Theoretical analysis of mutation hotspots and their DNA sequence context specificity

Mutation frequencies vary significantly along nucleotide sequences such that mutations often concentrate at certain positions called hotspots. Mutation hotspots in DNA reflect intrinsic properties of the mutation process, such as sequence specificity, that manifests itself at the level of interaction between mutagens, DNA, and the action of the repair and replication machineries. The hotspots might also reflect structural and functional features of the respective DNA sequences. When mutations in a gene are identified using a particular experimental system, resulting hotspots could reflect the properties of the gene product and the mutant selection scheme. Analysis of the nucleotide sequence context of hotspots can provide information on the molecular mechanisms of mutagenesis. However, the determinants of mutation frequency and specificity are complex, and there are many analytical methods for their study. Here we review computational approaches for analyzing mutation spectra (distribution of mutations along the target genes) that include many mutable (detectable) positions. The following methods are reviewed: derivation of a consensus sequence, application of regression approaches to correlate nucleotide sequence features with mutation frequency, mutation hotspot prediction, analysis of oligonucleotide composition of regions containing mutations, pairwise comparison of mutation spectra, analysis of multiple spectra, and analysis of "context-free" characteristics. The advantages and pitfalls of these methods are discussed and illustrated by examples from the literature. The most reliable analyses were obtained when several methods were combined and information from theoretical analysis and experimental observations was considered simultaneously. Simple, robust approaches should be used with small samples of mutations, whereas combinations of simple and complex approaches may be required for large samples. We discuss several well-documented studies where analysis of mutation spectra has substantially contributed to the current understanding of molecular mechanisms of mutagenesis. The nucleotide sequence context of mutational hotspots is a fingerprint of interactions between DNA and DNA repair, replication, and modification enzymes, and the analysis of hotspot context provides evidence of such interactions.     

UV mutation spectra in cell lines from patients with Cockayne's syndrome and ataxia telangiectasia, using the shuttle vector pZ189.

We have used the SV40-based shuttle vector pZ189 to determine ultraviolet mutation spectra in SV40-transformed cell lines from two patients with Cockayne's syndrome (CS) and ataxia telangiectasia (AT). The shuttle vector was UV-irradiated, transfected into the cells and recovered two days later, after many rounds of replication had occurred. Plasmid DNA was used to transform indicator bacteria in which plasmids containing a mutation in the supF gene resulted in white colonies. Mutant plasmids were analysed both by agarose gels and by DNA sequencing. In contrast to published spectra for xeroderma pigmentosum cells, the types of mutation induced by UV mutation in the CS and AT cell lines were similar to each other and to published spectra for normal cell lines. There were however, some differences in the sequence distribution of the mutations.     

Computational analysis of mutation spectra

Mutation frequencies vary along a nucleotide sequence, and nucleotide positions with an exceptionally high mutation frequency are called hotspots. Mutation hotspots in DNA often reflect intrinsic properties of the mutation process, such as the specificity with which mutagens interact with nucleic acids and the sequence-specificity of DNA repair/replication enzymes. They might also reflect structural and functional features of target protein or RNA sequences in which they occur. The determinants of mutation frequency and specificity are complex and there are many analytical methods for their study. This paper discusses computational approaches to analysing mutation spectra (distribution of mutations along the target genes) that include many detectable (mutable) positions. The following methods are reviewed: mutation hotspot prediction; pairwise and multiple comparisons of mutation spectra; derivation of a consensus sequence; and analysis of correlation between nucleotide sequence features and mutation spectra. Spectra of spontaneous and induced mutations are used for illustration of the complexities and pitfalls of such analyses. In general, the DNA sequence context of mutation hotspots is a fingerprint of interactions between DNA and DNA repair/replication/modification enzymes, and the analysis of hotspot context provides evidence of such interactions.     

UV light-induced cyclobutane pyrimidine dimers are mutagenic in mammalian cells.

