Effects of Ape1 overexpression on cellular resistance to DNA-damaging and anticancer agents

In vitro biochemical studies indicate that Ape1 is the major mammalian enzyme responsible for repairing abasic lesions in DNA and a significant factor in the processing of specific 3'-replication-blocking termini. Toward addressing the role of Ape1 in cellular resistance to specific DNA-damaging and anticancer agents, we constructed a chinese hamster ovary (CHO) cell line, AA8-Ape1, that exhibits a 7-fold higher Ape1-dependent nuclease activity; this overexpression is abolished upon exposure to tetracycline (Tc). In comparison to the AA8 parental control, our data indicates that Ape1 activity is not rate-limiting for the repair of cytotoxic damages induced by the alkylating agent methyl methanesulfonate (MMS), the oxidizing agent hydrogen peroxide (H2O2), or ionizing radiation (IR). AA8-Ape1 cells did exhibit increased resistance to bleomycin following a chronic 3-day exposure, but not to more acute challenges of 1 h. Most notably, the AA8-Ape1 line displayed approximately 1.7-fold elevated resistance to the replication-blocking nucleoside analog dioxolane cytidine (L-OddC); this improved resistance was abrogated by the addition of Tc to the medium. These studies demonstrate that Ape1 is not rate-limiting in the repair of MMS- or H2O2-induced DNA damage, that Ape1 may dictate the sensitivity of bleomycin, depending on dosing scheme, and for the first time, that Ape1 can influence cellular resistance to the anticancer/antiviral antimetabolite L-OddC.

XRCC2 and XRCC3, new human Rad51-family members, promote chromosome stability and protect against DNA cross-links and other damages

The phenotypically similar hamster mutants irs1 and irs1SF exhibit high spontaneous chromosome instability and broad-spectrum mutagen sensitivity, including extreme sensitivity to DNA cross-linking agents. The human XRCC2 and XRCC3 genes, which functionally complement irs1 and irs1SF, respectively, were previously mapped in somatic cell hybrids. Characterization of these genes and sequence alignments reveal that XRCC2 and XRCC3 are members of an emerging family of Rad51-related proteins that likely participate in homologous recombination to maintain chromosome stability and repair DNA damage. XRCC3 is shown to interact directly with HsRad51, and like Rad55 and Rad57 in yeast, may cooperate with HsRad51 during recombinational repair. Analysis of the XRCC2 mutation in irs1 implies that XRCC2's function is not essential for viability in cultured hamster cells.

Phenotypic heterogeneity in nucleotide excision repair mutants of rodent complementation groups 1 and 4

Rodent ultraviolet light (UV)-sensitive mutant cells in complementation groups (CGs) 1 and 4 normally are known for their extraordinary (approximately 80-100 x) sensitivity to mitomycin C (MMC), although some CG1 mutants with reduced MMC sensitivity were previously reported (Stefanini et al. (1987) Cytotechnology 1, 91). We report here new CG1 and CG4 mutants with only 1.6-10 x wild-type MMC sensitivity despite low unscheduled DNA synthesis (UDS) levels. Mutant UV140, in UV CG4, has approximately 3.8 x the UV sensitivity of parental line AA8, approximately 1.6 x wild-type MMC sensitivity, wild-type X-ray and ethyl methanesulfonate (EMS) sensitivity, and is only slightly (approximately 1.4 x) hypermutable to 8-azaadenine resistance by UV light. It has moderately decreased incision of UV-damaged DNA, has moderately decreased removal of (6-4) photoproducts, and is profoundly deficient in UDS after UV. After UV, it shows abnormally decreased DNA synthesis and persistently decreased RNA synthesis. In addition a cell-free extract of this mutant displays strongly reduced nucleotide excision repair synthesis using DNA treated with N-acetoxy-acetyl-amino-fluorene (AAF). The extract selectively fails to complement extracts of group 1 and 4 mutants consistent with the notion that the affected proteins, ERCC1 and ERCC4, are part of the same complex and that mutations in one subunit also affect the other component. Mutant UV212 is a CG1 mutant with approximately 3.3 x wild-type UV and approximately 5-10 x wild-type MMC sensitivity, with profoundly deficient UDS and hypermutability (approximately 5.8 x) by UV. Mutant UV201, probably in CG1, is only slightly (approximately 1.5 x) UV-sensitive and has near wild-type (1.02X) UV mutability. These unusual group 1 and 4 mutants demonstrate that the unique UV and MMC sensitivity phenotypes displayed by these groups can be separated and support the idea that they are the result of distinct repair functions of the corresponding ERCC1 and ERCC4 genes: nucleotide excision repair for UV lesions and a separate repair pathway for removal of interstrand crosslinks.

