CHO) K1, is defective in V(D)Jrecombination and DNA double-strand break (DSB) rejoining(Kemp et al., 1984; Pergola et al., 1993; Taccioli et al., 1994;Rathmell and Chu, 1994). The xrs mutants were initiallydescribed as the first mammalian cell mutants with a defect inDSB rejoining (Kemp et al., 1984). They were isolated by asimple technique involving transfer of heavily mutagenizedcells from single colonies, followed by irradiation with a dosedetermined not to affect survival of wild-type CHO K1 cells(Jeggo and Kemp, 1983). The xrs-5 cells, along with othermembers of the X-ray complementation group 5, also lackDNA end-binding activity (Rathmell and Chu, 1994; Getts andStamato, 1994; Singleton et al., 1997). The defective repair andrecombination phenotypes of the xrs-5 cells have been tracedto the 86 kDa subunit of the Ku protein, since Ku86 cDNArestored DNA-end binding, X-ray sensitivity and V(D)Jrecombination (Getts and Stamato, 1994; Rathmell and Chu,1994; Smider et al., 1994; Taccioli et al., 1994).

Ku (reviewed in Tuteja and Tuteja, 2000) is a heterodimericDNA-binding protein of 70 kDa and 86 kDa subunits (Mimoriet al., 1986). Ku binds avidly to DNA ends, whether blunt orwith 5′ or 3′ overhangs, as well as to other discontinuities inthe DNA structure (Mimori and Hardin, 1986; Taghva et al.,2002; Arosio et al., 2002). Sequence-specific binding of Ku hasalso been demonstrated, including binding to mammalian

origins of DNA replication (Giffin et al., 1996; Ruiz et al.,1999; Araujo et al., 1999; Novac et al., 2001). Ku is necessaryfor the processing of DNA DSBs that are induced by damagingagents such as ionizing radiation or oxidative reactions,by endogenous recombination processes and by certainchemotherapeutic drugs (reviewed in Featherstone andJackson, 1999; Tuteja and Tuteja, 2000; Doherty and Jackson,2001). An involvement of Ku has also been suggested intranscription (Giffin et al., 1996; Finnie et al., 1993; Kuhn etal., 1995; Generesch et al., 1995; Camara-Clayette et al.,1999), telomeric maintenance (Boulton and Jackson, 1996;Porter et al., 1996; Polotnianka et al., 1998; Haber, 1999;Baumann and Cech, 2000), replicative senescence (Woo et al.,1998; Lim et al., 2000), suppression of chromosomalaberrations, malignant transformation and tumour progression(Difilippantonio et al., 2000; Pucci et al., 2001), aging (Vogelet al., 1999; Cooper et al., 2000; Li and Comai, 2001) cell cycleregulation (Munoz et al., 2001), multidrug resistance (Um etal., 2001) and DNA replication (Ruiz et al., 1999; Novac et al.,2001).

A role for Ku in DNA replication and replication forkmovement was proposed a decade ago, on the basis of theobservation that Ku bound to DNA ends during S phase (Paillardand Strauss, 1991). Since then, evidence for the involvement ofKu in DNA replication has been accumulating. The HeLacell DNA-dependent ATPase, co-fractionating with a 21Smultiprotein complex capable of supporting SV40 in vitro DNA

111

The radiosensitive mutant xrs-5, a derivative of the Chinesehamster ovary (CHO) K1 cell line, is defective in DNAdouble-strand break repair and V(D)J recombination. Thedefective phenotypes of xrs-5 cells are complemented by the86 kDa subunit of Ku antigen. OBA is a protein, previouslypurified from HeLa cells, that binds in a sequence-specificmanner to mammalian origins of DNA replication. TheDNA-binding subunit of OBA has been identified as Ku86.We tested the xrs-5 cell line for its ability to replicate amammalian origin-containing plasmid, p186, in vivo andin vitro. In vivo, the p186 episomal DNA replication intransfected xrs-5 cells was reduced by 45% when comparedwith the CHO K1 cells transfected with p186. In vitro,although total and cytoplasmic cell extracts from xrs-5 cellsreplicated the p186 with the same efficiency as the parentalCHO K1 cell extracts, xrs-5 nuclear extracts did not possessany detectable replication activity. Addition of affinity-

purified OBA/Ku restored replication in the xrs-5 nuclearextract reaction. Western blot analyses showed that thelevels of other replication proteins (Orc2, PCNA, DNApolymerase ε and δ, Primase and Topoisomerase IIα) werecomparable in both the xrs-5 mutant and CHO K1 wild-type cell lines. In addition, the in vivo association of Ku withthe DHFR origin-containing sequence (oriβ) was examinedin both the CHO K1 and xrs-5 cell lines by a chromatinimmunoprecipitation (ChIP) assay. Anti-Ku antibodies didnot immunoprecipitate a detectable amount of Ku from thexrs-5 cells in the origin-containing sequence, in contrast tothe CHO K1 cells, wherein Ku was found to be associatedwith the oriβ origin. The data implicate Ku antigen in invivo and in vitro DNA replication and suggest the existenceof another protein with Ku-like functions in the xrs-5 cells.

Key words:xrs-5 cells, Ku antigen, DNA replication, ChIp

Summary

Analysis of the DNA replication competence of thexrs -5 mutant cells defective in Ku86Diamanto Matheos 1,2, Olivia Novac 1,2, Gerald B. Price 1 and Maria Zannis-Hadjopoulos 1,2,*1McGill Cancer Centre and 2Department of Biochemistry, McGill University, Montréal, Québec, Canada, H3G 1Y6*Author for correspondence (e-mail: [email protected])

Accepted 9 September 2002Journal of Cell Science 116, 111-124 © 2003 The Company of Biologists Ltddoi:10.1242/jcs.00156

Research Article

112

replication (Vishwanatha and Baril, 1990), is identical to Kuantigen (Cao et al., 1994). Ku binds to viral and eukaryoticorigins of DNA replication (de Vries et al., 1989; Toth et al.,1993; Araujo et al., 1999; Ruiz et al., 1999; Novac et al., 2001),to matrix attachment regions (MARs), which serve as initiationsites for DNA replication (Galande and Kohwi-Shigematsu,1999) and to Alu DNA, which provides cis-elements affectingchromatin structure during DNA replication (Tsuchiya et al.,1998). Ku also binds in vivo to genomic sequences comprisingorigins of DNA replication in a cell-cycle-dependent manner,with fivefold higher binding at the G1/S interphase bycomparison to G0 (Novac et al., 2001). Recently, Ku was shownto interact with WRN (Cooper et al., 2000), a DNA helicase with3′-5′ exonuclease activity that is part of the replication complex(Lebel et al., 1999) and interacts with replication protein A (RP-A) (Brosh et al., 1999). Furthermore, OBF2, a Saccharomycescerevisiae Ku homologue, supports the formation of a stablemultiprotein replication complex by binding to essentialreplication sequences (Shakibai et al., 1996), whereas a Ku70yeast mutant strain exhibits a high DNA content during mitoticgrowth (Barnes and Rio, 1997). Finally, the phenotypes of theKu knockout mice, such as their small size, the failure of thecells to proliferate in culture, their prolonged doubling time andtheir premature senescence may also suggest a role of Ku inDNA replication (Featherstone and Jackson, 1999). Extractsprepared from Ku86–/– cells, deficient in Ku protein, haveapproximately 70% lower in vitro DNA replication activitycompared with Ku86+/+ extracts (Novac et al., 2001).

We have previously shown that OBA (Ruiz et al., 1995), aHeLa cell activity whose DNA binding subunit has beenidentified as Ku86, is involved in mammalian DNA replication(Ruiz et al., 1999). The affinity-purified OBA (Ku) binds tomammalian origins of replication, including A3/4, a 36 bpsequence that is common to mammalian origins of DNAreplication including the Chinese hamster DHFR origin-containing sequence (oriβ), the human c-myc, dmnt-1 andlamin B2 origins and the monkey orssequences, among others(Araujo et al., 1999; Ruiz et al., 1999). Depletion of Ku, eitherby inclusion of an oligonucleotide comprising its binding site(A3/4) or by antibodies directed against the Ku protein,inhibited DNA replication to as low as 10-20% in a mammalianin vitro replication system (Ruiz et al., 1999).

The mammalian in vitro replication cell-free system used inthis study mimics nuclear, semi-conservative DNA replicationin vivo (Pearson et al., 1991; Pearson et al., 1994; Zannis-Hadjopoulos et al., 1994; Todd et al., 1995; Pelletier et al.,1997; Pelletier et al., 1999) and is comparable to othermammalian in vitro cell-free systems (Krude, 2000). It uses aplasmid containing a mammalian origin of DNA replicationand extracts from HeLa cells (Pearson et al., 1991) and is basedon the SV40 in vitro replication system, but it does not requirethe viral T antigen. This system has been used to study theeffects on DNA replication of cancer chemotherapeutic drugs(Diaz-Perez et al., 1996; Diaz-Perez et al., 1998), of variousproteins including the Oct-1 transcription factor (Matheoset al., 1998), the GATA-1 factor, Ku antigen and DNApolymerase δ (Ruiz et al., 1999; Matheos et al., 2002), amammalian polynucleotide kinase (Jilani et al., 1999), thecruciform binding 14-3-3 proteins (Novac et al., 2002) andthe replication competence of Ku86+/+ and Ku86–/– mouseembryonic fibroblasts (Novac et al., 2001).