We used a simian virus 40-based shuttle vector plasmid, pZ189, to determine the role of pyrimidine cyclobutane dimers in UV light-induced mutagenesis in monkey cells. The vector DNA was UV irradiated and then introduced into monkey cells by transfection. After replication, vector DNA was recovered from the cells and tested for mutations in its supF suppressor tRNA marker gene by transformation of Escherichia coli carrying a nonsense mutation in the beta-galactosidase gene. When the irradiated vector was treated with E. coli photolyase prior to transfection, pyrimidine cyclobutane dimers were removed selectively. Removal of approximately 90% of the pyrimidine cyclobutane dimers increased the biological activity of the vector by 75% and reduced its mutation frequency by 80%. Sequence analysis of 72 mutants recovered indicated that there were significantly fewer tandem double-base changes and G X C----A X T transitions (particularly at CC sites) after photoreactivation of the DNA. UV-induced photoproducts remained (although at greatly reduced levels) at all pyr-pyr sites after photoreactivation, but there was a relative increase in photoproducts at CC and TC sites and a relative decrease at TT and CT sites, presumably due to a persistence of (6-4) photoproducts at some CC and TC sites. These observations are consistent with the fact that mutations were found after photoreactivation at many sites at which only cyclobutane dimers would be expected to occur. From these results we conclude that UV-induced pyrimidine cyclobutane dimers are mutagenic in DNA replicated in monkey cells.     

Modulation of an ultraviolet mutational hotspot in a shuttle vector Xeroderma cells.

Ultraviolet mutagenesis of the shuttle vector plasmid pZ189 in Xeroderma Pigmentosum cells yields a mutational pattern marked by hotspots at photoproduct sites on both strands of the supF marker gene. In order to test the influence of strand orientation on the appearance of hotspots the mutagenesis study was repeated on a vector with the supF gene in the inverted orientation. We recovered a pattern the same as that in the earlier work and conclude that the nature of the DNA polymerase involved in the replication of specific strands is not a primary determinant of hotspot occurrence in this system. One of the hotspots lies in an 8 base palindrome while the corresponding site on the other strand was not a hotspot. These results were obtained with calcium phosphate transfection of the UV treated vector. When DEAE dextran was used as a transfection agent both sites in the palindrome were hotspots. In a mixing experiment the calcium phosphate pattern was recovered. Our data suggest that the sequence determinants of mutational probability at these two sites lie outside the 8 bases of the palindrome and that mutagenesis at one, but not the other, site is sensitive to perturbation of cellular calcium levels.     

Characterization of carbadox-induced mutagenesis using a shuttle vector pSP189 in mammalian cells

Carbadox, a quinoxaline 1,4-dioxide derivative, is a known mutagen with its functional mechanism yet to be well defined. In the present study we used a shuttle vector assay in vitro to uncover the functional details of carbadox-induced mutagenesis in mammalian cells. The plasmid DNA of a shuttle vector pSP189 was treated with different doses of carbadox at 37 degrees C for 1 or 2h with or without the presence of S9. The target gene SupF in the plasmid was sequenced after replication in Vero cells followed by amplification in Escherichia coli MBM7070 to evaluate mutation frequency. DNA sequencing analysis of recovered carbadox-induced mutations revealed 76.3% single base substitution, 7.9% single base insertion, 10.5% single base deletion and 5.3% large fragments deletion. All single base substitutions occurred at G:C base pairs, among which transversion and transition occurred at a 2:1 ratio. The mutations did not occur randomly in the supF gene, but had sequence specificity and hotspots instead: most substitutions were detected at the nucleotide N in a 5'-NNTTNN-3' sequence; 75% of base insertions were seen in the 5'-TCC-3' sequence; whereas all large fragments deletions occurred in the 5'-ANGGCCNAAA-3' sequence. Nucleotide 129, 141 and 155 in the supF gene of plasmid pSP189 were identified as the hotspots for carbadox-induced mutations that accounted for 65% of all single base substitutions. We conclude that carbadox and its metabolites induce sequence-specific DNA mutations at high frequencies, therefore its safe usage in animal husbandry should be seriously considered.     

DNA damage and mutations produced by chloroacetaldehyde in a CpG-methylated target gene

Chloroacetaldehyde (CAA) is a metabolite of the human carcinogen vinyl chloride. CAA produces several types of DNA adducts including the exocyclic base adducts 3,N(4)-ethenocytosine, 1,N(6)-ethenoadenine, N(2),3-ethenoguanine, and 1,N(2)-ethenoguanine. Adducts of CAA with 5-methylcytosine have not yet been characterized. Here we have analyzed the mutational spectra produced by CAA in the supF gene of the pSP189 shuttle vector when present in either an unmethylated or CpG-methylated state. The vectors were replicated in human nucleotide excision repair-deficient XP-A fibroblasts. The mutational spectra obtained with the unmethylated and methylated supF target genes were generally similar with a preponderance of C/G to T/A transitions and C/G to A/T transversions. CAA-induced DNA adducts were mapped along the supF gene by using thermostable thymine DNA glycosylase (TDG) in conjunction with ligation-mediated PCR or by a Taq polymerase stop assay. Prominent CAA-induced TDG-sensitive sites were seen at several CpG positions but were independent of methylation. Methylated CpG sites were sites of CAA-induced mutations but were not the major mutational hotspots. Taq polymerase arrest sites were observed at numerous sequence positions in the supF gene and reflected the rather broad distributions of mutations along the sequence. We conclude that methylated CpG sites are not preferential targets for chloroacetaldehyde-induced mutagenesis.     

