Portal de Periódicos da CAPES

Biological Consequences of 8-Methoxypsoralen-Photoinduced Lesions: Sequence-Specificity of Mutations and Preponderance of T to C and T to A Mutations

2004; Elsevier BV; Volume: 123; Issue: 6 Linguagem: Inglês

10.1111/j.0022-202x.2004.23502.x

1523-1747

Ahmad Besaratinia, Gerd P. Pfeifer,

Carcinogens and Genotoxicity Assessment

Psoriatic patients undergoing psoralen plus ultraviolet radiation (PUVA) therapy are susceptible for squamous cell carcinoma and melanoma of the skin. To investigate the etiological relevance of PUVA for these diseases, we performed mutation spectrometry on the cII transgene in mouse embryonic fibroblasts treated with a single or split PUVA dose (PUVA-I or PUVA-II, respectively). Both treatments were significantly mutagenic as they increased the cII mutant frequency up to 3.7-fold over background, and produced different mutational spectra from that derived spontaneously (p<0.01), but not from one another. The signature of induced mutations, i.e., T to C transitions and T to A transversions with significant site-specificities, i.e., adjacent to T bases at the 3′-neighboring side and to pyrimidines at the 5′-neighboring side, was more pronounced after PUVA-II treatment. Also, the overall mutations occurring at T bases with the same site-specificities were more prevalent after PUVA-II treatment. The characteristic PUVA-induced mutations predominate in the p53 mutational spectrum in controlled in vivo test systems or in high-dose PUVA-treated patients, and also are easily recognizable in the overall PUVA-treated patients. We conclude that PUVA-induced mutagenesis is initiated by PUVA-I treatment and subsequently, augmented by PUVA-II treatment, leaving a unique mutational signature on the cII transgene. The signature mutations of PUVA are discernible in the p53 mutational spectrum in PUVA-treated patients but complex exposure to other therapeutic/environmental carcinogens also leads to the frequent occurrence of other types of mutations in this population. Psoriatic patients undergoing psoralen plus ultraviolet radiation (PUVA) therapy are susceptible for squamous cell carcinoma and melanoma of the skin. To investigate the etiological relevance of PUVA for these diseases, we performed mutation spectrometry on the cII transgene in mouse embryonic fibroblasts treated with a single or split PUVA dose (PUVA-I or PUVA-II, respectively). Both treatments were significantly mutagenic as they increased the cII mutant frequency up to 3.7-fold over background, and produced different mutational spectra from that derived spontaneously (p 3.3.CO;2-GCrossref PubMed Scopus (0) Google Scholar). The easily recoverable cII transgene is relatively short in length (294 base pairs), thereby providing the opportunity for multiple DNA sequence analyses from minute amounts of DNA. Here, we performed the PUVA-I and PUVA-II treatments on Big Blue mouse embryonic fibroblasts, and investigated the consequent cytotoxicity and mutagenicity utilizing a lambda phage-based mutation detection system. Subsequently, we made a comparative analysis between our cII mutation spectrometry data and the existing data on p53 mutations in the PUVA-treated patients compiled from studies entered into the IARC Database of TP53 Somatic Mutations (http://www-p53.iarc.fr/p53DataBase.htm; R8 version) or otherwise listed in PubMed (Seidl et al., 2001Seidl H. Kreimer-Erlacher H. Back B. Soyer H.P. Hofler G. Kerl H. Wolf P. Ultraviolet exposure as the main initiator of p53 mutations in basal cell carcinomas from psoralen and ultraviolet A-treated patients with psoriasis.J Invest Dermatol. 2001; 117: 365-370https://doi.org/10.1046/j.0022-202x.2001.01413.xCrossref PubMed Scopus (41) Google Scholar;Stern et al., 2002Stern R.S. Bolshakov S. Nataraj A.J. Ananthaswamy H.N. p53 mutation in nonmelanoma skin cancers occurring in psoralen ultraviolet a-treated patients: Evidence for heterogeneity and field cancerization.J Invest Dermatol. 