Xeroderma pigmentosum variant (XPV) patients with mutations in the DNA polymerase η (pol η) gene are hypersensitive to sunlight and have greatly increased susceptibility to sunlight-induced skin cancer. Consistent with the ability of Pol η to efficiently bypass UV light–induced cyclobutane pyrimidine dimers, XPV cells lacking Pol η have diminished capacity to replicate UV-damaged DNA and are sensitive to UV light–induced killing and mutagenesis. To better understand these and other Pol η functions, we generated Pol η–deficient mice. Mice homozygous for a null mutation in pol η are viable, fertile, and do not show any obvious spontaneous defects during the first year of life. However, fibroblasts derived from these mutant mice are sensitive to killing by exposure to UV light, and all Pol η–deficient mice develop skin tumors after UV irradiation, in contrast to the wild-type littermate controls that did not develop such tumors. These results and biochemical studies of translesion synthesis by mouse Pol η indicate that Pol η–dependent bypass of cyclobutane pyrimidine dimers suppresses UV light–induced skin cancer in mice. Moreover, 37.5% of pol η heterozygous mice also developed skin cancer during 5 months after a 5-month exposure to UV light, suggesting that humans who are heterozygous for mutations in pol η may also have an increased risk of skin cancer. (Cancer Res 2006; 66(1): 87-94)

Xeroderma pigmentosum is a rare genetically heterogeneous autosomal recessive genetic disorder characterized by a 1,000-fold increase in the incidence of sunlight-induced skin cancer that results from defective processing of DNA photoproducts generated by the UVB component of sunlight (1, 2). Although most humans with xeroderma pigmentosum are defective in nucleotide excision repair (NER) of UV photoproducts, a subset of patients, designated xeroderma pigmentosum variant (XPV), have normal NER but are defective in replicating UV-damaged DNA due to inactivating mutations in the pol η gene (36). Human Pol η is a member of the Y family of DNA polymerases (7) that has the ability to bypass cis-syn cyclobutane pyrimidine dimers (CPD), common DNA photoproducts generated by exposure to UV light. Despite the DNA helix distortion caused by CPDs, Pol η bypasses thymine-thymine dimers (4, 6, 8), does so very efficiently (9), and usually inserts adenines opposite both thymines. Thus, Pol η participates in translesion DNA synthesis (TLS), a process that suppresses skin cancer by reducing the underlying cause, UV-induced mutagenesis (1012). The current hypothesis for the XPV phenotype is that when Pol η activity is absent, another polymerase(s) performs less accurate TLS, thus increasing sunlight-induced mutagenesis and skin cancer (1, 13, 14).

It has also been suggested (10) that Pol η may suppress internal cancers due to its ability to accurately bypass 8-oxo-G, a common lesion generated by oxidative stress. In addition, Pol η has been shown to bypass several other lesions, including O4-methyl thymine and O6-methyl guanine (15), adducts of acetylaminofluorene (11), adducts of cisplatin and oxaliplatin (11, 16), and adducts of benzopyrenediolepoxide (17). Compared with predominantly correct incorporation opposite thymine dimers, Pol η frequently inserts incorrect nucleotides opposite some lesions [e.g., (+ and −)-trans-anti-BPDE-dG (17), AAF-dG (11), and both residues of a cisplatin cross-linked di-guanine adduct (11)], leading to the suggestion that TLS by Pol η could increase mutagenesis resulting from some types of environmental stress. Moreover, Pol η lacks an intrinsic proofreading exonuclease activity and copies undamaged DNA templates with very low fidelity (18, 19). This low fidelity and the bias of Pol η towards generating errors at A-T base pairs (20) led to the hypothesis that error-prone synthesis by Pol η may contribute to somatic hypermutation of immunoglobulin genes (21). This hypothesis has been supported by error specificity studies during copying of immunoglobulin gene-coding sequences in vitro (22) and by studies showing that mutations at A-T base pairs in immunoglobulin genes are underrepresented in XPV patients (23, 24). More recently, Pol η has also been implicated in the process of class switch recombination of immunoglobulin genes (23).

The present study was motivated by the idea that a more complete understanding of the proposed biological roles of Pol η in mutagenesis, carcinogenesis, and maturation of the immune system would be facilitated by developing a mouse model for human XPV. We report on the generation and characterization of such a model. We first extend earlier biochemical studies (20), indicating that mouse Pol η has properties that are similar to its human homologue. We then report on the UV radiation induced phenotypes of mice that have a null mutation in Pol η, including the observation that heterozygous mice have increased susceptibility to UV radiation–induced skin cancer.

Polymerase and primer/templates. Mouse Pol η, purified as described in ref. (20), was provided by F. Hanaoka (Osaka University). Primer LBP-25 (5′-AATTTCTGCAGGTCGACTCCAAAGC) and undamaged template (5′-CCAGCTCGGTACCGGGTTAGCCTTTGGAGTCGACCTGCAGAAATT) oligonucleotides were purchased from Invitrogen (San Diego, CA). The cis-syn thymine-thymine dimer-containing template (5′-CCAGCTCGGTACCGGGTTAGCCTTTGGAGTCGACCTGCAGAAATT; dimer at underlined site) was provided by S. Iwai (Osaka University; ref. 9).

Lesion bypass efficiency measurements. Lesion bypass reactions were done as described (9), using 0.27 nmol/L polymerase and 133 nmol/L primer/template in a 30 μL reaction (500:1 substrate/enzyme ratio). Samples (6 μL) were removed at 3, 6, and 9 minutes and added to an equal volume of formamide loading buffer (95% deionized formamide, 25 mmol/L EDTA, 0.01% bromophenol blue, 0.01% xylene cyanol), heated, and resolved by 12% denaturing PAGE (19:1 acrylamide/bis-acrylamide, 8 mol/L urea, 1× Tris-borate EDTA). Gels were visualized and quantified by phosphoimagery as previously described (9, 25).