Since the xrs-5 mutant cells have severely reduced levels ofKu86 and Ku70 proteins, here we tested their potential forreplication in vivo and in an in vitro replication system. Wefound that the xrs-5 cells had a reduced ability to support invivo DNA replication upon transfection of p186, a mammalianorigin-containing plasmid. However, total and cytoplasmicextracts from xrs-5 cells replicated the origin-containingplasmid in vitro with the same efficiency as the wild-type CHOK1 cell extracts. By contrast, xrs-5 nuclear cell extracts did notpossess any detectable in vitro replication activity, but additionof affinity-purified OBA/Ku restored in vitro replicationactivity to wild-type levels; this is similar to the way that Ku86cDNA complements the defective repair and recombinationphenotypes (Smider et al., 1994). Also, the levels of otherreplication proteins such as Orc2, PCNA, DNA polymerase εand δ, Primase and Topoisomerase IIα were comparable inboth the xrs-5 mutant and CHO KI wild-type cell lines. Usinga ChIP assay, we also found that in xrs-5 cells no Ku could bedetected bound to the hamster DHFR oriβ in vivo, unlike theresult in CHO K1 cells, in which Ku was bound to this origin.Moreover, we identified a factor in the xrs-5 cytoplasmiccell extracts that bound in a sequence-specific manner to theA3/4 origin sequence, possibly accounting for the efficientreplication of the xrs-5 total and cytoplasmic cell extracts invitro. The data suggest that Ku participates in mammalianDNA replication in vivo and in vitro and that the factor presentin the cytoplasmic extracts can compensate for the lack of Kuin the nuclear extracts.

Materials and MethodsCell cultureThe xrs-5 cell line was derived from the CHO K1 cell line on the basisof its sensitivity to ionizing radiation (Jeggo and Kemp, 1983). CHOK1 and xrs-5 cells were cultured in RPMI medium (Gibco-BRL,Burlington, Ontario, Canada) supplemented with penicillin (100µg/ml), streptomycin (0.1 mg/ml), 1 mM glutamine and 10% fetalbovine serum (Gibco-BRL). Cells were maintained at 37°C in 5%CO2.

Extract preparationCell extracts from log phase CHO K1 and xrs-5 cell monolayers wereprepared as previously described by Pearson et al. (Pearson et al.,1991). Briefly, monolayers were washed twice with isotonic buffer (20mM Tris-HCl pH 7.4, 137 mM NaCl, 5 mM KCl, 1 mM CaCl2, 0.5mM MgCl2 and 250 mM glucose). The cells were collected and lysedin a Type B Dounce homogenizer. Nuclei were removed by 5 minutesof centrifugation at 1200 g and were subsequently used to prepare thenuclear extract. The supernatant was centrifuged for 1 hour at 100,000g in a Beckman type 50 Ti rotor, and the supernatant was used as thecytoplasmic extract. The nuclear pellet was suspended in hypertonicbuffer (hypotonic and 500 mM potassium acetate) in half the volumeused for the cytoplasmic extract. The mixture was incubated for 90minutes on ice, spun in a Beckman SW50.1 rotor at 300,000 g for 1hour and used as the nuclear extract.

Preparation of template DNA for replication assayPlasmid p186 consists of the minimal origin of ors8 (GenBankaccession number M26221). It contains the NdeI/RsaI sub-fragmentof ors8 cloned in the NruI site of pBR322 (Todd et al., 1995). PlasmidDNA was isolated using the QIAGEN-tip 500 column (QIAGEN,Mississauga, Ontario, Canada).

Journal of Cell Science 116 (1)

113DNA replication of xrs-5 cells

In vivo DNA replicationIn vivo DNA replication was performed as previously described(Landry and Zannis-Hadjopoulos, 1991; Pearson et al., 1991; Wu etal., 1993; Nielsen et al., 1994; Zannis-Hadjopoulos et al., 1994; Toddet al., 1995) with the following modification. HeLa cells, at a densityof 3×105, were seeded and transfected with 10 µg plasmid DNA (p186or equimolar amounts of pBR322) using the calcium phosphatetransfection kit (Invitrogen). Cells were also transfected with thepcDNA6.0 HislacZ to control for transfection efficiency. 72 hourspost-transfection, cells were lysed as previously described (Todd etal., 1995) and low molecular weight DNA was isolated and purifiedusing the QIAquick columns (QIAGEN). The DNA was then digestedwith 1 unit of DpnI for 1 hour to digest unreplicated (bacteriallymethylated) plasmid. The DpnI-digested and undigested DNAs wereused to transform TOP 10 E. coli cells (Invitrogen), as previouslydescribed (Landry and Zannis-Hadjopoulos, 1991). The in vivorelative DNA replication was determined by counting the number ofcolonies obtained with the transformation using the DpnI-digestedplasmid and correcting it for the amount of total plasmid DNAisolated as determined by counting the number of bacterial coloniesobtained with undigested plasmid DNA isolated by Hirt lysis. Thereplication level of the xrs-5 cells is expressed as a percentage of thewild-type CHO K1 cells.

Mammalian in vitro DNA replicationReplication assays were performed as previously described (Pearsonet al., 1991; Matheos et al., 1998; Novac et al., 2002), with slightmodifications. CHO K1 or xrs-5 nuclear cell extracts (16 µg),cytoplasmic cell extracts (45 µg) or nuclear and cytoplasmic cellextracts together (8:15 ratio) were added to the replication mixturewith 100 ng of input p186 plasmid DNA. Unmethylated pBluescriptKS+ was also typically included in each in vitro replication reactionas a control for DNA recovery and DpnI resistance. A reaction ofpBR322, a methylated, non-replicating plasmid, was also performedfor each in vitro replication experiment to show that the observedDNA replication was origin dependent.

Experiments involving the addition of A3/4-affinity-purified Ku(Ruiz et al., 1999) were performed by pre-incubating the nuclear cellextracts with 160 ng or 600 ng of affinity-purified Ku on ice for 20minutes, prior to the replication reaction. Replication reactions wereperformed at 30°C for 1 hour. The in vitro replication reactionproducts were purified using the QIAquick PCR purification kit(QIAGEN). Reaction samples were digested with DpnI (New EnglandBiolabs, Mississauga, Ontario, Canada), as previously described(Matheos et al., 1998), and resolved by electrophoresis on 1% agarosegel. Quantification was performed on DpnI-digested products bydensitometric measurements using a phoshorimager analyzer (FujiBAS2000, Stamford, CT), as described previously (Diaz-Perez et al.,1996; Matheos et al., 1998; Novac et al., 2001). The total amount ofDNA recovered from the in vitro replication reaction was determinedby quantitative analysis of the picture of the ethidium-bromide-stainedgel.

Immunoblotting assayNuclear or cytoplasmic extract proteins were resolved by 8% SDS-PAGE, transferred to Immobilon-P (Millipore, Mississauga, Ontario,Canada), probed with human anti-Ku70 (C-19), anti-Ku86 (C-20),anti-PCNA (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), anti-ORC2 (Medical and Biological Laboratories CO, Ltd, Watertown,MA), anti-DNA Polymerase δ (BD Transduction Laboratories, Inc.,Mississauga, Ontario), anti-DNA Polymerase ε (Clone 93G1A), anti-DNA Primase (p49) (Neomarkers, Union City, CA) or anti-Topoisimerase IIα (Ki-S1) (Chemicon International, Inc., Temecula,CA) followed by treatment with horseradish-peroxidase-conjugateddonkey anti-goat, anti-mouse or anti-rabbit IgG (Santa Cruz

Biotechnology, Inc). The blots were developed using the ECL westernblotting detection kit (Amersham-Pharmacia, Baie d’Urfé, Québec,Canada).

In vivo crosslinkingChIP assays were performed as described previously (Novac et al.,2001). Briefly, CHO K1 and xrs-5 cells were crosslinked with 1%formaldehyde (Sigma) in serum-free medium at room temperature for5 minutes. Cells were resuspended in lysis buffer, and the DNA wasfragmented by sonication to an average size of 500-1000 bp in length.Sheared chromatin extracts were precleared with protein G-agarosebeads (Roche Molecular Biochemicals) at 4°C for 1 hour and werethen incubated with either 50 µl of preimmune goat serum (Santa CruzBiotechnology) or 10 µg of anti-Ku70 (M-19) or anti-Ku86 (C-20)goat polyclonal antibodies (Santa Cruz Biotechnology) or anti-clone162 (against the Ku70/86 heterodimer) mouse monoclonalantibody (NeoMarker). The protein-DNA immune complexes werethen precipitated with 50 µl of protein G-agarose beads, washedthoroughly and eluted in 200 µl of extraction buffer (1% SDS in TE).Crosslinks were reversed overnight at 65°C, then incubated at 37°Cfor 2 hours with proteinase K (Roche Molecular Biochemicals). Theextracted DNA was purified by QIAquick PCR purification columns(QIAGEN).

Real-time PCR quantification analysis of immunoprecipitatedDNAPCR reactions were carried out as described previously (Novac etal., 2002), using the LightCycler Instrument (Roche MolecularBiochemicals), with some modifications. The PCR reaction contained1/20th of the immunoprecipitated material, 3 mM Mg2+ and 1 µMof each primer (oriβ F: 5′-ATCCTCCTAGCTCGGAGTCA-3′, 1534-1553 accession no. X94372; oriβ R: 5′-GGCTTATCTGCATCC-TATTC-3′, 1677-1696 accession no. X94372, AF028017F: 5′-GAG-CAGGTATAAGGGCCTTGG-3′, 138-158 accession no. AF028017;AF028017R: 5′-CGGTCTGGGTATGTTTAGCAAGAC-3′, 362-392accession no. AF028017) using the LightCycler capillaries (RocheMolecular Biochemicals) and the QuantiTect SYBR Green PCRreaction mix (QIAGEN). These two primer sets amplify a 152 bpfragment from the DHFR oriβ origin of DNA replication (Kobayashiet al., 1998) (Fig. 5A) or a 255 bp fragment 17 kb downstream fromthe DHFR gene (Kobayashi et al., 1998).

Quantification of the PCR products was assessed by theLightCycler instrument (Roche Molecular Diagnostics) using SYBRGreen I dye as detection format and LightCycler software version 3.5.An initial denaturation of 15 minutes at 95°C was followed by 40cycles with denaturation for 15 seconds at 95°C, annealing at 55°Cfor 10 seconds and polymerization for 20 seconds at 72°C. GenomicCHO K1 DNA (24 ng, 28 ng, 72 ng and 96 ng) was used to build thestandard curve necessary for the quantification of the PCR products(Fig. 4A). A melting curve analysis was also performed at the end ofthe PCR amplification to asses the specificity of the amplified product(Fig. 4B). The melting curve analysis cycle was composed of threesegments: 95°C for 0 seconds (temperature transition rate of20°C/second), 55°C for 30 seconds (temperature transition rate of20°C/second) and a final segment at 95°C for 0 seconds (temperaturetransition rate of 0.2°C/second). The specificity of the 152 bp and 255bp PCR products was also assessed by agarose gel electrophoresis andvisualized with ethidium bromide (Fig. 4C,D).

