Genomic Instability and Apoptosis Are Frequent in P53 Deficient Young Mice

Home , Dihydrofolate reductase, Interphase, Ribonucleotide reductase

Genomic instability and apoptosis are frequent in p53 de®cient young mice

Kenji Fukasawa1,2, Francis Wiener3,4, George F Vande Woude1 and Sabine Mai4

1ABL-Basic Research Program, NCI-Frederick Cancer Research and Development Center, P.O. Box B, Frederick, Maryland 21702-1201, USA; 2Department of Cell Biology, Neurobiology & Anatomy, University of Cincinnati College of Medicine, P.O. Box 670521, Cincinnati, OH, USA; 3Microbiology and Tumor Biology Center, Karolinska Institute, 17177 Stockholm, Sweden; 4Manitoba Institute of Cell Biology, Manitoba Cancer Treatment and Research Foundation, and Department of Physiology, University of Manitoba, 100 Olivia Street, Winnipeg, MB, R3E 0V9, Canada

The loss of p53 tumor suppressor functions results in at the G1/S (reviewed in Hartwell, 1992; Weinert and genetic instability, characteristically associated with Lydall, 1993; Hartwell and Kastan, 1994; Kastan et al., changes in chromosome ploidy and gene ampli®cation. 1995) and G2/M transitions (Stewart et al., 1995; Cross In vivo, we ®nd that cells from various organs of 4 to 6- et al., 1995; Guillouf et al., 1995; Orren et al., 1995; week old p53-nullizygous (p537/7) mice display aneu- Wahl et al., 1996) of the cell cycle. Thus, the absence of ploidy and frequent gene ampli®cation as well as p53 results in loss of checkpoint control, leading to evidence for apoptosis. Regardless of tissue types, many genome destabilization. p537/7 cells contain multiple centrosomes and abnor- p53 has recently been shown to be involved in the mally formed mitotic spindles. Thus, chromosome regulation of the centrosome duplication cycle; in instability in vivo may be associated with abnormal embryonic ®broblasts derived from p53-nullizygous centrosome ampli®cation. Moreover, we observed a mice (p537/7 mice), abnormal ampli®cation of signi®cant increase in the number of cells overexpressing centrosomes is frequently observed (Fukasawa et al., c-Myc in p537/7 mice. Consistent with previous studies 1996). During mitosis, centrosomes become spindle showing that c-Myc overexpression is associated with poles and play a vital role in bipolar spindle gene ampli®cation in vitro, many of the p537/7 cells formation and accurate chromosomal segregation. In exhibited, in the same cell, c-Myc overexpression and the presence of an abnormal number of centrosomes, ampli®ed c-myc, dihydrofolate reductase (DHFR), and chromosome instability occurs frequently due to carbamoyl-phosphate synthetase-aspartate transcarba- multipolar mitosis. Consistent with the proposed role moyl-dihydroorotase (CAD) genes. Furthermore, apop- of p53 in genome stability and that of genomic tosis was frequently observed in cells isolated from instability in tumorigenesis, p537/7 mice develop a p537/7 mice. The apoptotic cells contained abnormally variety of tumors early in life (Donehower et al., 1992; ampli®ed centrosomes, displayed aneuploidy, high levels Jacks et al., 1994). of c-Myc expression, as well as gene ampli®cation. These In addition to abnormal centrosome duplication, results indicate that a high number of aberrant cells is recent studies have shown that in the absence of p53, eliminated by p53-independent pathways in vivo. gene ampli®cation occurs at higher frequencies (Yin et al., 1992; Livingstone et al., 1992). Gene ampli®cation Keywords: p53; genomic instability; gene ampli®cation; aects several genomic loci, among them are oncogenes centrosome; Myc; apoptosis (Alitalo, 1985; Yokota et al., 1986; Van der Bliek et al., 1986; Stark, 1993; Feo et al., 1994). Gene ampli®cation is a dynamic process, and is often accompanied by the generation of double minutes (Wahl, 1989). Introduction The c-Myc protooncogene is frequently deregulated in tumors; translocation and ampli®cation of c-myc as The term genomic instability summarizes a variety of well as increased half life and overexpression of the genomic alterations which include the loss or gain of oncoprotein are observed in many tumors (for a chromosomes as well as genetic changes at the level of review, see Marcu et al., 1992; Bishop, 1995). More- single genes, such as