Generation of murine pol η targeting construct. The murine pol η gene-targeting vector was constructed using genomic DNA isolated from Mus musculus Strain 129S6/SvEvTac chromosome 17 BAC, RP22-187D17 (Genbank accession no. AC104518). A 9,730-bp XhoI-BamHI fragment covering pol η exons 2, 3, and 4 was subcloned onto XhoI and BamHI double-digested pSK vector. A loxP site was blunt-end inserted into the HpaI site, located at 6,738 bp downstream of XhoI site, upstream of exon 4, and the orientation of the loxP site is the same as the pol η gene as determined by sequencing. Then a NotI-digested loxP-Hygro-loxP cassette from pLoxP2-Hygro was blunt-end inserted into the above construct at the SalI site that is located downstream of exon 4 and 7,797 bp downstream of XhoI site. A clone that has hygromycin (Hygro) in the opposite orientation as pol η gene was selected, in which three loxP sites had the same orientation. Finally, an RsrII PGK-TK cassette from pKO-TK was inserted into the above plasmid at the XhoI site that is at the 5′ end of the long homologous arm (Fig. 1). Using this construct, a conditional knockout mouse line targeted at exon 4 can be generated. In this construct, there is a long 5′ homologous arm with 6.7 kb containing exons 2 and 3 without counting the sequence between the loxP inserted at the HpaI site and the loxP-Hygro-loxP inserted at the SalI site and a short 3′ arm with 1.9 kb (Fig. 1). The modified conditional knockout allele was designated pol η3LoxP. Removal of the 1,060-bp fragment containing the entire exon 4 flanked by loxP sites was achieved through Cre-mediated recombination. The resulting mutation was designated pol η-del4.

Figure 1.

Genetic modification of murine pol η gene in embryonic stem cells and mice. A, strategy for developing conditional knockout mutant mice. Restriction enzyme recognition sites, PCR primers, and probe used for hybridization. Partial map of the endogenous pol η gene locus, the pol η targeting vector, the modified pol η allele (pol η3loxP), and the pol η knockout allele (pol η-del4). B, examples of PCR genotyping of genomic DNA from offspring of pol η heterozygous intercross. The F7/R7 product corresponds to pol η wild-type allele and the F7/R14 product corresponds to the modified deletion allele. C, examples of Southern blot hybridization of genomic DNA from progenies of pol η heterozygous intercross. The 4.85-kb corresponds to the pol η wild-type allele, and the 3.55-kb band corresponds to the modified allele. WT, wild-type; mt, mutant.

Figure 1.

Genetic modification of murine pol η gene in embryonic stem cells and mice. A, strategy for developing conditional knockout mutant mice. Restriction enzyme recognition sites, PCR primers, and probe used for hybridization. Partial map of the endogenous pol η gene locus, the pol η targeting vector, the modified pol η allele (pol η3loxP), and the pol η knockout allele (pol η-del4). B, examples of PCR genotyping of genomic DNA from offspring of pol η heterozygous intercross. The F7/R7 product corresponds to pol η wild-type allele and the F7/R14 product corresponds to the modified deletion allele. C, examples of Southern blot hybridization of genomic DNA from progenies of pol η heterozygous intercross. The 4.85-kb corresponds to the pol η wild-type allele, and the 3.55-kb band corresponds to the modified allele. WT, wild-type; mt, mutant.

Close modal

Electroporation of embryonic stem cells and generation of pol η−/− mice. Mice with a modified pol η3LoxP locus were generated by homologous recombination in embryonic stem cells. NotI-digested (50 μg) pol η-CKO targeting vector was used to transfect J1 embryonic stem cells by electroporation. Colonies surviving positive (hygromycin) and negative (gancyclovir) selections were isolated and screened for targeted clones by PCR. Positive clones were confirmed by Southern blotting of SpeI restriction enzyme digested DNA using a 1-kb 3′-outside probe (Fig. 1). Two embryonic stem clones heterozygous for the pol η gene were used for microinjection into C57BL/6 blastocysts, and chimeric mice were generated by reimplantation. To generate both pol η knockout and conditional knockout mice with this single construct, chimeric male mice were bred with EIIa-Cre mice in C57BL/6 genetic background. Germ line transmission of the modified pol η3LoxP alleles was analyzed by Southern blotting and PCR analysis. Both male and female mice heterozygous for both EIIa-Cre and the pol η3LoxP allele were mated with C57BL/6 mice. Mice heterozygous for the modified pol η allele in which both exon 4 and loxP-Hygro-loxP (pol η-del4) were deleted by Cre-LoxP system were used for generating knockout mice (pol η-del4/pol η-del4 or pol η−/−). Subsequent genotyping was done by PCR using exon 4 deletion–specific primers with resultant differential fragment size amplification of the wild-type pol η gene versus the exon 4–deleted allele. The Pol η–deficient mice used were on the (129/Ola × C57BL/6) hybrid genetic background. Age- and sex-matched mice were used as controls in each experiment.

Isolation of mouse embryonic fibroblasts. Mouse embryonic fibroblast (MEF) cells were isolated from mouse embryos at day 13 of gestation essentially as described (26). Individual embryos was dissected to remove the head and soft tissues (liver, intestine, kidneys, lung, heart), which were used for DNA isolation for genotyping, using watchmaker forceps. The carcasses were washed once in PBS to deplete RBC. Individual decapitated embryo carcasses were put into a well of six-well plate and treated with 0.5 mL of 0.25% trypsin/EDTA. Small fragments of tissue were incubated in a 37°C incubator for 15 minutes followed by a second round of trypsin/EDTA (0.5 mL for an embryo) treatment and incubation. The digested tissues were washed with cell culture medium and plated out in a 150-mm dish in 25 mL of medium. The medium was changed once after 24-hour incubation and when the cells reached confluence. Cells were frozen and stored in liquid N2 for future experiments.

RNA isolation and reverse transcription-PCR. Total RNA was extracted from a 10-cm dish of MEF cells or testes by Trizol (Life Technologies Bethesda Research Laboratories, Gaithersburg, MD) following the manufacturer's protocol. RNA was resuspended in 100 μL of DEPC water. cDNA was made with Stratagene reverse transcription-PCR (RT-PCR) kit (La Jolla, CA) following the manufacturer's protocol. Briefly, 1 μL of total RNA sample was mixed with 3 μL of random primer and 11 μL of DEPC water and then incubated at 65°C for 10 minutes and chilled on ice for 10 minutes. After a short spin, 2 μL of reverse-transcript buffer, 1 μL of 10 mmol/L deoxynucleotide triphosphate (dNTP) mixture, 1 μL of RNAsin, and 1 μL of SuperScript II were added and incubated for 15 minutes at room temperature, 45 minutes at 42°C. The SuperScript II was inactivated at 65°C for 10 minutes. The cDNA product was used as template for PCR using primer sets, either XPV-F6/XPV-R7, XPV-F6/XPV-R15, or XPV-F15/XPV-R15 and TaqGold polymerase. Briefly, 2 μL of cDNA product were mixed with 33 μL of water, 5 μL of 10× TaqGold buffer, 8 μL of 25 mmol/L magnesium chloride, 0.5 μL of 25 mmol/L dNTP mixture, 0.5 μL of each primer (25 μmol/L), and 0.5 μL of TaqGold polymerase. The following TaqGold PCR program was applied. After 10 minutes of activation, it was amplified for 35 cycles with denaturation temperature of 94°C (45 seconds), annealing temperature of 57°C (45 seconds), and extension temperature of 72°C (1 minute). Two cycles of additional extension at 72°C for 7 minutes followed. The PCR products were examined in 2% of agarose gels. XPV-F6 primer sequence, 5′-GAATCGAGTGGTTGCTCTTGT-3′. XPV-R7 primer sequence, 5′-AAGTGCTTGGCAGCAAATCT-3′. XPV-F15 primer sequence, 5′-TTTGCTGCCAAGCACTTACA-3′. XPV-R15 primer sequence, 5′-CACTGACCCATGTGAGACCA-3′.