Replication fidelity of the supF gene integrated in the thymidine kinase locus of herpes simplex virus type 1

Recombinant viruses were constructed to have an Escherichia coli replicon containing a mutagenesis marker, the supF gene, integrated within the thymidine kinase locus (tk) of herpes simplex virus type 1. These viruses expressed either wild-type or mutant DNA polymerase (Pol) and were tested in a mutagenesis assay for the fidelity of their replication of the supF gene. A mutation frequency of approximately 10(-4) was observed for wild-type strain KOS-derived recombinants in their replication of the supF gene. However, recombinants derived from the PAA(r)5 Pol mutant, which has been demonstrated to have an antimutator phenotype in replicating the tk gene, had three- to fourfold increases in supF mutation frequency (P < 0.01), a result similar to that exhibited when the supF gene was induced to replicate as episomal DNA (Y. T. Hwang, B.-Y. Liu, C.-Y. Hong, E. J. Shillitoe, and C. B. C. Hwang, J. Virol. 73:5326-5332, 1999). Thus, the PAA(r)5 Pol mutant had an antimutator function in replicating the tk gene and was less accurate in replicating the supF gene than was the wild-type strain. The spectra of mutations and distributions of substituted bases within the supF genes that replicated as genomic DNA were different from those in the genes that replicated as episomal DNA. Therefore, the differences in sequence contents between the two target genes influenced the accuracy of the Pol during viral replication. Furthermore, the replication mode of the target gene also affected the mutational spectrum.     

An exploratory analysis of multiple mutation spectra

The last decade has witnessed a remarkable increase in the number of mutations identified both in human disease-related genes and mutation reporter genes including those in mammalian cells and transgenic animals. This has led to the curation of a number of computerised databases, which make mutation data freely available for analysis. A primary interest of both the clinical researcher and the genetic toxicologist is determination of location and types of mutation within a gene of interest. Collections of mutation data observed for a disease-related gene or, for a gene exposed to a particular chemical, permits discovery of regions of sequence along the gene prone to mutagenesis and may provide clues to the origin of a mutation. The principal tool for visualising the distribution pattern of mutant data along a gene is the mutation spectrum: the distribution and frequency of mutations along a nucleotide sequence. In genetic toxicology, the current wealth of mutation data available allows us to construct many mutation spectra of interest to investigate the mutagenic mechanisms and mutational sites for one or a group of mutagens. Using the multivariate statistical methods principal components analysis (PCA) and cluster analysis (CA) we have tested the ability of these methods to establish the underlying patterns within and between 60 UV-induced, mitomycin C-induced and spontaneous mutations in the supF gene. The spectra were derived from human, monkey and mouse cells including both repair efficient and repair deficient cell lines. We demonstrate and support the successful application of multivariate statistical methods for exploring large sets of mutation spectra to reveal underlying patterns, groupings and similarities. The methods clearly demonstrate how different patterns of spontaneous and UV-induced supF mutation spectra can result from variation in plasmid, culture medium, species origin of cell line and whether mutations arose in vivo or in vitro.     