Electrophoretic mobility-shift assay (EMSA)Nuclear or cytoplasmic cell extracts (10 µg) were incubated with0.5 ng of 32P-end-labeled A3/4 probe (5′ CCTCAAATGGTCTCC-AATTTTCCTTTGGCAAATTCC 3′) for 30 minutes on ice in thepresence of 1 µg poly dI-dC (Amersham-Pharmacia), used as non-

114

specific competitor, in a final volume of 20 µl including binding buffer(10 mM Tris-HCl pH 7.5, 80 mM NaCl, 1 mM EDTA, 10 mM 2-mercaptoethanol, 0.1% Triton X-100, 4% glycerol). The mixtureswere electrophoresed on a 6% PAGE gel at 180 Volts in TBE (45 mMTris-HCl pH 8.0, 45 mM Boric Acid, 1 mM EDTA), and the gel wasdried and subjected to autoradiography.

For electrophoretic mobility-shift competition assays, 0.5 ng of32P-labeled A3/4 was mixed with increasing molar excess amountsof A3/4 or non-specific (5′ TTCCGAATACCGCAAG 3′) coldcompetitor oligonucleotides. A 10 µg extract protein was then added,and the reaction was left to proceed as described above.

Electrophoretic mobility-supershift assayElectrophoretic mobility shift mixtures were prepared as describedabove, and after the standard 30 minutes incubation, 1.5 µg of clone162 antibody (Neomarkers, Union City, CA) or control goat IgG(Sigma, St Louis, MO) was added and further incubated for anadditional 3 hours on ice. The samples were then applied to native 6%PAGE and electrophoresed, as described above.

ResultsIn vivo replication activity of CHO K1 and xrs-5 cellsPrevious studies have shown that xrs-5 mutants are defective inKu86 (Taccioli et al., 1994; Getts and Stamato, 1994; Rathmelland Chu, 1994; Smider et al., 1994), a subunit of the Kuheterodimeric protein. This renders the mutants severelydefective in the repair and recombination processes. Since Ku86binds to origins of DNA replication and has been implicated inmammalian DNA replication (Ruiz et al., 1999; Novac et al.,2001), we tested the ability of xrs-5 cells to support the in vivoreplication of p186, a pBR322-based plasmid that contains theminimal monkey cell origin of ors8 (Todd et al., 1995). Thep186 plasmid replicates autonomously in vivo, whentransfected to mammalian cells and in a cell-free in vitro systemthat uses HeLa cell extracts. The replication of this plasmidinitiates within the ors sequence, is semi-conservative, bi-directional and dependent on the replicative DNA polymerasesα and δ (Pearson et al., 1991; Pearson et al., 1994; Todd et al.,1995). Furthermore, in vivo binding of Ku to this originsequence was recently demonstrated, with a five-fold higherbinding at the G1/S interphase than at G0 (Novac et al., 2001).

Supercoiled plasmid DNA from either p186 or pBR322 wastransfected into either xrs-5 or CHO K1 cells, and the ability ofthe plasmids to undergo autonomous replication was assayed bythe DpnI-resistance assay, to distinguish between input plasmidand plasmid replicated in the eukaryotic cells (Frappier andZannis-Hadjopoulos, 1987; Landry and Zannis-Hadjopoulos,1991). p186 DNA was isolated at 72 hours post-transfectionfrom both CHO K1 and xrs-5 cells and was digested with DpnI;the plasmids recovered from the xrs-5 cells yielded DpnI-resistant DNA, which transformed E. coliwith an efficiency thatwas 45% lower than that DNA recovered from the CHO K1 cells(Fig. 1). No bacterial colonies were obtained with DpnI-digestedDNA recovered from either the xrs-5 or CHO K1 cells that hadbeen transfected with pBR322 (data not shown) as expected,since this plasmid does not contain a mammalian origin of DNAreplication and thus is fully digested by DpnI. To verify that thedecreased recovery of replicated plasmids in the xrs-5 cells wasnot due to the introduction of breaks during transfection and therequirement of NHEJ to repair those breaks prior to replication,undigested DNA from p186 and pBR322 recovered from xrs-5

and CHO K1 cells was also used to transform bacteria. Nodifference in total DNA recovered was found (data not shown),indicating that the observed effect with DpnI digestion (Fig. 1)is due to a replication effect.

Absence of Ku protein from xrs-5 mutant cellsTo examine whether any residual Ku protein could be detectedin xrs-5 cell extracts that might account for their observed invitro DNA replication activity when using the total orcytoplasmic extracts, western immunoblot analyses wereperformed using anti-Ku antibodies (Fig. 2). An anti-humanKu86 antibody (C-20; Santa Cruz) was used that was able torecognize the hamster Ku86 protein, reacting with both theCHO K1 nuclear (K1 N) and cytoplasmic (K1 C) cell extracts(Fig. 2A, lanes 2 and 3). No material crossreacting with thisantibody was detectable either in the cytoplasmic (xrs-5 C) ornuclear (xrs-5 N) cell extracts derived from the xrs-5 cells (Fig.2A, lanes 4 and 5), which is in agreement with previous reportsstating that xrs-5 cells lack Ku86 protein, shown by westernblotting, and Ku86 RNA transcript, shown by northern blotting(Rathmell and Chu, 1994; Singleton et al., 1997). However, theKu86 transcript was detected in the xrs-5 mutants when themore sensitive technique of RT-PCR was used (Singleton et al.,1997). Each subunit of the Ku heterodimer is required tostabilize the other, and the absence of Ku86 in xrs-5 has beenshown to result also in the loss of the Ku70 subunit (Taccioliet al., 1994; Smider et al., 1994; Singleton et al., 1997). Toexamine the levels of Ku70 in the xrs-5 cell extracts, the samemembrane was probed with an anti-human Ku70 antibody (C-19; Santa Cruz) recognizing the Ku70 protein in the CHO K1nuclear and cytoplasmic cell extracts (Fig. 2B, lanes 2 and 3).By contrast, xrs-5 showed dramatically decreased levels ofKu70 (Fig. 2B, lanes 4 and 5). Upon longer exposures, a faintband crossreacting with the anti-Ku70 antibody was detectedin the xrs-5 cytoplasmic cell extracts.

Examination of the levels of replication protein in xrs-5vs CHO K1 Western blot analyses were also performed to verify the levelof other replication proteins in the mutant cell line and to


Fig. 1. In vivo DNA replication activity of CHO K1 and xrs-5 cells.CHO K1 or xrs-5 cells were transfected with either p186 or pBR322DNA. 72 hours post-transfection, plasmid DNA was isolated by themethod of Hirt, then purified and digested with DpnI. The DpnI-digested DNA was then used to transform E. coli. The number ofbacterial colonies produced was counted, corrected for the amount ofDNA recovered and related to the positive control reaction with theCHO K1 cells, which was taken as 100%. The bars represent theerror from the average of two experiments performed in triplicate.


compare it to the wild-type cells. Nuclear extracts from eitherCHO K1 or xrs-5 cells showed equivalent amounts of ORC2,PCNA, DNA polymerase ε, DNA polymerase δ, Primase orTopoisomerase IIα proteins in both the mutant and wild-typecell lines (Fig. 3A-F).

Ku associates in vivo with the DHFR oriβ in CHO K1cells but not in xrs-5 cellsTo analyze whether the Ku protein associates with replicationorigins in Chinese hamster cells, as it was found to do inmonkey (CV-1) cells (Novac et al., 2001), its interaction withthe DHFR replication origin, oriβ (Fig. 5A), was analyzed inwild-type CHO K1 cells and compared to that in the Ku86mutant cells (xrs-5)using a ChIP assay, as previously described(Novac et al., 2001). Genomic CHO K1 DNA was used to buildthe standard curve necessary for the quantification of theimmunoprecipitated DNA using the LightCycler instrument(Fig. 4A). The LightCycler allows the quantification of thePCR products (considered background level) that areundetectable by agarose gel electrophoresis stained withethidium bromide (Fig. 4C,D). To verify the specificity of thePCR products, a melting curve analysis was performed at theend of the amplification cycle, and no primer-dimers interferedwith the quantification (Fig. 4B). In addition, PCR productswere also separated on a 2% agarose gel to verify their size(Fig. 4C,D).

Anti-Ku70, anti-Ku86 and anti-clone 162 (directed againstthe Ku70/86 heterodimer) antibodies were used toimmunoprecipitate in vivo protein-DNA complexes that hadbeen crosslinked by treatment of the cells with formaldehyde(Novac et al., 2001). The isolated DNA was analyzed for anorigin-containing-sequence using the quantitative real-timePCR approach (Novac et al., 2001; Novac et al., 2002;

Ladenburger et al., 2002). In agreement with other ChIPanalyses (Alexandrow et al., 2002; Ladenburger et al., 2002;Novac et al., 2001; Novac et al., 2002), the formaldehydecrosslinking approach used to study protein-DNA interactionsin vivo largely prevents covalent binding of non-specificproteins to chromatin. With the CHO K1 cells, whenpreimmune goat serum (NGS) was used, the quantified DNAabundance in the origin-containing-sequence corresponded tobackground levels (3×103 molecules/2×108 cells) (Fig. 5B).This is considered to be the non-specific DNA pulled downwhen performing the ChIP assay. Similarly, the DNAimmunoprecipitated non-specifically with anti-clone162antibody in non-crosslinked CHO K1 and xrs-5 cells was alsoquantified and corresponded to background levels (Fig. 5B). In

Fig. 2. Immunoblot analysis of CHO K1 and xrs-5 cell extracts.Nuclear (N; panels A and B, lanes 3 and 5) or cytoplasmic (C; panelsA and B, lanes 2 and 4) cell extracts from CHO K1 or xrs-5 cellswere examined by western blotting using a Ku86 (A) or Ku70 (B)antibody. HeLa whole-cell extracts (lane 1) were used as positivecontrols.

