The Werner's Syndrome Protein Collaborates with REV1 to Promote

DNA Repair 9 (2010) 1064–1072

Contents lists available at ScienceDirect

DNA Repair

journal homepage: www.elsevier.com/locate/dnarepair

The Werner’s Syndrome protein collaborates with REV1 to promote replication fork progression on damaged DNA

Lara G. Phillips, Julian E. Sale ∗

Medical Research Council Laboratory of Molecular Biology, Hills Road, Cambridge, CB2 0QH UK article info abstract

Article history: DNA damage tolerance pathways facilitate the bypass of DNA lesions encountered during replication. Received 19 May 2010 These pathways can be mechanistically divided into recombinational damage avoidance and translesion Received in revised form 23 June 2010 synthesis, in which the lesion is directly bypassed by specialised DNA polymerases. We have recently Accepted 9 July 2010 shown distinct genetic dependencies for lesion bypass at and behind the replication fork in the avian Available online 5 August 2010 cell line DT40, bypass at the fork requiring REV1 and bypass at post-replicative gaps requiring PCNA ubiquitination by RAD18. The WRN helicase/exonuclease, which is mutated in the progeroid and cancer Keywords: predisposition disorder Werner’s Syndrome, has previously been implicated in a RAD18-dependent DNA DNA damage tolerance Translesion synthesis damage tolerance pathway. However, WRN has also been shown to be required to maintain normal Werner’s Syndrome replication fork progression on a damaged DNA template, a defect reminiscent of REV1-deﬁcient cells. PCNA ubiquitination Here we use the avian cell line DT40 to demonstrate that WRN assists REV1-dependent translesion WRN synthesis at the replication fork and that PCNA ubiquitination-dependent post-replicative lesion bypass REV1 provides an important backup mechanism for damage tolerance in the absence of WRN protein. © 2010 Elsevier B.V. All rights reserved.

1. Introduction in all eukaryotes studied to date [4–8] both through controlling the recruitment of translesion polymerases and by promoting recombi- During DNA replication, attempts to excise and repair DNA dam- national bypass. However, while more is now understood about the age are likely to result in replication fork collapse. Cells therefore mechanisms that control lesion bypass pathways, little is known usually opt to bypass damage encountered by replication forks, about the accessory factors that facilitate these processes. A num- instead deferring repair until replication has been completed. DNA ber of recent lines of evidence, discussed below, suggest that WRN damage bypass can take two general forms [for recent reviews is potentially one such factor. see [1,2]]. The most direct is translesion synthesis, in which the Werner’s Syndrome (WS, OMIM 277700) is an autosomal reces- stalled replicative polymerases are replaced by specialised transle- sive disorder caused by mutations in the WRN gene [9] and sion polymerases whose active sites are able to accommodate is characterised by a variety of disorders reminiscent of pre- adducted or distorted bases. Alternatively, cells can employ one mature ageing; cataracts, type II diabetes mellitus, osteoporosis, of the recombinational modes of bypass that require an undam- atherosclerosis as well as a high incidence of unusual cancers. aged template, usually the newly synthesised daughter strand on WRN encodes a member of the RecQ family of helicases [10] but the sister chromatid. In principle, lesion bypass can take place also contains a ‘helicase, RNaseD, C-terminal conserved’ (HRDC) either at the replication fork or at post-replicative gaps created domain, which has DNA-binding activity [11,12], and 3–5 exonu- when replication restarts downstream of an arrested fork. We have clease activity due to a motif located near the N terminal of the recently demonstrated that, in DT40, these temporally distinct protein [13]. lesion bypass pathways are genetically deﬁned by requirements The defects in WRN-deﬁcient cells, which include slow growth, for the C terminus of the Y-family DNA polymerase REV1 and the spontaneous genetic instability and sensitivity to a range of DNA monoubiquitination of the DNA sliding clamp PCNA respectively damaging agents, arise from the involvement of WRN in a very [3]. Monoubiquitination of PCNA by the E3 ubiquitin ligase RAD18 wide range of DNA transactions including base excision repair, non- [4–6] plays a central role in the control of DNA damage tolerance homologous end joining and transcription [reviewed in [14]]. Much evidence, however, points to its involvement in processes at or around the replication fork. Similarly to RecQ in E. coli [15] and ∗ Rqh1 in S. pombe [16], there is evidence that WRN helps to avoid Corresponding author at: Medical Research Council Laboratory of Molecular recombination [17], possibly by promoting damage tolerance, the Biology, Division of Protein & Nucleic Acid Chemistry, Hills Road, Cambridge, UK. Tel.: +44 1223 252941; fax: +44 1223 412178. set of mechanisms that promote DNA damage bypass during repli- E-mail address: [email protected] (J.E. Sale). cation. WRN has also been reported to play a role in homologous