Survival of MEFs after UV irradiation. Embryonic fibroblasts isolated from littermates of pol η+/+, pol η+/−, and pol η−/− mice were seeded at a density of 50,000 cells per well of a six-well tray (27). Cells were cultured overnight in DMEM, rinsed with PBS, aspirated, and irradiated with 254-nm UV light at the fluences indicated, using a Stratalinker. Following irradiation, cells were cultured in DMEM or in DEM containing 100 μg/mL caffeine. Surviving cells were counted 2 days after UV irradiation.

Mouse UV irradiation. Eight- to 12-week-old littermate mice were used, and their backs were shaved once a week for the UV irradiation experiment. Twelve littermates with at least one of three genotypes and an additional litter with four heterozygous littermates were irradiated thrice a week at a dose of 3.75 kJ/m2 each time with a bank of two UVB lamps (Blak-Ray lamp Model XX-15M, Ultraviolet Products, Inc., Upland, CA). UVB flux was measured by a UVX digital radiometer (Model UVX-31, Ultraviolet Products). The mice were examined at least once a week for their health and tumor development.

Histologic analysis of mouse tissues and skin tumors. Ear biopsies were collected by a small puncher after 2 and 3 months of UV treatment, and the tissues were processed for H&E staining. For whole mouse tissue and tumor analyses, mice were sacrificed by CO2 inhalation. The tissues and tumors were isolated and fixed in Bouin's solution and stained with H&E and examined.

Mouse pol η lesion bypass efficiency. Human Pol η bypasses thymine dimers more efficiently than the major replicative DNA polymerases Pol α, Pol δ, and Pol ε (6, 11, 28, 29). Moreover, human Pol η bypasses thymine dimers with an efficiency and processivity that is even higher than for copying undamaged thymines in the same sequence context, and dimer bypass occurs with low fidelity (9). Once bypass is completed, human Pol η switches to less processive synthesis (9), providing an opportunity for the major replicative polymerases to resume accurate replication of undamaged DNA (29). Like its human homologue, mouse Pol η has also been shown to bypass thymine dimers (30). However, the efficiency and fidelity of a complete dimer bypass reaction have not yet been quantified, and whether mouse Pol η switches to less processive synthesis following bypass has not been examined. We began this study by performing these measurements to determine if the interpretations derived with human Pol η can be generalized and to determine if studies in mice recapitulate the situation in humans.

The ability of mouse Pol η to bypass a cis-syn thymine-thymine cyclobutane dimer was examined in reactions incubated for 3 to 9 minutes using a 500-fold excess of primer/template over polymerase. These reaction conditions permit only a single cycle of processive DNA synthesis, as verified by the fact that termination probabilities at each template position remain constant with time, and quantification of product bands allows calculation of bypass efficiency (Fig. 2A). Mouse Pol η bypasses a thymine-thymine dimer with 180% of the efficiency with which it copies the equivalent undamaged thymine bases. This preferential bypass of damaged bases results from more efficient insertion opposite both Ts of the dimer and preferential extension from the primers generated by these insertions (Fig. 2A). The band intensities for thymine-thymine dimer bypass also reveal an increased termination probability following insertion opposite the second undamaged base after the dimer has been bypassed (Fig. 2A,, arrow). As for human Pol η (9), this represents a switch from synthesis that is more processive with damaged than undamaged DNA at the +1 position after bypass to synthesis that is less processive with damaged than undamaged DNA at the +2 position (Fig. 2B).

Figure 2.

Efficiency of translesion DNA synthesis by mouse Pol η. Lesion bypass reactions were done as described in Materials and Methods using conditions that result in a single cycle of processive DNA synthesis. A, gel images of reactions with mouse Pol η copying either undamaged or thymine-thymine (TT) dimer containing templates, with a partial template sequence given on the right (*, thymine-thymine dimer positions). The listed efficiencies for insertion opposite the lesion and surrounding sites are values relative to those observed with undamaged DNA and were calculated as described previously (25). B, graph of the termination probability for several template positions for undamaged (white columns) or thymine-thymine dimer (black columns) templates. Columns, average of six data points (three time points each for two separate reactions) calculated as described (25); bars, SD. Position where Pol η switches from processive (crossbar) to nonprocessive (arrow) synthesis on the thymine-thymine dimer template.

Figure 2.

Efficiency of translesion DNA synthesis by mouse Pol η. Lesion bypass reactions were done as described in Materials and Methods using conditions that result in a single cycle of processive DNA synthesis. A, gel images of reactions with mouse Pol η copying either undamaged or thymine-thymine (TT) dimer containing templates, with a partial template sequence given on the right (*, thymine-thymine dimer positions). The listed efficiencies for insertion opposite the lesion and surrounding sites are values relative to those observed with undamaged DNA and were calculated as described previously (25). B, graph of the termination probability for several template positions for undamaged (white columns) or thymine-thymine dimer (black columns) templates. Columns, average of six data points (three time points each for two separate reactions) calculated as described (25); bars, SD. Position where Pol η switches from processive (crossbar) to nonprocessive (arrow) synthesis on the thymine-thymine dimer template.

Close modal

Next, we used a lesion bypass assay (25) to determine the mouse Pol η error rates for a thymine dimer bypass reaction that is complete (i.e., one requiring all four dNTPs to be present and requiring both nucleotide misinsertions opposite the 3′ and/or 5′ thymine plus mismatch extension). The results (Table 1) indicate that for stable misincorporation of each of the three noncanonical nucleotides opposite the 3′ T of the dimer, mouse Pol η error rates are high, are similar to those for copying undamaged DNA, and are similar to the error rates previously observed with human Pol η (9). These biochemical properties imply that mouse Pol η behaves like its human homologue regarding thymine-thymine dimer bypass.