Fig. 3. Immunoblot analysis of replication proteins in CHO K1versus xrs-5 cells. Nuclear (N; panels A-F) cell extracts from CHOK1 or xrs-5 cells were examined by western blotting using anti-ORC2 (A), anti-PCNA (B), anti-DNA Polymerase ε (C), anti-Polymerase δ (D), anti-Primase (E) and anti-Topoisomerase IIα (F)antibodies.

116

addition, a primer set amplifying a fragment outside ofthe DHFR initiation zone (17 kb downstream from theDHFR gene) was also used as an additional control for non-specific DNA binding (Fig. 4A,B,D; Fig. 5B). The DNA

immunoprecipitated from that region with anti-clone162, anti-Ku86 or anti-Ku70 antibodies from crosslinked CHO K1 cellswas comparable to background levels (Fig. 5B). When theimmunoprecipitation was performed with anti-Ku70, anti-Ku-


Fig. 4.Ku is associated with the DHFRoriβ origin of DNA replication in CHO K1cells but not in xrs-5 mutant cells.(A) Standard curve, using CHO K1genomic DNA as template, used for thequantification of DNA abundance in theorigin-containing sequence (oriβ) or in thenon-origin-containing sequence (17 kbdownstream from the DHFRgeneamplified by AF028017 primer set) byreal-time PCR. The LightCycler software3.5 calculates the copy number ofmolecules, amplified by the respectiveprimer set, by plotting on the x-axis thelogarithm of fluorescence and on the y-axisthe cycle number and setting a baseline x-axis. (B) Melting peak analysis of the 152

bp or 255 bp PCR amplification products performed at the endof the PCR amplification cycle. Melting peaks were generatedby plotting the negative derivative of the SYBR Greenfluorescence with respect to temperature (–dF/dT) againsttemperature (°C). (C) LightCycler PCR amplification products,amplified with the oriβ primer set, were separated on a 2%agarose gel, visualized with ethidium bromide andphotographed with an Eagle Eye apparatus. Lane 1 represents a50 bp marker ladder (Amersham). Template DNA was asfollows. Lane 2, 1/20th of DNA recovered fromimmunoprecipitation with normal goat serum (NGS) incrosslinked CHO K1 cells. Lanes 3-6, CHO K1 genomic DNA

(S1-S4) from untreated cells used to build the standard curve in A. Lanes 7-14, 1/20th of DNA recovered from immunoprecipitation with anti-Ku70, anti-Ku86 or anti-clone162 from CHO K1 or xrs-5 cells crosslinked or untreated with formaldehyde. (D) As for C but with theAF028017 primer set. Template DNA was as follows. Lane 1 represents a 50 bp marker ladder (Amersham). Lanes 2-4, CHO K1 genomicDNA (S1-S3) from untreated cells used to build the standard curve in (A). Lane 5, water was used as template. Lanes 6-8, 1/20th of DNArecovered from immunoprecipitation with anti-Ku70 anti-Ku86, or anti-clone162 from CHO K1 cross-linked cells.


86 or anti-Ku70/86 antibodies in CHO K1 formaldehyde-crosslinked cells, the DNA abundance in the oriβ wasapproximately 13-fold, seven-fold or sixfold higher thanbackground level, respectively (Fig. 5B). By contrast, anti-Ku70, anti-Ku86 or anti-Ku70/86 antibodies precipitated theoriβ DNA in the xrs-5 cells at the background level (Fig. 5B),as expected. Thus, Ku associates with the origin of DNAreplication in vivo in CHO K1 cells but not in the xrs-5 mutantcells, where it is mutated.

A3/4 binding activity in CHO K1 and xrs-5 cell extractsElectrophoretic mobility shift and super-shift assays wereperformed to examine whether there was an A3/4-specificbinding activity present in the wild-type and mutant CHO cellextracts (Fig. 6A,B) that might account for the in vitroreplication results. First, titration of the amount of protein andoligonucleotide allowed us to determine the optimumamounts to use in order to avoid multiple bindings of Kumolecules, producing a ladder effect. When HeLa cellextracts were reacted with radiolabeled A3/4, two maincomplexes arose (Fig. 6A, lane B). The slower migratingcomplex (*) results from the interaction of A3/4 with the Kuheterodimer, Ku70/Ku86, whereas the faster migratingcomplex (**) arises from its interaction with the truncatedKu, Ku70/Ku69. Ku69 is a truncated form of Ku86 thatresults from site-specific proteolytic cleavage by a leupeptin-sensitive protease (Quinn et al., 1993; Han et al., 1996; Jenget al., 1999). The levels of the truncated complex vary

depending on the leupeptin concentration; at concentrationsabove 5 µM the truncated complex is not formed (Jeng et al.,1999) (D.M., O.N., G.B.P. and M.Z.-H., unpublished). Thiscomplex is not obtained with the CHO K1 or xrs-5 extractsowing to the higher leupeptin concentrations. However, aband immediately below the ** band was obtained, which isprobably due to non-specific binding since this complex isalso present with the control antibody. The migration of thesecomplexes was further retarded by clone 162 antibody(Fig. 6A, lane C, ***), which recognizes the Ku70-Ku86heterodimer, but not by the control antibody (Fig. 6A, laneD). The reaction of radiolabeled A3/4 oligonucleotide withthe CHO K1 nuclear cell extracts yielded a complex ofsimilar migration (Fig. 6A, lane E) to that obtained from itsreaction with the HeLa cell extracts (compare with Fig. 6A,lane B). The complexes obtained with the CHO extracts wereconsistently fainter. This might be due to the CHO extractshaving lower levels of Ku (Fig. 2) or because total HeLaextracts were used as opposed to either nuclear orcytoplasmic extracts used for the CHO extracts. This complexwas also supershifted by the clone 162 antibody (Fig. 6A,lane F) but not by the control antibody (Fig. 6A, lane G). TheEMSA reaction of radiolabeled A3/4 oligonucleotide with thexrs-5 nuclear cell extracts did not result in a Ku-A3/4complex (Fig. 6A, lanes H, J), consistent with the resultsobtained by the western blot analysis, in which neither Ku70nor Ku86 were detected (see Fig. 3). The EMSA reaction withxrs-5 cytoplasmic cell extracts (Fig. 6B) also resulted in acomplex with similar migration to the HeLa (Fig. 6A, lane B)

Fig. 5.Quantification of DNAabundance in the DHFRoriβregion by real-time PCR.(A) Map of the DHFR initiationzone, containing the β, β′ and γinitiation sites andencompassing the DHFRand2BE2121genes. The arrowsrepresent the location of theamplification product using aprimer set within the oriβregion. (B) Quantitative real-time PCR using the LightCyclerinstrument, using a primer setwithin the oriβ region or ~17 kbdownstream from the DHFRgene (AF028017), with DNAtemplate extracted from theimmunoprecipitation with anti-Ku70, anti-Ku86, anti-clone162antibodies or NGS fromcrosslinked or untreated CHOK1 or xrs-5 cells. Each barrepresents two experiments, andone standard deviation isindicated.

118

and CHO K1 nuclear (Fig. 6A, lane E) and CHO K1cytoplasmic (Fig. 6B, lane A) cell extracts. This complex,however, was not supershifted by the clone 162 antibody (Fig.6B, lane E), unlike the complex generated with the CHO K1cytoplasmic cell extracts (Fig. 6B, lane B; not visible),suggesting that the epitope that is normally recognized by thisantibody is either not present in the xrs-5cytoplasmic extractsor it is not accessible. Faster migrating complexes, probablydue to degradation, were also detected.

An A3/4-specific binding protein is present in CHO K1and xrs-5 cytoplasmic cell extractsThe binding specificity for the A3/4 oligonucleotide in thecomplexes formed with the xrs-5 cytoplasmic cell extracts wastested by competition bandshift assays, using increasing molarexcess of cold A3/4 (Fig. 7, lanes 3, 4, 8, 9) as a specificcompetitor and cold pBR322-derived oligonucleotide (Fig. 7,lanes 5, 6, 10, 11), which was used as non-specific competitor.The A3/4-Ku complex formed with both the CHO K1 (Fig. 7,lane 7) and xrs-5 (Fig. 7, lane 2) cytoplasmic cell extracts wasspecifically competed with increasing concentrations of cold

A3/4 but not with cold non-specific competitor, indicating thatthe complex formed with these extracts represents a specificprotein interaction with A3/4. Again some faster migratingcomplexes were also detected, which were probablyattributable to degradation products.

In vitro replication activity of CHO K1 and xrs-5 cellextractsSince the in vivo autonomous replication assay showed that thexrs-5 cells were impaired in DNA replication, we tested theirreplication activity in the mammalian in vitro replicationsystem, which allows the dissection and study of the proteinsrequired for DNA replication (Diaz-Perez et al., 1996; Diaz-Perez et al., 1998; Matheos et al., 1998; Ruiz et al., 1999; Jilaniet al., 1999; Novac et al., 2001; Novac et al., 2002; Matheos etal., 2002). The CHO cell extracts, both wild-type (K1) andmutant (xrs-5), yielded similar in vitro replication products(Fig. 8A) to those routinely obtained with the HeLa cellextracts (Pearson et al., 1991; Zannis-Hadjopoulos et al., 1994;Matheos et al., 1998), namely, relaxed circular (form II), linear(form III) and supercoiled (form I). However, in vitro


Fig. 6. A3/4 binding activity in CHO K1 and xrs-5 cell extracts. (A) Nuclear (N) cell extracts from CHO K1 or xrs-5 cells were mixed with aradiolabeled double-stranded A3/4 DNA probe. HeLa nuclear (N) and cytoplasmic (C) extracts were used as a positive control (HeLa NC).Following the binding reaction, clone 162 or control antibody was added to the mixture, as indicated. The DNA-protein complexes wereseparated by 6% PAGE. The Ku70/Ku86-A3/4 (*), Ku70/Ku69-A3/4 (**) and the supershifted complexes (***) are indicated. (B) As in panelA except that cytoplasmic (C) cell extracts from CHO K1 and xrs-5 cells were used.


replication of p186 DNA using the CHO cell extracts wasconsistently less efficient (approximately nine times) than withthe HeLa cell extracts (data not shown). This is consistentwith our previous findings and those of other laboratories,suggesting that HeLa cell extracts may be producing higherconcentrations of initiator proteins, resulting in more efficientreplication than that observed with CV-1 or COS-7 cell extracts(Pearson et al., 1991; Stillman and Gluzman, 1985; Wobbeet al., 1985; Li and Kelly, 1984; Guo et al., 1989). Otherlaboratories have also reported differences in in vitroreplication activities of cell extracts, depending on their source(Krude, 2000; Stoeber et al., 1998).