recombination itself at the level of the resolution of recombination first, followed by the 3 arm and then an antibiotic resistance cas- intermediates [18–21]. sette (either neomycin, puromycin, blasticidin or histidinol) was Evidence implicating WRN in DNA damage tolerance pathways introduced into the BamHI site. The neomycin, puromycin and blas- comes from a number of sources. WRN interacts functionally with ticidin resistance markers are removable using Cre recombinase as the Y-family translesion polymerases ␩, ␬ and ␫, stimulating their they are flanked by modified LoxP sites [29]. extension activity in the absence of PCNA [22]. WRN alleviates pausing of these polymerases at lesions by increasing the vmax of 2.3. Confirmation of WRN disruption polymerization, hence it was proposed that it may promote the progression of replication forks at the expense of increased muta- After targeting of the first and second alleles, disruption was genesis. This stimulatory effect of WRN on TLS polymerases was confirmed by Southern blot probing from the 3 end (Fig. 1A found to be present even in WRN mutants lacking either helicase and B). Probe DNA for Southern blotting was made by ampli- ␩ or exonuclease activities. WRN also co-localises with pol after fication of DT40 genomic DNA using the primers cWRKOPF1 UV-C irradiation, but formation of WRN foci did not depend on the [5-GTTAATACCGTGGCTTGCTGAAGCATTTCTTGAC] and cWRKOPR2 ␩ presence of pol [22] and a direct interaction between the proteins [5-GCTCAGCTGTAGGCTTTTGTTTTAATGAACACAAC], cloned into pCR has not been shown. 2.1-TOPO then excised as a SpeI, XhoI fragment. Genomic DNA from Genetic studies in DT40 suggest that WRN functions in a RAD18- transfected clones was digested with BamHI before Southern blot- dependent DNA damage tolerance pathway [23] with a wrn/rad18 ting. double mutant exhibiting sensitivity to NQO or MMS equal to that Loss of WRN transcript was also confirmed by qPCR, car- of the rad18 single mutant. From our previous work showing that ried out using a 7900HT Fast Real-Time PCR System (Applied RAD18 defines a post-replicative pathway of damage bypass [3], Biosystems). Reactions were performed in MicroAmp fast these data suggest that WRN too would act behind the replication optical 96-well reaction plate (Applied Biosystems) sealed fork. However, using DNA fibre labelling, WRN has been shown to with MicroAmp Optical Adhesive Film (Applied Biosystems) be required for normal replication fork progression on DNA dam- using SYBR GreenER qPCR SuperMix Universal (Invitrogen) aged with MMS [24], suggestive of a model in which WRN assists and ROX reference dye. All reactions were carried out using bypass of lesions at the replication fork. This phenotype is reminis- cDNA template with primers at 400 ␮M. Primers to amplify cent of REV1-deficient cells in similar fibre labelling assays [3,25] WRN cDNA were designed to hybridize to the exons 19–20 and suggests that WRN is required at the replication fork. and exons 22–23 junctions (Exon1920F has sequence 5- To resolve the question of which of the temporally distinct GCGGGATGAAATCCAGTGTGTTGTGG and Exon2322R has sequence bypass pathways, defined by RAD18-dependent PCNA ubiquitina- 5-GAGGTTGCCAAACAAGAGGTTCTGCACCTG) Relative amounts of tion and by REV1 [3], WRN is involved in we generated isogenic transcripts were determined using comparative quantitation rel- mutants in the chicken cell line DT40. Using a newly created wrn ative to wild type, using ␤-actin as the normaliser. Primers used to DT40 that lacks any detectable transcript, we show that WRN oper- amplify ␤-actin were BActinF1 [5-GAGAGAAGATGACACAGATCATG] ates in a pathway dependent on REV1 to maintain replication fork and BActinR1 [5-GACTCCATACCCAAGAAAGATGG]. progression following DNA damage and that RAD18-dependent post-replicative gap filling provides an important backup activity when WRN is lost. 2.4. Measurement of growth kinetics

Cells were diluted to a density of 1 × 104/ml in medium and 2. Materials and methods incubated (37 ◦C). Viable cells were counted using a Vi-CELL cell viability analyser (Beckman Coulter) at 24 h intervals for 96 h. 2.1. DT40 lines, culture and transfection 2.5. Cell cycle analysis Culture and transfection of DT40 was carried out as described previously [26]. The rev1 and pcnaK164R cell lines have also been Cell cycle analysis was performed as previously described [30]. previously described [26,27]. The rad18 mutants used were gen- Incorporated 5-bromo-2deoxyuridine was stained using a 1:200 erated in our laboratory using constructs described by Takeda and dilution of rat anti-BrdU antibody (BD Biosciences) followed by co-workers [28]. a 1:100 dilution of FITC-conjugated goat anti-rat antibody (BD Biosciences). FACS analysis was carried out using a FACScalibur 2.2. WRN locus targeting construct assembly flow cytometer (Beckman Coulter) and FlowJo 8.5.2 software (Tree Star/Stanford University). The WRN targeting construct was assembled in pBluescript SK+ using a custom multiple cloning site (KpnI-NotI-MluI- 2.6. Post-replication repair assay ApaI-EcoRI-BamHI-SacI) made by annealing and ligating the 5-CGCGGCCGCACGCGTGGGCCCGAAT- oligonucleotides cWRNMCSF [ Post-replication repair assays were carried out as previously TCGGATCCGAGCT 5-CGGATCCGAATTCGGGCCCAC- ] and cWRNMCSR [ described [3]. GCGTGCGGCCGCGGTAC]. The 2.89 kb 5 arm of homology was amplified by PCR from DT40 genomic DNA using primer pair cWRKO5F2 [5-CAGAGAGAATAACGCGTCATAAAGATTGTATCTAAATT- 2.7. Colony survival assays TCTAGTCTTC] and cWRKO5R2 [5-CTCAGGAAACAGCCACATACAAAAG- GGCCCGTAATAGTTTCCAGTTCC], which introduce MluI and ApaI sites Colony survival assays were performed as previously described at the 5 and 3 ends respectively. The 5.54 kb 3 arm of homol- [31]. ogy was amplified from genomic DNA using primers cWRKO3F1 [5-GGTGTTTGTTTGCCTGTGGCCGGGCATGG] and cWRKO3R1 [5- 2.8. DNA fibre spreading and labelling CCCATAAGCTCTGAGCTCTTGGGCCATGTTG] which introduce a BamHI site at the 5 end of the arm. An endogenous SacI site is present at DNA fibre spreading and labelling was carried out as described the 3 end. The 5 arm was inserted into the modified pBluescript [3]. 100 doubly labelled fibres were counted for each condition. 1066 L.G. Phillips, J.E. Sale / DNA Repair 9 (2010) 1064–1072

