Enzymatic restriction of mammalian cell DNA: evidence for double-strand breaks as potentially lethal lesions

Permeabilized Chinese hamster cells were treated with the restriction endonucleases Pvu II and Bam H1 which generate blunt-ended and cohesive-ended DNA double-strand breaks (dsb), respectively. Cells were then assayed for their clonogenic ability. These experiments were performed to test the hypothesis that mammalian cell death following X-ray exposure arises from the induction of dsb in DNA, and via the formation of chromosomal aberrations. It was shown previously that Pvu II induces chromosome aberrations whereas Bam H1 was ineffective in this respect. The results reported here show that Pvu II simulates X-ray exposure, in causing a dose-dependent loss of the reproductive integrity of mammalian cells. Dsb generated by Pvu II, i.e. with blunt ends, can therefore be regarded as potentially clastogenic as well as potentially lethal. Bam H1 was found not to reduce cell survival in the same enzyme dose range. These results support the notion that X-irradiated mammalian cells undergo a mode of death in which dsb in the DNA cause chromosomal aberrations which are lethal as a result of loss of genetic material in the form of chromosome fragments, or as a result of chromosome bridge formation.     

Molecular mechanisms involved in the production of chromosomal aberrations

Chinese hamster ovary cells (CHO cells) and mouse fibroblasts (PG 19) were permeabilized with inactivated Sendai virus, treated with different types of restriction endonucleases (Eco RV, Pvu II, Bam HI, Sma I, Asu III, Nun II), and studied for the occurrence of chromosomal aberrations at different times following treatment. The pattern of chromosomal aberrations observed was similar to that induced by ionizing radiations. Restriction endonucleases that induce blunt double-strand breaks (Eco RV, Pvu II) were more efficient in inducing chromosomal aberrations than those that induce breaks with cohesive ends (Bam HI, Nun II, Asu III). Ring types were very frequent among the aberrations induced by restriction enzymes. Cytosine arabinoside, an inhibitor of DNA repair, was found to increase the frequencies of aberrations induced by restriction enzymes, indicating its effect on ligation of double-strand breaks. The relevance of these results to the understanding of the mechanisms of chromosomal aberration formation following treatment with ionizing radiations is discussed.     

Evidence for induction of DNA double strand breaks in the bystander response to targeted soft X-rays in CHO cells

This study investigated the role of DNA double strand breaks and DNA base damage in radiation-induced bystander responses in Chinese hamster ovary (CHO) cell lines. Two CHO repair-deficient clones, xrs5 (DNA double strand break repair-deficient) and EM9 (DNA base excision repair-deficient) were used in addition to the wild type (CHO). The Gray Cancer Institute ultrasoft X-ray microprobe is a powerful tool for investigating the bystander response, because it permits the irradiation of only a single nucleus of a cell, as reported previously. In order to investigate the bystander effect in each repair-deficient cell line, we irradiated a single cell within a population and scored the formation of micronuclei. When a single nucleus in the population was targeted with 1 Gy, elevated numbers of micronuclei were induced in the neighbouring unirradiated cells in the EM9 and xrs5 cell lines, whereas induction was not observed in CHO. The induction of micronuclei in xrs5 was significantly higher than that in EM9. Under these conditions, the surviving fraction in the neighbouring cells was significantly lower in xrs5 than in the other cell lines, showing a higher cell killing effect in xrs5. To confirm that bystander factors secreted from irradiated cells caused these effects, we carried out medium transfer experiments using conventional X-irradiation. Medium conditioned for 24 h with irradiated cells was transferred to unirradiated cells and elevated induction of micronuclei was observed in xrs5. These results suggest that DNA double strand breaks rather than base damage are caused by factors secreted in the medium from irradiated cells.     