The role of DNA polymerase eta in UV mutational spectra

UV irradiation generates predominantly cyclobutane pyrimidine dimers (CPDs) and (6-4) photoproducts in DNA. CPDs are thought to be responsible for most of the UV-induced mutations. Thymine-thymine CPDs, and probably also CPDs containing cytosine, are replicated in vivo in a largely accurate manner by a DNA polymerase eta (Pol eta) dependent process. Pol eta is encoded by the POLH (XPV) gene in humans. In order to clarify the specific role of Pol eta in UV mutagenesis, we have used an siRNA knockdown approach in combination with a supF shuttle vector which replicates in mammalian cells. This strategy provides an advantage over studying mutagenesis in cell lines derived from normal individuals and XP-V patients, since the genetic background of the cells is identical. Synthetic RNA duplexes were used to inhibit Pol eta expression in 293T cells. The reduction of Pol eta mRNA and protein was greater than 90%. The supF shuttle vector was irradiated with UVC and replicated in 293T cells in presence of anti-Pol eta siRNA. The supF mutant frequency was increased by up to 3.6-fold in the siRNA knockdown cells relative to control cells confirming that Pol eta plays an important role in mutation avoidance and that the pol eta knockdown was efficient. UV-induced supF mutants were sequenced from siRNA-treated cells and controls. Surprisingly, neither the type of mutations nor their distribution along the supF gene were substantially different between controls and siRNA knockdown cells and were predominantly C to T and CC to TT transitions at dipyrimidine sites. The data are compatible with two models. (i) Incorrect replication of cytosine-containing photoproducts by a polymerase other than Pol eta produces similar mutations as when Pol eta is present but at a higher frequency. (ii) Due to lack of Pol eta or low levels of remaining Pol eta, lesion replication is delayed allowing more time for cytosine deamination within CPDs to occur. We provide proof of principle that siRNA technology can be used to dissect the in vivo roles of lesion bypass DNA polymerases in DNA damage-induced mutagenesis.     

Exonuclease-deficient polymerase mutant of herpes simplex virus type 1 induces altered spectra of mutations

The effect of exonuclease activity of the herpes simplex virus DNA polymerase (Pol) on DNA replication fidelity was examined by using the supF mutagenesis assay. The recombinants with exonuclease-deficient Pol, containing an integrated supF gene in the thymidine kinase locus (tk), exhibited supF mutation frequencies ranging from 0.14 to 5.6%, consistent with the tk mutation frequencies reported previously (Y. T. Hwang, B.-Y. Liu, D. M. Coen, and C. B. C. Hwang, J. Virol. 71:7791-7798, 1997). The increased mutation frequencies were 10- to 500-fold higher than those observed for wild-type Pol recombinants. The increased mutation frequencies also were significantly higher than those of supF mutant replicated by exonuclease-deficient Pols in the plasmid-borne assay. Furthermore, characterization of supF mutants demonstrated that recombinants with a defective exonuclease induced types and distributions of supF mutations different from those induced by wild-type Pol recombinants. The types of supF mutations induced by exonuclease-deficient recombinants differed between the plasmid- and genome-based assays. The spectra of supF mutations also differed between the two assays. In addition, exonuclease-defective viruses also induced different spectra of supF and tk mutations. Therefore, both the assay methods and the target genes used for mutagenesis studies can affect the repication fidelity of herpes simplex virus type 1 Pol with defective exonuclease activity.     

Test of models for the sequence specificity of UV-induced mutations in mammalian cells

We have used mathematical modeling and statistical analysis to examine the correlation between UV-induced DNA damage and resulting base-substitution mutations in mammalian cells. The frequency and site specificity of UV-induced photoproducts in the supF gene of the pZ189 shuttle vector plasmid were compared with the frequency and site specificity of base-substitution mutations induced upon passage of the UV-irradiated vector in monkey cells. The hypothesis that the observed mutational spectrum is due to a preferential insertion of adenosine opposite UV photoproducts in the DNA template was found to best explain the mutational data. Models in which it was postulated that only (6-4) photoproducts, and not cyclobutane dimers, are mutagenic, or that the relative frequency of photoproduct formation does not influence mutation frequencies, fit the data much less well. This analysis demonstrates that molecular mechanisms of mutagenesis in mammalian cells can be deduced from mutational data obtained with a shuttle vector system.     

An in vivo approach to identifying sequence context of 8-oxoguanine mutagenesis

Base substitution mutations are not distributed randomly in that most are located at a few specific hotspots sites. We have been studying 7,8-dihydro-8-oxoguanine mutagenesis in Escherichia coli in the supF gene carried in a plasmid. Among hotspots, guanine within the 5'-AGA-3' located in the anticodon site was susceptible to the induction of G:C-->T:A transversion. In this study, we constructed variants of the supF gene in which the hotspot 5'-AGA-3' was modified to 5'-AGT-3', 5'-AGG-3' and 5'-AGC-3' to determine the influence of 3' neighboring base on G:C-->T:A mutational activity. Using these variant supF genes propagated in a 7,8-dihydro-8-oxoguanine repair-deficient host, we found that guanine within 5'-AGA-3' and 5'-AGG-3' produce G:C-->T:A, but guanine within 5'-AGT-3' and 5'-AGC-3' reduce the formation of G:C-->T:A. These changes were thus due to the effect of sequence context on the efficiency of mutation formation at the sites of 7,8-dihydro-8-oxoguanine. We also observed a longer range base-pair effect on hotspot formation.     