2002; 119: 522-526https://doi.org/10.1046/j.1523-1747.2002.01814.xCrossref PubMed Scopus (69) Google Scholar). Both PUVA-I and PUVA-II treatment were cytotoxic in mouse embryonic fibroblasts (Figure 1). The severity of cytotoxicity was, however, significantly higher in PUVA-II treatment than in PUVA-I treatment (p<0.03). Whereas the UVA irradiation alone was considerably and dose-dependently cytotoxic, the 8-methoxypsoralen (8-MOP) alone treatment did not show any cytotoxic effects (Figure 1). Twenty-four hours after treatments, we performed an additional cytotoxicity examination, which showed a slightly but non-significantly lower viability for most treated cells (data not shown). Both PUVA-I and PUVA-II treatment were mutagenic in mouse embryonic fibroblasts as they significantly increased the cII mutant frequency (up to 3.7-fold relative to background; p 95% of all analyzed plaques. Of these, the vast majority were single-base substitutions and less frequently insertions, deletions, and multiple base substitutions/deletions/insertions, respectively (Figure 3a–d). In both induced and spontaneous mutational spectra, there were four "jackpot" mutations at nucleotide positions 179–184 [G insertion/deletion], 196 [G→A transition], 211 [G→C transversion] and 221 [T→G transversion] (Figure 3a–d). These jackpot mutations are common phenomena in transgenic model systems, and have been consistently reported by us (Besaratinia and Pfeifer, 2003Besaratinia A. Pfeifer G.P. Enhancement of the mutagenicity of benzo(a)pyrene diol epoxide by a nonmutagenic dose of ultraviolet A radiation.Cancer Res. 2003; 63: 8708-8716PubMed Google Scholar) as well as by others (Harbach et al., 1999Harbach P.R. Zimmer D.M. Filipunas A.L. Mattes W.B. Aaron C.S. Spontaneous mutation spectrum at the lambda cII locus in liver, lung, and spleen tissue of Big Blue transgenic mice.Environ Mol Mutagen. 1999; 33: 132-143https://doi.org/10.1002/(SICI)1098-2280(1999)33:2<132::AID-EM5>3.0.CO;2-2Crossref PubMed Scopus (0) Google Scholar). Presumably, these jackpot mutations occur in the early development of the transgenic rodents and undergo clonal expansion such that many cells from various tissues harbor the same type of mutations. Alternatively, they might represent actual hotspots of spontaneous mutagenesis (Swiger et al., 1999Swiger R.R. Cosentino L. Shima N. Bielas J.H. Cruz-Munoz W. Heddle J.A. The cII locus in the MutaMouse system.Environ Mol Mutagen. 1999; 34: 201-207https://doi.org/10.1002/(SICI)1098-2280(1999)34:2/3<201::AID-EM20>3.3.CO;2-GCrossref PubMed Scopus (0) Google Scholar). Regardless of their origin, it is methodologically appropriate to exclude all jackpot mutations from the comparative spectra analyses. Excluding the jackpot mutations, the spectra of mutations induced by PUVA-I or PUVA-II treatment were significantly different from that derived spontaneously, or produced by UVA irradiation alone (p 3.3.CO;2-GCrossref PubMed Scopus (0) Google Scholar). Therefore, we combined the strand mirror counterparts of all transitions and transversions, and compared the specific types of mutation between different treatment groups. As shown in Figure 4a, T→C+A→G transitions were the hallmark of mutations induced by both PUVA-I and PUVA-II treatment, relative to control. The increased frequency of T→C+A→G transitions initiated by PUVA-I treatment was augmented after PUVA-II treatment reaching a statistically significant level relative to control (p<0.03). The majority of these transitions occurred at sites adjacent to A/T bases at the 3′-neighboring side (approximately 66% in both PUVA-I and PUVA-II, and 39% in control; p<0.01). Also, for the most part, these transitions were preceded with a pyrimidine (T/C) as the 5′-neighboring base (77%, 81%, and 42% in PUVA-I, PUVA-II, and control, respectively). Likewise, of all mutations occurring at T/A sites induced by PUVA-I and PUVA-II treatments, approximately 56% and 69%, respectively, had A/T bases at the 3′-neighboring side, whereas the respective mutations in control were less than 29%. In addition, the overall mutations at T/A sites induced by PUVA-I and PUVA-II treatments had predominantly an adjacent pyrimidine at the 5′-side (p 3.0.CO;2-ICrossref PubMed Scopus (0) Google Scholar). The mutational spectra experimentally induced by the respective agents, i.e., C to T or CC to TT transitions at dipyrimidine sites and G to T transversions at methylated CpG dinucleotides, closely resemble those established for p53 mutations in skin and lung cancer patients, respectively (Pfeifer and Denissenko, 1998Pfeifer G.P. Denissenko M.F. Formation and repair of DNA lesions in the p53 gene: Relation to cancer mutations?.Environ Mol Mutagen. 