rearrangements, translocations, over, c-Myc overexpression has been implicated in gene ampli®cations, deletions and point mutations, and has ampli®cation (Denis et al., 1991; Mai, 1994; Mai et al., been considered to be a major driving force of 1996). multistep carcinogenesis (Nowell, 1976; Pienta et al., Here, we have studied genomic instability in vivo in 1989; Temin, 1988; Solomon et al., 1991). Genomic dierent organs of p537/7 mice (4 ± 6 weeks old), with integrity is maintained by checkpoint mechanisms; age-matched p53 homozygous (p53+/+) mice as when cells suer damage imposed by exposure to controls. We found that, in all p537/7 tissues genotoxic drugs or microtubule toxins, the cell cycle is examined, a substantial percentage of cells contained halted until the damage is repaired or apoptosis is abnormal numbers of centrosomes and displayed initiated (for reviews, see Hartwell, 1992; Weinert and aneuploidy. Moreover, c-Myc overexpression was Lydall, 1993; Hartwell and Kastan, 1994). The p53 observed in 5 ± 15% of p537/7 cells. In these cells, tumor suppressor protein exerts checkpoint functions dihydrofolate reductase (DHFR), carbamoyl-phosphate synthetase-aspartate transcarbamoyl-dihydroorotase (CAD) and c-myc genes exhibited gene ampli®cation. Correspondence: S Mai Apoptosis of cells, which displayed abnormal numbers Received 29 May 1997; revised 4 August 1997; accepted 4 August of centrosomes, aneuploidy, gene ampli®cation and c- 1997 Myc overexpression, was frequently observed. Genomic instability and apoptosis in p537/7 mice K Fukasawa et al 1296 1994), and the number of centrosomes per cell was Results scored. The numerical distribution of centrosomes in various organs of p53+/+ and p537/7 mice is Aneuploidy in p537/7 mice summarized in Figure 2a. More than 99% of the We examined genomic instability in dierent organs of interphase p53+/+ cells contained one or two centro- clinically healthy p537/7 mice (4 ± 6 weeks old) by somes, most likely depending on their duplication cytogenetically assessing chromosome ploidy. Age- cycle, while 20 ± 30% of the interphase p537/7 cells matched parental p53+/+ mice were used as controls. contained 42 centrosomes. A typical example of We characterized spleen-, thymus-, and bone marrow- normal vs aberrant centrosome duplication is shown derived cells, as well as skin- and spleen-derived in Figure 2b. Thus, as previously observed in vitro in ®broblasts. These analyses revealed aneuploidy; hyper- p537/7 MEFs, we found now in vivo that mitotic diploid, hypo-, hypertetraploid, and polyploid meta- p537/7 cells frequently displayed abberant spindles, phase plates were present in all organs examined organized by multiple copies of centrosomes. However, (Figure 1). The percentage of aneuploid mitotic plates as we have reported (Fukasawa et al., 1996) in some varied between individual mice; the mean frequencies mitotic p537/7 cells, abnormally ampli®ed centrosomes were 25.6% in the thymus, 34.8% in ®broblasts, 10% sequestered to the poles to form bipolarity (data not in the bone marrow, and 20% in the spleen (Figure 1). shown). These results imply that abnormal amplifica- No aneuploidy was observed in p53+/+ mice (Figure 1). tion of centrosomes occurs in vivo, and this may lead to Cytogenetic studies also revealed that ®ve percent of chromosome instability in p537/7 mice. all metaphases present in the p537/7 fetal liver hematopoietic cells were aneuploid (Figure 1). In Gene ampli®cation in p537/7 mice contrast, there was no evidence of aneuploidy in age- matched p53+/+ fetal liver hematopoietic cells. To evaluate the genomic (in)stability of single genes in p537/7 mice in vivo, we performed ¯uorescent in situ Abnormal centrosome ampli®cation in vivo in p537/7 mice a p53 has been implicated in the regulation of centrosome duplication, and multiple centrosomes are generated in p537/7 embryonic ®broblasts (MEFs) during a single cell cycle (Fukasawa et al., 1996). In the present work, we tested for the ®rst time whether the chromosome instability observed in p537/7 mice was associated with multiple centrosomes per cell in vivo. Spleen-, thymus-, and bone marrow-derived cells were immunostained with anti-g-tubulin antibody to identify centrosomes (reviewed in Oakley, 1992; Joshi,