The role of DNA polymerase iota in UV mutational spectra

UVB (280-320 nm) and UVC (200-280 nm) irradiation generate predominantly cyclobutane pyrimidine dimers (CPDs) and (6-4) photoproducts in DNA. CPDs are thought to be responsible for most of the UV-induced mutations. Thymine-thymine CPDs, and probably also CPDs containing cytosine, are replicated in vivo in a largely accurate manner by a DNA polymerase eta (Pol eta) dependent process. Pol eta is a DNA damage-tolerant and error-prone DNA polymerase encoded by the POLH (XPV) gene in humans. Another member of the Y family of error-prone DNA polymerases is POLI encoding DNA polymerase iota (Pol iota). In order to clarify the specific role of Pol iota in UV mutagenesis, we have used an siRNA knockdown approach in combination with a supF shuttle vector which replicates in mammalian cells, similar as we have previously done for Pol eta. Synthetic RNA duplexes were used to efficiently inhibit Pol iota expression in 293 T cells. The supF shuttle vector was irradiated with 254 nm UVC and replicated in 293 T cells in presence of anti-Pol iota siRNA. Surprisingly, there was a consistent reduction of recovered plasmid from cells with Pol iota knockdown and this was independent of UV irradiation of the plasmid. The supF mutant frequency was unchanged in the siRNA knockdown cells relative to control cells confirming that Pol iota does not play an important role in UV mutagenesis. UV-induced supF mutants were sequenced from siRNA-treated cells and controls. Neither the type of mutations nor their distribution along the supF gene were significantly different between controls and siRNA knockdown cells and were predominantly C to T and CC to TT transitions at dipyrimidine sites. These results show that Pol iota has no significant role in UV lesion bypass and mutagenesis in vivo and provides some initial data suggesting that this polymerase may be involved in replication of extrachromosomal DNA.     

Studies on direct and indirect effects of DNA damage on mutagenesis in monkey cells using an SV40-based shuttle vector

We are using an SV40-based shuttle vector, pZ189, to study mechanisms of mutagenesis in mammalian cells. The vector can be treated with mutagens in vitro and replicated in animal cells; resulting mutants can be selected and amplified in bacteria for DNA sequencing. This versatile vector system has allowed us to explore several different questions relating to the mutagenic process. We have studied the direct effects of template damage caused by UV or benzo[a]pyrene diolepoxide by treating vector DNA with these agents and then replicating the damaged DNA in monkey cells. Mutational mechanisms were deduced from the spectrum of mutations induced in the supF target gene of the vector DNA. To study the role of indirect effects of DNA damage on mutagenesis in mammalian cells, we have treated the cells and the vector DNA separately with DNA-damaging agents. We find that pretreatment of cells with DNA-damaging agents, or with conditioned medium from damaged cells, causes an enhancement of mutagenesis of a UV-damaged vector. Thus, DNA damage can act indirectly to enhance the mutagenic process. We also have preliminary evidence that pZ189 can be used in an in vitro DNA replication system to study the process of mutation fixation on the biochemical level. We believe that the pZ189 vector will prove to be as useful for in vitro studies of mutational mechanisms as it has been for in vivo studies.     

Singlet oxygen induced mutation spectrum in mammalian cells.

In order to characterize the molecular nature of singlet oxygen (1O2) induced mutations in mammalian cells, a SV40-based shuttle vector (pi SVPC13) was treated with singlet oxygen arising from the thermal decomposition of the water-soluble endoperoxide of 3,3'-(1,4-naphthylidene) dipropionate (NDPO2). After the passage of damaged plasmid through monkey COS7 cells, the vector was shuffled into E. coli cells, allowing the screening of supF mutants. The mutation spectrum analysis shows that single and multiple base substitutions arose in 82.5% of the mutants, the others being rearrangements. The distribution of mutations within the supF gene is not random and some hotspots are evident. Most of the point mutations (98.4%) involve G:C base pairs and G:C to T:A transversion was the most frequent mutation (50.8%), followed by G:C to C:G transversion (32.8%). These results indicate that mutagenesis in mammalian cells, mediated by 1O2-induced DNA damage, is targeted selectively at guanine residues.     

