Mutagen-induced recombination in mammalian cells in vitro

Radiation-induced molecular changes in rat mammary tissue: possible implications for radiation-induced carcinogenesis

PURPOSE

Ionizing radiation is a potent mammary gland carcinogen, yet the exact molecular etiology of radiation-induced breast cancer remains unknown.

MATERIALS AND METHODS

Our study utilized a rat model of breast carcinogenesis to analyse the molecular and epigenetic changes induced in mammary gland tissue upon exposure to ionizing radiation (IR). Using a methylation-sensitive cytosine extension assay we studied the IR-induced changes in DNA methylation. In parallel, we analysed the expression of proteins involved in DNA methylation, DNA repair and cell proliferation control. Molecular changes were related to cellular proliferation and apoptosis.

RESULTS

We found that IR led to a loss of genomic cytosine methylation in the exposed mammary tissue. Global DNA hypomethylation was paralleled by reduction in the levels of maintenance (DNMT1) and de novo (DNMT3a and 3b) DNA methyltransferases and methyl-binding protein MeCP2. The observed DNA hypomethylation was linked, at least in part, to activation of DNA repair processes. Concurrently, we observed increased levels of phosphorylated extracellular signal-regulated kinase (p-ERK1/2), phosphorylated AKT kinase (p-AKT), cyclin D1 and proliferating cells nuclear antigen (PCNA) proteins, suggesting IR alters intra-cellular signaling and cell cycle control mechanisms in mammary tissue. We also noted a significant induction of apoptosis in the exposed tissue 6 hours after irradiation. The observed apoptosis levels were paralleled by the slight elevation of cellular proliferation.

CONCLUSIONS

We have demonstrated that a single exposure to 5 Gy of X rays leads to noticeable epigenetic changes in the rat mammary gland that occurred in the context of activation of DNA damage repair and alterations in the pro-survival growth-stimulatory cellular signaling pathways. The possible cellular repercussions of the observed changes in relationship to breast carcinogenesis are discussed.

Interstrand crosslink-induced homologous recombination carries an increased risk of deletions and insertions

Homology directed repair (HDR) defends cells against the toxic effects of two-ended double strand breaks (DSBs) and one-ended DSBs that arise when replication progression is inhibited, for example by encounter with DNA lesions such as interstrand crosslinks (ICLs). HDR can occur via various mechanisms, some of which are associated with an increased risk of concurrent sequence rearrangements that can lead to deletions, insertions, translocations and loss of heterozygosity. Here, we compared the risk of HDR-associated sequence rearrangements that occur spontaneously versus in response to exposure to an agent that induces ICLs. We describe the creation of two fluorescence-based direct repeat recombination substrates that have been targeted to the ROSA26 locus of embryonic stem cells, and that detect the major pathways of homologous recombination events, e.g., gene conversions with or without crossing over, repair of broken replication forks, and single strand annealing (SSA). SSA can be distinguished from other pathways by application of a matched pair of site-specifically integrated substrates, one of which allows detection of SSA, and one that does not. We show that SSA is responsible for a significant proportion of spontaneous homologous recombination events at these substrates, suggesting that two-ended DSBs are a common spontaneous recombinogenic lesion. Interestingly, exposure to mitomycin C (an agent that induces ICLs) increases the proportion of HDR events associated with deletions and insertions. Given that many chemotherapeutics induce ICLs, these results have important implications in terms of the risk of chemotherapy-induced deleterious sequence rearrangements that could potentially contribute to secondary tumors.

"Recombomice": the past, present, and future of recombination-detection in mice

Homology directed repair (HDR) provides an efficient strategy for repairing and tolerating many types of DNA lesions, such as strand breaks, base damage, and crosslinks. Recombinational repair and lesion avoidance pathways that involve homology searching are integral to normal DNA replication. Indeed, it is estimated that at least ten HDR events take place each time a mammalian cell divides. HDR is associated with the transfer and exchange of DNA sequences. Usually, homologous sequences are aligned perfectly and flanking sequences are not exchanged. However, those sequence misalignments and exchanges that do occur can lead to rearrangements that contribute to cancer (e.g. deletions, inversions, translocations or loss of heterozygosity (LOH)). In order to reveal genetic and environmental factors that modulate HDR in mammals, several approaches have been used to detect recombination events in vivo. Here, we briefly review three methods for detecting homologous recombination in mice, namely: sister chromatid exchange (SCE), LOH, and recombination at tandem repeats. We conclude with a more detailed description of the recently developed "Fluorescent Yellow Direct Repeat" (FYDR) mouse model, which exploits enhanced yellow fluorescent protein (EYFP) for detecting mitotic homologous recombination in vivo. Applications of the FYDR mice are described, as well as the broader potential for using fluorescent proteins to detect recombination in various tissues/cell types in vivo.