Fig. 1. Generation of a WRN null DT40 line. (A) Schematic of the WRN gene targeting strategy. The binding sites of the primers used to generate the 5 and 3 arms of the targeting construct are shown above the locus map, along with the restriction sites used in their cloning. A selection cassette replaces exon 2 of the WRN transcript and which, following targeted integration, is predicted to lead to an out of frame splice event between exons 1 and 3 resulting in a stop codon 16 amino acids into exon 3. The Southern probe spanning exon 5 is indicated. (B) BamHI digested genomic DNA from wild type (WT), WRN+/−and WRN−/−(wrn) cells probed as indicated above. All lanes are from the same blot, but intervening lanes have been cut out between WT and +/−. (C) qPCR for exons 19–23 conﬁrming loss of WRN transcript downstream of the disruption in four clones. (D) Growth of wild type (WT) and wrn clone P6. The graph represents three independent experiments with the error bars indicating the standard deviation. (E) Comparison of the doubling time in four independent clones of wrn DT40. (F) Cell cycle distribution of cells in asynchronous cultures of wild type and wrn lines. The mean percentage of 3 independent determinations is shown and the error bars represent one standard deviation. L.G. Phillips, J.E. Sale / DNA Repair 9 (2010) 1064–1072 1067

Fig. 2. WRN and PCNA ubiquitination act in separate pathways. (A) Epistasis analysis of WRN and RAD18 for survival following treatment with 254 nm UV light (UV) and cisplatin (CDDP). (B) Epistasis analysis of WRN and cells carrying a point mutation of PCNA at K164, which prevents ubiquitination, for survival following treatment with 254 nm UV light (UV) and cisplatin (CDDP). Error bars represent the one standard deviation from the mean of three independent experiments. For clarity only the positive error bar is shown. C. Sister chromatid exchange in wrn and rad18 cells. SCE with (black bars) and without (light grey bars) treatment with 0.2 ng/ml NQO. The histogram indicates the percentage of metaphases with the number of SCE indicated on the X-axis. The mean number of SCE in each case, derived from blind scoring of at least 100 metaphases, is indicated above the histogram. For both untreated and NQO-treated cases, the difference between wild type and rad18 SCE is statistically signiﬁcant (unpaired t-test, p < 0.0001). There is no signiﬁcant difference between wild type and wrn or rad18 and wrn/rad18.

These were derived from c. 8 slides made from two independent substantial transcript (2.5 of 4.5 kb) as assessed by Northern labelling experiments. blot [32]. To achieve this we replaced exon 2 in both alleles of the DT40 WRN gene with a selectable marker by homol- 3. Results and discussion ogous recombination (Fig. 1A and B). Splicing of exon 1 to exon 3 results in an out of frame transcript and introduces a 3.1. Generation of a completely null wrn DT40 line stop codon 16 amino acids into exon 3. qPCR corresponding to exons 19–23 on cDNA from the targeted clones conﬁrmed For this study we wished to produce a completely null wrn the success of this strategy with no transcript detectable DT40 cell line, since an earlier wrn DT40 mutant retains a (Fig. 1C). 1068 L.G. Phillips, J.E. Sale / DNA Repair 9 (2010) 1064–1072

Fig. 3. WRN is not required for post-replicative gap ﬁlling. Upper panels: Wild type, rad18, wrn and rad18/wrn cells were irradiated with 4 J m−2 254 nm light, pulse labelled with [3H]-thymidine for 20 min and either lysed immediately (grey lines) or chased for 90’ in medium containing 10 ␮M thymidine before lysis (black lines). Lower panels: sham-irradiated cells.