b Centrosome DNA

p53 +/+

P53 –/–

Figure 2 Abnormal ampli®cation of centrosomes in p537/7 mice. (a) Number of centrosomes detected by immunostaining in bone marrow, spleen, and thymic cells isolated from p53+/+ and p537/7 mice. N1: one centromere; N2: two centromeres; N53: three or more centromeres (see text). (b) Representative picture of Figure 1 Genomic instability in p537/7 mice. Cytogenetic normal and aberrant centrosome numbers. The picture illustrates analysis of bone marrow, spleen, thymus-derived cells as well as the normal number of centrosomes as found in all organs of of ®broblasts (passage 0) isolated from ®ve p537/7 and ®ve p53+/+ mice (top panel) as well as aberrant numbers of parental p53+/+ (C57B1/6) mice, and of p537/7 and p53+/+ centrosomes as observed in all organs p537/7 mice in vivo fetal liver hematopoietic cells of 16 day old embryos (bottom panel) Genomic instability and apoptosis in p537/7 mice K Fukasawa et al 1297 hybridization (FISH). We chose to analyse the c-myc but not of the R1 gene within the same individual gene, which is frequently translocated and/or ampli®ed p537/7 bone marrow-, thymus-, spleen-derived cells as in tumors (for review, see Marcu et al., 1992; Bishop, in primary ®broblasts (Figure 4, panels a-a'', b-b''). 1995), and the DHFR gene, whose genomic instability Thus, locus-speci®c gene ampli®cation in p537/7 mice is altered following either drug selection (Stark, 1993), appears to be accompanied by c-Myc overexpression. growth factor (Huang and Wright, 1994) or c-Myc overexpression (Denis et al., 1991; Mai, 1994; Mai et Apoptosis of genomically altered cells in p537/7 mice al., 1996). We also examined the CAD and the ribonucleotide reductase R1 (R1) genes. The CAD Despite the extensive genomic alterations present in gene encodes a trifunctional enzyme of the pyrimidine fetal and neonatal life, p537/7 mice develop without biosynthesis and has been shown to be ampli®ed after apparent abnormalities (Donehower et al., 1992; Jacks exposure to PALA (N-(phosphonoacetyl)-L-aspartate) et al., 1994). One possibility is that cells with (Otto et al., 1989; Yin et al., 1992; Livingstone et al., deleterious genomic alterations may be eciently 1992). R1 forms a functional ribonucleotide reductase eliminated by apoptosis. We therefore determined molecule in association with a second subunit, apoptosis in organs and ®broblasts from p537/7 ribonucleotide reductase R2 (R2). The genomic mice. Apoptosis frequently occurred in tissues and stability of R1 is maintained even in malignant cells ®broblasts of p537/7 mice, with cells displaying (Mai et al., 1996) and served as a control. chromatin condensation characteristic of apoptosis In all p537/7 tissues examined, we detected an and a positive TUNEL reaction (Figure 5e and f). increase in ¯uorescent signals as observed by FISH for Cytogenetic and FISH analyses of p537/7 cells DHFR and c-myc as well as for CAD in both revealed the frequent presence of morphologically interphases and metaphases (Figure 3b ± e and Figure atypical chromosomes (Figure 5b ± d) in contrast to 4). In contrast, no increase in ¯uorescent signals for the aneuploid plates with the typical chromosome above genes was detected in age-matched control morphology (Figure 5a). Moreover, morphologically p53+/+ mice (Figure 3a). R1 was present as single copy atypical chromosomes exhibited a strong positive gene in both p537/7 and p53+/+ cells (data not shown). TUNEL reaction in all organs examined (Figure 5c). An increase of ¯uorescent signals as detected by Such morphologically atypical chromosomes were FISH may be due to two forms of genomic instability, exclusively observed in p537/7 mice. They seem to karyotypic instability (gain of chromosomes) and/or undergo chromatid fragmentation and degradation gene ampli®cation. The latter may also involve the and often showed gene ampli®cation as well (Figure formation of extrachromosomal elements which is 5d). Apoptotic cells characteristically displayed multi- frequently found in early stages of tumorigenesis ple copies of centrosomes (Figure 5g), aneuploidy, c- (Wahl, 1989). We observed extrachromosomal ele- Myc overexpression (Figure 5f), and/or gene amplifi- ments hybridizing with either DHFR or c-myc in all cations (Figure 5d). organs of 4 ± 6 week old p537/7 mice; a representative picture is shown in Figure 3e. Based on the data presented above, we suggest the Discussion presence of two types of genomic instability in p53 de®cient mice, namely gene ampli®cation and karyo- p537/7 mice develop a variety of tumors early in life typic instability (gain or loss of chromosomes). (Donehower et al., 1992; Jacks et al., 1994), but the mechanisms underlying this tumor predisposition have remained elusive. Here, we have shown that cells c-Myc overexpression and ampli®cation of c-myc, directly isolated from p537/7 mice display extensive DHFR and CAD in the same p537/7 cells aneuploidy and gene ampli®cation. This enhanced Since c-Myc overexpression has been shown to be genomic instability seems to be coupled with the high associated with locus speci®c gene ampli®cation, we susceptibility and frequency of tumor development in examined the levels of c-Myc protein in p537/7 bone these mice. marrow-, thymus-, spleen-derived cells as well as Genomic instability initiates during embryonic ®broblasts by quantitative ¯uorescent immunohisto- development and increases during life; haematopoietic chemistry (Materials and methods). Five to ®fteen cells of the fetal liver displayed 5% of aneuploidy, percent of the p537/7 cells in all tissues examined while all organs of young (4 ± 6 week old) exhibited showed a two- to eightfold increase in c-Myc higher levels of aneuploidy (Figure 1). expression, while less than 1% of p53+/+ cells expressed Abnormal ampli®cation of centrosomes occurs in detectable levels of c-Myc protein. To determine p537/7 mice in vivo, in all cell types tested. The adverse whether c-Myc deregulation and genomic instability of eects of abnormal centrosome ampli®cation are single genes occurred within the same p537/7 cell(s) in readily observed in cells undergoing mitosis. For vivo, we developed the Combined Protein/FISH instance, the formation of aberrant mitotic spindles Analysis (CPFA) (Materials and methods). This assay organized by multiple copies of centrosome was allows the simultaneous analysis of c-Myc protein levels frequently observed in p537/7 mice. Such events are and FISH hybridization signals in the identical cell(s) in likely to impair chromosomal segregation, and induce vivo. It therefore allows one to conclude whether c-Myc karyotypic instability and aneuploidy. overexpression is associated with genomic instability in Consistent with in vitro studies demonstrating that the same cell(s) in vivo. A similar technique has been c-Myc expression is negatively regulated by p53 independently established by Hessel et al. (1996). (Ragimov et al., 1993), we found that in the absence Using CPFA, we observed c-Myc overexpression of p53, all tissues became permissive to c-Myc and ampli®cation of DHFR,c-myc, and CAD genes, overexpression. However, only 5 ± 15% of p537/7 cells Genomic instability and apoptosis in p537/7 mice K Fukasawa et al 1298 expressed high levels of c-Myc protein, suggesting that The absence of p53 and the deregulation of c-Myc the additional event(s) may be required for c-Myc expression seem to cooperate during tumor develop- overexpression to occur and/or that cells overexpres- ment. For instance, c-Myc overexpression and lack of sing c-Myc undergo apoptosis (Figure 5; Evan et al., p53 have synergistic eects during lymphomagenesis in 1992; Packham and Cleveland, 1995). Emmyc/p53+/7 and CD2-myc/p537/7 mice (Blyth et al.,