Table 1.

Fidelity of human (9) and mouse (this study) Pol η during in vitro thymine-thymine dimer bypass

TemplatePlaques screened
Plaques sequencedError rate (×10−4)
TotalDark blueMutation frequency (%)3′ T to A3′ T to C3′ T to G
Mouse η Undamaged 6,839 170 2.5 51 16 370 16 
 Thymine-thymine Dimer 6,756 181 2.7 52 35 370 
Human η Undamaged 7,993 284 3.6 100 30 400 10 
 Thymine-thymine Dimer 6,486 184 2.8 97 19 390 10 
Control Gap only 9,357 <0.01     
TemplatePlaques screened
Plaques sequencedError rate (×10−4)
TotalDark blueMutation frequency (%)3′ T to A3′ T to C3′ T to G
Mouse η Undamaged 6,839 170 2.5 51 16 370 16 
 Thymine-thymine Dimer 6,756 181 2.7 52 35 370 
Human η Undamaged 7,993 284 3.6 100 30 400 10 
 Thymine-thymine Dimer 6,486 184 2.8 97 19 390 10 
Control Gap only 9,357 <0.01     

Generation of mice with a disrupted pol η gene. To determine if Pol η has a role in mouse development and to ascertain if a mouse model for UV-induced human skin cancer can be developed, we used gene targeting in embryonic stem cells to establish pol η conditional mutant mice. Mouse J1 embryonic stem cells were transfected with a gene-targeting construct and targeted cell clones with the predicted homologous recombination event were identified by PCR. These results were confirmed by Southern blot analysis (Fig. 1). Two lines of the modified embryonic stem cells were injected into host blastocysts to create chimeric mice. Male chimeric mice were mated to C57BL/6 (B6) females containing EIIa-cre transgene to generate offspring that transmitted both EIIa-cre transgene and the pol η3loxP mutant allele. Mice that are double heterozygous for EIIa-cre and pol η3loxP mutant alleles were crossed with B6 wild-type mice to produce mice heterozygous for pol η-del4 mutant allele. We examined 595 progeny from pol η-del4/+ intercrosses and found that among them, 33% were wild type, 45% were pol η-del4/+, and 22% were pol η-del4/pol η-del4. These results suggest that mice carrying a deletion in the pol η gene are capable of proceeding through normal development and are viable.

Pol η-del4/pol η-del4 mice have a null mutation. To examine the effect of the deletion mutation in pol η, total RNA was isolated from wild-type, pol η-del4/+, and pol η-del4/pol η-del4 primary embryonic fibroblasts (MEF) and the modified pol η-del4 gene was amplified by the PCR (RT-PCR). Using primers upstream and downstream of the deleted exon 4, as well as primers corresponding to sequences within exon 4, the expected products were detected for both primer sets in wild-type and pol η-del4/+ MEFs (Fig. 3). However, no product was found in the pol η-del4/pol η-del4 MEFs (Fig. 3). Using primers from exons 2 and 6, we detected the expected product from wild-type and pol η-del4/+ RNA. When we used the pol η-del4/pol η-del4 RNA, we did not detect the 712-bp fragment but observed a 497-bp fragment (Fig. 3C). The smaller fragments amplified from five individual pol η-del4/pol η-del4 RNA samples were isolated from the gel and sequenced. The sequencing results were identical for all RNA samples and matched exactly to the predicted sequence for pol η-del4 that result from splicing exons 3 to 5. The exon 4 deletion mutation results in a shift in the open reading frame in exon 5 and generates a stop codon 30 nucleotides downstream. (Fig. 3D). A mutant transcript would therefore encode a truncated polypeptide missing 621 of the normal 713 amino acids comprising Pol η, encompassing the majority of amino acids required for Pol η activity and interactions with other proteins.

Figure 3.

RT-PCR on RNA extracted from MEFs derived from embryos of pol η+/− intercross. Genotypes of the embryos. A, RT-PCR of F6/R7; F6 is on exon 2 and R7 on exon 4. No product was found in pol η-del4/pol η-del4 RNA samples, but a fragment with the expected size (∼421 bp) was amplified in both wild-type and heterozygous samples. B, RT-PCR of F15/R15; F15 is on exon 4, R15 on exon 6. No product was seen in pol η-del4/pol η-del4 RNA samples, but a band with the expected size (∼325 bp) was found in both wild-type or heterozygous samples. These data suggested that exon 4 was deleted. C, RT-PCR of F6/R15 was used to further prove the deletion of exon 4. One fragment of ∼712 bp was amplified in both wild-type and the heterozygous RNA samples but not in the homozygous samples. Instead, a smaller fragment of ∼497 bp, as predicted, was observed in the homozygous RNA templates. The product was gel purified and sequenced. D, summary of sequencing result of mutant band, which confirmed the deletion of exon 4, resulting in an open reading frameshift mutation resulting in a stop codon 30 bp downstream from the deletion point. Protocols. Total RNA and genomic DNA were isolated from MEFs by TRIzol (Life Technologies Bethesda Research Laboratories). DNA was used for genotyping. One micrograms of total RNA was used to reverse-transcribe to make first-strand cDNA using random primers (Superscript RT II kit, Life Technologies Bethesda Research Laboratories). cDNA products were used as templates for regular PCR amplification. The sequences of primers used were as follows: F6, GAATCGAGTGGTTGCTCTTGT; R7, AAGTGCTTGGCAGCAAATCT; F15, TTTGCTGCCAAGCACTTACA; R15, CACTGACCCATGTGAGACCA.

Figure 3.

