Altered CTP synthetase activity confers resistance to 5-bromodeoxyuridine toxicity and mutagenesis

A V79 Chinese hamster fibroblast cell line selected for resistance to the toxic effects of 5-fluorouracil (Kaufman, 1984b) was found to be cross-resistant to the toxic effects of the thymidine analog 5-bromodeoxyuridine (BrdUrd). When tested for sensitivity to BrdUrd mutagenesis, the fluorouracil-resistant cells were found to be resistant to mutagenesis induced by high concentrations of BrdUrd in the medium (INC mutagenesis) but not to mutagenesis induced by the replication of DNA containing 5-bromouracil (REP mutagenesis). Analyses of deoxyribonucleoside triphosphate pools indicated that high endogenous dCTP levels in the mutant prevented the high BrdUTP/dCTP ratio associated with INC mutagenesis. However, the mutant phenotype had no effect on the nucleotide pool imbalance associated with REP mutagenesis. This mutant provides further genetic evidence for the existence of two independent mechanisms for BrdUrd mutagenesis in mammalian cells.     

Analysis of mutagenesis and sister-chromatid exchanges induced by 5-bromo-2′-deoxyuridine in somatic hybrids derived from Syrian hamster melanoma cells and Chinese hamster ovary cells

Somatic cell hybrids were derived from the fusion of Chinese hamster ovary (CHO) cells and Syrian hamster melanoma cells (2E). These two cell lines had previously been shown to differ in their response to the induction of mutations and sister-chromatid exchanges (SCEs) by 5-bromo-2′-deoxyuridine (BrdUrd) (Kaufman, 1987). The parental cells and a number of representative, independent hybrid clones were tested for their response to both the INC and REP mutagenesis protocols. INC mutagenesis involves the incorporation of BrdUrd into DNA under conditions of deoxyribonucleoside triphosphate (dNTP) pool imbalance, while REP mutagenesis involves the replication of 5-bromouracil-substituted DNA in the presence of dNTP pool imbalance. When tested for the toxic effects of high concentrations of BrdUrd and for the induction of mutations by the INC protocol, the hybrid clones all expressed the 2E phenotype, i.e., sensitivity to relatively low concentrations of BrdUrd and thymidine for the induction of mutations, dNTP pool perturbation, and the toxic effects of BrdUrd. When the hybrid clones were tested for the induction of mutations and SCEs by the REP protocol, it was found that they expressed the 2E phenotype for the induction of mutations and the CHO phenotype for the induction of SCEs. Thus, various aspects of the 2E phenotype, such as high mutation frequencies associated with large dNTP pool perturbations, appeared to be dominantly expressed in the cell hybrids, while the lack of induction of SCEs by these mutagenic conditions in 2E cells was found to be a recessive characteristic.     

Analysis of mutagenesis and sister-chromatid exchanges induced by 5-bromo-2'-deoxyuridine in somatic hybrids derived from Syrian hamster melanoma cells and Chinese hamster ovary cells

Somatic cell hybrids were derived from the fusion of Chinese hamster ovary (CHO) cells and Syrian hamster melanoma cells (2E). These two cell lines had previously been shown to differ in their response to the induction of mutations and sister-chromatid exchanges (SCEs) by 5-bromo-2'-deoxyuridine (BrdUrd) (Kaufman, 1987). The parental cells and a number of representative, independent hybrid clones were tested for their response to both the INC and REP mutagenesis protocols. INC mutagenesis involves the incorporation of BrdUrd into DNA under conditions of deoxyribonucleoside triphosphate (dNTP) pool imbalance, while REP mutagenesis involves the replication of 5-bromouracil-substituted DNA in the presence of dNTP pool imbalance. When tested for the toxic effects of high concentrations of BrdUrd and for the induction of mutations by the INC protocol, the hybrid clones all expressed the 2E phenotype, i.e., sensitivity to relatively low concentrations of BrdUrd and thymidine for the induction of mutations, dNTP pool perturbation, and the toxic effects of BrdUrd. When the hybrid clones were tested for the induction of mutations and SCEs by the REP protocol, it was found that they expressed the 2E phenotype for the induction of mutations and the CHO phenotype for the induction of SCEs. Thus, various aspects of the 2E phenotype, such as high mutation frequencies associated with large dNTP pool perturbations, appeared to be dominantly expressed in the cell hybrids, while the lack of induction of SCEs by these mutagenic conditions in 2E cells was found to be a recessive characteristic.     