Figure 3 Genomic instability in p537/7 mice. Representative images of FISH analyses of DHFR and c-myc gene copies in p53+/+ and p537/7 mice. (a) Single copy DHFR signals (green) overlaid on DAPI staining in thymocytes of a parental p53+/+ C57BL/6 mouse. (b) Ampli®ed signals of DHFR (green) overlaid on DAPI staining in p537/7 splenocytes. (c) Ampli®ed signals of DHFR (pink) and c-myc (green) overlaid on DAPI staining in p537/7 thymocytes. (d) Ampli®ed signals of DHFR (red) and c- myc (green) overlaid on DAPI staining in p537/7 bone marrow cells. (e) Metaphase plate with extrachromosomal elements, indicative of early stages of gene ampli®cation (Wahl, 1989), hybridizing with c-myc and DHFR probes. The most intense hybridization signals are pointed at by arrows. Note that there are many tiny hybridization signals as well. Filled arrow point to DHFR signals (pink) and open arrows to c-myc signals (green) Genomic instability and apoptosis in p537/7 mice K Fukasawa et al 1299

Figure 4 Representative images showing c-Myc expression levels and gene ampli®cation within the same cells in vivo using CPFA analyses (Materials and methods). (a) p537/7 ®broblasts (passage 0) were immunostained with anti-c-Myc antibody. Two nuclei are shown that overexpress c-Myc fourfold. (a') The same cells show DHFR (red) and c-myc (green) ampli®cation by FISH analysis. The nuclei are stained with DAPI. (a'') This image allows one to visualize all FISH hybridization signals obtained for c-myc (green) and DHFR (red) in (a') in the absence of DAPI staining. (b) p537/7 fetal liver-derived hematopoietic cells were immunostained with anti-c-Myc antibody. Note a four- to ®vefold c-Myc overexpression in the nuclei. (b') shows the overlay image of c-Myc staining (red) and DAPI staining of the nuclei (blue) shown in (b). (b'') FISH analysis with a CAD probe (green) was performed on the same cells. The nuclei are counterstained with propidium iodide