RT-PCR on RNA extracted from MEFs derived from embryos of pol η+/− intercross. Genotypes of the embryos. A, RT-PCR of F6/R7; F6 is on exon 2 and R7 on exon 4. No product was found in pol η-del4/pol η-del4 RNA samples, but a fragment with the expected size (∼421 bp) was amplified in both wild-type and heterozygous samples. B, RT-PCR of F15/R15; F15 is on exon 4, R15 on exon 6. No product was seen in pol η-del4/pol η-del4 RNA samples, but a band with the expected size (∼325 bp) was found in both wild-type or heterozygous samples. These data suggested that exon 4 was deleted. C, RT-PCR of F6/R15 was used to further prove the deletion of exon 4. One fragment of ∼712 bp was amplified in both wild-type and the heterozygous RNA samples but not in the homozygous samples. Instead, a smaller fragment of ∼497 bp, as predicted, was observed in the homozygous RNA templates. The product was gel purified and sequenced. D, summary of sequencing result of mutant band, which confirmed the deletion of exon 4, resulting in an open reading frameshift mutation resulting in a stop codon 30 bp downstream from the deletion point. Protocols. Total RNA and genomic DNA were isolated from MEFs by TRIzol (Life Technologies Bethesda Research Laboratories). DNA was used for genotyping. One micrograms of total RNA was used to reverse-transcribe to make first-strand cDNA using random primers (Superscript RT II kit, Life Technologies Bethesda Research Laboratories). cDNA products were used as templates for regular PCR amplification. The sequences of primers used were as follows: F6, GAATCGAGTGGTTGCTCTTGT; R7, AAGTGCTTGGCAGCAAATCT; F15, TTTGCTGCCAAGCACTTACA; R15, CACTGACCCATGTGAGACCA.

Close modal

Pol η-del4/pol η-del4 mice are viable and fertile. Both male and female mice deficient for Pol η were fertile and showed no obvious defects in the laboratory environment other than those described below. For example, when examined over a 12-month period, they did not show any signs of spontaneous tumor development or other phenotypic abnormalities.

Pol η-del4/pol η-del4 mice are highly susceptible to UV-induced skin carcinomas. To determine whether the pol η-del4/pol η-del4 mutant mice are susceptible to developing skin carcinomas following exposure to UV light, 13 sets of littermates at 8 to 12 weeks of age were used for a UV irradiation experiment. Twelve pol η-del4/pol η-del4 mice (six males and six females), 24 pol η-del4/+ mice (14 males and 10 females), and 14 +/+ mice (seven males and seven females) were shaved on part of their backs with electronic clippers once a week and exposed to UV irradiation thrice a week at a dosage of 3.75 kJ/m2 each time.

After 2 weeks of UV exposure, all 12 pol η-del4/pol η-del4 mutant mice showed significantly altered appearance of the ears and to some extent the skin. Their ears were darker than those of pol η-del4/+ and wild-type littermates, which remained normal. Histologic analysis revealed that the darker ears in the homozygous mutant mice after UV irradiation resulted from melanocyte accumulation in dermis (data not shown). After 8 weeks of UV exposure, the ears of all pol η-del4/pol η-del4 mice became deformed and dry and exhibited significant atrophy, whereas pol η-del4/+ and wild-type littermates did not show any such abnormalities. Following 2 months of UV exposures, histologic analyses of ear biopsies were done on three homozygous, three heterozygous, and three wild-type littermates. The pol η-del4 homozygous mutant mice showed severe hyperplasia, acanthoris (finger-like projections down into the dermis), and mild dysplasia. No such changes were observed in +/+ or +/− mice. After about 10 weeks of UV irradiation, most of homozygous mice started to scratch themselves, possibly causing ulceration on the ears, neck, and face in some cases. One of the heterozygous mice also started to scratch after 5 months of UV exposure, whereas no scratches were found on any the wild-type control animals during the experiment.

After 3 months of UV exposure, all of the pol η-del4/pol η-del4 mice had severe lesions on their ears, whereas heterozygous and wild-type littermates were normal. Two homozygous mice with the most severe ear lesions and one each of their heterozygous and wild-type littermates were sacrificed. Histologic analysis of the affected ears of the homozygous pol η-del4/pol η-del4 mice revealed lesions that were squamous cell carcinomas in situ and carcinomas (Fig. 4). One homozygous mouse was sacrificed due to severe ulceration on the ears and neck and had a dermoid cyst lined with squamous epithelium and filled with keratin debris (data not shown). The remaining mice were examined once a week for tumor development (Fig. 5A). By the end of 18 weeks, all UV-treated pol η-del4 homozygous mutant mice had developed at least one skin carcinoma, confirmed by histologic analysis. No differential sensitivity was seen between male and female mutant mice. Most of the tumors were on the ears. The UV irradiation regimen was terminated after 5 months of UV exposure. At that time, no skin tumors were detectable in heterozygous and wild-type littermates, whereas some homozygous pol η-del4/pol η-del4 mice had skin tumors as large as 10 mm in diameter. Those mice were sacrificed and carefully examined for other tumors, and a full histologic analysis was done. Monitoring of mice with smaller tumors was continued, and by 2 months after termination of UV irradiation, all five remaining homozygous mice had multiple tumors on their ears, face, and the shaved part of the skin on their backs. Histologic analysis revealed no internal tumors in any of the homozygous pol η-del4/pol η-del4 mice or their eight heterozygous or four wild-type littermates. Histologic analysis revealed that all skin tumors were differentiated squamous cell carcinomas (Fig. 4). In contrast, mice heterozygous for the pol η-del4 allele and wild-type littermate did not develop any skin tumors at this stage. A few of the UV-irradiated pol η-del4/pol η-del4 mice were observed to have enlarged spleens. No other histologic abnormalities were found in other internal organs of the UV-irradiated homozygotes or heterozygous mice, and no metastasis to regional lymph nodes or other secondary organs was found in any of the mice with skin tumors.

Figure 4.

Histologic analysis of mouse ears following UV irradiation. A, wild-type animal ear without UV irradiation, showing normal ear. B, wild-type mouse ear with 3 months of UV exposure, showing no abnormality. C, wild-type mouse ear with 5 months of UV exposure, showing only mild hyperplasia. D and G, homozygous mutant mouse ear without UV exposure, showing no dramatic hyperplasia. E and H, homozygous mutant mouse ear after 3 months of UV exposure, showing squamous cell carcinoma. F and I, homozygous mutant mouse ear after 5 months of UV exposure, showing squamous cell carcinoma. Magnification, ×100 (A-F), ×400 (G-I).

Figure 4.

Histologic analysis of mouse ears following UV irradiation. A, wild-type animal ear without UV irradiation, showing normal ear. B, wild-type mouse ear with 3 months of UV exposure, showing no abnormality. C, wild-type mouse ear with 5 months of UV exposure, showing only mild hyperplasia. D and G, homozygous mutant mouse ear without UV exposure, showing no dramatic hyperplasia. E and H, homozygous mutant mouse ear after 3 months of UV exposure, showing squamous cell carcinoma. F and I, homozygous mutant mouse ear after 5 months of UV exposure, showing squamous cell carcinoma. Magnification, ×100 (A-F), ×400 (G-I).