We next performed in vitro DNA replication assays, usingeither cytoplasmic or nuclear extracts separately, or the twotogether, from either CHO K1 or xrs-5 cells (Fig. 8A,B). TheCHO K1 cytoplasmic cell extracts (Fig. 8A, lanes 3 and 4)replicated the p186 DNA as efficiently as the CHO K1 totalcell extracts (Fig. 8A, lanes 1 and 2), whereas the nuclear cellextracts alone (Fig. 8A, lanes 5 and 6) were approximatelysixfold less efficient. The profiles obtained for both the totalincorporation of radioactive precursor nucleotide (data notshown), indicating total incorporation owing to bothreplication and repair synthesis, and DpnI-resistance (Fig. 8A),indicating incorporation owing to replication alone, weresimilar for the two types of cell extracts. Furthermore, thequantification profiles of the DpnI-resistant bands, generated

by the two reactions and corresponding to DNA forms II andIII, were virtually the same (Fig. 8A). The xrs-5 cytoplasmiccell extracts (Fig. 8A, lanes 9 and 10) also showed comparablereplication activity to the xrs-5 total (Fig. 8A, lanes 7 and 8)and the CHO K1 total (Fig. 8A, lanes 1 and 2) and cytoplasmic(Fig. 8A, lanes 3 and 4) cell extracts. By contrast, the xrs-5nuclear cell extracts did not support the in vitro replication ofthe p186 DNA, as indicated by the failure to incorporate anyradioactive precursor (Fig. 8A, lanes 11 and 12). WhenpBR322 was used as template DNA for the in vitro replicationreaction, no DpnI-resistant products were obtained, althoughsome incorporation of radionucleotides into the DNA with boththe CHO K1 extracts (nuclear and cytoplamic) and the xrs-5cytoplasmic extracts was seen (data not shown), owing to DNArepair, as also observed previously (Pearson et al., 1991).

OBA/Ku restores replication activity of xrs-5 nuclear cellextractsTo determine whether Ku could restore replication activity inthe xrs-5 nuclear cell extracts, additional assays wereperformed where affinity-purified OBA/Ku was added to thein vitro replication reaction (Fig. 8C,D). The exogenousaddition of affinity-purified OBA/Ku increased replication ofthe xrs-5 nuclear cell extracts from 0% to approximately 60%,relative to the replication activity of the CHO K1 nuclearextract (Fig. 8D). Hence, Ku was able to complement the lackof replication activity in the xrs-5 nuclear extract, which issimilar to the way that Ku86 cDNA complements the defectiverepair and recombination phenotypes (Smider et al., 1994). Nosignificant effect was observed when affinity-purified OBA/Kuwas added to the CHO K1 nuclear cell extracts in the in vitroreplication reaction (Fig. 2D, lanes 2 and 3; Fig. 2E).

DiscussionIn the present study, we have examined the replication activityof the xrs-5 cells, defective in the 86 kDa subunit of Ku antigen,a protein known to be involved in numerous cellular metabolicprocesses, including DNA replication. The in vivo replicationresults (Fig. 1, Fig. 5B) clearly demonstrate an involvement ofKu in DNA replication, whereby in the absence of the proteinthe replication level was reduced by almost half whencompared to the control cells. Furthermore, Ku was foundassociated with the DHFR oriβ origin of DNA replication inCHO K1, but not in xrs-5 cells, in agreement with datapreviously obtained in vitro and in vivo with Ku86–/– cells(Novac et al., 2001) and with HeLa cells depleted of Ku antigen(Ruiz et al., 1999). The in vitro replication results with the xrs-5 nuclear extracts of the present study, showing a complete lossof replication activity (Fig. 8) that is restored upon addition ofaffinity-purified OBA/Ku protein (Fig. 8C,D), further supportsa role for Ku in DNA replication. Surprisingly, xrs-5cytoplasmic cell extracts (Fig. 8A,B) and xrs-5 total cellextracts (Fig. 8A,B) replicated the DNA as efficiently as thewild-type CHO K1 extracts. The discrepancy between the invivo and the in vitro replication results with total extracts maybe related to the absence of nuclear structure in vitro. An intactnuclear structure has been suggested to be required for theinitiation of DNA replication in human (Krude et al., 1997),yeast (Pasero et al., 1997) and Xenopus(Wu et al., 1997;

Fig. 7. Competition/bandshift assay using CKO K1 and xrs-5cytoplasmic cell extracts. The electrophoretic mobility shift assaywas performed by adding cytoplasmic (C) cell extracts from xrs-5(lanes 2-6) or CHO K1 (lanes 7-11) to a mixture containingradiolabeled A3/4 oligonucleotide (lane 1). Some reactions containedno additional competitor DNA (lanes 2 and 7). Lanes 3 and 8contained 50× molar excess of cold A3/4 competitor, lanes 4 and 9contained 500× cold A3/4 competitor, lanes 5 and 10 contained 50×non-specific cold competitor and lanes 6 and 11 contained 500× non-specific competitor. The positions of the protein-DNA complexes andof the free probe are indicated.

120

Dimitrova and Gilbert, 2000) cells. Furthermore, nuclei frommimosine-arrested HeLa cells do not require the nuclearmembrane for initiation of DNA replication in soluble cellextracts from proliferating human cells (Krude et al., 1997;Krude, 2000). One role of the nuclear structure may be inestablishing initiation sites. When sperm chromatin or nakedDNA is added to Xenopusegg extracts, DNA replication isinitiated randomly and at many sites along the DNA, whereas,if intact nuclei are used, site-specific initiation occurs (Gilbertet al., 1995). It is possible that some Ku protein sequestered inthe nuclear membrane of the xrs-5 cells (Yasui et al., 1999) isreleased into the cytoplasm during the extract preparation,thereby allowing the total and cytoplasmic xrs-5 cell extractsto replicate the DNA as efficiently as the wild-type cells.Alternatively, another protein, or a modified form of Ku

antigen, may exist in the Ku-deficient cells, which would allowthe cells or extracts to initiate replication. Such a protein withsequence-specific binding to the A3/4 sequence was apparentlypresent in the xrs-5 cytoplasmic extracts (Fig. 6A,B, Fig. 7A).Ku is an origin-binding protein involved in DNA replication,and in its absence, other protein(s) with a similar function maybe present, since (1) knockout mice are viable (Nussenzweiget al., 1996), (2) extracts from Ku86–/– cells or HeLa cells withdepleted Ku antigen have a basal level of replication in vitro(Ruiz et al., 1999; Novac et al., 2001) and (3) the in vivoreplication activity of xrs-5 cells is reduced by half (Fig. 1) butis not altogether abolished.

Upon examination of the Ku protein levels in the xrs-5 cellextracts by western blot analyses, it was found that xrs-5cytoplasmic or nuclear cell extracts did not have any detectable


Fig. 8. In vitro replication activityof the CHO K1 and xrs-5 cellextracts. (A) Typicalautoradiograph of DNA replicationproducts. p186 was incubated inreaction mixtures containing CHOK1 or xrs-5 total cell extracts. TheDNA was purified, concentratedand a sample was digested with 1unit of DpnI for 1 hour at 37°C.The DpnI-digested (lanes 1-12)samples were subjected toelectrophoresis on 1% agarose gel.The supercoiled (I), relaxedcircular (II) and linear (III) formsof the plasmid and the DpnI-digestion products are indicated.Duplicate samples are shown.(B) xrs-5 nuclear cell extracts donot replicate p186. In vitro DNAreplication assays were performedwith CHO K1 or xrs-5 nuclear andcytoplasmic (NC), cytoplasmic (C)or nuclear (N) cell extracts.Quantification of DNA replicationactivities of the cell extracts wasdone relative to the CHO K1 NCreaction. Each bar represents theaverage of four experiments andone standard deviation is indicated.(C) Ku restores replication activityto the xrs-5 nuclear cell extracts. Invitro DNA replication assays wereperformed with either K1 or xrs-5nuclear cell extracts in the presenceof A3/4-affinity-purified Ku. Anautoradiograph of the replicationproducts DNA forms II and III areindicated. Lane 1 and 4, 0 ngaffinity-purified Ku (OBA); lane 2and 5, 160 ng affinity-purified(OBA); lane 3 and 6, 600 ngaffinity-purified (OBA).(D) Quantification of DNAreplication activities of the cell nuclear extracts, relative to the CHO K1 N reaction using 0 ng of OBA, as described in C. Each bar representsthe average of three experiments, and one standard deviation is indicated.