Repair of DNA interstrand crosslinks may take place at the nuclear matrix

Host cell reactivation assay using Trioxsalen-crosslinked plasmid pEGFP-N1 showed that human cells were able to repair Trioxsalen interstrand crosslinks (ICL). To study the mechanism of this repair pathway, cells were transfected with the plasmids pEGFP-1, which did not contain the promoter of the egfp gene, and with pEGFP-G-, which did not contain the egfp gene. Neither of these plasmids alone was able to express the green fluorescent protein. After cotransfection with the two plasmids, 1%-2% of the cells developed fluorescent signal, which showed that recombination events had taken place in these cells to create DNA constructs containing the promoter and the gene properly aligned. When one or both of the plasmids were crosslinked with Trioxsalen, the recombination rate increased several fold. To identify the nuclear compartment where recombination takes place, cells were transfected with crosslinked pEGFP-N1 and the amount of plasmid DNA in the different nuclear fractions was determined. The results showed that Trioxsalen crosslinking increased the percentage of matrix attached plasmid DNA in a dose-dependent way. Immunoblotting experiments showed that after transfection with Trioxsalen crosslinked plasmids the homologous recombination protein Rad51 also associated with the nuclear matrix fraction. These studies provide a model system for investigating the precise molecular mechanisms that appear to couple repair of DNA ICL with nuclear matrix attachment.

In vivo Recombination After Chronic Damage Exposure Falls to Below Spontaneous Levels in “Recombomice”

All forms of cancer are initiated by heritable changes in gene expression. Although point mutations have been studied extensively, much less is known about homologous recombination events, despite its role in causing sequence rearrangements that contribute to tumorigenesis. Although transgenic mice that permit detection of point mutations have provided a fundamental tool for studying point mutations in vivo, until recently, transgenic mice designed specifically to detect homologous recombination events in somatic tissues in vivo did not exist. We therefore created fluorescent yellow direct repeat mice, enabling automated detection of recombinant cells in vivo for the first time. Here, we show that an acute dose of ionizing radiation induces recombination in fluorescent yellow direct repeat mice, providing some of the first direct evidence that ionizing radiation induces homologous recombination in cutaneous tissues in vivo. In contrast, the same total dose of radiation given under chronic exposure conditions suppresses recombination to levels that are significantly below those of unexposed animals. In addition, global methylation is suppressed and key DNA repair proteins are induced in tissues from chronically irradiated animals (specifically AP endonuclease, polymerase β, and Ku70). Thus, increased clearance of recombinogenic lesions may contribute to suppression of homologous recombination. Taken together, these studies show that fluorescent yellow direct repeat mice provide a rapid and powerful assay for studying the recombinogenic effects of both short-term and long-term exposure to DNA damage in vivo and reveal for the first time that exposure to ionizing radiation can have opposite effects on genomic stability depending on the duration of exposure.

TCDD-induced homologous recombination: the role of the Ah receptor versus oxidative DNA damage