Close modal
Figure 5.

Survival of mouse carcinogenesis and MEFs after UV irradiation. A, tumor incidence on the skin of UV-irradiated mice. The mice were exposed to UV thrice a week for 5 months. Rapid carcinogenesis occurred only on the ear skin and shaved back skin of all Pol η-deficient mice (n = 12) after 18 weeks of UV treatment; no tumor found in either heterozygous (n = 24) or wild-type (n = 14) littermate controls at this time. Six of 16 heterozygous mice developed tumors 4 months after termination of UV exposure, while none of 10 wild-type controls developed any tumor. All tumors observed were confirmed with histologic analysis. B, MEF survival. The experiment was done as described in Materials and Methods. polη+/+ (), polη+/− (), and polη−/− (). Points, mean for three independent experiments; bars, SD. C, MEF survival with caffeine. The experiment was done as described in Materials and Methods. polη+/+ (), polη+/− (), and polη−/− (). Points, mean for three independent experiments; bars, SD.

Figure 5.

Survival of mouse carcinogenesis and MEFs after UV irradiation. A, tumor incidence on the skin of UV-irradiated mice. The mice were exposed to UV thrice a week for 5 months. Rapid carcinogenesis occurred only on the ear skin and shaved back skin of all Pol η-deficient mice (n = 12) after 18 weeks of UV treatment; no tumor found in either heterozygous (n = 24) or wild-type (n = 14) littermate controls at this time. Six of 16 heterozygous mice developed tumors 4 months after termination of UV exposure, while none of 10 wild-type controls developed any tumor. All tumors observed were confirmed with histologic analysis. B, MEF survival. The experiment was done as described in Materials and Methods. polη+/+ (), polη+/− (), and polη−/− (). Points, mean for three independent experiments; bars, SD. C, MEF survival with caffeine. The experiment was done as described in Materials and Methods. polη+/+ (), polη+/− (), and polη−/− (). Points, mean for three independent experiments; bars, SD.

Close modal

Status of the pol ι gene in UV-irradiated mice. The pol η-del4 mice used here were generated using the mouse J1 embryonic stem cell line derived from a 129 mouse strain that has a spontaneous homozygous truncation mutation in the pol ι gene rendering it deficient in Pol ι (31), a Y family homologue of Pol η that may also have a role in translesion DNA synthesis. To examine the possibility that the increased susceptibility to UV-induced skin cancer of the pol η-del4 /pol η-del4 mice may partly result from Pol ι deficiency, we genotyped the 50 mice used in the analysis described above. These mice were derived from two rounds of backcrossing to B6 mice that are wild type for pol ι, suggesting that most of them should be wild type or heterozygous for pol ι. PCR genotyping (31) revealed that this was indeed the case. Among 14 mice that were wild type for pol η, four were wild type for pol ι, seven were heterozygous for pol ι, and three were homozygous for the pol ι mutation. Despite the differences in pol ι status, none of these 14 mice developed skin cancer following UV irradiation, and they did not exhibit differences in sensitivity to UV irradiation. Among 24 mice that were heterozygous for pol η, 10 were wild type for pol ι, 11 were heterozygous for pol ι, and three were homozygous pol ι mutants (see below for further discussion of the phenotypes of these pol η heterozygotes). Among 12 pol η-del4 /pol η-del4 animals, eight were wild type for pol ι and four were heterozygous for pol ι. The fact that the η-del4 /pol η-del4 mice all developed tumors regardless of their pol ι status clearly indicates that Pol η plays a major role in preventing UV-induced tumors. Because of the relatively small number of mice that are homozygous for the pol ι mutant allele, we were not able to assess the effect of such a mutation in the background of pol η mutations. Such experiments are currently under way.

Survival of mouse embryonic fibroblasts after UV irradiation. MEFs were prepared from two sets of pol η+/+, pol η+/−, and pol η−/− embryos from the same pregnancy. Fibroblasts of similar passages, all of which were confirmed to be pol ι+/+ (31), were cultured and irradiated with UV light. Fibroblasts from the pol η−/− mice (Fig. 4B,, circles) were observed to be more sensitive to killing by UV irradiation than were fibroblasts from wild-type littermates (squares). For all exposures of UV, the pol η−/− are more sensitive than wild-type cells, as has been reported with human cells of comparable genotypes (32). Interestingly, MEFs that are heterozygous for pol η-del4 were also more sensitive than the wild-type MEF control, although less so than the fibroblasts from pol η−/− mice. When this experiment was repeated for cells cultured in the presence of caffeine, pol η−/− MEFs were again observed to be more sensitive than wild-type cells, and pol η+/− MEFs exhibited intermediate sensitivity (Fig. 4C).

Heterozygous pol η-del4 mice are also susceptible to UV-induced skin carcinomas. The intermediate sensitivity of the heterozygous pol η-del4 fibroblasts to killing by UV irradiation prompted continued weekly examination of the remaining 16 heterozygous and wild-type mice for their health and for skin tumor development. Interestingly, one heterozygous mouse developed a skin tumor 3 months after completing the UV exposure regime; and 5 months after terminating UV exposure, six of 16 heterozygous mice had developed one or two tumors on the ear or skin. Among these six mice, two were wild type for pol ι, three were heterozygous for pol ι, and one was homozygous for the truncation mutation in pol ι. Among the 10 heterozygous pol η-del4 mice that did not develop skin tumors, four were wild type for pol ι, four were heterozygous for pol ι, and two were homozygous for the truncation mutation in pol ι. The skin tumors developed by the six heterozygous pol η-del4 mice were also squamous cell carcinomas. Similar to their homozygous littermate phenotypes, no histologic abnormalities were found in internal organs of the UV-irradiated heterozygous mice, and no metastasis to regional lymph nodes or other secondary organs was found in any of the heterozygous mice with skin tumors. Finally, 5 months after cessation of UV exposure, no skin tumors were found in any of the littermates that were wild type for pol η (three pol ι+/+, five pol ι+/−, and two pol ι−/−).