levels of Ku70 or Ku86 proteins (Fig. 2A,B). This is inagreement with previous studies in which neither Ku86 proteinnor Ku86 transcript were detected in xrs-5 cells or other celllines of the XRCC5 group, by western and northern analyses,respectively (Rathmell and Chu, 1994; Singleton et al., 1997;Errami et al., 1998). Since both Ku subunits are required forprotein stability (Errami et al., 1996; Singleton et al., 1997) andsince Ku86 probably regulates Ku70 levels post-transcriptionally (Chen et al., 1996), it is not surprising to findreduced Ku70 levels as a result of reduced Ku86 levels.However, studies with Ku knockout mice have demonstratedthat inactivation of Ku70 resulted in a phenotype that wasdistinct from that obtained with Ku86-knockout mice,suggesting that Ku70 and Ku86 have functions that areindependent of each other (Nussenzweig et al., 1996; Gu et al.,1997; Li et al., 1998) and that the remaining protein functionsas a monomer or a homodimer (reviewed in Featherstone andJackson, 1999). Low levels of wild-type Ku86 transcripts,however, have been detected in the xrs-5 cells by the moresensitive technique of RT-PCR (Singleton et al., 1997). Thelow level of wild-type transcript could be either due to thepresence of some revertants in the xrs-5 cell population or toa low level of transcription in each cell. By contrast, Yasui etal. have recently reported the presence of both Ku70 and Ku86proteins in the xrs-5 cells, albeit at low concentrations, usingwestern blot and 2D gel electrophoresis analyses (Yasui et al.,1999). Furthermore, by indirect immunofluorescence, Ku70and Ku86 were found distributed over both the cytoplasm andnuclei of CHO K1 and xrs-5 cells, with enhanced perinuclearlocalization of Ku86 in the xrs-5 cells (Yasui et al., 1999). Thesubcellular localization of Ku has been controversial, asdiscussed in Koike et al. (Koike et al., 1999). Reports of purelynuclear (Koike et al., 1999; Bakalkin et al., 1998), membrane(Dalziel et al., 1992), cytoplasmic (Bakalkin et al., 1998) andboth nuclear and cytoplasmic (Fewell and Kuff, 1996)localization of Ku have been published (reviewed in Koike etal., 1999). The discrepancy is probably due to differences inthe detection methods used or to a change in Ku’s subcellularlocalization during the cell cycle. Ku has been shown to bepresent in the cytoplasm of mitotic cells and at the peripheryof condensed chromosomes (Koike et al., 1999). The sameinvestigators found Ku70 and Ku86 associated as aheterodimer throughout the cell cycle. However, recentevidence suggests that Ku70 and Ku86 may not always bedimerized, since the nuclear translocation of Ku70 precedesthat of Ku86 at the late telophase/early G1 phase during thecell cycle (Koike et al., 1999). Although an enhancedperinuclear localization of only Ku86 in xrs-5 cells (Yasui etal., 1999) is surprising, the localization observed may be thatof the truncated form of Ku86, which is stable when notcomplexed to Ku70 (Singleton et al., 1997). Unlike the wild-type protein, truncated stable forms of Ku86 have been morefrequently detected without Ku70, perhaps because of proteinconformational changes or loss of degradation signalsequences (Singleton et al., 1997). Yasui et al. postulate thatthe concentration of Ku86 at the nuclear periphery of xrs-5cells keeps the Ku complex sequestered, thereby preventing itfrom accessing the DNA (Yasui et al., 1999). In terms of DNArepair, this might prevent the Ku protein complex fromaccessing DNA DSBs. Recently, it was shown that each Kusubunit can translocate to the nucleus independently of its own

NLS and that this translocation is dependent on its interactionwith the other subunit (Koike et al., 2001). Furthermore,irradiation of cells resulted in an upregulation of the cellularlevel of Ku70 but not Ku86 (Brown et al., 2000). The inabilityof the xrs-5 cells to replicate the DNA efficiently in vivo andof the xrs-5 nuclear cell extracts to replicate DNA in vitro maybe due to either a lack of or the presence of very low levels ofKu86. Alternatively, Ku86 might be localized in the nuclearperiphery and thus be unavailable for interaction with the DNAand the replication complex. The replication of the xrs-5cytoplasmic extract activity might be due to the presence ofanother protein with some affinity for the replication origin.Such a protein may also account for the observed in vivoreplication activity of xrs-5 cells of 55% (Fig. 1). It issurprising that this protein is found in the cytoplasm, but thismay reflect a more dynamic situation in the cells, as seen inthe in vitro replication reaction. That is, this protein maytranslocate to the nucleus and bind to the origin at G1/S onlyand then dissociate and exit from the nucleus. Since theextracts used for the in vitro replication reaction are preparedfrom log phase cells, the majority of this protein would bepresent in the cytoplasm, with undetectable levels in thenucleus (Figs 6, 7 and 8). This would also explain why the invivo association of Ku with the DHFR oriβ origin of DNAreplication, in logarithmically growing xrs-5 cells, wascomparable to background levels, whereas its association withthe same origin of DNA replication was approximatelyninefold higher in CHO K1 cells (Fig. 5B).

The lack of in vitro replication activity of the xrs-5 nuclearextracts was restored upon addition of affinity-purifiedOBA/Ku (Fig. 8C,D). The ability of affinity-purified OBA/Kuto rescue replication activity in the xrs-5 nuclear cell extractsfurther demonstrates that it plays a role in mammalian DNAreplication. Addition of OBA/Ku to the CHO K1 nuclear cellextracts had no effect on replication (Fig. 8D,E), suggestingthat the level of Ku in CHO K1 extracts was sufficient foroptimum replication.

Reaction of the xrs-5 cytoplasmic cell extracts withradiolabeled A3/4 produced a protein-DNA complex with asimilar migration to the Ku-A3/4 complex produced by theHeLa and CHO K1 cell extract (Fig. 6). Furthermore, thiscomplex was specific, as determined by competition with coldA3/4 DNA (Fig. 7). However, the xrs-5 cytoplasmic cellextracts A3/4 complex was not recognized by the anti-Ku(clone 162) antibody, which recognizes a conformationalepitope of the Ku70-Ku86 heterodimer (Fig. 6B) or by theindividual anti-Ku70 or anti-Ku86 antibodies (data not shown),suggesting that this epitope is not present. This implies that thecytoplasmic protein that recognizes A3/4 is another origin-specific binding protein with similar affinities for A3/4, or amodified form of Ku86 that is able to complex with Ku70,albeit in a manner that is not recognized by the clone 162antibody. Further studies are underway to address the nature ofthis protein.

Absence of the Ku86 protein, either in Ku86–/– mice orcells, results in hypersensitivity to ionizing radiation, defectiveDSB-repair pathways and lymphocyte development and earlyonset of an age-related phenotype including osteopenia,hepatocellular degeneration and shortened life span (Vogelet al., 1999; Nussenzweig et al., 1996; Gu et al., 1997).Furthermore, Ku86 knockout mice are half the size of their

122

heterozygous littermates, their cells have prolonged doublingtimes in culture, owing to rapid loss of proliferating cells, andthey exhibit replicative senescence (Nussenzweig et al., 1996;Gu et al., 1997). In vitro DNA replication using extractsprepared from Ku86–/– cells showed a 70% decrease in thereplication level, compared to the control (Novac et al., 2001).Depletion of Ku from HeLa cells also resulted in inhibition ofDNA replication to a level that was 10-20% of normal in vitroreplication (Ruiz et al., 1999). The fact that the knockout miceare viable and DNA replication is not completely abolishedwhen Ku is absent suggests that other mechanisms may takeover in the absence of Ku. We are presently investigating thesemechanisms in the xrs-5 and the Ku86–/– cells with the aim offurther understanding the involvement of Ku antigen in DNAreplication.

We thank Terry Chow (Montréal General Hospital, Montréal,Québec, Canada) for generously providing the xrs-5 cell line andMarcia T. Ruiz (McGill University Cancer Centre, Montréal, Québec,Canada) for providing A3/4 affinity-purified Ku. This work wassupported by funds from the Canadian Institute of Health Research(M.Z.-H.) and the Cancer Research Society (G.B.P.). D.M. is arecipient of studentships from the Cancer Research Society, theMcGill University Faculty of Medicine and the Défi CorporatifCanderel. O.N. is a recipient of studentship from the Fonds deRecherche en Sante du Quebec (FRSQ).

ReferencesAlexandrow, M. G., Ritzi, M., Pemov, A. and Hamlin, J. (2002). A

potential Role for mini-chromosome maintenance (MCM) proteins ininitiation at the dihydrofolate reductase replication origin. J. Biol. Chem.277, 2702-2708.

Araujo, F. D., Knox, J. D., Ramchandani, S., Pelletier, R., Bigey, P., Price,G., Szyf, M. and Zannis-Hadjopoulos, M. (1999). Identification ofinitiation sites for DNA replication in the human dnmt1 (DNA-methyltransferase) locus. J. Biol. Chem.274, 9335-9341.

Arosio, D., Cui, S., Ortega, C., Chovanec, M., di Marco, S., Baldini, G.,Falaschi, A. and Vindigni, A. (2002). Studies on the mode of Kuinteractions with DNA. J. Biol. Chem. 277, 9741-9748.

Bakalkin, G., Yakovleva, T., Hurd, Y. L., Nussenzweig, A., Li, G. C. andTerenius, L. (1998). Autoantigen Ku in the brain. Developmentally regulatedexpression and subcellular localization. Neuroreport9, 2147-2151.

Barnes, G. and Rio, D.(1997). DNA double-strand-break sensitivity, DNAreplication, and cell cycle arrest phenotypes of Ku-deficient Saccharomycescerevisiae. Proc. Natl. Acad. Sci. USA 94, 867-872.

Baumann, P. and Cech, T. R. (2000). Protection of telomeres by the Kuprotein in fission yeast. Mol. Biol. Cell11, 3265-3275.

Boulton, S. J. and Jackson, S. P.(1996). Identification of a SaccharomycescerevisiaeKu80 homologue: roles in DNA double strand break rejoiningand in telomeric maintenance. Nucleic Acids Res.24, 4639-4648.

Brosh, R. M., Jr, Orren, D. K., Nehlin, J. O., Ravn, P. H., Kenny, M. K.,Machwe, A. and Bohr, V. A. (1999). Functional and physical interactionbetween WRN helicase and human replication protein A. J. Biol. Chem.274,18341-18350.

Brown, K. D., Lataxes, T. A., Shangary, S., Mannino, J. L., Giadina, J. F.,Chen, J. and Baskaran, R.(2000). Ionizing radiation exposure results inup-regulation of Ku70 via a p53/Ataxia-Telangiectasia-mutated protein-dependent mechanism J. Biol. Chem.275, 6651-6656.

Camara-Clayette, V., Thomas, D., Rahuel, C., Barbey, R., Cartron, J. P.and Bertrand, O. (1999). The repressor which binds the –75 GATA motifof the GPB promoter contains Ku70 as the DNA binding subunit. NucleicAcids Res.27, 1656-1663.

Cao, Q. P., Pitt, S., Leszyk, J. and Baril, E. F.(1994). DNA-dependentATPase from HeLa cells is related to human Ku autoantigen. Biochemistry33, 8548-8557.

Chen, F., Peterson, S. R., Story, M. D. and Chen, D. J.(1996). Disruptionof DNA-PK in Ku80 mutant xrs-6 and the implications in DNA double-strand break repair. Mutat. Res.362, 9-19.