1998; 31: 197-205https://doi.org/10.1002/(SICI)1098-2280(1998)31:3<197::AID-EM1>3.0.CO;2-ICrossref PubMed Scopus (0) Google Scholar). In this study, we used mutation spectrometry to investigate the etiological relevance of photo-activated psoralens for squamous cell carcinoma and melanoma of the skin commonly seen in patients who have undergone PUVA therapy (Stern et al., 1997Stern R.S. Nichols K.T. Vakeva L.H. Malignant melanoma in patients treated for psoriasis with methoxsalen (psoralen) and ultraviolet A radiation (PUVA). The PUVA Follow-Up Study.N Engl J Med. 1997; 336: 1041-1045https://doi.org/10.1056/NEJM199704103361501Crossref PubMed Scopus (588) Google Scholar). Cytological examination of the cells that had undergone a single- or split-PUVA treatment (PUVA-I and PUVA-II, respectively) showed that both treatments were cytotoxic, the latter having significantly greater deteriorating effects (Figure 1). Both PUVA-I and PUVA-II treatment were significantly mutagenic in Big Blue mouse embryonic fibroblasts, the former inducing a slightly higher cII mutant frequency (Figure 2). The lower frequency of mutations induced by PUVA-II is concurrent with the higher cytotoxicity of this treatment. Although the PUVA-I treated cells had a high population-doubling rate and underwent one passaging during the 8-day growing period, the PUVA-II-treated cells could barely reach full confluency at the end of culturing. In agreement with the induction of mutant frequency, the spectra of mutations produced by PUVA-I and PUVA-II treatments were significantly different from the spontaneously derived and UVA-only derived mutational spectra. The overall spectra of mutations induced by PUVA-I and PUVA-II treatments were not significantly different from one another; however, the signature of induced mutations was more pronounced after PUVA-II treatment. Accordingly, T→C+A→G transitions and T→A+A→T transversions with significant site-specificities, i.e., adjacent to A/T bases at the 3′-neighboring side and to pyrimidines (T/C) at the 5′-neighboring side, were more discernible after PUVA-II treatment than after PUVA-I treatment. Also, the overall mutations occurring at T/A bases with the above-mentioned site-specificities were more distinguishable in the PUVA-II mutational spectrum than in the PUVA-I spectrum of mutations (Figure 3c,d). These observations give rise to the idea that PUVA-I treatment triggers mutagenesis via a specific pathway on which PUVA-II treatment impinges more intensely. It has previously been demonstrated that PUVA-I treatment induces a combination of furan/pyrone-side MA and ICL psoralen-adducts at different proportions depending on the test system, whereas PUVA-II treatment converts the majority of the cross-linkable furan-side MA thereby, giving rise to predominantly ICL psoralen-adducts (Islas et al., 1991Islas A.L. Vos J.M. Hanawalt P.C. Differential introduction and repair of psoralen photoadducts to DNA in specific human genes.Cancer Res. 1991; 51: 2867-2873PubMed Google Scholar;Yang et al., 1994Yang S.C. Lin J.G. Chiou C.C. Chen L.Y. Yang J.L. Mutation specificity of 8-methoxypsoralen plus two doses of UVA irradiation in the hprt gene in diploid human fibroblasts.Carcinogenesis. 1994; 15: 201-207Crossref PubMed Scopus (36) Google Scholar). Although MA may persist in replicated DNA, ICL psoralen-adducts represent an absolute block to the progression of the replication fork (Vos and Hanawalt, 1987Vos J.M. Hanawalt P.C. Processing of psoralen adducts in an active human gene: Repair and replication of DNA containing monoadducts and interstrand cross-links.Cell. 1987; 50: 789-799Abstract Full Text PDF PubMed Scopus (123) Google Scholar). The latter is reflected by the herein observed severe cytotoxicity of ICL psoralen-adducts (Figure 1). The ICL psoralen-adducts are also known to completely block the transcription machinery (Kupiec, 2000Kupiec M. Damage-induced recombination in the yeast Saccharomyces cerevisiae.Mutat Res. 2000; 451: 91-105Crossref PubMed Scopus (45) Google Scholar). Whilst MA are largely removed by nucleotide excision repair (NER) in a relatively error free fashion (Vos and Hanawalt, 1987Vos J.M. Hanawalt P.C. Processing of psoralen adducts in an active human gene: Repair