3.2. The phenotype of wrn DT40 resembles that of WS patient cells (Fig. 2A and B), showing that WRN acts largely independently of PCNA ubiquitination and, indeed, suggests that PCNA ubiquitina- Our wrn DT40 clones grew slowly, as has been reported for tion plays an important role in maintaining the viability of wrn cells human WS fibroblast lines [33]. We initially characterised four following genotoxic stress. independent wrn clones, all of which showed a comparable reduc- A striking feature of rad18 mutant lines is elevated spontaneous tion in doubling time. For example clone P6 had doubling time at sister chromatid exchange [28,43]. This arises from a combina- 37 ◦C of 20.2 ± 4.0 h compared to 11.0 ± 0.8 h for the parental wild tion of channelling DNA lesions away from RAD18-dependent DNA type, measured concurrently (Fig. 1D and E). Similar to human WS damage tolerance pathways to classical homologous recombina- fibroblasts [reviewed in [34]], this was largely explained by a pro- tion and from a role played by RAD18 in recombination itself [44] longed cell cycle rather than an increase in spontaneous apoptosis that is independent of PCNA ubiquitination [6]. In contrast, it has (Fig. 1F). The increase in the proportion of G1/G0 cells in an asyn- been reported that inactivation of WRN in human cells does not chronous culture of the wrn mutant to 22.7%, from 12.7% in wild result in elevated SCE [45]. As previously observed, rad18 DT40 cells type, corresponds to wrn cells spending an average of 4.5 h in G1 as exhibit elevated levels of SCE in the absence of any exogenously opposed to 1.4 h for wild type (Fig. 1F). Further, S-phase was also applied DNA damage, with a mean of 7 per metaphase (Fig. 2C). The prolonged from 6.4 h in wild type to 11.3 h in wrn cells, again in level of SCE rose further following treatment with NQO to 10 per agreement with studies of human WS cells [33,35,36]. Similar to metaphase (Fig. 2C). However, in agreement with studies in human human WS fibroblasts [37–42], wrn DT40 were modestly sensitive WS cells [45], wrn DT40 do not exhibit elevated spontaneous SCE to a number of agents expected to create replication blocks includ- or an exaggerated SCE response following exposure to NQO and 2 ing ultraviolet light [D10 8.3 vs. 4.8 J/m , for wild type (Fig. 2A)], the the wrn/rad18 double mutant behaves the same as the rad18 single adduct/cross-linking agent cisplatin (CDDP) [D10 34.1 vs. 45.7 ␮M mutant within the sensitivity limit of this assay (Fig. 2B). (Fig. 2A)], methylmethane sulphonate (MMS) [D10 99.8 vs. 142.1 parts per million (data not shown)] and nitroquinoline-1-oxide (NQO) [D10 23.7 vs. 41.5 nM (data not shown)]. 3.4. WRN is not required for post-replicative gap filling

In DT40, RAD18-mediated PCNA ubiquitination controls DNA 3.3. Complete WRN disruption is not epistatic with defective damage bypass at post-replicative gaps that are formed in conse- PCNA ubiquitination quence of replication fork arrest with downstream resumption of replication [3]. Thus, when PCNA ubiquitination is defective the rate Previous work has suggested that WRN functions in a RAD18- at which single stranded gaps are ﬁlled is reduced, i.e. PCNA ubiq- dependent damage avoidance pathway [23]. We re-examined this uitination mutants exhibit defective post-replication repair. A key relationship by disrupting the WRN loci using our construct in rad18 prediction of the notion that WRN functions in a RAD18-dependent [28] and pcnaK164R [27] DT40 lines. Epistasis analysis for sensi- damage avoidance pathway [23] is that WRN should also operate tivity to 254 nm UV light and cisplatin (CDDP) showed that both behind the fork. This can be monitored by following the incorpo- wrn/rad18 and wrn/pcnaK164R double mutants were markedly ration of a tract of tritiated label into high molecular weight single more sensitive than either of their respective single mutants stranded DNA by velocity sedimentation on an alkaline sucrose gra- L.G. Phillips, J.E. Sale / DNA Repair 9 (2010) 1064–1072 1069

Fig. 4. WRN is required to maintain replication fork progression on a damaged DNA template. (A) Schematic of DNA labelling experiments. Cells are pulse labelled first with chlorodeoxyuridine (CldU) for 20 min and then with iododeoxyuridine (IdU) for 20 min. DNA damage can be added along with the IdU to assess the effect on replication progression compared with the first labelling period without damage. Each period of labelling corresponds to synthesis of about 20 kb of DNA. Note the concentrations of drug used are supralethal in order to achieve a high probability of damage forming within c. 20 kb ahead of the fork [3]. (B) Fork progression on undamaged DNA in wrn cells is comparable to wild type (WT). Average fork rate during the first 20 min calculated using a previous calibration for this method [51] from 100 doubly labelled fibres for each case. The error bars indicate one standard deviation. There is no significant difference between WT and wrn (unpaired t-test). (C) Histograms of DNA fibre CldU:IdU ratios showing replication stalling in response to increasing doses of 4-nitroquinoline-1-oxide (NQO). Wild type: grey bars; wrn: red bars. (D–F) Cumulative percentage plots of DNA fibre CldU:IdU ratios. WT: dashed lines; wrn: solid lines. (D) 4-nitroquinoline-1-oxide (NQO). (E) Methylmethane sulphonate (MMS). (F) Cisplatin (CDDP). Statistical significance was assessed using the two-sample Kolmogorov–Smirnov (K–S) test and the probability that there is no difference between WT and wrn at each dose is indicated. dient [46]. We observed, as previously reported [3,47] that rad18 used in the earlier study [32]. Although the targeting approached cells exhibit delayed gap filling (Fig. 3). However, such a defect is not used in the work of Dong et al. will inactivate the helicase domain, seen in wrn cells and the defect in the wrn/rad18 line is compara- most cases of WS are a result of complete functional inactivation ble to that seen in the rad18 single mutant (Fig. 3). This suggests of the gene. Loss of just the helicase activity of WRN results has that WRN is not essential for timely post-replicative gap filling been previously shown to result in a distinct and milder pheno- and together these data do not support WRN acting in the PCNA type compared with complete inactivation of the protein [21]. Thus, ubiquitination-dependent post-replicative DNA damage tolerance while we cannot exclude that the helicase activity of WRN partic- pathway. ipates in PCNA ubiquitination-dependent damage avoidance, our Our conclusions are, on the face of it, in marked contrast to those results clearly demonstrate that the complete WRN protein ful- reached by Dong et al. [23]. We believe that the discrepancy likely fils an important function in the absence of PCNA ubiquitination. arises from the incomplete inactivation of WRN in the DT40 mutant To further address this issue, we have tried extensively to per- 1070 L.G. Phillips, J.E. Sale / DNA Repair 9 (2010) 1064–1072