1995; Hsu et al., 1995). Thus, c-Myc may play an enhanced frequency in p537/7 cells (Yin et al., 1992; important role in the overall tumor susceptibility of Livingstone et al., 1992). A role for c-Myc in PALA- p537/7 mice by contributing to genomic instability, induced CAD ampli®cation has not been described. such as locus-speci®c gene ampli®cation as well as Recent work suggests that c-Myc is involved in the incresaed proliferation rates (Karn et al., 1989). transcriptional regulation of the CAD gene (Boyd and It has been previously shown that cell lines Farnham, 1997). Since c-Myc acts both as a overexpressing c-Myc display DHFR ampli®cation, transcription and replication factor, one may propose independent of species and tissue origins (Mai et al., a role for c-Myc not only in the transcriptional 1996). Moreover, the c-myc gene was both translocated activation, but also in the replication/ampli®cation of and ampli®ed, and the protein was overexpressed in the CAD gene. Consistent with this idea, we have mouse plasmacytoma cells (Mai et al., 1995). In this recently observed the PALA-dependent upregulation of study, we tested the genomic stability of c-myc, DHFR, c-Myc protein levels in p537/7 ®broblasts. Moreover, and CAD, all of which were ampli®ed concomitant PALA-induced c-Myc overexpression occurred prior to with c-Myc overexpression in cells isolated from p537/7 CAD gene ampli®cation (SM, unpublished observa- mice. tion). c-Myc overexpression may thus precede the While no mutagens were used in this study, several genomic instability of the CAD gene as it does for groups have shown earlier that the administration of DHFR (Mai and Jalava, 1994; Mai et al., 1996), cyclin the drug PALA [N-(phosphonoacetyl)-L-aspartate] D2 (Mai et al., 1997) and ribonucleotide reductase R2 leads to the selection of drug-resistant cells with (Kuschak et al., 1997). Further investigation will show ampli®ed CAD genes (Otto et al., 1989; Yin et al., whether c-Myc deregulation is a necessary and limiting 1992; Livingstone et al., 1992), and this occurs with an molecular event in CAD gene ampli®cation. Genomic instability and apoptosis in p537/7 mice K Fukasawa et al 1300

Figure 5 Apoptotic p537/7 cells exhibit atypical chromsome morphology, gene ampli®cation, elevated c-Myc protein levels and abnormally ampli®ed centrosomes. Metaphase plates were prepared and evaluated as described (Mai, 1994; Mai et al., 1995, 1996). (a) shows a p537/7 thymus-derived Giemsa-stained aneuploid metaphase plate. (b) p537/7 spleen-derived chromosomes with atypical morphology. Such chromosomes were present in all organs examined. (c) TUNEL assay was performed on p537/7 bone marrow-derived morphologically atypical chromosomes. Extensive chromatid fragmentation was observed, as shown by a strong positive TUNEL reaction (yellowish green). DNA was counterstained with propidium iodide. Intact DNA stretches can be identi®ed in orange, and the fragmented chromatids can be recognized by their yellowish green staining. (d) FISH analysis of bone marrow- derived chromosomes with atypical morphology. Chromosomes were counterstained with DAPI. Both c-myc (green) and DHFR (red) were ampli®ed. (e) Spleen-derived interphase cells with chromatin condensation typical of apoptosis (arrows) and DNA fragmentation as determined by the TUNEL assay (dUTP-¯uorescein incorporated by TdT) (yellowish green, arrows). DNA was counterstained with PI. The number of apoptotic cells in p537/7 mice ranges between 2 and 11% in thymus, spleen, ®broblasts and bone marrow. (f) Representative image showing apoptosis in p537/7 thymocytes. p537/7 thymocytes were immunostained for c- Myc (red) along with TUNEL assay and DAPI staining (blue). The cell indicated by an arrow shows chromatin condensation typical of apoptotic cells, and expresses high levels of c-Myc. Fragmentaed nuclei and damaged DNA can be recognized by yellow/ yellowish green staining. (g) shows a p537/7 bone marrow cell displaying a typical apoptotic phenotype immunostained with anti-g- tubulin. The centrosomes are indicated by arrows