The role of deoxyribonucleotide metabolism in 5-bromo-2′-deoxyuridine mutagenesis in mammalian cells

The effects of deoxyribonucleoside triphosphate (dNTP) pool imbalance on the induction of mutations and siste-chromatid exchanges (SCEs) by 5-bromo-2′-deoxyuridine (BrdUrd) in mammalian cells is reviewed. The INC BrdUrd mutagenesis protocol involves the incorporation of BrdUrd into DNA under conditions of specific dNTP pool imbalance, while the REP BrdUrd mutagenesis protocol involves the replication of 5-bromouracil (BrUra)-substituted DNA in the presence of specific (but different) dNTP pool imbalance. Biochemical and genetic analyses of both the INC and REP mutagenesis protocols provided evidence that (1) INC mutagenesis resulted from errors of incorporation due to the mispairing of BrdUTP with a guanine residue in replicating DNA leading to GC to AT transitions and (2) REP mutagenesis resulted from errors of replication due to the mispairing of dGTP with a BrURA residue in replicating DNA leading to AT to GC transitions. Further analyses involving different cell lines has led to an hypothesis describing the role of mismatch repair in the induction of mutations and SCEs.     

Replication of DNA containing 5-bromouracil can be mutagenic in Syrian hamster cells.

A new protocol for inducing mutations in mammalian cells in culture by exposure to the thymidine analog 5-bromodeoxyuridine (BrdUrd) was established. This protocol, called "DNA-dependent" mutagenesis, involved the incorporation of BrdUrd into DNA under nonmutagenic conditions and the subsequent replication of the 5-bromouracil (BrUra)-containing DNA under mutagenic conditions but with no BrdUrd present in the culture medium. The mutagenic conditions were induced by allowing BrUra-containing DNA to replicate in the presence of high concentrations of thymidine. This generated high intracellular levels of dTTP and dGTP, causing nucleotide pool imbalance. The mutagenesis induced by this protocol was found to correlate with the level of BrUra substituted for thymine in DNA.     

Ribonucleotide reductase: A determinant of 5-bromodeoxyuridine mutagenesis in phage T4

Summary Mutagenesis by 5-bromodeoxyuridine (BrdUrd) can result from base-pairing errors either during replication of a BrdUrd-containing template or at the nucleotide incorporation step. Replication errors give rise predominantly to AT-to-GC transitions, while incorporation errors, in which 5-bromo-dUTP competes with dCTP at a template guanine site, should give rise to GC-to-AT transitions. The latter pathway should be sensitive to deoxyribonucleoside triphosphate (dNTP) pool fluctuations. Since dNTP pools are regulated through allosteric control of ribonucleotide reductase, the control of this enzyme should be a determinant of BrdUrd mutagenesis — if mutagenesis results largely from incorporation errors. Since T4 phage-encoded ribonucleotide reductase is insensitive to feedback inhibition, we established conditions under which phage DNA replication is dependent upon ribonucleotide reductase of the host, Escherichia coli. We examined BrdUrd mutagenesis of rII mutants known to revert to wild type either by AT-to-GC or GC-to-AT transition pathways. While both reversion pathways were stimulated under all conditions analyzed, the AT-to-GC pathway was stimulated more when the E. coli reductase was functioning, while the GC-to-AT pathway was more specifically enhanced when the T4 reductase was active. These results confirm that ribonucleotide reductase is a determinant of BrdUrd mutagenesis, but our observations, plus experiments showing that BrdUrd has relatively small effects upon dNTP pool sizes, indicate that the relationship between deoxyribonucleotide metabolism and BrdUrd mutagenesis is more complex than anticipated.     

