Nucleotide excision repair DNA synthesis by excess DNA polymerase : a potential source of genetic instability in cancer cells YVAN CANITROT,*,1 JEAN-SE´BASTIEN HOFFMANN,*,1 PATRICK CALSOU,† HIROSHI HAYAKAWA,‡ BERNARD SALLES,†,2 AND CHRISTOPHE CAZAUX*,2 *Groupe ‘Instabilite´ ge´ne´tique et cancer’, †Groupe ‘Toxico-re´sistance’, Institut de Pharmacologie et de Biologie Structurale, CNRS UMR 5089, 31077 Toulouse cedex 4, France; and ‡Department of Biochemistry, Medical school of Kyushu University, Fukuoka 812-8582, Japan ABSTRACT The nucleotide excision repair pathway contributes to genetic stability by removing a wide range of DNA damage through an error-free reaction. When the lesion is located, the altered strand is incised on both sides of the lesion and a damaged oligonucleotide excised. A repair patch is then synthesized and the repaired strand is ligated. It is assumed that only DNA polymerases ␦ and/or participate to the repair DNA synthesis step. Using UV and cisplatin-modified DNA templates, we measured in vitro that extracts from cells overexpressing the error-prone DNA polymerase  exhibited a fiveto sixfold increase of the ultimate DNA synthesis activity compared with control extracts and demonstrated the specific involvement of Pol  in this step. By using a 28 nt gapped, double-stranded DNA substrate mimicking the product of the incision step, we showed that Pol  is able to catalyze strand displacement downstream of the gap. We discuss these data within the scope of a hypothesis previously presented proposing that excess error-prone Pol  in cancer cells could perturb the well-defined specific functions of DNA polymerases during errorfree DNA transactions.—Canitrot, Y., Hoffmann, J.-S., Calsou, P., Hayakawa, H., Salles, B., Cazaux, C. Nucleotide excision repair DNA synthesis by excess DNA polymerase : a potential source of genetic instability in cancer cells. FASEB J. 14, 1765–1774 (2000)
Key Words: cell variants 䡠 DNA repair 䡠 mutagenesis
Genetic instability is one of the main characteristics that have been postulated to underlie neoplasia. In addition to large chromosomal changes involving thousands of base pairs, one form of genetic instability results in an increased mutation rate at the nucleotide level due to perturbation in nucleotide synthesis or DNA transactions like repair or replication (1, 2). During these transactions, cells seem to have developed a well-defined set of DNA poly0892-6638/00/0014-1765/$02.25 © FASEB
merases, each one being more or less specific to a given process. According to the current consensus, the role of DNA polymerase  (Pol ) is narrowly delimited to base excision repair (BER). Pol  is expressed at a constant low level throughout the cell cycle (3) and is inducible by some genotoxic treatments (4). Features that distinguish Pol  from other cellular polymerases are the lack of associated proofreading activity, its high infidelity in replicating DNA in vitro, and its poor ability to discriminate nucleotides at the level of binding (5). In accordance with its low accuracy, Pol  exhibits the lowest discrimination against mutagenic analogs of dGTP modified by endogenous processes (6). We recently demonstrated that an excess of Pol  in cells resulted in an increased mutagenesis (7). Moreover, we have shown by using a murine cell line mimicking the chronic myelogenous leukemia (CML) that Pol  activity is increased in this model (8). Excess Pol  could thus favor the genetic instability observed in this pathology, particularly during the blast crisis. These data suggest that enhanced level of Pol  may have a role in cancer predisposition and/or tumor progression (9). At the transcriptional level, Pol  was actually overexpressed in many cancer cells (10, 11). High levels of Pol  have also been detected at the protein level. For example, Pol  was found more than 10-fold over-represented in prostate, breast, or colon cancer tissues compared to adjacent normal tissues (12). We quote here as additive example that the intracellular level of Pol  is more than 10-fold increased in several ovarian cancer cell lines compared with normal ovarian tissue (Fig. 1). Based on these data, we hypothesize that an excess of the 1
Y.C. and J.S.H. contributed equally to the work. Correspondence: Groupe ‘Instabilite´ ge´ne´tique et cancer’, Groupe ‘Toxico-re´sistance’, Institut de Pharmacologie et de Biologie Structurale, CNRS UPR 9062, 205 route de Narbonne, 31077 Toulouse cedex 4, France. E-mail:
[email protected];
[email protected] 2
1765
evidence that overexpressed error-prone Pol  can compete with Pol ␦ and/or Pol ε in the DNA synthesis step and that this involvement enhances the probability for misincorporation. The putative effect in cancer cells of the DNA polymerase  imbalance on mutation accumulation is discussed.
MATERIALS AND METHODS Figure 1. Overexpression of Pol  in ovarian tumors. The human ovarian carcinoma cell lines used in this study included three serous cell types: NIH-OVCAR-3 (ATCC), OVCCR1 (63), 2008 (a generous gift from Dr. S. B. Howell, San Diego, Calif.); and two endometrioid cell types: IGROV1 (kindly provided by Dr. J. Be´nard, Villejuif, France) and SKOV-3 (ATCC, Rockville, Md.). The primary culture of human ovarian surface epithelium was obtained as described (64). Whole cell extracts (50 g protein) of normal and tumoral ovarian tissues and cell lines were separated on a 12% SDS-polyacrylamide gel electrophoresis and transferred to PVDF membrane. The membrane was blocked (TBS-T plus milk) and incubated with a polyclonal antibody to Pol . Signal was revealed using the ECL system (Amersham). Equal loading was checked using monoclonal antibody to actin. Quantification analysis was achieved by using the PhosphoImager Storm-system and Imagequant software.