Cooper, M. P., Machwe, A., Orren, D. K., Brosh, R. M., Ramsden, D. andBohr, V. A. (2000). Ku complex interacts with and stimulates the Wernerprotein. Genes Dev.14, 907-912.

Dalziel, R. G., Mendelson, S. C. and Quinn, J. P.(1992). The nuclearautoimmune antigen Ku is also present on the cell surface. Autoimmunity13, 265-277.

de Vries, E., van Driel, W., Bergsma, W. G., Arnberg, A. C. and van derVliet, P. C. (1989). HeLa nuclear protein recognizing DNA termini andtranslocating on DNA forming a regular DNA-multimeric protein complex.J. Mol. Biol.208, 65-78.

Diaz-Perez, M. J., Wainer, I. W., Zannis-Hadjopoulos, M. and Price, G. B.(1996). Application of an in vitro system in the study of chemotherapeuticdrug effects on DNA replication. J. Cell. Biochem.61, 444-451.

Diaz-Perez, M. J., Zannis-Hadjopoulos, M., Price, G. B. and Wainer, I. W.(1998). Receptor independent effects on DNA replication by steroids. J.Cell. Biochem.70, 323-329.

Difilippantonio, M. J., Zhu, J., Chen, H. T., Meffre, E., Nussenzweig, M.C., Max, E. E., Ried, T. and Nussenzweig, A.(2000). DNA repair proteinKu80 suppresses chromosomal aberrations and malignant transformation.Nature404, 510-514.

Dimitrova, D. S. and Gilbert, D. M. (2000). Temporally coordinatedassembly and disassembly of replication factories in the absence of DNAsynthesis. Nat. Cell Biol. 2, 686-694.

Doherty, A. J. and Jackson, S. P.(2001). DNA repair: How Ku makes endsmeet. Curr. Biol. 11, 920-924.

Errami, A., Smider, V., Rathmell, W. K., He, D. M., Hendrickson, E. A.,Zdzienicka, M. Z. and Chu, G. (1996). Ku86 defines the genetic defectand restores X-ray resistance and V(D)J recombination to complementationgroup 5 hamster cell mutants. Mol. Cell. Biol.16, 1519-1526.

Errami, A., Finnie, N. J., Morolli, B., Jackson, S. P., Lohman, P. H. andZdzienicka, M. Z. (1998). Molecular and biochemical characterization ofnew X-ray-sensitive hamster cell mutants defective in Ku80. Nucleic AcidsRes.26, 4332-4338.

Featherstone, C. and Jackson, S. P.(1999). Ku, a DNA repair protein withmultiple cellular functions? Mutat. Res.434, 3-15.

Fewell, J. W. and Kuff, E. L. (1996). Intracellular redistribution of Kuimmunoreactivity in response to cell-cell contact and growth modulatingcomponents in the medium. J. Cell Sci.109, 1937-1946.

Finnie, N., Gottlieb, T., Hartley, K. and Jackson, S. P. (1993). Transcriptionfactor phosphorylation by the DNA-dependent protein kinase. Biochem. Soc.Trans. 21, 930-935.

Frappier, L. and Zannis-Hadjopoulos, M. (1987). Autonomous replicationof plasmids bearing monkey DNA origin-enriched sequences. Proc. Natl.Acad. Sci. USA84, 6668-6672.

Galande, S. and Kohwi-Shigematsu, T. (1999). Poly(ADP-ribose)polymerase and Ku autoantigen form a complex and synergistically bind tomatrix attachment sequences. J. Biol. Chem.274, 20521-20528.

Genersch, E., Eckerskorn, C., Lottspeich, F., Herzog, C., Kuhn, K. andPoschl, E.(1995). Purification of the sequence-specific transcription factorCTCBF, involved in the control of human collagen IV genes: subunits withhomology to Ku antigen. EMBO J.14, 791-800.

Getts, R. C. and Stamato, T. D.(1994). Absence of a Ku-like DNA endbinding activity in the xrs double-strand DNA repair-deficient mutant. J.Biol. Chem.269, 15981-15984.

Giffin, W., Torrance, H., Rodda, D. J., Prefontaine, G. G., Pope, L. andHache, R. J.(1996). Sequence-specific DNA binding by Ku autoantigenand its effects on transcription. Nature380, 265-268.

Gilbert, D. M., Neilson, A., Miyazawa, H., DePamphilis, M. L. andBurhans, W. C. (1995). Mimosine arrests DNA synthesis at replicationforks by inhibiting deoxyribonucleotide metabolism. J. Biol. Chem. 270,9597-9606.

Gu, Y., Jin, S., Gao, Y., Weaver, D. T. and Alt, F. W. (1997). Ku70-deficientembryonic stem cells have increased ionizing radiosensitivity, defectiveDNA end-binding activity, and inability to support V(D)J recombination.Proc. Natl. Acad. Sci. USA 94, 8076-8081.

Guo, Z. S., Gutierrez, C., Heine, U., Sogo, J. M. and Depamphilis, M. L.(1989). Origin auxiliary sequences can facilitate initiation of simian virus40 DNA replication in vitro as they do in vivo. Mol. Cell. Biol. 9, 3593-3602.

Han, Z., Johnston, C., Reeves, W. H., Carter, T., Wyche, J. H. andHendrickson, E. A. (1996). Characterization of a Ku86 variant protein thatresults in altered DNA binding and diminished DNA-dependent proteinkinase activity. J. Biol. Chem.271, 14098-14104.

Jeggo, P. A. and Kemp, L. M.(1983). X-ray-sensitive mutants of Chinese



hamster ovary cell line. Isolation and cross-sensitivity to other DNA-damaging agents. Mutat. Res.112, 313-327.

Jeng, Y. W., Chao, H. C., Chiu, C. F. and Chou, W. G.(1999). Senescenthuman fibroblasts have elevated Ku86 proteolytic cleavage activity. Mutat.Res.435, 225-232.

Jilani, A., Slack, C., Matheos, D., Zannis-Hadjopoulos, M. and Lasko, D.D. (1999). Purification of a polynucleotide kinase from calf thymus,comparison of its 3′-phosphatase domain with T4 polynucleotide kinase, andinvestigation of its effect on DNA replication in vitro. J. Cell. Biochem.73,188-203.

Kobayashi, T., Rein, T. and DePamphilis, M. L.(1998). Identification ofprimary initiation sites for DNA replication in the hamster dihydrofolatereductase gene initiation zone. Mol. Cell. Biol.18, 3266-3277.

Kemp, L. M., Sedgwick, S. G. and Jeggo, P. A.(1984). X-ray sensitivemutants of Chinese hamster ovary cells defective in double-strand breakrejoining. Mutat. Res. 132, 189-196.

Koike, M., Ikuta, T., Miyasaka, T. and Shiomi, T. (1999). Ku80 cantranslocate to the nucleus independent of the translocation of Ku70 using itsown nuclear localization signal. Oncogene18, 7495-7505.

Koike, M., Shiomi, T. and Koike, A. (2001). Dimerization and nuclearlocalization of Ku proteins. J. Biol. Chem. 276, 11167-11173.

Krude, T. (2000). Initiation of human DNA replication in vitro using nucleifrom cells arrested at an initiation-competent state. J. Biol. Chem. 275,13699-13707.

Krude, T., Jackman, M., Pines, J. and Laskey, R. A.(1997). Cyclin/Cdk-dependent initiation of DNA replication in a human cell-free system. Cell88, 109-119.

Kuhn, A., Gottlieb, T. M., Jackson, S. P. and Grummt, I.(1995). DNA-dependent protein kinase: a potent inhibitor of transcription by RNApolymerase I. Genes Dev.9, 193-203.

Ladenburger, E.-M., Keller, C. and Knippers, R. (2002). Identification ofa binding region for human origin recognition complex proteins 1 and 2that coincides with an origin of DNA replication. Mol. Cell. Biol.22, 1036-1048.

Landry, S. and Zannis-Hadjopoulos, M.(1991). Classes of autonomouslyreplicating sequences are found among early-replicating monkey DNA.Biochim. Biophys. Acta 1088, 234-244.

Lebel, M., Spillare, E. A., Harris, C. C. and Leder, P.(1999). The Wernersyndrome gene product co-purifies with the DNA replication complex andinteracts with PCNA and topoisomerase I. J. Biol. Chem.274, 37795-37799.

Li, B. and Comai, L. (2001). Requirements for the nucleolytic processing ofDNA ends by the Werner syndrome protein Ku70/80 complex. J. Biol.Chem. 276, 9896-9902.

Li, G. C., Ouyang, H., Li, X., Nagasawa, H., Little, J. B., Chen, D. J., Ling,C. C., Fuks, Z. and Cordon-Cardo, C.(1998). Ku70: a candidate tumorsuppressor gene for murine T cell lymphoma. Mol. Cell 2, 1-8.

Li, J. J. and Kelly, T. J. (1984). Simian virus 40 DNA replication in vitro.Proc. Natl. Acad. Sci. USA81, 6973-6977.

Lim, D. S., Vogel, H., Willerford, D. M., Sands, A. T., Platt, K. A. andHasty, P. (2000). Analysis of ku80-mutant mice and cells with deficientlevels of p53. Mol. Cell. Biol.20, 3772-3780.

Matheos, D. D., Ruiz, M. T., Price, G. B. and Zannis-Hadjopoulos, M.(1998). Oct-1 enhances the in vitro replication of a mammalianautonomously replicating DNA sequence. J. Cell. Biochem.68, 309-327.

Matheos, D., Ruiz, M. T., Price, G. B. and Zannis-Hadjopoulos, M. (2002).OBA/Ku86, an origin-specific binding protein that associates withreplication proteins, is required for mammalian DNA replication. Biochim.Biophys. Acta 1578, 59-72.

Mimori, T. and Hardin, J. A. (1986). Mechanism of interaction between Kuprotein and DNA. J. Biol. Chem.261, 10375-10379.

Mimori, T., Hardin, J. A. and Steitz, J. A. (1986). Characterization of theDNA-binding protein antigen Ku recognized by autoantibodies frompatients with rheumatic disorders. J. Biol. Chem.261, 2274-2278.