UV-induced mutation in bacteriophage T4.

Two late gene am mutants of bacteriophage T4 that can be induced to revert by UV were crossed to a temperature-sensitive ligase mutant. In the double mutants, UV-induced reversion was eliminated at a semirestrictive temperature. When the single am mutants were irradiated and then allowed a single passage in a permissive host, the UV-induced reversion frequency was increased by 15- to 25-fold. This increased mutagenesis was also abolished by the presence of the ligase allele. When the UV-irradiated single am mutants multiply infected a permissive host, allowing multiplicity reactivation to occur, the induced reversion frequency was reduced similarly to the reduction in lethality. The mutagenesis that remained was again abolished by the presence of the ligase allele. It is concluded that UV induces mutations in phage T4 through the action of a pathway that includes polynucleotide ligase. The increase in mutation frequency after growth in a permissive host implies that mutagenesis can occur at more than one stage of the infection rather than only in an early stage before expression of the mutant genome. The process of multiplicity reactivation appears to be error-free since it overcomes lethal lesions without inducing new mutations.     

Uncoupling of the induction of mutations and sister-chromatid exchanges by the replication of 5-bromouracil-substituted DNA

The REP mutagenesis protocol, which involves the replication of 5-bromouracil (BrUra)-substituted DNA in the presence of deoxyribonucleoside triphosphate (dNTP) pool imbalance, has been shown to induce both mutations and sister-chromatid exchanges (SCEs) in Chinese hamster ovary (CHO) cells. However, when a Syrian hamster melanoma-derived cell line, called 2E, which was selected for its ability to replace all of the thymine residues in DNA with BrUra, was subjected to the REP mutagenesis protocol, the correlation between the induction of mutations and SCEs was no longer observed. The 2E cells were found to be much more sensitive to the induction of mutations by REP mutagenesis than were the CHO cells. This increased sensitivity to REP mutagenesis was found to correlate with increased perturbations of the dNTP pools that have been shown to be involved in the mutagenic mechanism of this protocol. In contrast, when the induction of SCEs by the REP protocol was measured, it was found that although a baseline level of SCEs was detected in 2E cells, no significant induction of SCEs due to dNTP pool perturbation was observed. It was shown that high levels of SCEs were readily induced in 2E cells by other agents, e.g. mitomycin C. A model, which discusses the fate of mismatched bases thought to be generated by the REP mutagenesis protocol as the determining factor for the induction of mutations of SCEs, is proposed to explain the uncoupling of mutagenesis and SCE induction in 2E cells.     

Induction of sister-chromatid exchanges by the replication of 5-bromouracil-substituted DNA under conditions of nucleotide-pool imbalance

The induction of sister-chromatid exchanges (SCEs) by the replication of 5-bromouracil(BrUra)-containing DNA under conditions of nucleotide-pool imbalance was investigated. A modification of a protocol developed for the induction of mutations under these conditions (E.R. Kaufman, Mol. Cell. Biol., 4, 2449-2454, 1984) was used. To induce SCEs, Chinese hamster ovary cells were grown under non-mutagenic conditions which allowed the uniform incorporation of BrUra into their DNA at specific levels of substitution for thymine residues (25, 50 and 75% BrUra substitution). After 4 and 5 days of growth, the cells, which had incorporated BrUra into their DNA, were washed free of 5-bromodeoxyuridine (BrdUrd) and provided with fresh culture medium supplemented with various concentrations of thymidine (10 microM to 3 mM) and no BrdUrd. The cells were allowed to replicate their BrUra-containing DNA under these conditions, in the absence of BrdUrd, for two rounds of DNA synthesis to achieve sister-chromatid differentiation, and second-division metaphases were scored for SCEs. The results of these studies indicated that the SCEs observed were proportional to the level of BrUra substituted for thymine in the cellular DNA, were induced by increasing concentrations of thymidine in the culture medium during replication of the BrUra-containing DNA, correlated well with the induction of mutations to thioguanine resistance and to ouabain resistance, correlated with increases in the intracellular levels of dTTP and dGTP generated by the high concentrations of thymidine. These findings provide direct evidence for the induction of SCEs by the replication of BrUra-containing DNA and for the importance of the pools of nucleoside triphosphate precursors for DNA replication in these processes. When the effects of 3-aminobenzamide, a potent inhibitor of poly(ADP-?ibose) synthesis, were tested, it was found that 3-aminobenzamide significantly increased SCEs, but it had no effect on mutations induced.     