Fig. 5. WRN acts in a REV1-dependent pathway to promote replication fork progression on damaged DNA. (A) Fork progression on undamaged DNA in wild type (WT), wrn, rev1 and wrn/rev1 cells. There is no significant difference between any of the lines (unpaired t-test). (B) Cumulative percentage plots of DNA fibre CldU:IdU ratios with 10 ␮g/ml NQO added during the second (IdU) labelling period for wild type (black line), wrn (red line), rev1 (blue line) and wrn/rev1 green line. Statistical significance assessed by the two-sample K-S test is indicated. (C and D) Epistasis analysis of WRN and REV1 (C) and REV3 (D) for survival following treatment with 254 nm UV light. E.

Summary of the replicating plasmid assay for lesion bypass. For full details see Szüts et al. [50]. Briefly, the plasmid pQTS, which can support replication in DT40, contains staggered T–T (6-4) photoproducts placed opposite a GpC mismatch, since this is an uncommon insertion opposite this lesion. This arrangement allows determination of TLS and recombinational bypass by template switching. Since GpC opposite the lesion would also arise as a result of nucleotide excision repair, the assay is carried out in an xpa background. (F) The proportion of TLS vs. error-free bypass in pQTs sequences recovered from xpa (n = 130) and xpa/wrn cells (n = 70), shown as percentage of the total. (G) The pattern of nucleotide incorporation opposite the T–T (6-4) photoproduct. The percentage of each nucleotide incorporated at each position is indicated by the size of the letter of the nucleotide in the column; del: deletion. L.G. Phillips, J.E. Sale / DNA Repair 9 (2010) 1064–1072 1071 form complementation of our wrn mutant. However, we have been is in marked contrast to the relationship between WRN and PCNA unable to obtain clones ectopically expressing the WRN protein ubiquitination (Fig. 2B). Together, these data suggest that WRN despite using a number of different promoter systems. It will there- and REV1 can operate in a common pathway for the alleviation fore be necessary to further these studies by creating endogenous of replication arrest. mutations in the WRN locus. We have previously shown that REV3 is essential for translesion synthesis across (6-4) T–T photoproducts, a major UV-induced DNA 3.5. WRN is required to maintain replication fork progression on lesion, in vivo in a replicating plasmid assay that can monitor the damaged DNA relative use of TLS and recombinational bypass (Fig. 5E) [50].We also showed that loss of REV1 not only diminished the frequency of Recent data from WRN-depleted human cells have implicated use of TLS but also resulted in reduced TLS frame fidelity. To exam- WRN in maintaining replication fork progression on a damaged ine the impact of the wrn mutation in this assay, we inactivated DNA template [24]. Fork progression is monitored by labelling DNA nucleotide excision repair, as previously described, by disrupting in vivo with halogenated nucleotides and then spreading the DNA the XPA locus [50]. Since the lesion is placed opposite a G–C mis- fibres on a glass slide. The tracts of labelled DNA are revealed by match, this allows the distinction of recombinational bypass from denaturation and staining with antibodies that recognise halo- excision repair. Interestingly, wrn/xpa cells exhibited a very similar genated bases. In our version of this assay [3], we sequentially frequency of TLS usage (Fig. 5E) and no alteration in the muta- labelled cells with chlorodeoxyuridine (CldU) and iododeoxyuri- tion spectrum created by the use of TLS compared with xpa cells dine (IdU). During the second labelling period we damaged the DNA (Fig. 5F). This suggests that WRN does not directly influence the such that the advancing fork will encounter DNA damage and may TLS reaction itself or impact mutagenesis, at least in this context. stall (Fig. 4A). We can then monitor, in a large number of fibres, the Rather it appears to be required for efficient use of TLS at a subset of length of tract replicated before and after the application of DNA events at the fork. It will ultimately be interesting to determine the damage. circumstances under which WRN is required. Nonetheless, in the In the absence of DNA damage the replication rate of wrn cells absence of WRN, we suggest that bypass is more likely to take place was comparable to wild type at around 1 kb/min (Fig. 4B). To assess behind the fork explaining the marked reliance of wrn mutant cells the effect of disruption of WRN on fork progression following DNA on PCNA ubiquitination. damage we performed titrations in which increasing doses of DNA damaging agent were added during the second labelling period. Conflicts of interest Damage arresting replication during the second labelling period will result in shortening of the IdU labelled tract and an increase The authors declare that there are no conflicts of interest. in the ratio of CldU:IdU tract lengths. As the position of the stall is random, this ratio increases stochastically resulting in both an Acknowledgements increase in the mean CldU:IdU ratio and in the spread of values (Fig. 4C). A convenient method to represent this data is as a cumu- Work in our lab is supported by the Medical Research Council lative percentage of forks at each ratio (Fig. 4D–F). For all doses and Association for International Cancer Research. We would like of NQO, methyl methane sulphonate and cisplatin wrn cells (solid to thank Shunichi Takeda and Jean-Marie Buerstedde for sharing lines) exhibited a more profound inability to maintain replication DT40 mutants and targeting constructs, Shige Iwai for preparing the fork progression on a damaged template compared with wild type T–T(6-4) photoproduct containing oligonucleotides, Dávid Szüts (dashed lines), in agreement with the recent work on human cells for assembling the replicating plasmid and members of the group [24]. for discussions.