error-prone Pol  can perturb the well-defined specific functions of error-free DNA polymerases in some cancer cells. We have already proposed that Pol -mediated gap fillings during DNA transactions such as repair, replication, or recombination pathways can introduce mutations (9). Such a mutagenic interference could thus participate to the genetic variability of primary tumors whose consequence is known to be the emergence of invasive cell variants resistant to the therapeutic treatment. Among other DNA repair cellular processes, nucleotide excision repair (NER) is a pathway involved when DNA is exposed to physical or chemical agents, leading to bulky or helix-altering lesions (13). This repair pathway operates by recognition of lesion, incision of the damaged strand on either side of the targeted lesion, and excision of an ⬃30-nucleotide fragment carrying the lesion. Finally, the filling of the resulting gap is mediated by DNA polymerases in a proliferating cell nuclear antigen (PCNA) -dependent manner and is followed by ligation. In mammalian cell extracts, the absolute dependence of NER on PCNA (14) suggests the participation of DNA polymerase ␦ (Pol ␦) or DNA polymerase ε (Pol ε) in this last step since these enzymes are known to be stimulated by PCNA (15). Biochemical studies showed that the entire assay could be reconstituted in the presence of either Pol ␦ or Pol ε (16). To date, no evidence has shown an involvement of Pol  in NER of mammalian cells. Here we show that an enhanced level of Pol  can perturb the wellorchestrated steps in NER synthesis. We present 1766
Vol. 14
September 2000
Cell culture Construction of Pol -overexpressing plasmid pUTpol  and establishment of stably transfected Chinese hamster ovary (CHO) overexpressing Pol  has been described in detail elsewhere (17). PUTpol  harbors a fusion between the pol  cDNA and the bleSh gene conferring resistance to zeocin. Control cells were transfected with a similar pUT526⌬ plasmid (Cayla, France) harboring only the bleSh gene. Cells were grown at 37°C in MEM medium (BioMedia) supplemented with glutamine, fetal calf serum (10%), and antibiotics (penicillin/streptomycin) in a humidified atmosphere 95% CO2/5% O2. The pol  cDNA and the bleSh gene being transductionally fusioned, the stability of the transfectants in terms of continuity of Pol  overexpression was ensured by the continuous presence in the culture medium of 250 g/ml zeocin (Cayla). Ovarian carcinoma cell lines were growth in RPMI 1640 medium (BioMedia) supplemented with glutamine, fetal calf serum (10%), and antibiotics (penicillin/ streptomycin). Preparation of radiolabeled 8-oxo-dGTP Radiolabeled 8-oxodGTP was prepared as already described (18). Briefly, 6 mM dGTP was incubated at 37°C for 2 h in the dark in a reaction mixture containing 100 mM sodium potassium (pH 6.8), 30 mM ascorbic acid, and 100 mM H2O2. Fifty microliters of the reaction mixture was loaded onto a high-performance liquid chromatography Spherisorb SAX column equilibrated with 150 mM potassium phosphate (pH 5.5) and chromatographic separation was carried out at 25°C with the same buffer at a flow rate 1 ml/min. 8-Oxo-dGTP was monitored with a UV detector (254 nm) and eluted with a retention ⬃1 min later than for dGTP. Fractions containing 8-oxo-dGTP were combined, applied to a DEAE-MemSep 1000 cartridge equilibrated with 50 mM triethylammonium hydrogen carbonate (pH 7.0), and eluted with a linear gradient (50 –500 mM) of triethyl ammonium hydrogen carbonate (pH 7.0) at a flow rate 1 ml/min. The fractions containing 8-oxo-dGTP were lyophilized, dissolved in 20 mM sodium phosphate (pH 6.8), and stored at ⫺20°C. Purity and quantity of the preparation were examined in electrophoretical measurements and UV spectrum analyses. In vitro DNA repair reactions Whole cell extract preparation CHO extracts preparation was performed according to the previously described protocol (19). Briefly, CHO cells were washed with ice-cold phosphate-buffered saline, harvested by centrifugation, and the cell pellet was suspended in hypotonic buffer (10 mM Tris-HCl, pH 7.5, 10 mM KCl, 10 mM MgCl2, 1 mM DTT) containing protease inhibitors. The cells were disrupted with 20 strokes of the tight-fitting pestle in a dounce homogenizer during incubation at 2°C. Nuclei were
The FASEB Journal
CANITROT ET AL.
harvested by centrifugation and the nuclear proteins were extracted in hypotonic buffer containing 350 mM NaCl. Cytosolic and nuclear extracts proteins were precipitated by addition of ammonium sulfate. The precipitates were resuspended in dialysis buffer (50 mM Tris-HCl, pH 7.5, 1 mM EDTA, 100 mM mono-K glutamic acid, and 10% glycerol) and dialyzed for 2 h at 4°C. The extracts were frozen in liquid nitrogen and stored at ⫺70°C.
Quantification of repair Gels were dried and processed on a PhosphorImager (Storm System, Molecular Dynamics, Sunnyvale, Calif.) and radioactivity was quantified (Image quant 1.1 software). Photographs of the ethidium bromide-stained gels were scanned and processed to normalize for plasmid DNA recovery. Strand displacement assay
Preparation of plasmids and treatment with damaging agents The 2959 bp pBS (pBluescript KS⫹; Stratagene, San Diego, Calif.) and the related 3738 bp pHM14 plasmid (gift from R. D. Wood, ICRF, U.K.) were prepared by alkaline lysis method from Escherichia coli JM109. Both plasmids were carefully purified by one cesium chloride and two neutral sucrose gradient centrifugation as described (20). pBluescript KS⫹ plasmid was treated with either cisplatin according to Hansson and Wood (21) or with UV light (300 J/m2) at 254 nm. pHM plasmids were treated with 10 mM MMS as described (22). In vitro repair synthesis assay The repair synthesis assay was performed as described (23). The reaction mixture (50 l) contained 300 ng each of damaged pBluescript KS⫹ and untreated pHM14 closed circular plasmids, 2 Ci ␣ [32P] dCTP (800 Ci/mmol), 200 g cell extract protein, and 70 mM potassium glutamate in reaction buffer containing 45 mM HEPES-KOH (pH 7.8), 7.4 mM MgCl2, 0.9 mM DTT, 0.4 mM EDTA, 2 mM ATP, 20 M each dGTP, dATP and dTTP, 4 M dCTP, 40 mM phosphocreatine, 2.5 g phosphocreatine kinase, 3.4% glycerol, and 18 g bovine serum albumin (BSA) as described. When specified, 10 M radioactive [␣32P] 8-oxo-dGTP, was added to a similar reaction mixture containing 20 M each dATP, dCTP and dTTP. When noticed, rat Pol , purified as described (24), was added. One unit of Pol  corresponds to 1 nmol of dNTP incorporated into acid-insoluble materials at 37°C in 60 min by using as a substrate an activated calf thymus DNA preincubated with DNAase I. Incubation was 3 h at 30°C. Before the electrophoresis on a 1% agarose gel containing 0.5 g/ml ethidium bromide, plasmid DNA was purified from reaction mixtures and linearized with EcoRV, except for the measurement of the complete gap filling activity. In vitro repair incision assay The incision assay was performed as already reported (25, 26). Briefly, repair reaction took place as described above except that deoxyribonucleotides were omitted and 4.5 M aphidicolin was added. After 2 h at 30°C, the reaction was stopped by the addition of 0.5% sodium dodecyl sulfate (SDS) and 0.1 M EDTA. Plasmids were treated by gentle manual agitation with 200 g/ml proteinase K (42°C, 1 h), purified by phenol-chloroform-isoamyl alcohol extraction, ethanol precipitated, washed in 70% ethanol, dried, and redissolved in TE buffer. DNA was incubated (10 min at room temperature) in reaction mixture containing 90 mM HEPESKOH (pH 6.6), 10 mM MgCl2, 0.2 Ci of [␣32P] dCTP, 2 mM dithiothreitol, 20 M of each of dGTP, dATP, dTTP and 1 unit of E. coli DNA polymerase I large fragment. The reaction was stopped by addition of EDTA 50 mM. DNA was extracted, precipitated, washed, linearized by EcoRV, and electrophoresed on a 1% agarose gel containing 0.5 g/ml ethidium bromide. EFFECT OF POL EXCESS ON NUCLEOTIDE EXCISION REPAIR
To serve as a DNA template, an 85-mer oligonucleotide was hybridized to a 5⬘-32P-labeled 20-mer primer and a looped 43-mer (see Fig. 6A). This 28 nucleotide-gapped, doublestranded DNA template (5 ng) was replicated for 1 h in vitro by various amounts of purified Pol  in reactions (15 l) containing 45 mM HEPES-KOH (pH 7.8), 7 mM MgCl2, 1 mM DTT, 0.4 mM EDTA, 3.4% glycerol, 65 mM potassium glutamate, 18 g of BSA, 200 M each dATP, dCTP, dGTP, and dTTP. At the end of the reaction, 5 l of stopping buffer (90% formamide/0.1% xylene cyanol/0.1% bromophenol blue/0.1 mM EDTA) was added. Samples were denatured for 10 min at 70°C and loaded onto a 15% polyacrylamide/7 M urea/30% formamide gel.