The environmental toxicant 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) elicits numerous biological responses including carcinogenicity. The molecular mechanism by which TCDD exerts its tumorigenic effects is unclear, since it does not directly damage DNA. TCDD-initiated toxicity can be mediated by the aryl hydrocarbon receptor (AhR) pathway and/or via increased oxidative stress. DNA damage, including DNA oxidation, can induce DNA double-strand breaks, which can be repaired through homologous recombination. Excessive DNA double-strand breaks may promote aberrant DNA recombination, which can lead to detrimental genetic changes and ultimately to carcinogenesis. TCDD has been shown to induce homologous recombination but the molecular mechanism mediating these events are unknown. To investigate the role of the AhR and oxidative DNA damage in mediating TCDD-induced homologous recombination we used a Chinese hamster ovary (CHO) cell line containing a neo direct repeat recombination substrate (CHO 3-6). CHO 3-6 cells were exposed to TCDD (50, 500 or 1000 pM) in the presence or absence of an AhR antagonists (0.1 microM alpha-naphthoflavone (alpha-NF)) for 6 or 24 h and 2 weeks later homologous recombination frequencies were determined by counting the number of neo expressing, G418-resistant colonies per live cells plated. TCDD-initiated DNA oxidation was determined by measuring the formation of 8-hydroxy-2'-deoxyguanosine via HPLC and electrochemical detection. Exposure to 500 pM TCDD for 24 h significantly increased the frequency of homologous recombination. Southern blot analysis on G418-resistant colonies determined that TCDD induced both conservative gene conversion events and deletion events. DNA oxidation was not increased in cells exposed to TCDD for either 6 or 24 h. However, alpha-naphthoflavone exposure resulted in a significant decrease in TCDD-induced homologous recombination frequency. These results suggest that TCDD-initiated homologous recombination in CHO 3-6 cells is mediated by the AhR and not via increased oxidative stress.

Spontaneous mitotic homologous recombination at an enhanced yellow fluorescent protein (EYFP) cDNA direct repeat in transgenic mice

A transgenic mouse has been created that provides a powerful tool for revealing genetic and environmental factors that modulate mitotic homologous recombination. The fluorescent yellow direct-repeat (FYDR) mice described here carry two different copies of expression cassettes for truncated coding sequences of the enhanced yellow fluorescent protein (EYFP), arranged in tandem. Homologous recombination between these repeated elements can restore full-length EYFP coding sequence to yield a fluorescent phenotype, and the resulting fluorescent recombinant cells are rapidly quantifiable by flow cytometry. Analysis of genomic DNA from recombined FYDR cells shows that this mouse model detects gene conversions, and based on the arrangement of the integrated recombination substrate, unequal sister-chromatid exchanges and repair of collapsed replication forks are also expected to reconstitute EYFP coding sequence. The rate of spontaneous recombination in primary fibroblasts derived from adult ear tissue is 1.3 +/- 0.1 per 106 cell divisions. Interestingly, the rate is approximately 10-fold greater in fibroblasts derived from embryonic tissue. We observe an approximately 15-fold increase in the frequency of recombinant cells in cultures of ear fibroblasts when exposed to mitomycin C, which is consistent with the ability of interstrand crosslinks to induce homologous recombination. In addition to studies of recombination in cultured primary cells, the frequency of recombinant cells present in skin was also measured by direct analysis of disaggregated cells. Thus, the FYDR mouse model can be used for studies of mitotic homologous recombination both in vitro and in vivo.

Gamma-irradiation stimulates homology-directed DNA double-strand break repair in Drosophila embryo

To test the DNA double-strand break (DSB) repair activities present in Drosophila early embryos, we have analyzed the circularization of a microinjected linear plasmid. In order to study repair by homologous recombination, the linear plasmid was injected with an homologous fragment encompassing the break. After extraction from embryos, repair products were analyzed directly by PCR and after their cloning into bacteria. We demonstrate, in addition to the repair by homologous recombination, the presence of an efficient end-joining activity in embryos. Plasmid circularization by end-joining was accompanied by short deletions frequently associated with non-random insertions. Most importantly, pre-irradiation of embryos specifically enhanced the accurate repair by homologous recombination. Such a stimulation is described for the first time in the context of a whole higher organism.

Analysis of intrachromosomal homologous recombination in mammalian cell, using tandem repeat sequences

In all the organisms, homologous recombination (HR) is involved in fundamental processes such as genome diversification and DNA repair. Several strategies can be devised to measure homologous recombination in mammalian cells. We present here the interest of using intrachromosomal tandem repeat sequences to measure HR in mammalian cells and we discuss the differences with the ectopic plasmids recombination. The present review focuses on the molecular mechanisms of HR between tandem repeats in mammalian cells. The possibility to use two different orientations of tandem repeats (direct or inverted repeats) in parallel constitutes also an advantage. While inverted repeats measure only events arising by strand exchange (gene conversion and crossing over), direct repeats monitor strand exchange events and also non-conservative processes such as single strand annealing or replication slippage. In yeast, these processes depend on different pathways, most of them also existing in mammalian cells. These data permit to devise substrates adapted to specific questions about HR in mammalian cells. The effect of substrate structures (heterologies, insertions/deletions, GT repeats, transcription) and consequences of DNA double strand breaks induced by ionizing radiation or endonuclease (especially the rare-cutting endonuclease ISce-I) on HR are discussed. Finally, transgenic mouse models using tandem repeats are briefly presented.