3.6. WRN functions at the replication fork to facilitate References REV1-dependent translesion synthesis [1] P.L. Andersen, F. Xu, W. Xiao, Eukaryotic DNA damage tolerance and translesion This phenotype is strongly reminiscent of cells lacking REV1 synthesis through covalent modifications of PCNA, Cell Res. 18 (2008) 162–173. [3,25]. REV1 plays a crucial role in coordinating translesion synthe- [2] M. Budzowska, R. Kanaar, Mechanisms of dealing with DNA damage-induced replication problems, Cell Biochem. Biophys. 53 (2009) 17–31. sis at the replication fork by coordinating, through its C terminus, [3] C.E. Edmunds, L.J. Simpson, J.E. Sale, PCNA Ubiquitination and REV1 define tem- interactions between PCNA and translesion polymerases [3,48,49]. porally distinct mechanisms for controlling translesion synthesis in the avian We therefore wished to know whether this phenotypic similarity cell line DT40, Mol. Cell 30 (2008) 519–529. [4] C. Hoege, B. Pfander, G.L. Moldovan, G. Pyrowolakis, S. Jentsch, RAD6- between WRN and REV1 was a reflection of their operating in com- dependent DNA repair is linked to modification of PCNA by ubiquitin and SUMO, mon damage bypass pathway at the replication fork. To do this we Nature 419 (2002) 135–141. created a wrn/rev1 double mutant. In the absence of damage these [5] P.L. Kannouche, J. Wing, A.R. Lehmann, Interaction of human DNA polymerase ␩ with monoubiquitinated PCNA: a possible mechanism for the polymerase cells show no defect in DNA replication rate on undamaged DNA switch in response to DNA damage, Mol. Cell 14 (2004) 491–500. (Fig. 5A). To examine the effect of DNA damage we chose a dose [6] L.J. Simpson, A.L. Ross, D. Szuts, C.A. Alviani, V.H. Oestergaard, K.J. Patel, J.E. Sale, of NQO (10 ␮g/ml) at which we would expect to be able to read- RAD18-independent ubiquitination of proliferating-cell nuclear antigen in the avian cell line DT40, EMBO Rep. 7 (2006) 927–932. ily detect an additive effect of the double disruption (see Fig. 4D). [7] J. Frampton, A. Irmisch, C.M. Green, A. Neiss, M. Trickey, H.D. Ulrich, K. Furuya, In a side-by-side comparison, rev1 cells exhibited a more profound F.Z. Watts, A.M. Carr, A.R. Lehmann, Postreplication repair and PCNA Modifica- defect in fork progression following damage than wrn cells (Fig. 4B), tion in Schizosaccharomyces pombe, Mol. Biol. Cell (2006). but importantly the double wrn/rev1 mutant was no worse than [8] C.A. Leach, W.M. Michael, Ubiquitin/SUMO modification of PCNA promotes replication fork progression in Xenopus laevis egg extracts, J. Cell Biol. 171 rev1 on its own. This suggests that WRN operates at a subset of repli- (2005) 947–954. cation impediments at the fork that require REV1 for their bypass [9] C.E. Yu, J. Oshima, Y.H. Fu, E.M. Wijsman, F. Hisama, R. Alisch, S. Matthews, J. or resolution. Nakura, T. Miki, S. Ouais, G.M. Martin, J. Mulligan, G.D. Schellenberg, Positional cloning of the Werner’s syndrome gene, Science 272 (1996) 258–262. We next examined the genetic relationship between WRN and [10] M.D. Gray, J.C. Shen, A.S. Kamath-Loeb, A. Blank, B.L. Sopher, G.M. Martin, J. REV1 and REV3 for sensitivity to UV light. These experiments Oshima, L.A. Loeb, The Werner syndrome protein is a DNA helicase, Nat. Genet. revealed a clear epistatic relationship between WRN and REV1/3- 17 (1997) 100–103. [11] V. Morozov, A.R. Mushegian, E.V. Koonin, P. Bork, A putative nucleic acid- dependent translesion synthesis, the double mutants exhibiting no binding domain in Bloom’s and Werner’s syndrome helicases, Trends Biochem. additional sensitivity to the single TLS mutants (Fig. 5C and D). This Sci. 22 (1997) 417–418. 1072 L.G. Phillips, J.E. Sale / DNA Repair 9 (2010) 1064–1072