ERCC4 (XPF) encodes a human nucleotide excision repair protein with eukaryotic recombination homologs

ERCC4 is an essential human gene in the nucleotide excision repair (NER) pathway, which is responsible for removing UV-C photoproducts and bulky adducts from DNA. Among the NER genes, ERCC4 and ERCC1 are also uniquely involved in removing DNA interstrand cross-linking damage. The ERCC1-ERCC4 heterodimer, like the homologous Rad10-Rad1 complex, was recently found to possess an endonucleolytic activity that incises on the 5' side of damage. The ERCC4 gene, assigned to chromosome 16p13.1-p13.2, was previously isolated by using a chromosome 16 cosmid library. It corrects the defect in Chinese hamster ovary (CHO) mutants of NER complementation group 4 and is implicated in complementation group F of the human disorder xeroderma pigmentosum. We describe the ERCC4 gene structure and functional cDNA sequence encoding a 916-amino-acid protein (104 kDa), which has substantial homology with the eukaryotic DNA repair and recombination proteins MEI-9 (Drosophila melanogaster), Rad16 (Schizosaccharomyces pombe), and Rad1 (Saccharomyces cerevisiae). ERCC4 cDNA efficiently corrected mutants in rodent NER complementation groups 4 and 11, showing the equivalence of these groups, and ERCC4 protein levels were reduced in mutants of both groups. In cells of an XP-F patient, the ERCC4 protein level was reduced to less than 5%, consistent with XPF being the ERCC4 gene. The considerable identity (40%) between ERCC4 and MEI-9 suggests a possible involvement of ERCC4 in meiosis. In baboon tissues, ERCC4 was expressed weakly and was not significantly higher in testis than in nonmeiotic tissues.

Genomic sequence comparison of the human and mouse XRCC1 DNA repair gene regions

The XRCC1 (X-ray repair cross complementing) gene is involved in the efficient repair of DNA single-strand breaks formed by exposure to ionizing radiation and alkylating agents. The human gene maps to chromosome 19q13.2, and the mouse homologue maps to the syntenic region on chromosome 7. Two cosmids (approximately 38 kb each) containing the human and mouse genes were sequenced to an average 8-fold clonal redundancy. The XRCC1 gene spans a genomic distance of 26 kb in mouse and 31.9 kb in human. Both genes contain 17 exons, are 84% identical within the coding regions, and are 86% identical at the amino acid sequence level. Intron and exon lengths are highly conserved. For the human cosmid, a total of 43 Alu repetitive elements are present, a density of 1.1 Alu/kb, but due to clustering, the local density is as high as 1.8 Alu/kb. In addition, we observed a statistically significant bias for insertion of these elements in the 3'-5' orientation relative to the direction of XRCC1 transcription, predominantly in the second and third introns. This bias may indicate that XRCC1 is more accessible to Alu retroposition events during transcription than genes not expressed during spermatogenesis. The density of B1 and B2 elements in the mouse is 0.4/kb, integrated primarily in the 5'-3' orientation. The human chromosome 19-specific minisatellite PE670 was present in the same orientation in 3 introns in the human gene, and a similar repeat was found at 3 different locations in the mouse cosmid.(ABSTRACT TRUNCATED AT 250 WORDS)