Genomic instability in p537/7 embryos and young normally. However, the yield of p537/7 ospring from mice aects several genetic loci, includes c-Myc heterozygous crosses has been reported to be only overexpression, abnormal centrosome numbers and *60% of the expected yield (Jacks et al., 1994), aneuploidy. Further genomic and molecular altera- suggesting that some homozygous lethality occurs tions, such as changes in the expression and/or half life during embryogenesis. Indeed, we observed genomic of additional oncogenes, cell cycle related genes, alterations in a fraction of embryonic cells (Fukasawa growth factor-mediated signaling, DNA repair, etc., et al., 1996, and this study). Since c-Myc has been are conceivable, but have not been examined here. implicated in apoptosis (reviewed in, Amati and Land, During the multistage process of carcinogenesis, all of 1994; Harrington et al., 1994; Packham and Cleveland, the above events may ultimately contribute to the 1995) and apoptosis occurs in the presence of elevated selection and evolution of neoplastic cell(s). c-Myc expression in p537/7 mice, we are currently Despite extensive genomic alterations, p537/7 mice examining the possible role of c-Myc in apoptosis of mature without discernible abnormalities until tumors genomically unstable p537/7 cells. start to appear (Donehower et al., 1992; Jacks et al., 1994). We show here a high incidence of apoptosis in all p537/7 tissues. In addition to interphase cells Materials and methods exhibiting chromosome condensation and a positive TUNEL staining, we detected chromosomes with Mice, cell suspensions, cell culture atypical morphology. These chromosomes stained in the modi®ed TUNEL reaction and thus showed Mice were obtained from Taconic farms (CA, USA), with chromatid fragmentation. It is noteworthy that these the exception of two p53+/+ parental C57B1/6 mice that apoptotic cells usually contained abnormally ampli®ed were obtained from Charles River (Quebec, Canada). Nine p537/7 and six p53+/+ mice between 4 and 6 weeks old centrosomes, expressed high levels of c-Myc protein, were used to prepare single cell suspensions of spleen, bone and displayed aneuploidy and gene ampli®cations. marrow, and thymus. Ten-day-old skin- and spleen-derived Thus, many of the genetically abberant cells may be primary ®broblasts (passage 0) were also obtained from eliminated through p53-independent apoptosis. This these mice by explanting subcutaneous skin tissue pieces may explain how p537/7 mice seemingly develop and spleen fragments into RPMI 1640 medium supple- Genomic instability and apoptosis in p537/7 mice K Fukasawa et al 1301 mented with 10% fetal calf serum, 2 mM L-glutamin, and Combined protein/FISH analysis (CPFA) 1mM sodiumpyruvate. Hematopoietic cells were prepared To visualize c-Myc protein expression and genomic from fetal livers of 16-day-old embryos. Ten fetal livers (in)stability within the same cells, we combined immunohis- were pooled for further analyses. tochemical analysis and FISH. Cells immobilized on slides were ®xed (3.7% formaldehyde), permeabilized (0.2% g-tubulin immunostaining Triton X-100), incubated with 20 ng/slide of the anti-c- Myc monoclonal antibody (3C7, Evan et al., 1985), followed Cells isolated from mice were seeded onto the slides. by the secondary Texas Red-conjugated goat anti-mouse Samples were ®xed in 3.7% formaldehyde in phosphate IgG (Southern Biotechnology Associates, Inc, USA) (10 ng/ buered saline (PBS) for 20 min at room temperature slide). The nuclei were counterstained with DAPI (1 mg/ml in (RT). Cells were then incubated in the blocking solution PBS), photographed, and the positions were recorded. (10% normal goat serum in PBS) for 1 h at RT. The Thereafter, the slides were processed for FISH analysis as samples were then incubated with anti-g-tubulin antibody described (Mai, 1994; Mai et al., 1996). Cells in the recorded raised against CSREIVQQLIDEYHAATRPDYISWGTQ positions were re-evaluated for their respective gene copy for 1 h at 378C. The samples were then washed extensively numbers. A microscopic ®eld of 104 to 105 cells was analysed, in PBS, followed by incubation with ¯uorescein isothio- and 100 cells were evaluated per sample. As above, relative cyanate (FITC)-conjugated goat anti-rabbit immunoglobu- ¯uorescent intensity per pixel was determined using the lin G (IgG) for 30 min at RT. The samples were then IPLabSpectrum software (Signal Analytics, USA). washed extensively in Tris buered saline (TBS), followed by 4',6' diamidino-2-phenylindole (DAPI) DNA staining (1 mg/ml). For each cell type, 4400 cells were examined. Apoptosis assays The TUNEL assay was performed on interphase cells as c-Myc determination described (Li et al., 1995). A modi®ed TUNEL assay was designed to visualize gaps and DNA fragmentation on Quantitative ¯uorescent immunohistochemistry was used metaphase chromosomes. Metaphase chromosomes were to determine c-Myc protein levels in single cells directly prepared as described (Mai, 1994; Mai et al., 1996). Brie¯y, isolated from the mice. Following immunohistochemistry the slides underwent ®xation, RNAse and pepsin treat- with 20 ng/slide of a monoclonal anti c-Myc antibody 3C7 ments, and post®xation. Thereafter, the apoptotic assay (Evan et al., 1985) and a secondary antibody (anti-mouse was performed using dUTP-¯uorescein and the terminal IgG-Texas Red; Southern Biotechnology Associates, Inc., deoxynucleotidyl transferase (TdT) enzyme according to USA) at 10 ng/slide, images were acquired using a Zeiss the suppliers (Boehringer Mannheim, Canada). Chromo- Axiophot microscope, coupled to a CDD camera (Optikon/ somes were counterstained with propidium iodide (1 mg/ml) Photometrics), and the relative ¯uorescence intensity per and analysed as described (Mai, 1994; Mai et al., 1996). pixel (one pixel=6.8 mm)wasanalysedonaPowerMac Chromatin condensation was visualized with DAPI (1 mg/ 8100, using IPLabSpectrum and Multiprobe softwares, ml). One hundred to three hundred cells were evaluated per version 3.1 (Signal Analytics, USA). One hundred to three sample. hundred cells were evaluated per sample.