[12] C. von Kobbe, V.A. Bohr, A nucleolar targeting sequence in the Werner syndrome [33] F. Takeuchi, F. Hanaoka, M. Goto, M. Yamada, T. Miyamoto, Prolongation of S protein resides within residues 949–1092, J. Cell Sci. 115 (2002) 3901–3907. phase and whole cell cycle in Werner’s syndrome fibroblasts, Exp. Gerontol. 17 [13] S. Huang, B. Li, M.D. Gray, J. Oshima, I.S. Mian, J. Campisi, The premature age- (1982) 473–480. ing syndrome protein, WRN, is a 3’– > 5’ exonuclease, Nat. Genet. 20 (1998) [34] J.M. Sidorova, Roles of the Werner syndrome RecQ helicase in DNA replication, 114–116. DNA Repair (Amst.) 7 (2008) 1776–1786. [14] M.L. Rossi, A.K. Ghosh, V.A. Bohr, Roles of Werner syndrome protein in protec- [35] Y. Fujiwara, T. Higashikawa, M. Tatsumi, A retarded rate of DNA replication and tion of genome integrity, DNA Repair (Amst.) 9 (2010) 331–344. normal level of DNA repair in Werner’s syndrome fibroblasts in culture, J. Cell. [15] K. Hanada, T. Ukita, Y. Kohno, K. Saito, J. Kato, H. Ikeda, RecQ DNA helicase is Physiol. 92 (1977) 365–374. a suppressor of illegitimate recombination in Escherichia coli, Proc. Natl. Acad. [36] M. Poot, H. Hoehn, T.M. Runger, G.M. Martin, Impaired S-phase transit of Sci. U.S.A. 94 (1997) 3860–3865. Werner syndrome cells expressed in lymphoblastoid cell lines, Exp. Cell Res. [16] E. Stewart, C. Chapman, F. Al-Khodairy, A. Carr, T. Enoch, rqh1+, a fission yeast 202 (1992) 267–273. gene related to the Bloom’s and Werner’s syndrome genes, is required for [37] M. Poot, K.A. Gollahon, M.J. Emond, J.R. Silber, P.S. Rabinovitch, Werner reversible S phase arrest, EMBO J. 16 (1997) 2682–2692. syndrome diploid fibroblasts are sensitive to 4-nitroquinoline-N-oxide and 8- [17] A. Franchitto, L. Pirzio, E. Prosperi, O. Sapora, M. Bignami, P. Pichierri, Replica- methoxypsoralen: implications for the disease phenotype, FASEB J. 16 (2002) tion fork stalling in WRN-deficient cells is overcome by prompt activation of a 757–758. MUS81-dependent pathway, J. Cell Biol. 183 (2008) 241–252. [38] P.R. Prince, C.E. Ogburn, M.J. Moser, M.J. Emond, G.M. Martin, R.J. Monnat Jr., Cell [18] P. Pichierri, A. Franchitto, P. Mosesso, F. Palitti, Werner’s syndrome protein is fusion corrects the 4-nitroquinoline 1-oxide sensitivity of Werner syndrome required for correct recovery after replication arrest and DNA damage induced fibroblast cell lines, Hum. Genet. 105 (1999) 132–138. in S-phase of cell cycle, Mol. Biol. Cell 12 (2001) 2412–2421. [39] C.E. Ogburn, J. Oshima, M. Poot, R. Chen, K.E. Hunt, K.A. Gollahon, P.S. [19] P. Prince, M. Emond, R.J. Monnat, Loss of Werner syndrome protein func- Rabinovitch, G.M. Martin, An apoptosis-inducing genotoxin differentiates tion promotes aberrant mitotic recombination, Genes Dev. 15 (2001) 933– heterozygotic carriers for Werner helicase mutations from wild-type and 938. homozygous mutants, Hum. Genet. 101 (1997) 121–125. [20] Y. Saintigny, K. Makienko, C. Swanson, M. Emond, R.J. Monnat, Homologous [40] M. Poot, J.S. Yom, S.H. Whang, J.T. Kato, K.A. Gollahon, P.S. Rabinovitch, Werner recombination resolution defect in werner syndrome, Mol. Cell. Biol. 22 (2002) syndrome cells are sensitive to DNA cross-linking drugs, FASEB J. 15 (2001) 6971–6978. 1224–1226. [21] C. Swanson, Y. Saintigny, M.J. Emond, R.J. Monnat Jr., The Werner syndrome pro- [41] K.K. Dhillon, J. Sidorova, Y. Saintigny, M. Poot, K. Gollahon, P.S. Rabinovitch, R.J. tein has separable recombination and survival functions, DNA Repair (Amst.) 3 Monnat Jr., Functional role of the Werner syndrome RecQ helicase in human (2004) 475–482. fibroblasts, Aging Cell 6 (2007) 53–61. [22] A.S. Kamath-Loeb, L. Lan, S. Nakajima, A. Yasui, L.A. Loeb, Werner syndrome [42] J.A. Harrigan, D.M. Wilson 3rd, R. Prasad, P.L. Opresko, G. Beck, A. May, S.H. protein interacts functionally with translesion DNA