Molecular cloning of the human nucleotide-excision-repair gene ERCC4

ERCC4 was previously identified in somatic cell hybrids as a human gene that corrects the nucleotide-excision-repair deficiency in mutant hamster cells. The cloning strategy for ERCC4 involved transfection of the repair-deficient hamster cell line UV41 with a human sCos-1 cosmid library derived from chromosome 16. Enhanced UV resistance was seen with one cosmid-library transformant and two secondary transformants of UV41. Cosmid clones carrying a functional ERCC4 gene were isolated from a library of a secondary transformant by selecting in Escherichia coli for expression of a linked neomycin-resistance gene that was present in the sCos-1 vector. The cosmids mapped to 16p13.13-p13.2, the location assigned to ERCC4 by using somatic cell hybrids. Upon transfection into UV41, six cosmid clones gave partial correction ranging from 30% to 64%, although all appeared to contain the complete gene. The capacity for in vitro excision of thymine dimers from a plasmid by transformant cell extracts correlated qualitatively with enhanced UV resistance.

Isolation and characterization of mouse Xrcc-1, a DNA repair gene affecting ligation

Human DNA repair gene XRCC1 complements the strand-break rejoining defect in Chinese hamster mutant EM9 and encodes a protein that is apparently required for optimal activity of DNA ligase III. Toward the goal of producing transgenic mice that carry a mutation in the Xrcc-1 locus, the murine homolog of XRCC1 was cloned from both cosmid genomic and cDNA libraries. Upon transfection into EM9 cells, cosmids containing the functional mouse gene efficiently corrected (94-100%) the high sister-chromatid-exchange defect. Mouse Xrcc-1 is 26 kb in length, contains 17 exons, and maps by metaphase in situ hybridization to the 7A3-7B2 region of mouse chromosome 7. Isolated cDNA clones were highly truncated and were extended by anchored polymerase chain reactions. The 1893-bp open reading frame of mouse Xrcc-1 encodes 631 amino acids, compared with 633 for the human homolog. The predicted mouse Xrcc-1 protein of 69.1 kDa and pI of 5.95 is 86% identical and 93% similar to human XRCC1.

Molecular cloning of the human XRCC1 gene, which corrects defective DNA strand break repair and sister chromatid exchange

We describe the cloning and function of the human XRCC1 gene, which is the first mammalian gene isolated that affects cellular sensitivity to ionizing radiation. The CHO mutant EM9 has 10-fold-higher sensitivity to ethyl methanesulfonate, 1.8-fold-higher sensitivity to ionizing radiation, a reduced capacity to rejoin single-strand DNA breaks, and a 10-fold-elevated level of sister chromatid exchange compared with the CHO parental cells. The complementing human gene was cloned from a cosmid library of a tertiary transformant. Two cosmid clones produced transformants that showed approximately 100% correction of the repair defect in EM9 cells, as determined by the kinetics of strand break repair, cell survival, and the level of sister chromatid exchange. A nearly full-length clone obtained from the pcD2 human cDNA expression library gave approximately 80% correction of EM9, as determined by the level of sister chromatid exchange. Based on an analysis of the nucleotide sequence of the cDNA insert compared with that of the 5' end of the gene from a cosmid clone, the cDNA clone appeared to be missing approximately 100 bp of transcribed sequence, including 26 nucleotides of coding sequence. The cDNA probe detected a single transcript of approximately 2.2 kb in HeLa polyadenylated RNA by Northern (RNA) blot hybridization. From the open reading frame and the positions of likely start sites for transcription and translation, the size of the putative XRCC1 protein is 633 amino acids (69.5 kDa). The size of the XRCC1 gene is 33 kb, as determined by localizing the endpoints on a restriction endonuclease site map of one cosmid clone. The deduced amino acid sequence did not show significant homology with any protein in the protein sequence data bases examined.