Cytogenetics and ¯uorescent in situ hybridization (FISH) Abbreviations Metaphase spreads for cytogenetic analysis, FISH and a CAD, carbamoyl-phosphate synthetase-aspartate transcar- modi®ed TUNEL assay (see below) were performed bamoyl-dihydroorotase; DHFR, dihydrofolate reductase; according to standard protocols (Mai, 1994; Mai et al., FITC, ¯uorescein isothiocyanate; DAPI, 4',6'diamindino-2- 1996) using cells directly isolated from the mice. Analyses phenylindole; FISH, ¯uorescent in situ hybridization; of interphase cells by FISH, CPFA (see below) and the CPFA, Combined Protein/FISH Analysis; R1, ribonucleo- TUNEL assay (Li et al., 1995) were carried out on cytospin tide reductase R1; PALA, N-(phosphonoacetyl)-L-aspar- preparations. FISH determination of metaphase chromo- tate. somes and interphase cells was performed as described (Mai, 1994; Mai et al., 1996). All probes used have been described elsewhere (Mai et al., 1996), except the CAD Acknowledgements probe (a generous gift from Dr George Stark, The We thank Drs M Mowat, N Stewart, AH Greenberg, JA Cleveland Clinic Foundation, OH). The number of Wright (Winnipeg), D Litch®eld (London, Ontario), G metaphase plates evaluated for cytogenetic analyses was Brandner and RD Hess (Freiburg), and K Vousden 50 per organ and 100 per fetal liver cells. The number of (Frederick) for stimulating discussions and critical reading metaphase plates and interphases evaluated for FISH of this manuscript. We also thank Drs JL Hamerton, P analyses was 100 per sample. Using the IPLabSpectrum McAlpine (Winnipeg) and H Klinger (New York) for and Multiprobe softwares (Signal Analytics, USA), discussions of morphologically atypical chromosomes. We ampli®ed signals were determined with the `Line Measure- also thank T Copeland for the production of anti-g-tubulin ment' function; relative ¯uorescent intensities per pixel antibody. This work was supported by National Cancer (one pixel=6.8 mm) were measured for single and ampli®ed Institute of Canada/Terry Fox Development Fund and hybridization signals. As signal is classi®ed as ampli®ed if Manitoba Health and Research Council grant (SM) and the ratio between the relative ¯uorescent intensity per pixel sponsored in part by the National Cancer Institute, of ampli®ed vs the relative ¯uorescent intensity per pixel of Department of Human Health Service, under contract single copy signals is 42 in one hundred interphase cells. with ABL (KF and GFV).