A comparison of calcium phosphate coprecipitation and electroporation. Implications for studies on the genetic effects of DNA damage

Plasmid-based transfection assays provide a rapid means to measure homologous and nonhomologous recombination in mammalian cells. Often it is of interest to examine the stimulation of recombination by DNA damage induced by radiation, genotoxic chemicals, or nucleases. Transfection is frequently performed by using calcium phosphate coprecipitation (CPP), because this method is well suited for handling large sample sets, and it does not require expensive reagents or equipment. Alternative transfection methods include lipofection, microinjection, and electroporation. Since DNA strand breaks are known to stimulate both homologous and nonhomologous recombination, the induction of nonspecific damage during transfection would increase background recombination levels and thereby reduce the sensitivity of assays designed to detect the stimulation of recombination by experimentally induced DNA damage. In this article, we compare the stimulatory effects of nuclease-induced double-strand breaks (DSBs) on homologous and nonhomologous recombination for molecules transfected by CPP and by electroporation. Although electroporation yielded fewer transfectants, both nonhomologous and homologous recombination were stimulated by nuclease-induced DSBs to a greater degree than with CPP. Ionizing radiation is an effective agent for inducing DNA strand breaks, but previous studies using CPP generally showed little or no stimulation of homologous recombination among plasmids damaged with ionizing radiation. By contrast, we found clear dose-dependent enhancement of recombination with irradiated plasmids transfected using electroporation. Thus, electroporation provides a higher signal-to-noise ratio for transfection-based studies of damage-induced recombination, possibly reflecting less nonspecific damage to plasmid DNA during transfection of mammalian cells.

Overexpression of Rad51 protein stimulates homologous recombination and increases resistance of mammalian cells to ionizing radiation

Rad51 proteins share both structural and functional homologies with the bacterial recombinase RecA. The human Rad51 (HsRad51) is able to catalyse strand exchange between homologous DNA molecules in vitro . However the biological functions of Rad51 in mammals are largely unknown. In order to address this question, we have cloned hamster Rad51 cDNA and overexpressed the corresponding protein in CHO cells. We found that 2-3-fold overexpression of the protein stimulated the homologous recombination between integrated genes by 20-fold indicating that Rad51 is a functional and key enzyme of an intrachromosomal recombination pathway. Cells overexpressing Rad51 were resistant to ionizing radiation when irradiated in late S/G2phase of the cell cycle. This suggests that Rad51 participate in the repair of double-strand breaks most likely by homologous recombination involving sister chromatids formed after the S phase.

Ectopic mitotic recombination in Drosophila probed with bacterial beta-galactosidase gene-based reporter transgenes

Plasmids were constructed to investigate homologous mitotic recombination in Drosophila cells. Heteroalleles containing truncated but overlapping segments of the bacterial beta-galactosidase gene (lacZ) were positioned either on separate plasmids or as direct repeats on the same chromosome. Recombination reconstituted a functional lacZgene leading to expression of LacZ+activity detectable by histochemical staining. High extrachromosomal recombination (ECR) frequencies between unlinked heteroalleles were observed upon transient co-transfection into Drosophila melanogaster Schneider line 2 (S2) cells. Stably transfected cells containing the lacZ heteroalleles linked on a chromosome exhibited intrachromosomal recombination (ICR) frequencies two orders of magnitude lower than ECR frequencies. Recombination was inducible by exposing the cells to ethyl methanesulphonate or mitomycin C. Recombination products were characterized by multiplex PCR analysis and unequal sister chromatid recombination was found as the predominant mechanism reconstituting the lacZ gene. To investigate recombination in vivo imaginal disc cells from transgenic larvae carrying the reporter gene on the X chromosome were isolated and stained for LacZ+ activity. The presence of a few LacZ+ clones indicated that mitotic recombination events occurred at frequencies two orders of magnitude lower than the corresponding event in cultured cells and late during larval development.