Recent progress with the DNA repair mutants of Chinese hamster ovary cells

Repair-deficient mutants of Chinese hamster ovary (CHO) cells are being used to identify human genes that correct the repair defects and to study mechanisms of DNA repair and mutagenesis. Five independent tertiary DNA transformants were obtained from the EM9 mutant, which is noted for its very high sister-chromatid exchange frequencies. In these clones a human DNA sequence was identified that correlated with the resistance of the cells to chlorodeoxyuridine (CldUrd). After EcoRI digestion, Southern transfer, and hybridization of transformant DNAs with the BLUR-8 Alu family sequence, a common fragment of 25-30 kilobases (kb) was present. Since the DNA molecules used to produce these transformants were sheared to less than 50 kb in size, the correcting gene should be small enough to clone in a cosmid vector. Using drug-resistance markers to select for hybrids after fusion, we have done complementation experiments with ultraviolet light (u.v.)-sensitive mutants and have identified a sixth complementation group, line UV61. Additionally, CHO mutants UV27-1 and MMC-2, isolated in other laboratories, were found to belong to UV group 3, which is represented by line UV24. To study the behaviour of transfected DNA molecules in repair-deficient cells, we treated plasmid pSV2gpt with either u.v. radiation or cis-diamminedichloroplatinum(II) (cis-DDP) and introduced the damaged DNA into normal CHO cells (AA8) and mutants UV4 and UV5. Unrepaired damage to the plasmid was indicated by loss of colony-forming ability of the transfected cells in selective medium containing mycophenolic acid. With u.v. damage, the differential survival of the cell lines was similar to that seen when whole cells are treated with u.v. However, with cis-DDP damage, mutant UV4 did not exhibit the extreme hypersensitivity (50-fold) that occurs when cells are treated. This result suggests that UV4 cells may be able to repair cross-links in transfected DNA.

Genetic complementation between UV-sensitive CHO mutants and xeroderma pigmentosum fibroblasts

The purpose of this study was to determine the feasibility of doing complementation analysis between DNA-repair mutants of CHO cells and human fibroblasts based on the recovery of hybrid cells resistant to DNA damage. Two UV-sensitive CHO mutant lines, UV20 and UV41, which belong to different genetic complementation groups, were fused with fibroblasts of xeroderma pigmentosum in various complementation groups. Selection for complementing hybrids was performed using a combination of ouabain to kill the XP cells and mitomycin C to kill the CHO mutants. Because the frequency of viable hybrid clones was generally less than 10(-6) and the frequency of revertants of each CHO mutant was approximately 2 X 10(-7), putative hybrids required verification. The hybrid character of clones was established by testing for the presence of human DNA in a dot-blot procedure. Hybrid clones were obtained from 9 of the 10 different crosses involving 5 complementation groups of XP cells. The 4 attempted crosses with 2 other XP groups yielded no hybrid colonies. Thus, a definitive complementation analysis was not possible. Hybrids were evaluated for their UV resistance using a rapid assay that measures differential cytotoxicity (DC). All 9 hybrids were more resistant than the parental mutant CHO and XP cells, indicating that in each case complementation of the CHO repair defect by a human gene had occurred. 3 hybrids were analyzed for their UV-radiation survival curves and shown to be much more resistant that the CHO mutants but less resistant than normal CHO cells. With 2 of these hybrids, sensitive subclones, which had presumably lost the complementing gene, were found to have similar sensitivity to the parental CHO mutants. We conclude that the extremely low frequency of viable hybrids in this system limits the usefulness of the approach. The possibility remains that each of the nonhybridizing XP strains could be altered in the same locus as one of the CHO mutants.