RESULTS DNA repair synthesis activity of Sh and Sh::pol  cell extracts NER activity was evaluated by an in vitro repair synthesis assay (23). Control extracts from CHO UV4 and UV135 mutant cells (respectively deficient for the incisions 5⬘ and 3⬘-side of the lesion) exhibited no NER activity under our assay conditions (data not shown). Thus, the repair activity of Sh and Pol::Sh cell extracts can confidently be attributed to true NER. In addition, mixing both extracts in a same reaction restored a signal close to that of wild-type cell extracts (data not shown), showing that our preparations were proficient. As shown in Fig. 2, protein extracts obtained from pol ::Sh cells (11) constitutively overexpressing ⬃7–10 times Pol  (7) exhibited a five- to sixfold higher repair synthesis activity on UVC-treated plasmid than extracts from control cells (Sh). To ensure that this increased activity was specific of the Pol -mediated action, we used the chain terminator ddCTP, which has been demonstrated to be a good substrate of Pol  but a poor substrate of the replicative DNA polymerases ␦ and ε (11, 27). The presence of ddCTP in the pool of nucleotides during the NER step of synthesis decreased the signal of incorporation from Pol -overexpressing extracts on UVC-treated plasmids (Fig. 2). On the other hand, ddCTP had no effect on the NER process mediated by Sh cell extracts, showing predominance of the action of DNA polymerases ␦ and ε in control cells. This result suggests a specific contribution of Pol  in the enhanced signal obtained with Pol -overexpressing cell extracts. A complementary demonstration was obtained by in1767
sis might be due to NER, BER, or both. To evaluate the residual participation of BER in the yield of radiolabeled incorporation, we performed two sets of control experiments. First, a methyl-methane sulfonate (MMS) damaged plasmid DNA was incubated in a repair reaction since MMS provokes the formation of base methylation repaired by the BER but not the NER pathway. As shown in Fig. 3A, no significant variation in the repair synthesis signal was observed with extracts from Pol ::Sh cells compared to the extracts from Sh cell extracts. This result is in accordance with cellular survival data showing that Pol ::Sh cells are as sensitive to MMS as Sh cells (7) and shows that the action of Pol  in BER is not the limiting step. In other terms, since it has been shown that null pol  mutants are sensitive to MMS (28), the
Figure 2. DNA repair synthesis on UVC-damaged plasmid with extracts from Sh and pol ::Sh cells. Cell extracts from Sh and pol ::Sh (200 g per reaction) were incubated for 3 h at 30°C with damaged (pBS UV) and undamaged (pHM) plasmid in a repair synthesis reaction. When indicated, 20 or 50 M of dideoxyCTP (ddCTP) was added to the normal pool of nucleotides. When specified, purified Pol  was added to control Sh cell extracts. A) Photograph of ethidium bromide stained gel (upper) and autoradiogram of the dried gel (lower). B) Specific incorporation expressed as the ratio of radiolabel incorporation in the damaged vs. undamaged plasmid. The value of [␣32P] dCMP incorporation was normalized to the same amount of DNA as determined from the fluorograph. Each value corresponds to the mean ⫾ sd of three separate experiments performed with three independent preparations of cell extracts. Signal for Sh cell extracts was arbitrary set at 1.
cubating control Sh cell extracts with an increased amount of purified Pol . The addition of four units of purified protein, which correspond to the specific activity measured in Pol -overexpressing cell extracts (data not shown), increased the incorporation of radiolabeled nucleotides on UV-treated plasmids (Fig. 2). BER activity of Sh and pol ::Sh cell extracts Because it is known that UVC induced the formation of pyrimidine dimers recognized by the NER pathway but also of minor lesions such as thymine glycols recognized by the BER, the increased repair synthe1768
Vol. 14
September 2000
Figure 3. A) DNA repair activity on MMS-damaged plasmids. Cell extracts from Sh and pol ::Sh (200 g per reaction) were incubated for 2 h at 30°C with damaged (pHM MMS) and undamaged (pBS) plasmids in a repair synthesis reaction. Photograph of ethidium bromide stained gel (upper) and autoradiogram of the dried gel (lower) are presented. B) DNA repair synthesis on cisplatin-damaged plasmid with extracts from Sh and pol ::Sh cells. Cell extracts from Sh and pol ::Sh (200 g per reaction) were incubated for 3 h at 30°C with damaged (pBS Pt) and undamaged (pHM) plasmid in a repair synthesis reaction. When indicated, ddCTP or purified Pol  was added to the reaction. Photograph of ethidium bromide stained gel (upper) and autoradiogram of the dried gel (lower). Specific incorporation expressed as the ratio of radiolabeled incorporation in the damaged vs. undamaged plasmid.