polymerases, Proc. Natl. Wilson, V.A. Bohr, The Werner syndrome protein operates in base excision Acad. Sci. U.S.A. 104 (2007) 10394–10399. repair and cooperates with DNA polymerase ␤, Nucleic Acids Res. 34 (2006) [23] Y. Dong, M. Seki, A. Yoshimura, E. Inoue, S. Furukawa, S. Tada, T. Enomoto, WRN 745–754. functions in a RAD18-dependent damage avoidance pathway, Biol. Pharm. Bull. [43] S. Tateishi, H. Niwa, J. Miyazaki, S. Fujimoto, H. Inoue, M. Yamaizumi, Enhanced 30 (2007) 1080–1083. genomic instability and defective postreplication repair in RAD18 knockout [24] J. Sidorova, L. Nianzhen, A. Folch, R.J. Monnat Jr., The RecQ helicase WRN is mouse embryonic stem cells, Mol. Cell. Biol. 23 (2003) 474–481. required for normal replication fork progression after DNA damage or replica- [44] D. Szüts, L.J. Simpson, S. Kabani, M. Yamazoe, J.E. Sale, Role for RAD18 in homol- tion fork arrest, Cell Cycle 7 (2008) 796–807. ogous recombination in DT40 cells, Mol. Cell. Biol. 26 (2006) 8032–8041. [25] J. Jansen, A. Tsaalbi-Shtylik, G. Hendriks, H. Gali, A. Hendel, F. Johansson, K. [45] E. Gebhart, R. Bauer, U. Raub, M. Schinzel, K.W. Ruprecht, J.B. Jonas, Spontaneous Erixon, Z. Livneh, L. Mullenders, L. Haracska, N. de Wind, Separate domains of and induced chromosomal instability in Werner syndrome, Hum. Genet. 80 Rev1 mediate two modes of DNA damage bypass in mammalian cells, Mol. Cell. (1988) 135–139. Biol. 29 (2009) 3113–3123. [46] W.D. Rupp, P. Howard-Flanders, Discontinuities in the DNA synthesized in an [26] L.J. Simpson, J.E. Sale, Rev1 is essential for DNA damage tolerance and non- excision-defective strain of Escherichia coli following ultraviolet irradiation, J. templated immunoglobulin gene mutation in a vertebrate cell line, EMBO J. 22 Mol. Biol. 31 (1968) 291–304. (2003) 1654–1664. [47] S. Tateishi, Y. Sakuraba, S. Masuyama, H. Inoue, M. Yamaizumi, Dysfunction of [27] H. Arakawa, G.L. Moldovan, H. Saribasak, N.N. Saribasak, S. Jentsch, J.M. Buer- human Rad18 results in defective postreplication repair and hypersensitivity stedde, A role for PCNA ubiquitination in immunoglobulin hypermutation, PLoS to multiple mutagens, Proc. Natl. Acad. Sci. U.S.A. 97 (2000) 7927–7932. Biol. 4 (2006) e366. [48] C. Guo, P.L. Fischhaber, M.J. Luk-Paszyc, Y. Masuda, J. Zhou, K. Kamiya, C. Kisker, [28] Y.M. Yamashita, T. Okada, T. Matsusaka, E. Sonoda, G.Y. Zhao, K. Araki, S. E.C. Friedberg, Mouse Rev1 protein interacts with multiple DNA polymerases Tateishi, M. Yamaizumi, S. Takeda, RAD18 and RAD54 cooperatively contribute involved in translesion DNA synthesis, EMBO J. 22 (2003) 6621–6630. to maintenance of genomic stability in vertebrate cells, EMBO J. 21 (2002) [49] A.L. Ross, L.J. Simpson, J.E. Sale, Vertebrate DNA damage tolerance requires the 5558–5566. C-terminus but not BRCT or transferase domains of REV1, Nucleic Acids Res. 33 [29] H. Arakawa, D. Lodygin, J.M. Buerstedde, Mutant loxP vectors for selectable (2005) 1280–1289. marker recycle and conditional knock-outs, BMC Biotechnol. 1 (2001) 7. [50] D. Szüts, A.P. Marcus, M. Himoto, S. Iwai, J.E. Sale, REV1 restrains DNA [30] R. Franklin, J.E. Sale, 2D cell cycle analysis, Subcell. Biochem. 40 (2006) 405–408. polymerase ␨ to ensure frame fidelity during translesion synthesis of UV pho- [31] L.J. Simpson, J.E. Sale, Colony survival assay, Subcell. Biochem. 40 (2006) toproducts in vivo, Nucleic Acids Res. 36 (2008) 6767–6780. 387–391. [51] D.A. Jackson, A. Pombo, Replicon clusters are stable units of chromosome struc- [32] O. Imamura, K. Fujita, C. Itoh, S. Takeda, Y. Furuichi, T. Matsumoto, Werner ture: evidence that nuclear organization contributes to the efficient activation and Bloom helicases are involved in DNA repair in a complementary fashion, and propagation of S phase in human cells, J. Cell Biol. 140 (1998) 1285–1295. Oncogene 21 (2002) 954–963.