Munoz, P., Zdzienicka, M. Z., Blanchard, J. M. and Piette, J.(1998).Hypersensitivity of Ku-deficient cells toward the DNA topoisomerase IIinhibitor ICRF-193 suggests a novel role for Ku antigen during the G2 andM phases of the cell cycle. Mol. Cell. Biol.18, 5797-5808.

Munoz, P., Baus, F. and Piette, J. (2001). Ku antigen is required to relieveG2 arrest caused by inhibition of DNA topoisomerase II activity by thebisdioxopiperazine ICRF-193. Oncogene20, 1990-1999.

Nielsen, T., Bell, D., Lamoureux, C., Zannis-Hadjopoulos, M. and Price,G. (1994). A reproducible method for identification of human genomic DNAautonomously replicating sequences. Mol. Gen. Genet.242, 280-288.

Novac, O., Matheos, D., Araujo, F. D., Price, G. B. and Zannis-

Hadjopoulos, M. (2001). In vivo association of Ku with mammalian originsof DNA replication. Mol. Biol. Cell12, 3386-3401.

Novac, O., Alvarez, D., Pearson, C. E., Price, G. B. and Zannis-Hadjopoulos, M. (2002). The human cruciform binding protein, CBP, isinvolved in DNA replication and associates in vivo with mammalianreplication origins. J. Biol. Chem. 277, 11174-11183.

Nussenzweig, A., Chen, C., da Costa Soares, V., Sanchez, M., Sokol, K.,Nussenzweig, M. C. and Li, G. C.(1996). Requirement for Ku80 in growthand immunoglobulin V(D)J recombination. Nature 382, 551-555.

Paillard, S. and Strauss, F.(1991). Analysis of the mechanism of interactionof simian Ku protein with DNA. Nucleic Acids Res.19, 5619-5624.

Pasero, P., Braguglia, D. and Gasser, S. M.(1997). ORC-dependent andorigin-specific initiation of DNA replication at defined foci in isolated yeastnuclei. Genes Dev. 11, 1504-1518.

Pearson, C. E., Frappier, L. and Zannis-Hadjopoulos, M.(1991). Plasmidsbearing mammalian DNA-replication origin-enriched (ors) fragmentsinitiate semiconservative replication in a cell-free system. Biochim. Biophys.Acta.1090, 156-166.

Pearson, C. E., Shihab-el-Deen, A., Price, G. B. and Zannis-Hadjopoulos,M. (1994). Electron microscopic analysis of in vitro replication products ofors 8, a mammalian origin enriched sequence. Somat. Cell. Mol. Genet.20,147-152.

Pelletier, R., Mah, D., Landry, S., Matheos, D., Price, G. B. and Zannis-Hadjopoulos, M. (1997). Deletion analysis of ors12, a centromeric, earlyactivated, mammalian origin of DNA replication. J. Cell. Biochem.66, 87-97.

Pelletier, R., Price, G. B. and Zannis-Hadjopoulos, M.(1999). Functionalgenomic mapping of an early-activated centromeric mammalian origin ofDNA replication. J. Cell. Biochem.74, 562-575.

Pergola, F., Zdzienicka, M. Z. and Lieber, M. R. (1993). V(D)Jrecombination in mammalian cell mutants defective in DNA double-strandbreak repair. Mol. Cell. Biol.13, 3464-3471.

Polotnianka, R. M., Li, J. and Lustig, A. J.(1998). The yeast Ku heterodimeris essential for protection of the telomere against nucleolytic andrecombinational activities. Curr. Biol. 8, 831-834.

Porter, S. E., Greenwell, P. W., Ritchie, K. B. and Petes, T. D.(1996). TheDNA-binding protein Hdf1p (a putative Ku homologue) is required formaintaining normal telomere length in Saccharomyces cerevisiae. NucleicAcids Res.24, 582-585.

Pucci, S., Mazzarelli, P., Rabitti, C., Giai, M., Gallucci, M., Flammia, G.,Alcini, A., Altomare, V. and Fazio, V. M. (2001). Tumor specificmodulation of Ku70/80 DNA binding activity in breast and bladder humanbiopsies. Oncogene 20, 739-747.

Quinn, J. P., Simpson, J. and Farna, A. R.(1993). The Ku compex ismodulated in response to viral infection and other cellular changes. Biochim.Biophys. Acta.1131, 181-187.

Rathmell, W. K. and Chu, G. (1994). Involvement of the Ku autoantigen inthe cellular response to DNA double-strand breaks. Proc. Natl. Acad. Sci.USA91, 7623-7627.

Ruiz, M. T., Pearson, C. E., Nielsen, T., Price, G. B. and Zannis-Hadjopoulos, M. (1995). Cofractionation of HeLa cell replication proteinswith ors-binding activity. J. Cell. Biochem.58, 221-236.

Ruiz, M. T., Matheos, D., Price, G. B. and Zannis-Hadjopoulos, M.(1999).OBA/Ku86: DNA binding specificity and involvement in mammalian DNAreplication. Mol. Biol. Cell10, 567-580.

Shakibai, N., Kumar, V. and Eisenberg, S.(1996). The Ku-like protein fromSaccharomyces cerevisiaeis required in vitro for the assembly of a stablemultiprotein complex at a eukaryotic origin of replication. Proc. Natl. Acad.Sci. USA 93, 11569-11574.

Singleton, B. K., Priestley, A., Steingrimsdottir, H., Gell, D., Blunt, T.,Jackson, S. P., Lehmann, A. R. and Jeggo, P. A.(1997). Molecular andbiochemical characterization of xrs mutants defective in Ku80. Mol. Cell.Biol. 17, 1264-1273.

Smider, V., Rathmell, W. K., Lieber, M. R. and Chu, G. (1994). Restorationof X-ray resistance and V(D)J recombination in mutant cells by Ku cDNA.Science 266, 288-291.

Stillman, B. W. and Gluzman, Y. (1985). Replication and supercoiling ofsimian virus 40 DNA in cell extracts from human cells. Mol. Cell. Biol.5,2051-2060.

Stoeber, K., Mills, A. D., Kubota, Y., Krude, T., Romanowski, P.,Marheineke, K., Laskey, R. A. and Williams, G. H. (1998). Cdc6 proteincauses premature entry into S phase in a mammalian cell-free system.EMBO J. 17, 7219-7229.

Taccioli, G. E., Gottlieb, T. M., Blunt, T., Priestley, A., Demengeot, J.,

124

Mizuta, R., Lehmann, A. R., Alt, F. W., Jackson, S. P. and Jeggo, P. A.(1994). Ku80: product of the XRCC5 gene and its role in DNA repair andV(D)J recombination. Science265, 1442-1445.

Taghva, A., Ma, Y. and Lieber, M. R. (2002). Analysis of the kinetic andequilibrium binding of Ku protein to DNA. J. Theor. Biol.214, 85-97.

Todd, A., Landry, S., Pearson, C. E., Khoury, V. and Zannis-Hadjopoulos,M. (1995). Deletion analysis of minimal sequence requirements forautonomous replication of ors8, a monkey early-replicating DNA sequence.J. Cell. Biochem.57, 280-289.

Toth, E. C., Marusic, L., Ochem, A., Patthy, A., Pongor, S., Giacca, M. andFalaschi, A.(1993). Interactions of USF and Ku antigen with a human DNAregion containing a replication origin. Nucleic Acids Res.21, 3257-3263.

Tsuchiya, T., Saegusa, Y., Taira, T., Mimori, T., Iguchi-Ariga, S. M. andAriga, H. (1998). Ku antigen binds to Alu family DNA. J. Biochem. (Tokyo)123, 120-127.

Tuteja, R. and Tuteja, N. (2000). Ku autoantigen: a multifunctional DNA-binding protein. Crit. Rev. Biochem. Mol. Biol.35, 1-33.

Um, J. H., Kang, C. D., Lee, B. G., Kim, D. W., Chung, B. S. and Kim, S.H. (2001). Increased and correlated nuclear factor kappa B and Ku antigenactivities are associated with development of multidrug resistance.Oncogene20, 6048-6056.

Vishwanatha, J. K. and Baril, E. F.(1990). Single-stranded-DNA-dependentATPase from HeLa cells that stimulates DNA polymerase alpha-primase

activity: purification and characterization of the ATPase. Biochemistry29,8753-8759.

Vogel, H., Lim, D. S., Karsenty, G., Finegold, M. and Hasty, P.(1999).Deletion of Ku86 causes early onset of senescence in mice. Proc. Natl. Acad.Sci. USA 96, 10770-10775.

Wobbe, C. R., Dean, F., Weissbach, L. and Hurwitz, J.(1985). In vitroreplication of duplex circular DNA containing the simian virus 40 DNAorigin site. Proc. Natl. Acad. Sci. USA 82, 5710-5714.

Woo, R. A., McLure, K. G., Lees-Miller, S. P., Rancourt, D. E. and Lee,P. W. (1998). DNA-dependent protein kinase acts upstream of p53 inresponse to DNA damage. Nature394, 700-704.

Wu, C., Zannis-Hadjopoulos, M. and Price, G. B.(1993). In vivo activityfor initiation of DNA replication resides in a transcribed region of the humangenome. Biochim. Biophys. Acta.1174, 258-266.

Wu, J. R., Yu, G. and Gilbert, D. M. (1997). Origin-specific initiation ofmammalian nuclear DNA replication in a Xenopus cell-free system.Methods13, 313-324.

Yasui, L. S., Lotshaw, D. P., Tumuluri, S. and Schwartz, J. (1999).Detection of Ku70 and Ku80 in CHO K1 and xrs-5 cells. Scanning21, 110-111.

Zannis-Hadjopoulos, M., Nielsen, T. O., Todd, A. and Price, G. B. (1994).Autonomous replication in vivo and in vitro of clones spanning the region ofthe DHFR origin of bidirectional replication (ori beta). Gene151, 273-277.


Documents

Analysis of the DNA replication competence of the xrs-5 ...1994; Smider et al., 1994; Taccioli et al., 1994). Ku (reviewed in Tuteja and Tuteja, 2000) is a heterodimeric DNA-binding