Elevated homologous recombination activity in fanconi anemia fibroblasts

It is widely believed that Fanconi anemia cells possess a reduced ability to repair inter-strand DNA cross-links. While the mechanism through which inter-strand DNA cross-links are removed from mammalian chromosomes is unknown, these lesions are repaired via homologous recombination in lower eukaryotes and bacteria. Based on the hypothesis that a similar mechanism of DNA repair functions in mammalian somatic cells, we measured homologous recombination activity in diploid fibroblasts from healthy donors, and Fanconi anemia patients. Somewhat surprisingly, homologous recombination levels in nuclear protein extracts prepared from Fanconi anemia cells were nearly 100-fold higher than in extracts prepared from control cells. We observed a similar increase in the activity of a 100-kDa homologous DNA pairing protein in extracts from Fanconi anemia cells. Transfection studies confirmed that plasmid homologous recombination levels in intact Fanconi anemia cells were substantially elevated, compared with control cells. These results suggest that inappropriately elevated levels of homologous recombination activity may contribute to the genomic instability and cancer predisposition that characterize Fanconi anemia.

Cell cycle dependence of radiation-induced homologous recombination in cultured monkey cells

A lacZ transgene recombination system that reports homologous recombination events involving duplicated lacZ segments was used to study recombination in monkey cells exposed to ionizing radiation at different points in the cell cycle. With this system, recombination events can be detected in single cells by histochemical staining soon after exposure of cells to DNA-damaging treatment. Ionizing radiation rapidly induced recombination 5-10-fold in cells that were at the mitosis stage of the cell cycle. Irradiation either of cells at other points in the cell cycle or of nonsynchronized cells had less of an effect on recombination between lacZ segments.

Perturbation of nuclear architecture by long-distance chromosome interactions

SUMMARY

Position-effect variegation (PEV) describes the stochastic transcriptional silencing of a gene positioned adjacent to heterochromatin. Using FISH, we have tested whether variegated expression of the eye-color gene brown in Drosophila is influenced by its nuclear localization. In embryonic nuclei, a heterochromatic insertion at the brown locus is always spatially isolated from other heterochromatin. However, during larval development this insertion physically associates with other heterochromatic regions on the same chromosome in a stochastic manner. These observations indicate that the brown gene is silenced by specific contact with centromeric heterochromatin. Moreover, they provide direct evidence for long-range chromosome interactions and their impact on three-dimensional nuclear architecture, while providing a cohesive explanation for the phenomenon of PEV.

'Non-genotoxic' carcinogens evaluated using the white-ivory assay of Drosophila melanogaster

Seven carcinogenic compounds (urethane, ethionine, auramine O, safrole, amitrole, acetamide and thioacetamide) were tested using the white-ivory (Wi) assay of Drosophila melanogaster. These compounds were chosen because they were considered as Ames-test negative but produced positive results in the yeast DEL assay, which estimates the introduction of intrachromosomal recombination. Only one compound, urethane, produced positive results in the Wi assay, while the remaining were classified as negative. These results indicate that, in contrast with which has been postulated in yeast, these carcinogens do not induce any event associated to intrachromosomal recombination in D. melanogaster.

Mutagenic effects at hprt locus and in minisatellite sequences induced in V79 cells by treatments with UV and methyl-nitro-nitroso guanidine

DNA alterations induced in V79 cells treated with UV light or methyl-nitro-nitrosoguanidine were analyzed by the mutagenicity test at the hprt locus and by DNA fingerprint analysis. Treated and control cells were seeded in the presence or absence of 6-thioguanine to determine mutant frequency and cell survival. From clonal cultures of the same cell populations we isolated a number of clones and grew them up individually to obtain appropriate amounts of DNA. High molecular weight DNA was digested with HinfI or HaeIII and hybridized with 32P-labelled 33.15 multilocus probe. The induction of 6-thioguanine resistant cells depended on the mutagen dose. The highest value of mutant frequency obtained was 7475 x 10(-6) (MNNG, 27 microM), corresponding to 0.7 percent of clonable cells. DNA fingerprint analysis carried on the same treated cells showed that DNA rearrangements occurred at minisatellites much more frequently than in transcribed sequences. UV irradiation produced the highest frequency of variation, modifying minisatellite patterns in about 50 percent of the analyzed clones.