Comparative mutagenic efficiencies of the DNA adducts from the cooked-food-related mutagens Trp-P-2 and IQ in CHO cells

The relationship between DNA-adduct formation and mutagenicity of two heterocyclic aromatic amines associated with cooked foods was determined in a CHO cell strain lacking nucleotide excision repair. Cells were exposed to tritiated IQ (2-amino-3-methylimidazo[4,5-f]quinoline) or Trp-P-2 (3-amino-1-methyl-5H-pyrido[4,3-b]indole) supplemented with hamster S9 microsomal fraction for metabolic activation. DNA from nuclei was isolated by DNAase-mediated elution from polycarbonate filters after RNAase and proteinase treatment. The presumed metabolites of both compounds bound to DNA in a dose-dependent fashion. Although the dose required to produce 50% cell killing was 15 times higher for IQ than Trp-P-2, the amount of radioactive material bound to DNA at that dose was about 10-fold lower with IQ. When mutations at the hprt and aprt loci were compared with the estimated levels of adducts, the calculated mutagenic efficiency of the adducts was about 4 mutations per 1000 adducts for both compounds, assuming a target sequence of 1000 base pairs for either locus. We conclude that IQ is acting as a weak mutagen in this system because its extracellular metabolites either do not reach or do not react efficiently with the DNA of the CHO cells.

DNA-mediated transfer of a human DNA repair gene that controls sister chromatid exchange

The Chinese hamster cell line mutant EM9, which has a reduced ability to repair DNA strand breaks, is noted for its highly elevated frequency of sister chromatid exchange, a property shared with cells from individuals with Bloom's syndrome. The defect in EM9 cells was corrected by fusion hybridization with normal human fibroblasts and by transfection with DNA from hybrid cells. The transformants showed normalization of sister chromatid exchange frequency but incomplete correction of the repair defect in terms of chromosomal aberrations produced by 5-bromo-2'-deoxyuridine.

Repair of DNA adducts in asynchronous CHO cells and the role of repair in cell killing and mutation induction in synchronous cells treated with 7-bromomethylbenz[a]anthracene

CHO cells of normal or UV-sensitive phenotypes were analyzed for their ability to remove DNA adducts produced by the carcinogen 7-BrMeBA. At a dose of 0.1 microM, which reduced the survival of the normal AA8 cells to approximately 90% and the mutant UV5 cells to approximately 20%, the frequency of adducts was 5-6 per 10(6) nucleotides for both cell types, and AA8 cells removed approximately 30% of the adducts in 8 h and approximately 55% in 24 h. In contrast, UV5 and mutants from four other genetic complementation groups had no significant removal. Binding of 7-BrMeBA did not vary through the cell cycle in synchronous cultures. At a dose of mutagen (0.07 microM) resulting in approximately 25% survival of asynchronous UV5, the survival of synchronous cultures rose about threefold from early G1 to early S phase and then decreased somewhat in late S/G2. At a dose (0.28 microM) producing similar survival of asynchronous cultures, AA8 cells differed qualitatively in that survival decreased progressively by 5- to 10-fold between early G1 and the early part of S, and rose steeply through late S/G2 to give a 10- to 20-fold increase. We conclude that DNA repair is the major determinant of variations in survival through the cycle in normal cells. The patterns observed are consistent with a mechanism of killing in AA8 cells in which adducts disrupt DNA replication, while in UV5 cells transcriptional blocks or other effects may govern lethality. Induced mutations at the aprt and hprt loci showed changes through the cycle in both AA8 and UV5 cells, and the patterns were not readily explainable by the action of repair.