The FASEB Journal
CANITROT ET AL.
intracellular concentration of endogenous Pol  seems to be saturating in normal cells. Second, a cisplatin-damaged plasmid DNA was incubated in the repair reaction since adducts produced by this drug are mostly recognized in vitro by the NER pathway (29). As expected, a five- to sixfold increase in the repair synthesis activity was found with extracts from pol ::Sh cells by comparison to the control Sh cells (Fig. 3B). This repair signal was dependent on addition of ddCTP or four units of purified Pol  as reported with the UVC-damaged plasmid DNA (Fig. 2). However, the signal restitution by purified Pol  is to a lesser extent, maybe because recombinant Pol  purified from E. coli is less efficient with regard to cisplatin than that extracted from eukaryotic cells. NER incision activity of Sh and pol ::Sh cell extracts The variation of the repair synthesis activity in the assay might be due to either an increased activity in incision of the damaged DNA, an increase of the length of the repair patch, or both. To alleviate the ambiguity, an incision assay (25) was performed with extracts from Sh and pol ::Sh cells. No difference of incision activity in a NER reaction with cisplatin or UVC-damaged plasmid DNA was observed (Fig. 4). Another control was realized by measuring the NER activity using NER-deficient extracts from CHO UV4 and UV135 mutant cells. No significant signal was detected with such extracts, complemented or not with purified Pol  (data not shown), showing as expected that this DNA polymerase is not able to complement the UV4/UV135 defects and does not play a role in incision. Taken together, our data demonstrate that when overexpressed into the cell, Pol  participates in the NER reaction through the DNA polymerization step. Complete gap filling activity by Sh and pol ::Sh cell extracts In vitro, the NER reaction proceeds up to the ligation step as already shown (23). Therefore, it was of interest to examine whether Pol  could promote a complete gap filling when involved in the repair patch synthesis. We conducted similar experiments as described in Figs. 2 and 3B, except that plasmids were not linearized after incubation. By using such substrates, the production of an only partial gap DNA synthesis would result in an impediment for ligation process and therefore in an accumulation of the radioactivity in open circular forms. In contrast, a majority of radioactive materials in closed circular forms would reflect a complete gap filling DNA synthesis and subsequent ligation. As observed in Fig. 5, higher incorporation of radioactive nucleoEFFECT OF POL EXCESS ON NUCLEOTIDE EXCISION REPAIR
Figure 4. DNA repair incision activity by extracts from Sh and pol ::Sh cells. Whole cell extracts from Sh and pol ::Sh (200 g per reaction) were incubated for 2 h at 30°C with cisplatin or UV-treated plasmid in a repair synthesis reaction in the absence of dNTP and with 4.5 M aphidicolin. Subsequent gap filling by Klenow fragment was performed in the presence of [␣32P] dCTP. A) Photograph of ethidium bromidestained gel (upper) and autoradiogram of the dried gel (lower) from one representative experiment. B) Values of specific incorporation were calculated by subtraction of unspecific dCMP incorporation into damaged pBS plasmid. Quantification was the mean ⫾ sd of three separate experiments performed with three independent preparations of cell extracts. Arbitrary the incorporation value from Sh cell extracts was set at 1.
tide in the closed circular form of the repaired plasmids damaged with cisplatin or UVC light was observed for both Sh and pol ::Sh cell extracts, strongly suggesting that the DNA polymerization step went to completion in the absence or presence of excess of Pol . Furthermore, addition of four units of purified Pol  to normal cell extracts after the 3 h repair reaction for an additional hour did not result in an increased incorporation of radioactivity in open or closed circular forms, indicating that the enhanced repair synthesis activity in pol ::Sh cell 1769
completion of the gap filling with increased size of the repair patch. DNA strand displacement by Sh and pol ::Sh cell extracts.
Figure 5. Complete NER activity by extracts from Sh and pol ::Sh  cells. Repair reactions were conducted as described in Figs. 1 and 3 with cisplatin-(A) and UV-(B) damaged plasmids except that after incubation, plasmids were not linearized. OC, open circular form; CC, closed circular form.
extracts was not due to completion of repair reactions blocked at intermediate steps (data not shown). Taken together, these data show that Pol  when substituting for Pol ␦/ε is able to promote
To investigate the molecular basis of the enhanced repair synthesis activity mediated by an excess of Pol , we measured the ability of pol ::Sh cell extracts to displace the DNA strand opposite the template after completion of the gap filling. We thus synthesized a 28 nt gapped duplex DNA mimicking the substrate of the DNA polymerases after the NER excision step. To discriminate between a ligation process and a continuous DNA synthesis mediated by a strand displacement, we constructed a 85/20/43 substrate (see Fig. 6A) in which the 43-mer strand 5⬘ to the gap, annealed to the 85/20 primed template strand, possesses a loop of 6 extra bases placed in the hybridized part of the duplex. We reasoned that gap filling plus ligation would yield a 91-mer product whereas strand displacement accompanying DNA synthesis without ligation process would result in a 85-mer product. The outcome of these experiments is shown in Fig. 6B, where one can distinctly see a Pol -mediated appearance of the 85-mer by addition of purified Pol  in cell extracts (Fig. 6B). In conclusion, strand displacement DNA synthesis catalyzed by excess Pol  may account for the enhanced repair synthesis in recombinant cells.
Figure 6. Strand displacement by Pol  during gapped DNA synthesis. A) Sequences of the 28 nucleotides gapped duplex DNA used in the in vitro replication assay. Gap filling and subsequent ligation yield a 91-mer product (20 ⫹ 28 ⫹ 43) whereas strand displacement accompanying DNA synthesis without ligation process results in a 85-mer product. B) Primer extension experiments by Sh cell extracts in absence or in presence of 2 and 4 purified Pol . Reaction conditions were as indicated in Materials and Methods. Positions of the 20-mer (primer), 85-mer (full-size strand displacement products), and 91-mer (ligation reaction products) as well as the position of the 28 nucleotides gap are indicated by arrows.
1770
Vol. 14
September 2000
The FASEB Journal
CANITROT ET AL.