We have generated a line of Pol η mutant mice. There are several lines of evidence that suggest that the homozygous and heterozygous pol η mutant mice are valid models for investigating the multiple proposed functions of mammalian DNA polymerase η. Based in its biochemical properties, mouse Pol η seems to have evolved well for its proposed roles in TLS (8) and somatic hypermutation (21, 24). Mouse Pol η copies a cis-syn thymine-thymine dimer with higher processivity and higher efficiency than it copies undamaged thymines in the same sequence context. Dimer bypass occurs with fidelity at the 3′ thymine of the dimer that is no worse than for copying an undamaged thymine, albeit with error rates that are much higher (e.g., 370 × 10−4 for T-dGMP; Table 1) than those for DNA synthesis by the more accurate replicative DNA polymerases. Moreover, when copying undamaged DNA, mouse Pol η preferentially generates base substitutions at A-T base pairs in specific sequence contexts (20) that are reminiscent of substitutions at A-T base pairs generated during somatic hypermutation of immunoglobulin genes (22, 24). Finally, following thymine dimer bypass, mouse Pol η switches to less processive synthesis, potentially allowing DNA polymerase δ and/or ε to resume accurate replication of undamaged DNA (29). Each of these properties of mouse Pol η is shared by human Pol η, implying that the mouse and human enzymes have similar biological roles.

The pol η exon 4 deletion introduced into the mutant mouse line results in a mRNA transcript that would encode a truncated polypeptide lacking most of the NH2-terminal amino acids required for polymerase activity and lacking all of the COOH-terminal amino acids required for important functional interactions with other proteins. This inactivating mutation thus mimics the majority of pol η mutations identified in XPV patients, which also lead to NH2-terminal truncations of Pol η (4, 5, 33, 34). Male and female mice homozygous for the exon 4 deletion were viable and fertile, showing that neither Pol η protein nor polymerization activity is essential for normal mouse development. In addition, none of the 12 pol η-del4 homozygous mice developed spontaneous tumors of the skin or internal organs. This lack of spontaneous internal tumors in Pol η–deficient mice is interesting, given the suggestion that Pol η suppresses internal cancers by performing error-free bypass of template 8-oxo-guanosine residues in DNA resulting from oxidative stress (10). Future studies using larger numbers of mice can further test this suggestion, as well as the possibility that a defect in Pol η may affect carcinogenesis induced by other physical or chemical agents that generate damaged DNA templates requiring TLS. When irradiated with UV light, the ears and shaved backs of homozygous mutant mice accumulate skin pigments much faster and earlier than do wild-type littermates, and the homozygous mutant mice develop UV radiation-induced skin tumors, whereas the wild-type mice do not. Embryonic fibroblasts from the homozygous mutant mice are also more sensitive to killing by UV irradiation than are fibroblasts from wild-type littermates. These features of this homozygous Pol η–deficient mouse model are similar to those of human XPV patients. Beyond these phenotypes for homozygotes, it is very interesting that several months after UV irradiation, 6 of 16 (38%) mice that were heterozygous for the mutant pol η allele developed UV-induced squamous cell carcinomas. None of 10 wild-type littermates developed skin tumors, indicating that pol η-del4/+ mice have an increased predisposition to UV-induced skin cancer compared with wild-type littermates. The tumor susceptibility in animals is consistent with the observation that MEF cells from heterozygotes were more sensitive than wild-type MEF cells to killing by UV radiation. This finding suggests that humans who are heterozygous for mutations that inactivate Pol η may be at increased risk of developing sunlight-induced skin cancer. The observations reported here clearly indicate that the UV-dependent phenotypes of the homozygous and heterozygous Pol η–deficient mice and fibroblasts correlate with the status of pol η but not with the genotypes of the pol ι gene. Nonetheless, substantial evidence suggests that other TLS polymerases, possibly including Pol ι, may participate in bypass of UV photoproducts (3537). Therefore, it should be informative to generate mouse models that are deficient in Pol η in combination with defects in one or more of the other TLS enzymes. The generation and phenotypic analysis of pol η and pol ι double knockout mice is in progress.

As we were preparing our article for submission, two other groups have independently generated Pol η–deficient mice and reported on their immunoglobulin hypermutation analysis (38, 39). Both of these groups have generated XPV null mutant mice and examined the effect of this mutation on immunoglobulin gene hypermutation. They have not reported on the properties of these mice with respect to UV-induced skin carcinogenesis in either pol η homozygous or heterozygous mutant mice.

Grant support: NIH grants CA (R. Kucherlapati) and EH (R. Kucherlapati and T.A. Kunkel).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Lian Yu for embryonic stem cell transfection, Hong Liu and Hongfeng Ma for blastocyst injection, and Li Zhang, Hua Chang, Andrew Thompson, and Jinbo Li for histologic analysis assistance.