Hypersensitivity to cell killing and mutation induction by chemical carcinogens in an excision repair-deficient mutant of CHO cells

A strain of Chinese hamster ovary cells that is deficient in nucleotide excision repair, strain UV5, was compared with the normal parental CHO cells in terms of cytotoxicity and mutagenesis after exposure to several chemical carcinogens that are known to produce bulky, covalent adducts in DNA. Induced mutations were measured at the hprt locus using thioguanine resistance and at the aprt locus using azaadenine resistance. The compounds tested that required metabolic activation (using rat or hamster microsomal fractions) were 7,12-dimethylbenz(a)anthracene, 3-methylcholanthrene, benzo(a)pyrene, aflatoxin B1, 2-acetylaminofluorene, and 2-naphthylamine. The direct-acting compounds (+/-)-r-7,t-8-dihydroxy-t-9,10-epoxy-7,8,9,10-tetrahydrobenzo(a)pyrene, N-acetoxy-2-acetylaminofluorene, and N-OH-2-naphthylamine were also studied. For all compounds except 2-naphthylamine and its active metabolite, the repair-deficient cells were significantly more sensitive to killing than the normal CHO cells. Mutation induction at both loci was also more efficient in UV5 cells in each instance where enhanced cytotoxicity was observed. By using tritium-labeled N-acetoxy-2-acetylaminofluorene, normal and mutant cells were shown to bind mutagen to their nuclear DNA with similar efficiency, and a greater amount of adduct removal occurred in the normal cells. From this study it is concluded that the use of excision repair-deficient CHO cells provides enhanced sensitivity for detecting mutagenesis and that a positive differential cytotoxicity response gives an indication of repairable, potentially lethal genetic damage.

Hypersensitivity to mutation and sister-chromatid-exchange induction in CHO cell mutants defective in incising DNA containing UV lesions

Five UV-sensitive mutant strains of CHO cells representing different genetic complementation groups were analyzed for their ability to perform the incision step of nucleotide excision repair after UV exposure. The assay utilized inhibitors of DNA synthesis to accumulate the short-lived strand breaks resulting from repair incisions. After 6 J/m2, each of the mutants showed less than 10% of the incision rate of the parental AA8 cells. After 50 J/m2, the rate in AA8 was similar to that at 6 J/m2, but the rates in the mutants were significantly higher (approximately 20% of the rate of AA8). Thus by this incision assay the mutants were phenotypically indistinguishable. Each of the mutants were hypersensitive to mutation induction at both the hprt and aprt loci by a factor of 10, and in the one strain tested ouabain resistance was induced sevenfold more efficiently than in AA8 cells. Sister chromatid exchange was also induced with sevenfold increased efficiency in the two mutant strains examined. Thus, these CHO mutants resemble xeroderma pigmentosum cells in terms of their incision defects and their hypersensitivity to DNA damage by UV.

A CHO-cell strain having hypersensitivity to mutagens, a defect in DNA strand-break repair, and an extraordinary baseline frequency of sister-chromatid exchange

A mutant of CHO cells (strain EM9) previously isolated on the basis of hypersensitivity to killing by ethyl methanesulfonate (EMS) is approx. 10-fold more sensitive than the parental line, AA8, to killing by both EMS and MMS. It is also hypersensitive to killing by other alkylating agents (ethyl nitrosourea and N-methyl-N'-nitro-N-nitrosoguanidine), X-rays, and ultraviolet radiation. The production and repair of DNA single-strand breaks (SSB) were studied using the technique of alkaline elution of DNA from filters. After exposure to 4 Gy of X-rays at 0 degrees C and subsequent incubation at 25 degrees C, SSB were repaired within 12 min in AA8, but little repair occurred in EM9. Similarly, with doses of EMS or MMS that produced comparable numbers of SSB in AA8 and EM9 at the end of a 10-min exposure, repair of SSB occurred more rapidly in AA8 than in EM9, suggesting that individual SSB are longer lived in EM9. EM9 was found to be hypersensitive also to the induction of mutations and sister-chromatid exchanges (SCE) by EMS; per unit dose the mutant had twice as many mutations to thioguanine resistance, 3 times as many mutations to azaadenine resistance, and a 7-fold enhancement in SCE, compared to AA8. Moreover, the baseline frequency of SCE in the mutant was extraordinarily high, i.e., 8.6 +/- 0.6 vs. 107 +/- 5 SCE/cell for AA8 and EM9, respectively, with 10 microM BrdUrd in the medium. The high SCE frequency in EM9 did not vary significantly with BrdUrd concentration over the range examined from 2.5 to 20 microM, and the percentage of 5-bromouracil substitution in the DNA was the same in EM9 and AA8 under these conditions. These data, however, do not rule out the possibility that the high SCE frequency in EM9 is a consequence of an altered sensitivity to incorporated BrdUrd. Thus, EM9 may carry a pleiotropic mutation affecting some function in DNA replication and/or DNA repair and causing the variety of phenotypic properties described in this study.