Relative effectiveness of HZE iron-56 particles for the induction of cytogenetic damage in vivo

One of the risks of prolonged manned space flight is the exposure of astronauts to radiation from galactic cosmic rays, which contain heavy ions such as (56)Fe. To study the effects of such exposures, experiments were conducted at the Brookhaven National Laboratory by exposing Wistar rats to high-mass, high-Z, high-energy (HZE) particles using the Alternating Gradient Synchrotron (AGS). The biological effectiveness of (56)Fe ions (1000 MeV/nucleon) relative to low-LET gamma rays and high-LET alpha particles for the induction of chromosome damage and micronuclei was determined. The mitotic index and the frequency of chromosome aberrations were evaluated in bone marrow cells, and the frequency of micronuclei was measured in cells isolated from the trachea and the deep lung. A marked delay in the entry of cells into mitosis was induced in the bone marrow cells that decreased as a function of time after the exposure. The frequencies of chromatid aberrations and micronuclei increased as linear functions of dose. The frequency of chromosome aberrations induced by HZE particles was about 3.2 times higher than that observed after exposure to (60)Co gamma rays. The frequency of micronuclei in rat lung fibroblasts, lung epithelial cells, and tracheal epithelial cells increased linearly, with slopes of 7 x 10(-4), 12 x 10(-4), and 11 x 10(-4) micronuclei/binucleated cell cGy(-1), respectively. When genetic damage induced by radiation from (56)Fe ions was compared to that from exposure to (60)Co gamma rays, (56)Fe-ion radiation was between 0.9 and 3.3 times more effective than (60)Co gamma rays. However, the HZE-particle exposures were only 10-20% as effective as radon in producing micronuclei in either deep lung or tracheal epithelial cells. Using microdosimetric techniques, we estimated that 32 cells were hit by delta rays for each cell that was traversed by the primary HZE (56)Fe particle. These calculations and the observed low relative effectiveness of the exposure to HZE particles suggest that at least part of the cytogenetic damage measured was caused by the delta rays. Much of the energy deposited by the primary HZE particles may result in cell killing and may therefore be "wasted" as far as production of detectable micronuclei is concerned. The role of wasted energy in studies of cancer induction may be important in risk estimates for exposure to HZE particles.     

High chromosomal sensitivity of Chinese hamster xrs 5 cells to restriction endonuclease induced DNA double-strand breaks

The cytogenetic effects of restriction endonucleases (RE) and X-rays were examined in the radiosensitive mutant Chinese hamster cell line xrs 5 and its normal parental line CHO K1. Cells were permeabilized with Sendai virus and exposed to Pvu II and Eco RV which induce blunt-ended double-strand breaks (dsb) in the DNA of cells, or Bam H1 and Eco R1 which induce cohesive-ended dsb with a four-base overlap. Treated cells were then assayed for the presence of metaphase chromosomal aberrations by sampling at multiple fixation times and in experiments where cells were exposed to graded series of RE concentrations. Exposure to X-rays or RE causing blunt-ended dsb was found to be between two and three times more effective in xrs 5 than in CHO K1 cells. We interpret this higher chromosomal sensitivity of xrs 5 cells as reflecting the reported defect in dsb repair in xrs 5. Both xrs 5 and CHO K1 cells yielded less aberrations after exposure to Bam H1 or Eco R1 than after exposure to Pvu II or Eco RV, confirming our previous results and demonstrating that cohesive-ended dsb are less damaging than blunt-ended dsb. Multiple fixation time experiments showed that the higher sensitivity of xrs 5 was evident at several different sampling times after treatment. Similarly the low yield of aberrations after exposure of cells to Bam H1 was evident at all sampling times. Overdispersion of chromosomal aberrations was observed in samples exposed to RE. This is thought to be due to a non-uniform permeabilization of the cell population to RE. Our results indicate that RE-induced dsb are handled by cells in a similar way to those arising during X-ray exposure.     