and replication of DNA containing monoadducts and interstrand cross-links.Cell. 1987; 50: 789-799Abstract Full Text PDF PubMed Scopus (123) Google Scholar), the ICL psoralen-adducts undergo mostly recombinational repair, which might be a mutagenic pathway (Yandell et al., 1994Yandell M.D. Edgar L.G. Wood W.B. Trimethylpsoralen induces small deletion mutations in Caenorhabditis elegans.Proc Natl Acad Sci USA. 1994; 91: 1381-1385Crossref PubMed Scopus (105) Google Scholar;Grossmann et al., 2001Grossmann K.F. Ward A.M. Matkovic M.E. Folias A.E. Moses R.E. S. cerevisiae has three pathways for DNA interstrand crosslink repair.Mutat Res. 2001; 487: 73-83Crossref PubMed Scopus (87) Google Scholar). Transcription-coupled repair and "post-replicative repair" are also implicated in the processing of both MA and ICL psoralen-adducts. Mechanistically, these repair processes involve initial damage recognition followed by excision of the lesion and translesion synthesis with error-prone or error-free polymerases across the gap. Although the complex pathway of recombinational repair for ICL psoralen-adducts is not fully delineated, it is thought to include incision by endonucleases, polymerization, homologous recombination, excision of the cross-link, a second step of repair synthesis, and finally ligation. Such sequential recombinogenic events have the potential to induce mutations (Grossmann et al., 2001Grossmann K.F. Ward A.M. Matkovic M.E. Folias A.E. Moses R.E. S. cerevisiae has three pathways for DNA interstrand crosslink repair.Mutat Res. 2001; 487: 73-83Crossref PubMed Scopus (87) Google Scholar;Cohen et al., 2002Cohen Y. Dardalhon M. Averbeck D. Homologous recombination is essential for RAD51 up-regulation in Saccharomyces cerevisiae following DNA crosslinking damage.Nucleic Acids Res. 2002; 30: 1224-1232Crossref PubMed Scopus (13) Google Scholar). Previously, we have shown that early stage Big Blue mouse embryonic fibroblasts possess a quantifiable NER to repair a variety of bulky DNA adducts (You et al., 1999You Y.H. Li C. Pfeifer G.P. Involvement of 5-methylcytosine in sunlight-induced mutagenesis.J Mol Biol. 1999; 293: 493-503https://doi.org/10.1006/jmbi.1999.3174Crossref PubMed Scopus (77) Google Scholar). Scant data are available on the ability of mouse fibroblasts to perform recombinational DNA repair. In the Big Blue cI/cII mutation detection system, large deletions/insertions along the cII locus that change the structure of the adjacent O gene, which is essential for lytic growth, will not be phenotypically expressed. The occurrence of such events, however, can be noticed from a low packaging efficiency because the affected LIZ construct loses its optimal size and as a result, cannot be efficiently packaged into the viable phages (Stratagene Instruction Manual). This scenario is very improbable to have occurred in this study because both PUVA-I- and PUVA-II-treated cells yielded comparably high packaging efficiency. Notable is, however, the occurrence of few small tandem base deletions consequent to both PUVA-I and PUVA-II treatment. A similar pattern of small tandem base deletions was observed in transgenic mouse embryonic fibroblasts irradiated with a mutagenic dose of 18 J per cm2 of UVA, an agent known to cause DNA strand breaks possibly through recombinational DNA repair (Besaratinia et al., 2004Besaratinia A. Synold T.W. Xi B. Pfeifer G.P. G-to-T transversions and small tandem base deletions are the hallmark of mutations induced by ultraviolet A radiation in mammalian cells.Biochemistry. 2004; 43: 8169-8177Crossref PubMed Scopus (73) Google Scholar). A single base deletion occurred frequently at a homopolymeric run of adenines preceded with a thymine (nucleotide position: 239–246) after both PUVA-I and PUVA-II treatment (Figure 3). The relative frequency of this single deletion was, however, not different between PUVA-I- and PUVA-II-treated cells. This together with the similar spectra of mutations induced by PUVA-I and PUVA-II treatments implies that in our system, an individual repair machinery handles both PUVA-I- and PUVA-II-induced lesions. Such machinery of repair is likely to recruit translesion synthesis using error-prone-polymerases to bypass both MAs and ICL-psoralen-adducts. To see the significance of our findings for cancer epidemiology, we analyzed the existing data on p53 mutations in PUVA-treated patients (http://www-p53.iarc.fr/p53DataBase.htm; R8 version) (Seidl et al., 2001Seidl H. Kreimer-Erlacher H. Back B. Soyer H.P. Hofler G. Kerl H. Wolf P. Ultraviolet exposure as the main initiator of p53 mutations in basal cell carcinomas from psoralen and ultraviolet A-treated patients with psoriasis.J Invest Dermatol. 