References

Alitalo K. (1985). TIBS, 10, 194 ± 197. Blyth K, Terry A, O'Hara M, Baxter EW, Campbell M, Amati B and Land H. (1994). Curr. Opin. Genet. & Dev., 4, Stewart M, Donehower LA, Onions DE, Neil JC and 102 ± 108. Cameron ER. (1995). Oncogene, 10, 1717 ± 1723. Bishop JM. (1995). Genes & Dev., 9, 1309 ± 1315. Genomic instability and apoptosis in p537/7 mice K Fukasawa et al 1302 Boyd KE and Farnham PJ. (1997). Mol. Cell. Biol., 17, Livingstone LR, White A, Sprouse J, Livanos F, Jacks F and 2529 ± 2537. Tlsty TD. (1992). Cell, 70, 923 ± 935. Cross SM, Sanchez CA, Morgan CA, Schmike MK, Ramel Mai S. (1994). Gene, 148, 253 ± 260. S, Idzerda RL, Raskind WH and Reed BJ. (1995). Science, Mai S and Jalava A. (1994). Nucl. Acid Res., 22, 2264 ± 2273. 267, 1353 ± 1356. Mai S, Hanley-Hyde J, Coleman A, Siwarski D and Huppi Denis N, Kitzis A, Kruh J, Dautry F and Crocos D. (1991). K. (1995). Genome, 38, 780 ± 785. Oncogene, 6, 1453 ± 1457. Mai S, Hanley-Hyde J and Fluri M. (1996). Oncogene, 12, Donehower LA, Harvey M, Slagle B, McArthur MJ, 277 ± 288. Montgomery Jr CA, Butel JS and Bradley A. (1992). Mai S, Hanley-Hyde J, Rainey GJ, Fluri M, Littlewood TD, Nature, 356, 215 ± 221. Stevens LM, Henderson DW and Mushinski JF. (1997). Evan GI, Lewis GK, Ramsay G and Bishop JM. (1985). Mol. Submitted. Cell. Biol., 5, 3610 ± 3616. Marcu KB, Bossone SA and Patel AJ. (1992). Ann. Rev. Evan GI, Wyllie AH, Gilbert CS, Littlewood TD, Land H, Biochem., 61, 809 ± 860. Brooks M, Waters CM, Penn LZ and Hancock DC. Nowell P. (1976). Science, 194, 23 ± 28. (1992). Cell, 69, 119 ± 128. Oakley BR. (1992). Trends Cell Biol., 2, 1±5. Feo S, Liegro CD, Jones T, Read M and Fried M. (1994). Orren DK, Petersen LN and Bohr VA. (1995). Mol. Cell. Oncogene, 9, 955 ± 961. Biol., 15, 3722 ± 3730. Fukasawa K, Choi T, Kuriyama R, Rulong S and Vande Otto E, McCord S and Tlsty TD. (1989). J. Biol. Chem., 264, Woude GF. (1996). Science, 271, 1744 ± 1747. 3390 ± 3396. Guillouf C, Rosselli F, Krishnaraju K, Moustacchi E, Packham G and Cleveland JL. (1995). Biochim. Biophys. Homan B and Liebermann DA. (1995). Oncogene, 10, Acta, 1242, 11 ± 18. 2263 ± 2270. Pienta JK, Partin AW and Coey DS. (1989). Cancer Res., Harrington EA, Fanidi A and Evan GI. (1994). Curr. Opin. 49, 2525 ± 2532. Genet. & Dev., 4, 120 ± 129. RagimovN,KrauskopfA,NavotN,RotterV,OrenMand Hartwell L. (1992). Cell, 71, 543 ± 646. Aloni Y. (1993). Oncogene, 8, 1183 ± 1193. Hartwell LH and Kastan MB. (1994). Science, 266, 1821 ± Solomon E, Borrow J and Goddard AD. (1991). Science, 1828. 254, 1153 ± 1160. Hessel H, Mittermuller J, Zitzelsberger H, Weier H-U and Stark G. (1993). Adv. Cancer Res., 48, 87 ± 113. Bauchinger M. (1996). Histochem. Cell Biol., 106, 481 ± Stewart N, Hicks GG, Paraskevas F and Mowat M. (1995). 485. Oncogene, 10, 109 ± 115. Hsu B, Marin MC, El-Naggar AK, Stephens LC, Brisbay S Temin HM. (1988). Cancer Res., 48, 1697 ± 1701. and McDonnell TJ. (1995). Oncogene, 11, 175 ± 179. Wahl GM. (1989). Cancer Res., 48, 1333 ± 1340. Huang A and Wright JA. (1994). Oncogene, 9, 491 ± 499. Wahl AF, Donaldson KL, Fairchild C, Lee FYF, Foster SA, Jacks T, Remington L, Williams BO, Schmidt EM, Halachmi Demers GW and Galloway DA. (1996). Nat. Med., 2, 72 ± S, Bronson RT and Weinberg R. (1994). Curr. Biol., 4, 1± 79. 7. Weinert T and Lydall D. (1993). Sem. Cancer Biology, 4, Joshi HC. (1994). Curr. Opin. Cell Biol., 6, 54 ± 62. 129 ± 140. Karn J, Watson JV, Lowe AD, Green SM and Vedeckis W. Van der Bliek AM, Van der Velde-Koerts T, Ling V and (1989). Oncogene, 4, 773 ± 787. Borst P. (1986). Mol. Cell. Biol., 6, 1671 ± 1678. Kastan MB, Canman CE and Leonard CJ. (1995). Cancer Yin Y, Tainsky MA, Bischo FZ, Strong L and Wahl GM. and Metastasis Rev., 14, 3 ± 15. (1992). Cell, 70, 937 ± 948. Kuschak TI, Taylor C, Henderson DW, Mushinski JF, Yokota Y, Tsunetsugu-Yokota Y, Battifora CL and Cline Wright JA and Mai S. (1997). Submitted. MJ. (1986). Science, 231, 261 ± 265. Li X, Traganos F, Melamed MR and Darzynkiewicz Z. (1995). Cytometry, 20, 172 ± 180.