DISCUSSION Biological significance of Pol  overexpression We recently demonstrated an increased mutation frequency in cells displaying excess Pol  (7), a specific context that has been shown to occur in various tumors such as glioma and lymphoma (10, 11), breast, colon, and prostate (12) as well as in ovarian tumor cell lines (Fig. 1). Because the genome of tumor cells has been shown to carry increased level of DNA damage, an up-regulation of DNA repair proteins may be a first response. The accumulation of alkylation intermediates can, for example, lead to the activation (30) in some cancer tissues of the pol  promoter through an increased DNA-alkylating agent-dependent phosphorylation of the cAMP response element binding protein 1. It could also be envisaged an up-regulation of the Pol  expression with the aim of complementing a deficiency of replicative DNA polymerases such as Pol ␦, which is altered in several cancer cells (31, 32). Moreover, we have shown that Pol -mediated mutation accumulation is dependent on the cell proliferation state, being about twice as important during stationary than in log-phase (data not shown). Thus, it can be assumed (9) that Pol  could complement replicative DNA polymerases, which would be downregulated during the numerous nonproliferating states of the long carcinogenesis process (33) where the tumor is prevented from growing because of physiological barriers, deficiencies in blood supply, inadequate pH, etc. Aneuploidy processes found in nearly all major human tumors (2) could also account for the overexpression. For instance, three copies of pol-containing chromosome 8 can be found in acute myeloid leukemia (34). A more subtle regulation of the Pol  expression is also likely to take place during the carcinogenesis process. For example, we have shown that the presence of the bcr-abl tyrosine kinase, the hallmark of chronic myelogenous leukemia is concomitant to an increase of both the level and activity of Pol  (8).
across intrastrand cross-links (19, 38), and 4) its distributive mode of DNA synthesis (37). The high accuracy of the DNA polymerases involved in the patch synthesis of NER pathway (Pol ␦ and ε) results in an error-free restitution of the nucleotide sequence (13). Using UV-treated DNA plasmids and cell extracts, we present evidence that the inaccurate Pol  can play a key role in NER in the specific context of its overexpression (Fig. 1). We have verified that this interference is due neither to the BER-associated action of Pol  (Figs. 2, 3) nor to a putative involvement of Pol  in the incision step (Fig. 4). Our data show that Pol  in excess acts during the final NER gap filling step (Fig. 5). The consequence of this involvement is the displacement of the DNA strand downstream of the gap (Fig. 6). Considering the particular Pol  features, we also tested the extent of misincorporation of the oxidative base 8-oxo-G, known to potentially lead to the 8oxoG:A mispairing after DNA replication (39), by including 32P-radiolabeled 8-oxo-dGTP into the cell extract-mediated repair reaction of UV treated plasmids. Cell extracts from Pol  overexpressing cells showed a higher ability to incorporate radiolabeled 32 P 8-oxo-dGTP than cell extracts from control cells (data not shown). Taken together, these results lead us to propose that in excess, Pol  could compete with or substitute for error-free DNA polymerases ␦ and ε during the accurate NER DNA synthesis, switching it to an error-prone process (Fig. 7). A model for Pol -mediated NER DNA synthesis In the base excision repair reaction, Pol  is involved not only in the single-nucleotide patch repair but
Incidence of Pol  excess on NER synthesis DNA polymerase  is a 39 kDa protein with both nucleotidyl-transferase and 5⬘-deoxyribose phosphodiesterase activities involved in the BER pathway as well as in meiosis (28, 35). It is expressed at a constant low level throughout the cell cycle (3) and is inducible by some genotoxic treatments (4). It has been described as the least accurate mammalian DNA polymerase, resulting from many features that characterize the enzyme: 1) the lack of associated proofreading activity (36), 2) its poor ability to discriminate nucleotides (37), 3) its capacity to efficiently catalyze error-prone translesion synthesis EFFECT OF POL EXCESS ON NUCLEOTIDE EXCISION REPAIR
Figure 7. Hypothetical model of competition between Pol  and Pol ␦/ε during the NER DNA synthesis step. In view of the multiprotein character of the NER complex, Pol  in excess could substitute replicative DNA polymerases and their cofactors (PCNA , RFC, RPA) instead of act jointly. Substitution could take place either at the initiation step of the DNA synthesis (A) or during the elongation action of Pol ␦/ε (B). 1771
also in long patch repair although its role in this pathway remains unclear. The role of Pol  in the long patch repair was shown by using pol  null cell extracts that do not repair a reduced abasic site normally repaired through the long patch pathway (40). The reaction can be reconstituted with either DNA polymerase  or ␦, suggesting that both DNA polymerases can substitute for each other in long patch BER (41). In addition, it has been shown recently by using cell extracts and a uracil-containing plasmid DNA substrate that the excision of a 3 nt oligonucleotide produced after strand displacement and flap incision is highly dependent on Pol  (42). Indeed, the oligonucleotide release was strongly reduced in the pol  null cell extract (43). The reconstitution of long gap filling in pol  null cell extracts was also achieved in a dose-dependent manner by adding purified Pol  (42). Finally, a very small amount of purified Pol  was sufficient to restore the defect (42), showing that Pol  plays an essential role in long patch BER. Taken together, these experiments indicate that DNA repair synthesis performed by Pol  is required for excision in long patch BER. If this 2- to 8-nucleotide long patch BER route is minor in mammalian cells, a context of Pol  overexpression could thus reinforce this route. We hypothesize here that a comparable situation in presence of excess Pol  may occur in the NER 27–29 nt DNA synthesis. Our experiments indicate that NER synthesis, in which Pol  is not involved in normal conditions (16, 44), can be accomplished by overexpressed Pol  as in long patch BER. It could substitute for the PCNAcontaining replication machinery by promoting the initiation of the DNA synthesis (hypothesis A in the model presented in the Fig. 7) or by switching the elongation complex and complete the gap (hypothesis B in Fig. 7). If we consider that the intervention of Pol  in NER could be the mirror image of what happens in the long patch pathway BER, it is likely that Pol  acts alone in the NER DNA elongation step instead of as part of a protein complex. It could thus promote the DNA synthesis, maybe by forming homo oligomers along the DNA substrate, the protein being able to undergo indefinite self-association (45). Indeed, if the BER long patch pathway is performed by several proteins such as Pol , Pol ␦, or Pol ε, PCNA , RPA, FEN1 structure-specific endonuclease, and DNA ligase I (41, 46 – 48), it has not been proved that these partners act through the formation of a physical complex. No scaffold protein such as XRCC1, involved in the short-patch pathway, was actually detected in the long patch route (49). In addition, Pol -protein analysis showed physical interactions and functional coordination only between Pol  and BER proteins such as human apurinic Ape endonuclease 1772
Vol. 14
September 2000
(50), DNA ligase I (48), and the scaffold protein XRCC1 (49), which functions only in the short-patch complex containing also DNA ligase III and PARP proteins (51). Last, DNA polymerase  is present neither in multiprotein replication complexes/replication foci (52) nor in RPA/PCNA-containing replication complexes recently underscored close to the replication fork in the long patch BER synthesis (47). According to its biochemical features, we have shown here that involvement of Pol  in the NER gap filling reaction resulted in strand displacement, resulting in the observed enhanced labeled materials incorporation in our NER assay. PCNA , engaged in the formation of dynamic complexes with a number of alternative proteins, which facilitates the excision during long patch BER (53), could promote the release of the generated flap by stimulating the FEN-1 endonuclease (54), the subsequent ligation step being mediated by the NER DNA ligase I (Fig. 7). This hypothetical model matches a recent observation in yeast showing that UV-induced lesions can be repaired by a process involving the flap endonuclease FEN-1 (55). However, we cannot exclude a PCNA requirement at a step preceding DNA resynthesis, as it has been shown in DNA mismatch repair (56). Incidence on DNA mutagenesis of the error-prone Pol  interference Early studies of the fidelity of DNA synthesis by Pol  using single-stranded templates demonstrated that the average Pol  error rates for single-base substitution (7⫻10⫺4) and single-base deletion (3–9⫻10⫺4) were substantially higher than those measured for the replicative DNA polymerases, consistent with its lack of 3⬘ 3 5⬘ proofreading exonuclease activity (36). This low-fidelity mode of DNA synthesis was recently examined during the filling of gapped DNA, generating clustered multiple substitutions (57). Considering that such types of errors were observed in experiments using UV-irradiated shuttle vectors (58), these data suggest that these errors could have resulted partially from gap filling synthesis in vivo by Pol  through the NER pathway. Obstacles such as nutritional requirement, inadequate blood supply, and barriers generated by normal matrices limit the expansion of cancer cells during the long-time of cancer process. During these successive periods of nonproliferation, where DNA replication is affected, Pol , whose level is constant throughout the cell cycle, may play a key but mutagenic role by becoming a predominant DNA polymerase. In bacteria, DNA recombination has been proposed to be associated to mutational events occurring during the plateau phase (59). Since it is a
The FASEB Journal
CANITROT ET AL.