2001; 117: 365-370https://doi.org/10.1046/j.0022-202x.2001.01413.xCrossref PubMed Scopus (41) Google Scholar;Stern et al., 2002Stern R.S. Bolshakov S. Nataraj A.J. Ananthaswamy H.N. p53 mutation in nonmelanoma skin cancers occurring in psoralen ultraviolet a-treated patients: Evidence for heterogeneity and field cancerization.J Invest Dermatol. 2002; 119: 522-526https://doi.org/10.1046/j.1523-1747.2002.01814.xCrossref PubMed Scopus (69) Google Scholar). As shown in Figure 4b, the established signature of PUVA-mutagenesis is easily distinguishable in the p53 mutational spectrum of the PUVA-treated patients (n=120; from tumor specimens). Stratified for PUVA dose, the characteristic PUVA-induced mutations predominate in patients with high exposure to psoralen and/or UVA (Stern et al., 2002Stern R.S. Bolshakov S. Nataraj A.J. Ananthaswamy H.N. p53 mutation in nonmelanoma skin cancers occurring in psoralen ultraviolet a-treated patients: Evidence for heterogeneity and field cancerization.J Invest Dermatol. 2002; 119: 522-526https://doi.org/10.1046/j.1523-1747.2002.01814.xCrossref PubMed Scopus (69) Google Scholar). In addition, UV-associated transitions "C to T and CC to TT at dipyrimidine sites" are readily noticeable in this population (Figure 4b). It needs to be taken into account that PUVA-treated patients are likely to have prior therapeutic/environmental exposure to a wide range of physical/chemical carcinogens, e.g., UV light, ionizing radiation, topical tar preparations, methotrexate, cyclosporine, etc. It is also known that these patients intentionally seek sunlight exposure because of its palliative effects. As such, PUVA might act as a promoter and not as an initiator in patients previously exposed to other carcinogens. PUVA could also have co-carcinogenic effects in patients with simultaneous exposure to other carcinogenic agents (seeGasparro, 2000Gasparro F.P. The role of PUVA in the treatment of psoriasis. Photobiology issues related to skin cancer incidence.Am J Clin Dermatol. 2000; 1: 337-348Crossref PubMed Scopus (44) Google Scholar for detailed information). Interestingly, under well-controlled experimental conditions, the mutational signature of PUVA was readily detectable in the p53 gene in mouse skin tumors (Nataraj et al., 1996Nataraj A.J. Black H.S. Ananthaswamy H.N. Signature p53 mutation at DNA cross-linking sites in 8-methoxypsoralen and ultraviolet A (PUVA)-induced murine skin cancers.Proc Natl Acad Sci USA. 1996; 93: 7961-7965https://doi.org/10.1073/pnas.93.15.7961Crossref PubMed Scopus (64) Google Scholar) and in yeast harboring a human p53 cDNA (Monti et al., 2000Monti P. Inga A. Aprile A. et al.p53 mutations experimentally induced by 8-methoxypsoralen plus UVA (PUVA) differ from those found in human skin cancers in PUVA-treated patients.Mutagenesis. 2000; 15: 127-132https://doi.org/10.1093/mutage/15.2.127Crossref PubMed Scopus (12) Google Scholar), as well as in the adenine phosphoribosyltransferase gene in Chinese hamster ovary cells (Sage et al., 1993Sage E. Drobetsky E.A. Moustacchi E. 8-Methoxypsoralen induced mutations are highly targeted at crosslinkable sites of photoaddition on the non-transcribed strand of a mammalian chromosomal gene.EMBO J. 1993; 12: 397-402Crossref PubMed Scopus (49) Google Scholar). Such observations in experimental systems versus humans are elaborated upon in our recent work where we demonstrated that the mutational signature of mixed or combined carcinogens might not always be straightforward due to the synergistic and/or multiplicative effects of the carcinogenic agents (Besaratinia and Pfeifer, 2003Besaratinia A. Pfeifer G.P. Enhancement of the mutagenicity of benzo(a)pyrene diol epoxide by a nonmutagenic dose of ultraviolet A radiation.Cancer Res. 2003; 63: 8708-8716PubMed Google Scholar). Lastly, it needs to be acknowledged that genes other than p53 might be of influence in the development of neoplasia associated