Frameshift mutagenesis in Escherichia coli by reversible DNA intercalators: sequence specificity

The mutagenic potency of the simple reversible intercalators isopropyl-OPC (iPr-OPC) and 9-aminoacridine (9-AA) is assessed in E. coli using reversion assays based on plasmids derived from pBR322 carrying various frameshift mutations within the tetracycline resistance gene in repetitive sequences: +/- 2 frameshift mutations within alternating GC sequences; +/- 1 frameshift mutation at runs of guanines. The results obtained show that iPr-OPC and 9-AA have a sequence specificity for mutagenesis: they revert +1 and -1 frameshift mutations within runs of monotonous G:C base pairs. The precise determination of the size of a small restriction fragment which contains the mutation allowed us to demonstrate that reversion occurred by -1 deletions for the +1 frameshift mutations and by +1 additions for the -1 frameshift mutations. The possible relations of this specific reversion with the base sequence specificity of the mutagenesis are briefly discussed.     

Reversion of nonsense mutants induced by 4-nitroquinoline-1-oxide in Schizosaccharomyces pombe.

We have studied the reversion of 8 nonsense alleles located in 7 different genes of Schizosaccharomyces pombe using 4-nitroquinoline-1-oxide (NQO) as a mutagenic agent. The nonsense mutants of S. pombe have been classified according to their suppressibility by defined opal and ochre suppressors into a class of efficiently suppressed opal and a class of inefficiency suppressed ochre mutants. The UGA alleles tested all revert consistently with NQO, in agreement with the high specificity of this mutagen for G-residues reported for bacteria and yeast. The UAA alleles show a lack or a low level of reversion with NQO. This low level of reversion is due to the low level of non-G-specific transversions at A sites of the UAA triplet. Within each class of nonsense mutants the extent of induction is site-dependent. We conclude that NQO acts predominantly on G-residues in S. pombe.     

Imbalanced deoxyribonucleoside triphosphate pools and spontaneous mutation rates determined during dCMP deaminase-defective bacteriophage T4 infections

DNA precursor imbalances are known to be mutagenic in both eukaryotic and prokaryotic systems. Almost certainly, such mutagenesis involves competition between correctly and incorrectly base-paired precursors at replication sites. Since other factors may be involved, it is important to identify specific mutations induced by specific pool imbalances. Using bacteriophage T4, we have developed a system for such analysis. We prepare double mutants of T4; one mutation affects a phage-coded enzyme of deoxyribonucleoside triphosphate (dNTP) metabolism, while the second is an rII mutation known to revert along a specific pathway. We determine dNTP pools in infection by such a mutant and measure both the spontaneous reversion rate of the rII mutation and, in some cases, the nucleotide sequence at the mutant site. In this paper we analyze mutations induced by a deficiency of T4-encoded deoxycytidylate deaminase. This causes pools of 5-hydroxymethyl-dCTP to expand some 30-fold, while dTTP pools contract. This specifically stimulates AT-to-GC reversion. One of the four AT-to-GC reverters tested, rIIUV215, increases its reversion rate at least 1000-fold under these pool-imbalance conditions, while the other mutants tested show increases of only about 10-fold. Therefore, factors other than dNTP competition, including local DNA sequence environment, must be invoked to fully explain mechanisms of dNTP pool imbalance-induced mutagenesis. We discuss models for this, and we also report unexpected effects of the dCMP deaminase deficiency upon pools of ribonucleoside triphosphates.     