Possible relationship between micronucleated and binucleated cells induced by cisplatin in cultured CHO cells

Chinese hamster ovary (CHO) cells were treated with a single dose (10 μg/ml) of cis-diammino-dichloroplatinum (II) (cisplatin) for 1 h and the effect of the drug on the kinetics of proliferation of the cultures was studied. It was found that the drug produces a delay in the proliferation rates of the treated cultures.The induction of micronuclei and binucleated cells (BC) at different times after treatment have also been studied, and the ability of these cells to undergo DNA synthesis (measured as the ability to incorporate [³H]thymidine) is shown.It was found that cisplatin induced a particular type of BC that contains one or more micronuclei rather than a pure population of BC. The results obtained show a possible relationship between micronuclei and BC. The possibility that some of the micronucleated cells evolve in subsequent cell divisions to BC with micronuclei is suggested.     

2-Hydroxy-1,4-naphthoquinone, the natural dye of Henna, is non-genotoxic in the mouse bone marrow micronucleus test and does not produce oxidative DNA damage in Chinese hamster ovary cells

2-Hydroxy-1,4-naphthoquinone (HNQ) has been found positive in a previous chromosome aberration test in Chinese hamster ovary (CHO) cells and in a mouse bone marrow micronucleus test at 72h after oral administration (vehicle: DMSO). However it was negative at 24 and 48h sampling times, and in subsequent micronucleus tests that used 0.5% aqueous methyl cellulose (MC) as vehicle. We performed a bone marrow micronucleus test in male and female NMRI BRL/BR mice at oral doses of 75, 150 and 300mg/kg in two vehicles (DMSO and 0.5% aqueous MC), evaluated micronuclei at 24, 48 and 72h, plasma levels of HNQ at 0.5, 1 and 4h, and haematology parameters at 72h after administration. The mechanism of in vitro clastogenic activity of HNQ was investigated by evaluation of the potential of HNQ to produce oxidative DNA damage after treatment of CHO with 10mM HNQ, followed by quantification of DNA fragments using the comet assay. In the micronucleus test, HNQ at 300mg/kg produced mortality and clinical signs at similar incidence and severity for both vehicles. Levels of HNQ in the plasma of treated mice were dose-related, of similar magnitude for both vehicles, but higher in females than in males. Maximum concentrations were found at 0.5 or 1h. At 300mg/kg, HNQ slightly affected RBC parameters suggesting haematotoxicity. No increase in the frequency of micronuclei was observed for any dose, vehicle or time point, whereas the positive control substance (CPA) produced a clear positive response. No evidence of HNQ-induced oxidative DNA damage was found at clastogenic concentrations in vitro, whereas the positive control substance (H(2)O(2)) produced a clear increase. In conclusion, HNQ was negative for induction of bone marrow micronuclei in mice up to 72h after administration in two different vehicles, and its in vitro clastogenicity was not due to oxidative damage. These results confirm that HNQ poses no or negligible genotoxic risk.     

Germ-cell mutagenicity of etoposide: induction of meiotic micronuclei in cultured rat seminiferous tubules

Mutagen effects on male germ cells can be quantified by meiotic micronucleus induction in vitro. Late pachytene and diakinetic primary spermatocytes are able to differentiate through meiotic divisions in vitro and develop to round spermatids. In the presence of mutagens micronucleus induction reflects the potential of the chemical to induce chromosome breakage or uneven chromosome distribution. In this study we have investigated the mutagenicity of etoposide (VP-16) and its ability to induce micronuclei S-independently in meiosis by the meiotic micronucleus method in vitro. Our results indicate that etoposide is able to cause a statistically significant increase in the frequency of micronuclei at a concentration range as low as 0.5–8 μmole/1. The meiotic micronucleus method in vitro seems to be a feasible and sensitive test system of male germ-cell mutagenesis.     