1
Cleaver JE. Common pathways for ultraviolet skin carcinogenesis in the repair and replication defective groups of xeroderma pigmentosum.
J Dermatol Sci
2000
;
23
:
1
–11.
2
van Steeg H, Kraemer KH. Xeroderma pigmentosum and the role of UV-induced DNA damage in skin cancer.
Mol Med Today
1999
;
5
:
86
–94.
3
Lehmann AR, Kirk-Bell S, Arlett CF, et al. Xeroderma pigmentosum cells with normal levels of excision repair have a defect in DNA synthesis after UV-irradiation.
Proc Natl Acad Sci U S A
1975
;
72
:
219
–33.
4
Masutani C, Kusumoto R, Yamada A, et al. The XPV (xeroderma pigmentosum variant) gene encodes human DNA polymerase η.
Nature
1999
;
399
:
700
–4.
5
Johnson RE, Kondratick CM, Prakash S, Prakash L. hRAD30 mutations in the variant form of xeroderma pigmentosum.
Science
1999
;
285
:
263
–5.
6
Johnson RE, Prakash S, Prakash L. Efficient bypass of a thymine-thymine dimer by yeast DNA polymerase, Pol η.
Science
1999
;
283
:
1001
–4.
7
Ohmori H, Friedberg EC, Fuchs RP, et al. The Y-family of DNA polymerases.
Mol Cell
2001
;
8
:
7
–8.
8
Masutani C, Araki M, Yamada A, et al. Xeroderma pigmentosum variant (XP-V) correcting protein from HeLa cells has a thymine dimer bypass DNA polymerase activity.
EMBO J
1999
;
18
:
3491
–501.
9
McCulloch SD, Kokoska RJ, Masutani C, Iwai S, Hanaoka F, Kunkel TA. Preferential cis-syn thymine dimer bypass by DNA polymerase h occurs with biased fidelity.
Nature
2004
;
428
:
97
–100.
10
Haracska L, Yu SL, Johnson RE, Prakash L, Prakash S. Efficient and accurate replication in the presence of 7,8-dihydro-8-oxoguanine by DNA polymerase η.
Nature Gene
2000
;
25
:
458
–61.
11
Masutani C, Kusumoto R, Iwai S, Hanaoka F. Mechanisms of accurate translesion synthesis by human DNA polymerase h.
EMBO J
2000
;
19
:
3100
–9.
12
Zhang Y, Yuan F, Wu X, et al. Error-prone lesion bypass by human DNA polymerase η.
Nucleic Acids Res
2000
;
28
:
4717
–24.
13
Johnson RE, Washington MT, Haracska L, Prakash S, Prakash L. Eukaryotic polymerases ι and ζ act sequentially to bypass DNA lesions.
Nature
2000
;
406
:
1015
–9.
14
Zhu F, Zhang M. DNA polymerase z: new insight into eukaryotic mutagenesis and mammalian embryonic development.
World J Gastroenterol
2003
;
9
:
1165
–9.
15
Glick E, Vigna KL, Loeb LA. Mutations in human DNA polymerase η motif II alter bypass of DNA lesions.
EMBO J
2001
;
20
:
7303
–12.
16
Bassett E, Vaisman A, Tropea KA, et al. Frameshifts and deletions during in vitro translesion synthesis past Pt-DNA adducts by DNA polymerases b and h.
DNA repair
2002
;
1
:
1003
–16.
17
Zhang Y, Wu X, Guo D, Rechkoblit O, Geacintov NE, Wang Z. Two-step error-prone bypass of the (+)- and (−)-trans-anti-BPDE-N2-dG adducts by human DNA polymerases η and κ.
Mutat Res
2002
;
510
:
23
–35.
18
Matsuda T, Bebenek K, Masutani C, Hanaoka F, Kunkel TA. Low fidelity DNA synthesis by human DNA polymerase-η.
Nature
2000
;
404
:
1011
–3.
19
Johnson RE, Washington MT, Prakash S, Prakash L. Fidelity of human DNA polymerase η.
J Biol Chem
2000
;
275
:
7447
–50.
20
Matsuda T, Bebenek K, Masutani C, Rogozin IB, Hanaoka F, Kunkel TA. Error rate and specificity of human and murine DNA polymerase η.
J Mol Biol
2001
;
312
:
335
–46.
21
Rogozin IB, Pavlov YI, Bebenek K, Matsuda T, Kunkel TA. Somatic mutation hotspots correlate with DNA polymerase η error spectrum.
Nature Immunology
2001
;
2
:
530
–6.
22
Pavlov YI, Rogozin IB, Galkin AP, et al. Correlation of somatic hypermutation specificity and A-T base pair substitution errors by DNA polymerase η during copying of a mouse immunoglobulin κ light chain transgene.
PNAS
2002
;
99
:
9954
–9.
23
Zeng X, Negrete GA, Kasmer C, Yang WW, Gearhart PJ. Absence of DNA polymerase η reveals targeting of C mutations on the non transcribed strand in immunoglobulin switch regions.
J Exp Med
2004
;
199
:
917
–24.
24
Zeng X, Winter DB, Kasmer C, Kraemer KH, Lehmann AR, Gearhart PJ. DNA polymerase η is an A-T mutator in somatic hypermutation of immunoglobulin variable genes.
Nat Immunol
2001
;
2
:
537
–41.
25
Kokoska RJ, McCulloch SD, Kunkel TA. The efficiency and specificity of apurinic/apyrimidinic site bypass by human DNA polymerase η and Sulfolobus solfataricus Dpo4.
J Biol Chem
2003
;
278
:
50537
–45.
26
Roberson EJ. Teratocarcinomas and embryonic stem cells. Oxford:IRL;1987.
27
Horton JK, Prasad R, Hou E, Wilson SH. Protection against methylation-induced cytotoxicity by DNA polymerase β-dependent long patch base excision repair.
J Biol Chem
2000
;
275
:
2211
–8.
28
Kusumoto R, Masutani C, Shimmyo S, Iwai S, Hanaoka F. DNA binding properties of human DNA polymerase η: implications for fidelity and polymerase switching of translesion synthesis.
Genes Cells
2004
;
9
:
1139
–50.
29
McCulloch SD, Kokoska RJ, Chilkova O, et al. Enzymatic switching for efficient and accurate translesion DNA replication.
Nucleic Acids Res
2004
;
32
:
4665
–75.
30
Yamada A, Masutani C, Iwai S, Hanaoka F. Complementation of defective translesion synthesis and UV light sensitivity in xeroderma pigmentosum variant cells by human and mouse DNA polymerase η.
Nucleic Acids Res
2000
;
28
:
2473
–80.
31
McDonald JP, Frank EG, Plosky BS, et al. 129-derived strains of mice are deficient in DNA polymerase ι and have normal immunoglobulin hypermutation.
J Exp Med
2003
;
198
:
635
–43.
32
Maher VM, Ouellette LM, Curren RD, McCormick JJ. Caffeine enhancement of the cytotoxic and mutagenic effect of ultraviolet irradiation in a xeroderma pigmentosum variant strain of human cells.
Biochem Biophys Res Commun
1976
;
71
:
228
–34.
33
Broughton BC, Cordonnier A, Kleijer WJ, et al. Molecular analysis of mutations in DNA polymerase η in xeroderma pigmentosum-variant patients.
Proc Natl Acad Sci U S A
2002
;
99
:
815
–20.
34
Kannouche P, Broughton BC, Volker M, Hanaoka F, Mullenders LH, Lehmann AR. Domain structure, localization, and function of DNA polymerase η, defective in xeroderma pigmentosum variant cells.
Genes Dev
2001
;
15
:
158
–72.
35
Friedberg EC, Fischhaber PL, Kisker C. Error-prone DNA polymerases: novel structures and the benefits of infidelity.
Cell
2001
;
107
:
9
–12.
36
Bebenek K, Kunkel TA, editors. Functions of DNA polymerases in DNA repair and replication, 2004.
37
Prakash S, Prakash L. Translesion DNA synthesis in eukaryotes: a one- or two-polymerase affair.
Genes Dev
2002
;
16
:
1872
–83.
38
Delbos F, De Smet A, Faili A, Aoufouchi S, Weill JC, Reynaud CA. Contribution of DNA polymerase h to immunoglobulin gene hypermutation in the mouse.
J Exp Med
2005
;
201
:
1191
–6.
39
Martomo SA, Yang WW, Wersto RP, et al. Different mutation signatures in DNA polymerase η- and MSH6-deficient mice suggest separate roles in antibody diversification.
Proc Natl Acad Sci U S A
2005
;
102
:
8656
–61.