Properties of r Mutants of Bacteriophage T4 Photodynamically Induced in the Presence of Thiopyronin and Psoralen

About 4 x 10(-4)r mutants were induced per lethal hit, a frequency characteristic of weak mutagens. Collections of mutants produced in the presence of either dye were indistinguishable in most of their properties. The rII mutants differed sharply from spontaneous mutants in their mutational spectra (fine scale map distribution) and their reversion responses to specific mutagens. Few or none of the induced mutants were induced to revert with proflavine (sign mutants; reading frame shift mutants). A majority were induced to revert with base analogues (base pair substitution mutants), and about half of these also responded to the hydroxymethylcytosine-specific agent hydroxylamine. A large minority of the mutants reverted spontaneously but failed to respond either to proflavine or to base analogues. We believe these mutants, as well as some of the mutants which did respond to base analogues, to be transversions (base pair substitutions which reverse the purine-pyrimidine orientation).     

Unstable ascospore color mutants of Ascobolus immersus

Summary New unstable mutants of Ascobolus immersus involving the color or size of ascospores were sought among spontaneous mutants. Among the 34 unstable mutants isolated, 31 had white spores, 2 had pink spores and 1 had a large sized spores. The unstable mutants involve 11 loci whose mutation leads to white spores and 2 loci whose mutations give pink spores, among the 19 loci known to be implicated in this character; 1 locus is defined by only one large spore mutant. All these genes are localized on at least 7 different chromosomes. Unstable mutants of the same locus may correspond to several different sites, but the number of these sites is very limited.The frequency of unstable mutations was estimated: among 36 mutants belonging to 8 different genes, 20 were stable and 16 were unstable. This high frequency of unstable mutants is undoubtedly underestimated. The moment of reversion of 23 of these new mutants was also sought: 15 of them revert as does mutant B, previously studied, in the very young mycelium, at high temperature and with a reversion frequency of 0.004 to 0.34, according to the mutant; 5 of them revert as mutant 301, also previously studied, during the development of the fruit-body and with a frequency of 0.009 to 0.035; two of these mutants revert very early in the ascospore as soon as the first mitoses or in the very young mycelium at 22° C, with a very high reversion frequency that may reach 1.0; finally, the last mutant studied reverts in the fruit-body with a frequency reaching 0.40, but with modalities different from mutant 301. The mutants of the same locus may revert with different modalities. The same modality may correspond to different sites of the same gene.In unstable double-mutant strains involving two different genes, the reversion of one is independent of the reversion of the other, whether or not the reversion modailities of each mutant are identical.Results indicate the existence of inducers common to several unstable mutants which present the same modalities of reversion.These data support the previously formulated hypothesis of transposable elements.     

SOS mutator effect in E. coli mutants deficient in mismatch correction.

We have used bacteriophage lambda to characterize the mutator effect of the SOS response induced by u.v. irradiation of Escherichia coli. Mutagenesis of unirradiated phages grown in irradiated or unirradiated bacteria was detected by measuring forward mutagenesis in the immunity genes or reversion mutagenesis of an amber codon in the R gene. Relative to the wild-type, the SOS mutator effect was higher in E. coli mismatch correction-deficient mutants (mutH, mutL and mutS) and lower in an adenine methylation-deficient mutant ( dam3 ). We conclude that a large proportion of SOS-induced 'untargeted' mutations are removed by the methyl-directed mismatch correction system, which acts on newly synthesized DNA strands. The lower SOS mutator effect observed in E. coli dam mutants may be due to a selective killing of mismatch-bearing chromosomes resulting from undirected mismatch repair. The SOS mutator effect on undamaged lambda DNA, induced by u.v. irradiation of the host, appears to result from decreased fidelity of DNA synthesis.     