Enzymatic restriction of mammalian cell DNA using Pvu II and Bam H1: evidence for the double-strand break origin of chromosomal aberrations

Permeabilized Chinese hamster cells were treated with the restriction enzymes Pvu II and Bam H1 which generate blunt-ended with cohesive-ended double-strand breaks in the DNA respectively. Cells were then allowed to progress to the first mitosis, where chromosomal aberrations were scored. It was found that blunt-ended double-strand breaks induced both chromosome and chromatid aberrations of exchange and deletion types, including a high frequency of tri-radials. The total aberration frequency at high enzyme concentrations was more than ten times the control background frequency. Treatment with Bam H1 on the other hand did not induce aberrations above the background rate. This may indicate that the cohesive ends generated by this enzyme may be easily repaired by the cell due to the stabilization of the hydrogen bonding at the site of the double-strand break. Measurements using the unwinding method showed that the enzymes caused strand breaks in the DNA of permeabilized cells, and an approximate X-ray dose equivalent of the restriction-enzyme-induced breaks could be calculated. This indicated that restriction-induced blunt-ended double-strand breaks are relatively inefficient in causing chromosomal aberrations. This may be because of the presence of 'clean ends' at the site of a double-strand break, which may be repaired by ligation. The method of introducing restriction enzymes into cells opens up a new model approach for the study of the conversion of double-strand breaks into chromosome aberrations.     

Mitotic dynamics of micronuclei induced by amiprophos-methyl and prospects for chromosome-mediated gene transfer in plants

Summary Mitotic dynamics and the kinetics of mass induction of micronuclei after treatment of Nicotiana plumbaginifolia cell suspensions with the spindle toxin amiprophos-methyl (APM) are reported. The addition of APM to suspension cells resulted in the accumulation of a large number of metaphases. The course of mitosis was strikingly different from normal. Metaphase chromosomes showed neither centromere division nor separation of chromatids. Single chromosomes and groups of 2 or more chromosomes were scattered over the cytoplasm. After 5–6 h of APM treatment, chromosomes decondensed and formed micronuclei. When treatment duration was increased, the frequency of cells with micronuclei as well as those showing lobed micronuclei increased. Similarly, with an increase in APM concentration the frequency of cells with micronuclei increased. After removal of APM, chromosome grouping disappeared, cells showing lobed micronuclei further increased and mitoses with doubled chromosome numbers appeared in the next cell division. Cytological observations and DNA measurements revealed that several sub-diploid micronuclei containing 1 or a few chromosomes can be obtained, and that flow cytometry can detect and sort out these micronuclei. The applications of micronuclei for genetic manipulation of specific chromosomes and gene mapping are indicated.     

The relationship between colony-forming ability, chromosome aberrations and incidence of micronuclei in V79 Chinese hamster cells exposed to microwave radiation

Cultured V79 Chinese hamster fibroblast cells were exposed to continuous radiation, frequency 7.7 GHz, power density 0.5 mW/cm2 for 15, 30 and 60 min. The effect of microwave radiation on cell survival and on the incidence and frequency of micronuclei and structural chromosome aberrations was investigated. The decrease in the number of irradiated V79 cell colonies was related to the power density applied and to the time of exposure. In comparison with the control samples there was a significantly higher frequency of specific chromosome aberrations such as dicentric and ring chromosomes in irradiated cells. The presence of micronuclei in irradiated cells confirmed the changes that had occurred in chromosome structure. These results suggest that microwave radiation can induce damage in the structure of chromosomal DNA.     

Enhanced sensitivity of lymphoblastoid cells from individuals carrying the mutation for the fragile X syndrome to the clastogenic effects of FUdR

Individuals known to carry the mutation for the fragile X syndrome can sometimes be identified cytogenetically by the presence of a fragile site on the X chromosome at q27.3. The frequency of cells bearing this fragile site is known to be enhanced by culturing the cells in folic acid deficient medium and/or by introducing folic acid metabolism inhibitors such as FUdR. In this study FUdR induction of chromosomal aberrations other than the fragile X was investigated. Lymphoblastoid cells from an obligate carrier, a mentally retarded male and a control were cultured in folic acid deficient medium in the presence of FUdR and harvested at various times after culture initiation. The frequency of chromosome and chromatid breaks was found to be higher in cells from the individuals carrying the mutation for the fragile X syndrome. The frequency of micronuclei, an indirect index of chromosome breakage, was also more elevated in cells from these individuals than in cells from the control. These findings are of potential importance to carrier detection of this common genetic disorder.     