Differential repair of UVB-induced cyclobutane pyrimidine dimers in cultured human skin cells and whole human skin

Cyclobutane pyrimidine dimers (CPDs) and pyrimidine (6-4) pyrimidone photoproducts (6-4PPs) are the two main classes of mutagenic DNA damages induced by UVB radiation. Numerous studies have been devoted so far to their formation and repair in human cells and skin. However, the biochemical methods used often lack the specificity that would allow the individual study of each of the four CPDs and 6-4PPs produced at TT, TC, CT and CC dinucleotides. In the present work, we applied an HPLC-mass spectrometry assay to study the formation and repair of CPDs and 6-4PPs photoproducts in primary cultures of human keratinocytes and fibroblasts as well as in whole human skin. We first observed that the yield of dimeric lesions was slightly higher in fibroblasts than in keratinocytes. In contrast, the rate of global repair was higher in the last cell type. Moreover, removal of DNA photoproducts in skin biopsies was found to be slower than in both cultured skin cells. In agreement with previous works, the repair of 6-4PPs was found to be more efficient than that of CPDs in the three types of samples, with no observed difference between the removal of the TT and TC derivatives. In contrast, a significant influence of the nature of the two modified pyrimidines was observed on the repair rate of CPDs. The decreasing order of removal efficiency was the following: C<>T>C<>C>T<>C>T<>T. These data, together with the known intrinsic mutational properties of the lesions, would support the reported UV mutation spectra. A noticeable exception concerns CC dinucleotides that are mutational hotspots with an UV-specific CC to TT tandem mutation, although related bipyrimidine photoproducts are produced in low yields and efficiently repaired.     

Peroxyl radical mediated oxidative DNA base damage: implications for lipid peroxidation induced mutagenesis

Endogenous DNA damage induced by lipid peroxidation is believed to play a critical role in carcinogenesis. Lipid peroxidation generates free radical intermediates (primarily peroxyl radicals, ROO(*)) and electrophilic aldehydes as the principal genotoxicants. Although detailed information is available on the role of aldehyde base adducts in mutagenesis and carcinogenesis, the contribution of peroxyl radical mediated DNA base damage is less well understood. In the present study we have mapped oxidative base damage induced by peroxyl radicals in the supF tRNA gene and correlated this information with peroxidation-induced mutations in several human fibroblast cell lines. Nearly identical patterns of oxidative base damage were obtained from reaction of DNA with either peroxidizing arachidonic acid (20:4omega6) or peroxyl radicals generated by thermolysis of ABIP in the presence of oxygen. Oxidative base damage primarily occurred at G and C. Transversions at GC base pairs in the supF gene were the major base substitution detected in all cell lines. Peroxyl radical induced tandem mutations were also observed. Many mutation hot spots coincided with sites of mapped oxidative lesions, although in some cases hot spots occurred adjacent to the damaged base. Evidence is presented for the involvement of 8-oxodG in the oxidation of DNA by ROO(*). These results are used to interpret some key features of previously published mutation spectra induced by lipid peroxidation in human cells.     

DNA adducts formed from 4-hydroxytamoxifen are more mutagenic than those formed by alpha-acetoxytamoxifen in a shuttle vector target gene replicated in human Ad293 cells

The drug tamoxifen, used to treat breast cancer, causes liver cancer in rats and endometrial cancer in women. Tamoxifen forms liver DNA adducts in both short- and long-term dosing of rodents, and DNA adducts have also been reported in tissues of women undergoing tamoxifen therapy. It is not known if the induction of endometrial cancer in women is through these DNA adducts or through the estrogenic nature of the drug. In this study, we have investigated the mutagenicity of two model reactive intermediates of tamoxifen, alpha-acetoxytamoxifen and 4-hydroxytamoxifen quinone methide (4-OHtamQM). These form the same DNA adducts as those found in tamoxifen-treated rats. The two compounds were used to treat the pSP189 plasmid containing the supF gene, which was replicated in Ad293 cells before being screened in indicator bacteria. Plasmid reacted with 4-OHtamQM was more likely to be mutated (2-7-fold increase) than that reacted with alpha-acetoxytamoxifen, despite having a lower level of DNA damage (12-20-fold less), as assayed by (32)P-postlabeling. The two compounds induced statistically different mutation spectra in the supF gene. The majority of mutations in alpha-acetoxytamoxifen-treated plasmid were GC -->TA transversions while GC-->AT transitions were formed in 4-OHtamQM-treated plasmid. 4-OHTamQM-treated DNA induced a larger proportion of multiple mutations and large deletions compared to alpha-acetoxytamoxifen. Sites of mutational hotspots were observed for both compounds. In conclusion, the quantitatively minor DNA adduct of tamoxifen (dG-N(2)-4-hydroxytamoxifen) is more mutagenic than the major tamoxifen DNA adduct (dG-N(2)-tamoxifen).