part of recombination nodules (60), Pol  could be involved, for example, in the gap filling of recombination intermediates. In the light of our data showing that an excess DNA polymerase  results in a mutator phenotype in mammalian cells (7), we proposed that in Pol -overexpressing cells such as the above-described cancer cells, Pol  could enhance genetic instability (9). We hypothesized that an excess of this error-prone enzyme could perturb error-free DNA transactions by interfering with the action of the accurate and processive DNA polymerases involved in these processes. Here we showed that nucleotide excision repair could be affected. By using as a model the nucleoside 8-oxo-dG, we also demonstrated the increased capacity of the enzyme for misincorporation of mutagenic base analogs. The involvement in cancer cells of Pol  in such a pathway could therefore result in a mutagenic response toward some therapeutic treatments based on NER-repaired genotoxic drugs, such as cisplatin, melphalan, or mechlorethamine (7). Variant cells could thus emerge from the treated tumor mass, some of them being likely to gain a growth advantage and consequently to be selected for their invasive and chemoresistant potentials. We cannot rule out that a ‘non-natural’ intervention of the error-prone Pol  (through the context of its overexpression) in error-free DNA transactions pathways may be generalized to processes other than NER involving a long patch DNA synthesis, such as DNA homologous recombination, mismatch repair or DNA replication. Further investigations regarding these hypotheses will allow a better understanding of one of the multiple ways that confer genetic variability during tumor progression. Pol  belongs to a special group of enzymes, recently the focus of attention of many groups, also called ‘mutases’ (61), which are part of stressinducible processes. This enzyme group also includes DNA polymerases Pol , Pol , Pol IV, and Pol V (62). A better knowledge of such enzymes could represent a novel approach to find pharmaceutical means, such as mutase inhibitors, for protecting people exposed to DNA damaging agents from mutagenic alterations during the course of cancer therapies. This work was partly supported by grants from l’Association de Recherche sur le Cancer (grant 9469 to J.S.H., grant 5446 to C.C., and grant to B.S.), l’Universite´ Paul Sabatier (BQR 1999 to C.C.), and la Re´gion Midi-Pyre´ne´es (grant 99001081 to C.C.). Y.C. was supported by Ligue Nationale contre le Cancer and Electricite´ de France.
2. 3.
4.
5. 6.
7.
8.
9. 10.
11. 12. 13. 14. 15. 16. 17.
18.
19.
20. 21. 22.
REFERENCES 23. 1.
Hoffmann, J. S., and Cazaux, C. (1997) DNA synthesis, mismatch repair, and cancer. Int. J. Oncol. 12, 377–382
EFFECT OF POL EXCESS ON NUCLEOTIDE EXCISION REPAIR
Lengauer, C., Kinzler, K. W., and Vogelstein, B. (1998) Genetic instabilities in human cancers. Nature (London) 396, 643– 648 Zmudzka, B. Z., Fornace, A. J., Collins, J., and Wilson, S. H. (1988) Characterization of DNA polymerase beta mRNA: cellcycle and growth response in cultured human cells. Nucleic Acids Res. 16, 9589 –9596 Fornace, A. J., Zmudzka, B., Hollander, M. C., and Wilson, S. H. (1989) Induction of beta-polymerase mRNA by DNAdamaging agents in Chinese hamster ovary cells. Mol. Cell. Biol. 9, 851– 853 Kunkel, T. A. (1986) Frameshift mutagenesis by eukaryotic DNA polymerases in vitro. J. Biol. Chem. 261, 13581–13587 Kamath-Loeb, A. S., Hizi, A., Kasai, H., and Loeb, L. A. (1997) Incorporation of the guanosine triphosphate analogs 8-oxodGTP and 8-NH2-dGTP by reverse transcriptases and mammalian DNA polymerases. J. Biol. Chem. 272, 5892–5898 Canitrot, Y., Cazaux, C., Fre´chet, M., Bouayadi, K., Lesca, C., Salles, B., and Hoffmann, J. S. (1998) Overexpression of DNA polymerase beta in cell results in a mutator phenotype and a decreased sensitivity to anticancer drugs. Proc. Natl. Acad. Sci. USA 95, 12586 –12590 Canitrot, Y., Lautier, D., Laurent, G., Fre´chet, M., Ahmed, A., Turhan, A. G., Salles, B., Cazaux, C., and Hoffmann, J. S. (1999) Mutator phenotype of BCR-ABL transfected Ba/F3 cell lines and its association with enhanced expression of DNA polymerase . Oncogene 18, 2676 –2680 Canitrot, Y., Fre´chet, M., Servant, L., Cazaux, C., and Hoffmann, J. S. (1999) Overexpression of DNA polymerase beta: a genomic instability enhancer process. FASEB J. 13, 1107–1111 Gomi, A., Shinoda, S., Sakai, R., Hirai, H., Ozawa, K., and Masuzawa, T. (1996) Elevated expression of DNA polymerase beta gene in glioma cell lines with acquired resistance to 1-(4-&mino-2-methyl-5-pyrimidinyl)methyl-3-(2-chloroethyl)-3nitrosourea. Biochem. Biophys. Res. Commun. 227, 558 –563 Scanlon, K. J., Kashani-Sabet, M., and Miyachi, H. (1989) Differential gene expression in human cancer cells resistant to cisplatin. Cancer Invest. 