Saccharomyces cerevisiae mutants with enhanced induced mutation and altered mitotic gene conversion

We have developed a method to isolate yeast (Saccharomyces cerevisiae) mutants with enhanced induced mutagenesis based on nitrous acid-induced reversion of the ade2-42 allele. Six mutants have been isolated and designated him (high induced mutagenesis), and 4 of them were studied in more detail. The him mutants displayed enhanced reversion of the ade2-42 allele, either spontaneous or induced by nitrous acid, UV light, and the base analog 6-N-hydroxylaminopurine, but not by gamma-irradiation. It is worth noting that the him mutants turned out not to be sensitive to the lethal effects of the mutagens used. The enhancement in mutation induced by nitrous acid, UV light, and 6-N-hydroxylaminopurine has been confirmed in a forward-mutation assay (induction of mutations in the ADE1, ADE2 genes). The latter agent revealed the most apparent differences between the him mutants and the wild-type strain and was, therefore, chosen for the genetic analysis of mutants, him mutations analyzed behaved as a single Mendelian trait; complementation tests indicated 3 complementation groups (HIM1, HIM2, and HIM3), each containing 1 mutant allele. Uracil-DNA glycosylase activity was determined in crude cell extracts, and no significant differences between the wild-type and him strains were detected. Spontaneous mitotic gene conversion at the ADE2 locus is altered in him1 strains, either increased or decreased, depending on the particular heteroallelic combination. Genetic evidence strongly suggests him mutations to be involved in a process of mismatch correction of molecular heteroduplexes.     

Mutations as possible replication errors in bacteria growing under conditions of thymine deficiency

When Escherichia coli or Bacillus subtilis cells having inhibited thymidylate synthetase activity were incubated for a long time on solid medium supplemented with a limiting concentration of thymine or thymidine (0.1–0.3 μg/ml) most of them became mutants for one or more genetic markers. This “overall mutagenesis” was detected both in Thy^? bacteria and in prototrophs for thymine (Thy⁺) with thymidylate synthetase inhibited by the addition of 5-fluorodeoxyuridine (FUdR) to the growth medium. When thymine (or thymidine) was present in very low amounts (10^?3 μg/ml) or was totally absent, the efficiency of mutagenesis decreased some 100-fold. The solid growth medium is essential because it supports the filamentous cells grown under conditions of limiting thymine.For some of the mutants with identified deficiency their ability to revert under the action of different mutagens was studied. Most efficient was 5-bromouracil (BU). This reversion is the characteristic response of mutations due to AT → GC transitions. In addition to single mutants, many multiple mutants were induced. The repair-defective strain of E. coli pol A1^? and strains Rec A^? and Exr A^?, which are also defective in UV-induced mutagenesis, showed a high level of mutation induction under the conditions described. All these results are in accord with the hypothesis that overall low-thymine mutagenesis reflects the accumulation of replication errors in DNA under the conditions of a precursor deficiency.     

The efficiency and extent of mutagenic activity of some new mutagens of base-analogue type.

N4-Hydroxycytidine, 5-methyl-N4-hydroxydeoxycytidine and 2-amino-N6-hydroxyadenine were tested for their mutagenic activity in S. typhimurium and E. coli cells. Reversion analysis of different markers was applied in a plate-test system, and 2-aminopurine was used as a reference mutagen. (i) 2-Amino-N6-hydroxyadenine was the most potent mutagen. In some cases it gave more than 1000 colonies of revertants per plate. (ii) N6-Hydroxycytidine was the least specific mutagen. Almost all the tested markers were inducible to revert by this analogue. (iii) The mutagenic specificity of 5-methyl-N4-hydroxydeoxycytidine seemed to be opposite to that of 2-aminopurine. This suggests that the former can induce transition of CG to TA. (iv) A comparison of the mutagenic actions of N4-hydroxycytidine and 5-methyl-N4-hydroxy-deoxycytidine showed that deoxyriboside analogues are not necessarily more efficient mutagens than ribonucleosides. (v) No purine or pyrimidine deficiency was needed for mutagenesis to occur for any of the mutagens investigated. (vi) The results on bacteria with different repair abilities suggest that base-analogue mutagenesis (except perhaps for BrdUrd) occurs mainly during replication of nucleic acids containing substituted nucleosides with bi-functional specificity.     