A recombination-based transgenic mouse system for genotoxicity testing

It is well established that mutagens induce recombination in cultured cells and experimental organisms. Presumably, this is a consequence of the DNA-damage-triggering cellular-repair mechanisms. The relationship between recombination and mutagenicity has been exploited in submammalian organisms, such as yeast, to assay the ability of chemical agents and radiation to induce a form of recombination called gene conversion--the non-reciprocal transfer of genetic information. This work has demonstrated the efficacy of predicting mutagenicity on the basis of recombination induction. Here, we describe the utilization of a transgenic mouse system for efficient detection of germ-line gene-conversion events as a mutagen-screening tool. These mice contain two mutually defective reporter (lacZ) genes under the regulatory control of a spermatogenesis-specific promoter. A particular intrachromosomal gene conversion event must occur for the generation of functional lacZ activity. Conversion events are visualized by histochemical staining or flow cytometric analysis of transgenic spermatids. The highly mutagenic compound chlorambucil induced a several fold percentage-wise increase of lacZ-positive spermatids, whereas acrylamide, a weak genotoxin, produced no marked increase in converted spermatids. The results indicate that recombination-based transgenic mouse models for genotoxin screening present a viable option for inexpensive and rapid whole-animal mutagen testing. The particular mice we describe may ultimately prove to be a useful tool for identifying agents which can cause heritable genetic mutations in humans.

Demonstration by DNA fingerprint analysis of genomic instability in mouse BALB 3T3 cells during cell transformation

We employed DNA fingerprint analysis to monitor DNA rearrangements in BALB 3T3 cells transformed spontaneously or by treatment with 3-methylcholanthrene (MCA) and UV-C. The effect of 12-O-tetradecanoyl-phorbol-13-acetate (TPA) in combination with MCA was also examined. Twenty-three spontaneously transformed cells, 28 induced transformed cells (18 by 1 microgram/ml MCA, six by 5 micrograms/ml MCA, and four by UV-C), and 31 non-transformed subclones were isolated from parental BALB 3T3 A31-1-1 cells. The DNAs were digested with HinfI and subjected to DNA fingerprint analysis with three multi-locus minisatellite probes, Per-6, Core, and Ins. Per-6 was the most effective probe for detecting DNA rearrangements. Rearranged bands detected by the Per-6 probe were observed in 9/31 (29%) of non-transformed subclones, 14/23 (61%) of spontaneously transformed cells, 16/18 (89%) of cells transformed by 1 microgram/ml of MCA, 6/6 (100%) of cells transformed by 5 micrograms/ml MCA, and 4/4 (100%) of UV-C-transformed cells. Higher numbers of DNA rearrangements (> or = 3) occurred most frequently in the induced transformed cells. TPA enhanced the frequency of DNA rearrangements in cells transformed by MCA. These data indicate that (1) genomic DNA in BALB 3T3 cells is unstable and susceptible to rearrangement, (2) its instability is elevated during cell transformation, and (3) MCA and UV-C induce DNA rearrangements, and TPA enhances the effect of the former, probably via the recombination process. DNA fingerprint analysis is valuable for monitoring genomic instability during cell transformation.

Induction of genetic recombination: consequences and model systems

Radiation and many chemicals have been found to induce homologous genetic recombination. Experimental systems that allow the detection and characterization of recombinagens exist in organisms as diverse as bacteria, fungi, plants, insects, and mammals. Recombination plays an important role in many biological processes, and studies of recombinagens can provide insight into underlying mechanisms. Studies of recombinagens are also of applied interest in genetic toxicology, because recombinational events in somatic cells can contribute to human disease. Clear connections have been established between mitotic recombination and the etiology of some cancers. This article briefly reviews two aspects of the induction of genetic recombination by radiation and chemicals--the health implications of recombinagenic effects and assays for detecting recombinagens.

Spontaneous and induced homologous recombination between lacZ chromosomal direct repeats in CV-1 cells

A lacZ substrate for intrachromosomal homologous recombination was generated at a specific site within the genome of CV-1 cells by FLP recombinase-mediated gene targeting. A histochemical stain was used to detect cells that contained recombined lacZ genes. The spontaneous rate of homologous recombination was approximately 1 x 10(-5) events per cell generation. Recombination was induced 30-fold in cells following exposure to mitomycin C (MMC) and by serum starvation. These results demonstrate the utility of the FLP recombinase in modifying the genome of mammalian cells in a predetermined manner and show that homologous recombination between direct repeats is increased in cells as a result of the withdrawal of serum growth factors.