7, 581–587 Srivastava, D. K., Husain, I., Arteaga, C. L., and Wilson, S. H. (1999) DNA polymerase beta expression differences in selected human tumors and cell lines. Carcinogenesis 20, 1049 –1054 Araujo, S. J., and Wood, R. D. (1999) Protein complexes in nucleotide excision repair. Mutat. Res. 435, 23–33 Shijvi, M. K. K., Kenny, M. K., and Wood, R. D. (1992) Proliferating cell nuclear antigen is required for DNA excision repair. Cell 69, 367–374 Kelman, Z. (1997) PCNA: structure, functions and interactions. Oncogene 14, 629 – 640 Wood, R. D., and Shivji, M. K. K. (1997) Which polymerases are used for DNA-repair in eukaryotes? Carcinogenesis 18, 605– 610 Bouayadi, K., Hoffmann, J., Fons, P., Tiraby, M., Reynes, J., and Cazaux, C. (1997) Overexpression of DNA polymerase  sensitizes mammalian cells to 2⬘,3⬘-dideoxycytidine and 3⬘-azido-3⬘deoxythymidine. Cancer Res. 57, 110 –116 Mo, J. Y., Maki, H., and Sekiguchi, M. (1992) Hydrolytic elimination of a mutagenic nucleotide, 8-oxodGTP, by human 18-kilodalton protein: sanitization of nucleotide pool. Proc. Natl. Acad. Sci. USA 89, 11021–11025 Hoffmann, J. S., Pillaire, M. J., Lesca, C., Burnouf, D., Fuchs, R. P. P., Defais, M., and Villani, G. (1996) Fork-like DNA templates support bypass replication of lesions that block DNA synthesis on single-stranded templates. Proc. Natl. Acad. Sci. USA 93, 13766 –13769 Biggerstaff, M., Robins, P., Coverley, D., and Wood, R. D. (1991) Effect of exogenous DNA fragments on human cell extract mediated DNA repair synthesis. Mutat. Res. 254, 217–224 Hansson, J., and Wood, R. D. (1989) Repair synthesis by human cell extracts in DNA damaged by cis and trans-diamminedichloroplatinum(II). Nucleic Acids Res. 20, 8073– 8091 Frit, P., Calsou, P., Chen, D. J., and Salles, B. (1998) Ku70/Ku80 protein complex inhibits the binding of nucleotide excision repair proteins on linear DNA in vitro. J. Mol. Biol. 284, 963–973 Wood, R. D., Robins, P., and Lindahl, T. (1988) Complementation of the Xeroderma pigmentosum DNA repair defect in cell-free extracts. Cell 53, 97–106
1773
24.
25. 26. 27.
28.
29.
30.
31.
32.
33. 34. 35. 36.
37. 38.
39. 40.
41.
42.
43.
44. 45.
1774
Kumar, A., Widen, S. G., Williams, K. R., Kedar, P., Karpel, R. L., and Wilson, S. H. (1990) Studies of the domain structure of mammalian DNA polymerase beta. Identification of a discrete template binding domain. J. Biol. Chem. 265, 2124 –2131 Calsou, P., and Salles, B. (1994) Measurement of damage specific DNA incision by nucleotide excision repair in vitro. Biochem. Biophys. Res. Commun. 202, 788 –795 Calsou, P., and Salles, B. (1994) Properties of damage-dependent DNA incision by nucleotide excision-repair in human cell-free extracts. Nucleic Acids Res. 22, 4937– 4942 Copeland, W. C., Chen, M. S., and Wang, T. S. F. (1992) Human DNA polymerases ␣ and  are able to incorporate anti-HIV deoxynucleotides into DNA. J. Biol. Chem. 267, 21459 –21464 Sobol, R. W., Horton, J. K., Ku¨hn, R., Gu, H., Singhal, R. K., Prasad, R., Rajewsky, K., and Wilson, S. H. (1996) Requirement of mammalian DNA polymerase-beta in base-excision repair. Nature (London) 379, 183–186 Calsou, P., Frit, P., and Salles, B. (1992) Repair synthesis by human cell extracts in cisplatin-damaged DNA is preferentially determined by minor adducts. Nucleic Acids Res. 20, 6363– 6368 Narayan, S., He, F., and Wilson, S. H. (1996) Activation of the human DNA polymerase beta promoter by a DNA-alkylating agent through induced phosphorylation of the cAMP response element-binding protein-1. J. Biol. Chem. 271, 18508 –18513 Da Costa, L., Liu, B., El-Diery, W., Hamilton, S., Kinzler, K., Vogelstein, B., Markowitz, S., Willson, J., De la Chapelle, A., and Downey, K. (1995) Polymerase delta variants in RER colorectal tumours. Nature (London) Genetics 9, 10 –11 Flohr, T., Dai, J., Buttner, J., Popanda, O., Hagmuller, E., and Thielmann, H. (1999) Detection of mutations in the DNA polymerase delta gene of human sporadic colorectal cancers and colon cancer cell lines. Int. J. Cancer 80, 919 –929 Loeb, L. (1997) Many mutations in cancers. Cancer Surveys 28, 329 –342 Macdonald, F., and Ford, C. H. J. (1997) Molecular Biology of Cancer, BIOS Scientific Publishers, Oxford, U.K. Plug, A. W., Clairmont, C. A., Sapi, E., Ashley, T., and Sweasy, J. B. (1997) Evidence for a role of DNA polymerase  in mammalian meiosis. Proc. Natl. Acad. Sci. USA 94, 1327–1331 Kunkel, T. A. (1985) The mutational specificity of DNA polymerase-beta during in vitro DNA synthesis. Production of frameshift, base substitution, and deletion mutations. J. Biol. Chem. 260, 5787–5796 Kornberg, A., and Baker, T. A. (1992) DNA Replication, New York Hoffmann, J. S., Pillaire, M. J., Maga, G., Podust, V., Hu¨bscher, U., and Villani, G. (1995) DNA polymerase  bypasses in vitro d(GpG)-cisplatin adduct placed on codon 13 of H-ras gene. Proc. Natl. Acad. Sci. USA 92, 5356 –5360 Grollman, A., and Moriya, M. (1993) Mutagenesis by 8-oxoguanine: an enemy within. Trends Genet. 9, 246 –249 Biade, S., Sobol, R. W., Wilson, S. H., and Matsumoto, Y. (1998) Impairment of proliferating cell nuclear antigen-dependent apurinic/apyrimidinic site repair on linear DNA. J. Biol. Chem. 273, 898 –902 Klungland, A., and Lindahl, T. (1997) Second pathway for completion of human DNA base excision-repair: reconstitution with purified proteins and requirement for DNase IV (FEN1). EMBO J. 16, 3341–3348 Dianov, G. L., Prasad, R., Wilson, S. H., and Bohr, V. A. (1999) Role of DNA polymerase beta in the excision step of long patch mammalian base excision repair. J. Biol. Chem. 274, 13741– 13743 Dianov, G., Prasad, R., Wilson, S., and Bohr, V. (1999) Role of DNA polymerase beta in the excision step of long patch mammalian base excision repair. J. Biol. Chem. 274, 13741– 13743 Wood, R. D. (1997) Nucleotide excision repair in mammalian cells. J. Biol. Chem. 272, 23465–23468 Dimitriadis, E. K., Prasad, R., Vaske, M. K., Chen, L., Tomkinson, A. E., Lewis, M. S., and Wilson, S. H. (1998) Thermodynamics of human DNA ligase I trimerization and association with DNA polymerase beta. J. Biol. Chem. 273, 20540 –20550
Vol. 14
September 2000
46. 47.