Transfer-deficient mutants of the narrow-host-range plasmid R91-5 of Pseudomonas aeruginosa.

Three methods have been successful in the isolation of transfer-deficient mutants of the narrow-host-range R plasmid R91-5 of Pseudomonas aeruginosa: (i) selection for donor-specific phage resistance; (ii) direct screening after mutagenic treatment with either ethyl methane sulfonate or N-methyl-N'-nitro-N-nitrosoguanidine; (iii) in vitro mutagenesis of plasmid DNA by hydroxylamine followed by transformation and direct screening. The majority of transfer-deficient mutants were donor-specific phage resistant, supporting the view that sex pili and other surface components are essential for conjugal transfer (since the phages PRD1 and PR4 adsorb to these sites). Some of the transfer-deficient mutants were also unable to inhibit the replication of phage G101 or lost entry exclusion or both phenotypes. The ability to revert these pleiotropic mutants to wild type implicates the latter two functions in R91-5 transfer. Suppressor mutations in P. aeruginosa enabled the detection of suppressor-sensitive, transfer-deficient mutants. Such mutants should prove useful in conjugational complementation tests for the identification of the transfer cistrons of R91-5.     

Effect of bromodeoxyuridine on the proliferation and growth of ethyl methanesulfonate-exposed P3 cells: Relationship to the induction of sister-chromatid exchanges

Although sister-chromatid exchange (SCE) analysis is recognized as an indicator of exposure to DNA-damaging agents, the results of these analyses have been confounded by the use of bromodeoxyuridine (BrdUrd) to differentially label the sister chromatids. Not only does BrdUrd itself induce SCE, it also modulates the frequency of SCE induced by certain DNA-damaging agents. In order to examine this effect of BrdUrd on SCE frequency, an indirect method which lends itself to measurements both with and without BrdUrd was employed. Human teratocarcinoma-derived (P3) cells were exposed to ethyl methanesulfonate (EMS) and cultured with increasing concentrations of BrdUrd for lengths of time corresponding to one, two, and three generations of cell growth. At each time point, the distribution of nuclei among the phases of the cell-cycle and cell growth were evaluated for each concentration and chemical. A statistical model was employed which tested both for the main effects of chemicals and culture times and for interactions between these factors. Both EMS and BrdUrd significantly affected the percentages of nuclei within the cell-cycle. Exposure to EMS resulted in decreases in the percentages of nuclei in G0 + G1 and increases in the G2 + M compartment. Exposure to BrdUrd affected the size of the G0 + G1 compartment as well as the percentage of S-phase nuclei. Cell growth was reduced as a consequence of increasing EMS concentration and as a function of BrdUrd concentration; the effects of these chemicals were more readily apparent at the later time points. Most importantly, for both the cell-cycle kinetics data and the cell growth data, no evidence of an interaction between the effects of EMS and the effects of BrdUrd was detected statistically. These results may be interpreted to mean that while both EMS and BrdUrd affect the induction of SCE, under the conditions of this experiment, the effects are additive rather than interactive.Abbreviations: EMS, ethyl methanesulfonate - BrdUrd, bromodeoxyuridine - BrdUTP, bromodeoxyuridine triphosphate - dCTP, deoxycytidine triphosphate - SCE, sister-chromatid exchange - P3, human teratocarcinoma derived - HBSS, Hank's Balanced Salt Solution - HOUR, culture time - REP, replicate