with PUVA therapy. In summary, we have demonstrated a unique mutagenicity of PUVA treatment in the cII transgene in mouse embryonic fibroblasts. The established signature of mutations induced by PUVA is not, however, exclusively manifested in the p53 mutational spectrum of all PUVA-treated patients possibly due to the complexity of carcinogen exposure in these individuals. Early passage Big Blue mouse embryonic fibroblasts (prepared from 13.5-day-old embryos) were grown to monolayer confluence (∼70%) in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fetal bovine serum. The cells were kept in phenol red and serum free medium, Opti-MEM I (Invitrogen, Carlsbad, California), at least 12 h prior to treatments. The cells were treated with 10 μM 8-MOP (Sigma-Aldrich, Saint Louis, Missouri) or control solvent, dimethylsulfoxide, for 30 min in the dark. Following multiple washes with phosphate-buffered saline (PBS), the cells were irradiated (in Opti-MEM I) with a single dose of 1.05 J per cm2 UVA (PUVA-I treatment), or a double dose of 1.05+2.1 J per cm2 UVA (PUVA-II treatment) with thorough PBS washes in between irradiations. The UVA source consisted of two black-light lamps, F15T8.BLB, 15W (each), 365 nm (General Electric, Los Angeles, California), being filtered through a 3 mm window glass, yielding an average fluence rate of 1.75 mW per cm2 determined by a UVX radiometer (Ultraviolet Products, Upland, California). For homogeneous irradiation of the cells, the culture Petri dishes were placed on the filter glass and were rotated every 2–3 min during the course of irradiation (10–30 min). Immediately after the PUVA-I and PUVA-II treatments, the cells were harvested by trypsinization and evaluated for survival using the trypan blue dye exclusion technique. Alternatively, the cells were cultured in complete growth medium for an additional 8 d after which, they were analyzed for mutant frequency and mutational spectrum of the cII transgene. The 8-d growing period is essential for the fixation of all induced/spontaneously derived mutations into the genome. Except for the PUVA-II-treated cell cultures, all other cultures were passed once or twice during the 8-d growing period. The latter cell cultures underwent three to five population doublings. All experiments were run in triplicate. Genomic DNA was isolated using a standard phenol and chloroform extraction and ethanol precipitation protocol (Besaratinia and Pfeifer, 2003Besaratinia A. Pfeifer G.P. Enhancement of the mutagenicity of benzo(a)pyrene diol epoxide by a nonmutagenic dose of ultraviolet A radiation.Cancer Res. 2003; 63: 8708-8716PubMed Google Scholar). The DNA was dissolved in TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 7.5) and kept at -80°C until further analysis. The cII mutant frequency and mutational spectrum were determined using the previously published protocols (Besaratinia and Pfeifer, 2003Besaratinia A. Pfeifer G.P. Enhancement of the mutagenicity of benzo(a)pyrene diol epoxide by a nonmutagenic dose of ultraviolet A radiation.Cancer Res. 2003; 63: 8708-8716PubMed Google Scholar) (available online as Supplementary Material). Results are expressed as medians±SE. All variables between two treatment groups were compared by the Wilcoxon signed rank test. The entire mutational spectra and the specific types of mutation in different treatment groups were compared by the hypergeometric test of Adams and Skopek (Adams and Skopek, 1987Adams W.T. Skopek T.R. Statistical test for the comparison of samples from mutational spectra.J Mol Biol. 1987; 194: 391-396https://doi.org/10.1016/0022-2836(87)90669-3Crossref PubMed Scopus (261) Google Scholar) and χ2 test, respectively. Values of p≤0.05 were considered statistically significant. The experiments were approved by the institutional animal care committee of the Beckman Research Institute of the City of Hope (City of Hope, Duarte, CA). The institutional and national guide for the care and use of laboratory animals was followed thoroughly. We thank Steven Bates for assistance in cell culturing. This work was supported by a grant from the National Institute of Environmental Health Sciences (ES06070) to G. P. P. The supplementary material is available from http://www.blackwellpublishing.com/products/journals/suppmat/JID/JID23502/JID23502sm.htm