48.
49.
50.
51.
52.
53. 54.
55.
56.
57.
58.
59.
60. 61. 62. 63. 64.
DeMott, M. S., Zigman, S., and Bambara, R. A. (1998) Replication protein A stimulates long patch DNA base excision repair. J. Biol. Chem. 273, 27492–27498 Otterlei, M., Warbrick, E., Nagelhus, T. A., Haug, T., Slipphaug, G., Akbari, M., Aas, P. A., Steinbekk, K., Bakke, O., and Krokan, H. E. (1999) Post-replicative base excision repair in replication foci. EMBO J. 18, 3834 –3844 Prasad, R., Singhal, R. K., Srivastava, D. K., Molina, J. T., Tomkinson, A. E., and Wilson, S. H. (1996) Specific interaction of DNA polymerase beta and DNA ligase I in a multiprotein base excision repair from bovine extracts. J. Biol. Chem. 271, 16000 –16007 Kubota, Y., Nash, R. A., Klungland, A., Scha¨r, P., Barnes, D., and Lindahl, T. (1996) Reconstitution of DNA base excision-repair with purified human proteins: interaction between DNA polymerase beta and the XRCC1 protein. EMBO J. 15, 6662– 6670 Bennet, R. A., Wilson, D. M. R., Wong, D., and Demple, B. (1997) Interaction of human apurinic endonuclease and DNA polymerase beta in the base excision repair pathway. Proc. Natl. Acad. Sci. USA 94, 7166 –7169 Marintchev, A., Mullen, M. A., Maciejewski, M. W., Pan, B., Gryk, M. R., and Mullen, G. P. (1999) Solution structure of the single-strand break repair protein XRCC1 N-terminal domain. Nat. Struct. Biol. 6, 884 – 893 Applegren, N., Hickey, R., Kleinschmidt, A. M., Zhou, Q., Coll, J., Wills, P., Swaby, R., Wei, Y., Quan, J., and Lee, M. E. A. (1995) Further characterization of the human cell multiprotein DNA replication complex. J. Cell. Biochem. 59, 91–107 Gary, R., Kim, K., Cornelius, H. L., Park, M. S., and Matsumoto, Y. (1999) Proliferating cell nuclear antigen facilitates excision in long-patch base excision repair. J. Biol. Chem. 274, 4354 – 63 Wu, X., Li, J., Li, X., Hsieh, C. I., Bueger, P. M. J., and Lieber, M. R. (1996) Processing of branched DNA intermediates by a complex of human FEN-1 and PCNA. Nucleic Acids Res. 24, 2036 –2043 Yoon, J. H., Swiderski, P. M., Kaplan, B. E., Takao, M., Yasui, A., Shen, B. H., and Pfeifer, G. P. (1999) Processing of UV damage in vitro by FEN-1 proteins as part of an alternative DNA excision repair pathway. Biochemistry 38, 4809 – 4817 Umar, A., Buermeyer, A. B., Simon, J. A., Thomas, D. C., Clark, A. B., Liskay, R. M., and Kunkel, T. A. (1996) Requirement for PCNA in DNA mismatch repair at a step preceding DNA resynthesis. Cell 87, 65–73 Osheroff, W. P., Jung, H. K., Beard, W. A., Wilson, S. H., and Kunkel, T. A. (1999) The fidelity of DNA polymerase beta during distributive and processive DNA synthesis. J. Biol. Chem. 274, 3642–3650 Hauser, J., Seidman, M. M., Sidur, K., and Dixon, K. (1986) Sequence specificity of point mutations induced during passage of a UV-irradiated shuttle vector plasmid in monkey cells. Mol. Cell. Biol. 6, 277–285 Torkelson, J., Harris, R., Lombardo, M., Nagendran, J., Thulin, C., and Rosenberg, S. (1997) Genome-wide hypermutation in a subpopulation of stationary-phase cells underlies recombination-dependent adaptive mutation. EMBO J. 16, 3303–3311 Plug, A., Clairmont, C., Sapi, E., Ashley, T., and Sweasy, J. (1997) Evidence for a role of DNA polymerase  in mammalian meiosis. Proc. Natl. Acad. Sci. USA 94, 1327–1331 Radman, M. (1999) Enzymes of evolutionary change. Nature (London) 401, 866 – 868 Lindahl, T., and Wood, R. (1999) Quality control by DNA repair. Science 286, 1897–1905 Jozan, S., Roche´, H., Cheutin, F., Carton, M., and Salles, B. (1992) New human ovarian cell line OVCCR1/sf in serum-free medium. In vitro Cell. Dev. Biol. 28A, 687– 689 Barboule, N., Mazars, P., Baldin, V., Vidal, S., Jozan, S., Martel, P., and Valette, A. (1995) Expression of p21WAF1/CIP1 is heterogeneous and unrelated to proliferation index in human ovarian carcinoma. Int. J. Cancer 27, 611– 615
The FASEB Journal
Received for publication December 21, 1999. Revised for publication March 13, 2000.
CANITROT ET AL.