The correlation between molecular cellular parameters and immune status in the blood lymphocytes of chernobyl nuclear accident cleanup workers

Biophysics - Genome damage, namely, the frequency of cells with micronuclei (MN) and chromosome aberrations, the level of DNA double-strand breaks, and concentration of reactive oxygen species...     

Chromosomal malsegregation and micronucleus induction in vitro by the DNA topoisomerase II inhibitor fisetin

The plant flavonol fisetin is a common dietary component that has a variety of established biological effects, one of which is the inhibition of the enzyme DNA topoisomerase II (topo II). Compounds that inhibit topo II can exert genotoxic effects such as DNA double strand breaks, which can lead to the induction of kinetochore- or CREST-negative micronuclei. Despite reports that fisetin is an effective topoisomerase II inhibitor, its genotoxic effects have not yet been well characterized. Genotoxicity testing of fisetin was conducted in TK6 and HL60 cell lines and the cells were analyzed for malsegregating chromosomes as well as for the induction of micronuclei. Using the cytokinesis-blocked CREST micronucleus assay to discriminate between micronuclei formed from chromosomal breakage (CREST-negative) and chromosomal loss (CREST-positive), a statistically significant increase in CREST-positive micronuclei was seen for all doses tested in both cell lines. CREST-negative micronuclei, however, were significantly increased at the higher test concentrations in the TK6 cell line. These data indicate that at low concentrations fisetin is primarily exerting its genotoxic effects through chromosomal loss and that the induction of DNA breaks is a secondary effect occurring at higher doses. To confirm these results, the ability of fisetin to inhibit human topoisomerase II-alpha was verified in an isolated enzyme system as was its ability to interfere with chromosome segregation during the anaphase and telophase periods of the cell cycle. Fisetin was confirmed to be an effective topo II inhibitor. In addition, significant increases in the number of mis-segregating chromosomes were observed in fisetin-treated cells from both cell lines. We conclude that fisetin is an aneugen at low concentrations capable of interfering with proper chromosomal segregation and that it is also an effective topo II inhibitor, which exerts clastogenic effects at higher concentrations.     

Substrate recognition by the Pvu II endonuclease: binding and cleavage of CAG5mCTG sites.

The Pvu II restriction endonuclease (R. Pvu II) cleaves CAG downward arrowCTG sequences as indicated, leaving blunt ends. Its cognate methyltransferase (M. Pvu II) generates N4-methylcytosine, yielding CAGN4mCTG, though the mechanism by which this prevents cleavage by R. Pvu II is unknown. The heterologous 5-methylcytosinemethylation CAG5mCTG has also been reported to prevent cleavage by R. Pvu II and this has been used in some cloning methods. Since this heterologousmethylation occurs at the native methylated base, it can provide insights into the detection of DNAmethylation by R. Pvu II. We found that the cloned gene for R. Pvu II could not stably transform cells protected only by M. Alu I (AG5mCT) and then determined that R. Pvu II cleaves CAG5mCTG in vitro, even when both strands are methylated. DNase I footprint analysis and competition experiments reveal that R. Pvu II binds to CAG5mCTG specifically, though with reduced affinity relative to the unmethylated sequence. These results provide biochemical support for the publishedstructures of R. Pvu II complexed with DNA containing CAGCTG and CAG5-iodoCTG and support a model for how methylation interferes with DNA cleavage by this enzyme.     

Lack of interference of DNA single-strand breaks with the measurement of double-strand breaks in mammalian cells using the neutral filter elution assay.