Role of DNA repair in mutagenesis of Chinese hamster ovary cells by 7-bromomethylbenz[a]anthracene

The role of DNA repair in mutagenesis was studied in normal, repair-proficient Chinese hamster ovary cells and in two mutant strains that are deficient in excision repair. By using the mutagen 7-bromomethylbenz[a]anthracene (7-BrMeBA) and the technique of alkaline elution of DNA, the mutants were found to be defective at or before the incision step of excision repair. Dose--responses were determined for cell killing, mutation induction at three loci, and sister chromatid exchanges over a survival range of 1.0--0.1 after 7-BrMeBA treatment. The mutants were 5-fold more sensitive to killing than were the normal cells, but the degree of hypersensitivity to mutation induction varied depending on the mutant strain, the genetic marker, and the dose of mutagen. In each instance, the dose--response curve for mutations was essentially linear in the repair-deficient cells. In the normal cells, however, the curves for induced resistance to thioguanine and azaadenine were complex and were curvilinear with increasing slope at low doses. This behavior may be attributable to saturation of the excision repair system. No difference was seen in the efficiency of inducing ouabain-resistant mutations in the repair-deficient cells compared to the normal cells, indicating a qualitatively different behavior of this marker. These results are consistent with excision repair of 7-BrMeBA damage being error-free in Chinese hamster ovary cells. Sister chromatid exchange, another manifestation of DNA damage, also was induced with greater efficiency in the repair-deficient cells.

Failure of the phorbol ester 12-O-tetradecanoylphorbol-13-acetate to enhance sister chromatid exchange, mitotic segregation, or expression of mutations in Chinese hamster cells

The potent tumor promoter 12-O-tetradecanoylphorbol-13-acetate (TPA) was tested for its ability (a) to induce sister chromatid exchange, (b) to increase the rate of transition at the adenine phosphoribosyltransferase (apt) locus from the presumptive heterozygous state ((+/- to the homozygous state (-/ - or -), and (c) to enhance the frequency of mutations expressed after ultraviolet radiation mutagenesis. We have found no significant effect of TPA in any of these experiments. Sister chromatid exchange frequencies in both V79 and Chinese hamster ovary cells remained unchanged by TPA treatment under various conditions, a result inconsistent with the hypothesis that an important effect of TPA might be to increase the rate of chromosomal mitotic recombination (and hence segregation of recessive mutations) in a manner akin to increased chromatid recombination. We have also been unable to obtain evidence for mitotic recombination affecting the aprt locus in Chinese hamster ovary cells for which the rate of change to a high level of resistance to azaadenine was measured. The rate of 8.6 X 10(-7) mutation (and/or segregations) per cell generation assessed by fluctuation analysis was not increased by the continuous presence of TPA, 4 microgram/ml, in the medium. In the third set of experiments, mutant frequencies in Chinese hamster ovary cells after ultraviolet mutagenesis were measured for the markers ouabain resistance, thioguanine resistance, and azaadenine resistance, under conditions with and without pretreatment with TPA before mutant selection. No convincing enhancement in mutation expression was observed. In summary, these results argue that promotion by TPA does not proceed by a mechanism involving genetic recombination or the altered expression of newly mutated alleles.