In this study, the effect of DNA single strand breaks (ssb) on the neutral (pH 9.6) filter elution of DNA from Chinese hamster ovary (CHO K1) cells containing DNA double strand breaks (dsb) was investigated. Protein associated ssb were induced by the inhibition of DNA topoisomerase I with camptothecin (cpt). Protein associated dsb were introduced by treating cells with the DNA topoisomerase II poison; etoposide (VP-16). Protein associated ssb and dsb were converted to ssb and dsb by proteinase K present in the lysis solution. In some experiments dsb were generated by the restriction endonuclease Pvu II. It was found that elution of DNA in the presence and absence of ssb was similar under neutral conditions. This finding is consistent with the view that the fast component of the bi-phasic repair kinetics observed in irradiated mammalian cells with the neutral filter elution technique is not attributable to the interference of ssb with the measurement of dsb, and thus suggests that the two components of repair observed with the neutral filter elution elution technique may represent two different types of dsb or modes of repair of dsb.     

Genotoxicity of 2-amino-6-N-hydroxyadenine (AHA) to mouse lymphoma and CHO cells.

2-Amino-6-N-hydroxyadenine (AHA) treated L5178Y/TK (+/-)-3.7.2C mouse lymphoma cells were evaluated for mutations at the tk, hgprt, and Na+/K+ ATPase loci, as well as for gross chromosome aberrations and induction of micronuclei. In addition, AHA was evaluated for its ability to induce HGPRT mutants in CHO cells. AHA was found to induce mutations at all evaluated loci and in both cell types. The TK mutants were primarily large colonies although a few small colonies were also induced, particularly at the higher concentrations. Preliminary cytogenetic analysis of AHA-treated mouse lymphoma cells indicated that some gross aberrations but not micronuclei were induced. The 20 small-colony TK mutants evaluated by banded karyotype indicate that only a small fraction (2 of 20) showed chromosome 11 abnormalities. From these studies, it appears that AHA may be one of a very few chemicals that is capable of inducing multi-locus point mutations, with only slight clastogenic activity. Particularly at the higher concentrations, some of the mutants may contain multi-locus point mutations that result in slow growth.     

Inhaled radon-induced genotoxicity in Wistar rat,Syrian hamster,and Chinese hamster deep-lung fibroblasts in vivo

This study was performed (1) to provide a comparison of the genotoxin effects of inhaled radon and radon progeny, referred to as radon in this paper, among three species of rodents: Wistar rats, Syrian hamsters, and Chinese hamsters; (2) to determine if initial chromosome damage was related to the risk of induction of lung cancer; and (3) to evaluate the tissue repair and long-term presence of cytogenetic damage in respiratory tract cells. These species were selected because Syrian hamsters are very resistant to radon induction of lung cancer and Wistar rats are sensitive; no literature is available on the in vivo effects of radon in the Chinese hamster. Exposure-response relationships were established for the rats and Syrian hamsters while the Chinese hamsters received a single exposure of radon. At 4 h (0.2 days), 15 days, and 30 days after the highest WLM exposure to radon, Wistar rats, Chinese hamsters, and Syrian hamsters were killed, and lung fibroblasts were isolated and grown in culture to determine the frequency of induced micronuclei. Animals at each level of exposure showed an increase in the frequency of micronuclei relative to that in controls (P < 0.05). The exposure-response relationship data for rats and Syrian hamsters killed 0.2 days after the end of exposure were fit to linear equations (micronuclei/1000 binucleated cells = 15.5±14.4+0.53±0.06 WLM and 38.3±15.1+0.80±0.08 WLM, respectively). For the single exposure level used (496 WLM) in Chinese hamsters killed at 0.2 days after exposure, the frequency of micronuclei/1000 binucleated cells/WLM was 1.83±0.02. A comparison of the sensitivity for induction of micronuclei/WLM illustrated that Chinese hamsters were three times more sensitive than rats. The Syrian hamsters also showed a significantly elevated response (P < 0.05) relative to rats. These data suggest that initial chromosome damage is not the major factor responsible for the high rate of radon-induced cancer in rats relative to Syrian hamsters. The frequency of micronuclei in radon-exposed rats, Syrian hamsters, and Chinese hamsters significantly decreased (P < 0.05) as a function of time after the exposure. The rate of loss of damaged cells from the lung was greatest in the Chinese hamsters, followed by Wistar rats and Syrian hamsters, respectively. Our experiments demonstrated that the mammalian lung fibroblast/micronucleus method has the potential to (1) detect species differences in the induction of in vivo genotoxic damage in the lungs by inhaled environmentalal agents; (2) evaluate exposure-response relationships for in vivo induction of genetic damage; and (3) determine the persistence in vivo of preclastogenic and premutagenic lesions in cell populations.     

Induction of micronucleated and binucleate cells in Chinese hamster ovary (CHO) cells by cis-diamminedichloroplatinum (II): a morphological and morphometric study

The mutagenicity and cytotoxicity of cis-diamminedichloroplatinum (II) (cisplatin) at doses of 5, 10 and 20 micrograms/ml in Chinese hamster ovary (CHO) cells have been examined. A morphological characterization of several cell types induced by cisplatin was carried out. The frequencies of both cells with micronuclei and binucleate cells as a time-dependent parameter have also been studied. Whilst the number of cells with micronuclei was found to decrease with time, the number of binucleate cells increased. The possible kinetic mechanism for the production of binucleate cells and cells with micronuclei is discussed. A morphometric analysis was also performed. The nuclear area in both treated and control nuclei was measured with the IBAS image analysis system. The results of this analysis show that a continuous reduction in the nuclear size in the control cells is produced. However the size of the treated cells increased after treatment.     

Effect of p53 haploinsufficiency on melphalan-induced genotoxic effects in mouse bone marrow and peripheral blood

Mice heterozygous for a p53 null mutation develop tumours induced by genotoxic carcinogens with a shorter latency than wild type mice and have been proposed as an alternate animal model for carcinogenicity testing. Some literature data suggest that p53+/- mice might also be more sensitive to the short-term effects of genotoxic agents and manifest a haploinsufficiency phenotype that could contribute to the higher tumour susceptibility. We have compared the induction of micronuclei in bone marrow and blood of p53+/- and p53+/+ isogenic mice after treatment with a single or multiple doses of melphalan (MLP), a crosslinking genotoxic carcinogen. We have also characterized the mechanism of micronucleus induction with CREST staining of kinetochore proteins to distinguish between chromosome break- and chromosome loss-induced micronuclei. Significant increases of micronucleated bone marrow polychromatic erythrocytes and blood reticulocytes were induced under all MLP exposure conditions. The frequency of micronucleated blood erythrocytes increased linearly with duration of exposure. Micronuclei were essentially a consequence of chromosome break events. After a single MLP dose, a significant reduction of the frequency of polychromatic erythrocytes in bone marrow of p53+/+ animals suggested the induction of cytotoxicity/cell cycle delay. This effect was not observed in p53+/- mice. We believe this finding to provide some evidence of a haploinsufficiency phenotype in the modulation of cell cycle/apoptotic pathways mediated by the p53 protein. In bone marrow of wild type mice, an increased effect of multiple MLP doses was detected over that of a single administration, whereas, in p53+/- mice, no differential effect was found of different exposure durations. Possibly, the probability of micronucleus formation increased under chronic exposure because of increased cell division in response to peripheral anemia and a reduction of p53 protein level had a small effect on cell cycle modulation and on such indirect mechanism of micronucleus induction. However, pairwise comparisons between the frequencies of cells with micronuclei in wild type and p53+/- mice under all exposure conditions did not show statistically significant differences, suggesting that the observed effects of p53 haploinsufficiency were weak and temporary and a higher/faster induction of irreversible chromosome damage could not account for the increased susceptibility of p53+/- mice to MLP-induced tumours.