ABSTRACT
Defects in DNA mismatch repair predispose cells to the development of several types of malignant disease. The absence of Msh2 or Mlh1, two key molecules that mediate mismatch repair in eukaryotic cells, increases the frequency of mutation and also alters the response of some cells to apoptosis and cell cycle arrest. To understand the way these changes contribute to cancer predisposition, we examined the effects of defective mismatch repair on the multistep process of pre-B-cell transformation by Abelson murine leukemia virus. In this model, primary transformants undergo a prolonged apoptotic crisis followed by the emergence of fully transformed cell lines. The latter event is correlated to a loss of function of the p53 tumor suppressor protein and down-modulation of the p53 regulatory protein p19Arf. Analyses of primary transformants from Msh2 null mice and their wild-type littermates revealed that both types of cells undergo crisis. However, primary transformants from Msh2 null animals recover with accelerated kinetics, a phenomenon that is strongly correlated to the appearance of cells that have lost p53 function. Analysis of the kinetics with which p53 function is lost revealed that this change provides the dominant stimulus for emergence from crisis. Therefore, the absence of mismatch repair alters the molecular mechanisms involved in transformation by affecting a gene that controls apoptosis and cell cycle progression, rather than by affecting these processes directly.
DNA mismatch repair (MMR) is a critical cellular defense mechanism which can correct endogenous errors in DNA replication or errors resulting from DNA damage (reviewed in references 15 and 34). In mammalian cells, MMR is orchestrated by six proteins: Msh2, Msh6, and Msh3, which recognize mispaired nucleotides, and Mlh1, Pms2, and Pms1, which facilitate the assembly of factors required for DNA excision and resynthesis (reviewed in reference 27). The importance of this system in regulating genomic stability has been emphasized by the finding of frequent loss of Msh2 and Mlh1 function in hereditary nonpolyposis colorectal cancer and sporadic malignancies, including colorectal, endometrial, and ovarian cancers (4, 16, 29, 37). In addition, targeted inactivation of the Msh2,Mlh1, Pms1, or Pms2 gene in mice results in a high susceptibility to tumor formation (12, 13, 40, 45).
Consistent with an absence of MMR, cells lacking Msh2 or Mlh1 exhibit dramatic instability in microsatellite sequences (5, 44) and show elevated rates of mutations in endogenous genes, including genes encoding transforming growth factor β receptor II (TGF-β RII), APC, and Bax, which are involved in control of cellular growth and apoptosis (23, 30, 32, 42, 46). In addition to having effects on genomic stability, Msh2 appears to play a direct role in mediating apoptosis and regulating the cell cycle. For example, Msh2-deficient cells are highly resistant to apoptosis induced by a number of agents (13, 17, 52), and overexpression of Msh2 or Mlh1 induces apoptosis in both MMR-deficient and MMR-competent cells (57). Cells lacking Msh2 or Mlh1 activity also fail to undergo cell cycle arrest in response to DNA damage (6, 10, 20), and both proteins interact with proliferating cell nuclear antigen (53), a multifunctional protein that is intimately involved with several components of the cell cycle machinery (reviewed in reference 25). Despite these data and the clear effects of Msh2 loss on cancer susceptibility, the way in which effects on mutation, cell cycle control, and apoptosis contribute to malignancy is not fully understood.
Unraveling the role of Msh2 in tumor development is complicated by the fact that cells are often examined after they have acquired a tumorigenic phenotype, making it difficult to separate causes of tumor development from changes that are acquired during oncogenesis. One experimental model that circumvents this problem is pre-B-cell transformation by Abelson murine leukemia virus (Ab-MLV). This highly oncogenic retrovirus contains the v-abl oncogene and transforms pre-B cells in vitro and in vivo (49). Although transformation requires the protein tyrosine kinase activity of the v-Abl oncoprotein, Ab-MLV-mediated oncogenesis is a multistep process (18, 19, 41, 54, 55). In vitro, primary transformants undergo a prolonged apoptotic crisis, after which a fraction of the transformants emerge as fully malignant cell lines (41, 54). About 50% of those which emerge from crisis have acquired p53 mutations (51); most of the others have down-regulated p19Arf (41), a protein that regulates p53 levels via interaction with Mdm2 (reviewed in reference50). The importance of these changes is highlighted by the observation that primary transformants from p53 null or Ink4a/Arf null mice do not undergo crisis (41, 54).
To probe the function of MMR in Ab-MLV pre-B-cell transformation, primary pre-B-cell transformants were derived fromMsh2 null animals. Although Msh2 does not affect the initial frequency of transformation or the onset of crisis, the absence of MMR affects recovery from crisis. Cells lackingMsh2 emerge from crisis faster and at a higher frequency than do wild-type cells, a change which coincides with the acquisition of p53 mutations. Systematic analysis of the kinetics with which cells expressing mutant p53 accumulate in the population demonstrated that acquisition of p53 mutations provides the dominant stimulus for emergence from crisis. Therefore, the absence of MMR alters the outcome of Ab-MLV infection through effects on a gene involved in the control of apoptosis and cell cycle progression, rather than by altering these processes directly.
MATERIALS AND METHODS
Cells and mice.Pre-B cells transformed by Ab-MLV were maintained in RPMI 1640 medium supplemented to contain 10% heat-inactivated fetal calf serum (Sigma), 50 μM 2-mercaptoethanol, 2 mM l-glutamine, 50 μg of streptomycin per ml, and 50 U of penicillin per ml at 37°C in a 6% humidified CO2atmosphere. The established transformed cell lines used included 38B9, which expresses wild-type p53 (41), L1-2, which is null for p53 (56), and IA-17, a transformant derived fromInk4a/Arf null mice (41). Primary transformants were derived by infecting bone marrow cells with Ab-MLV-P160 and plating the cells in soft agar as described previously (48). Macroscopically visible colonies of primary transformants were isolated from agar 10 days later and plated in 24-well plates in supplemented RPMI 1640 medium containing 20% fetal calf serum. The cultures were monitored daily for cell density and viability and subcultured when cells filled the wells. When the viability of the cells was >85%, the cells were transferred to a 35-mm dish and subcultured as before. When the cells were consistently >90% viable and could be subcultured routinely, they were considered established (41, 54). In some experiments, cells were seeded at 106/ml and the number and viability of cells in the culture were determined daily by counting cells after trypan blue staining. The cultures were reseeded when the population had doubled and the process was repeated until the population maintained high viability and had a predictable doubling time. During this period, samples were treated with approximately 1,000 rads of gamma irradiation by using a cesium 137 Gammacell-100 source, and cultures were monitored for viability 10 to 12 h later by visual inspection or trypan blue staining (41, 51).Msh2−/− mice with a mixed 129/Ola and C57BL/6J background (45) were bred by mating heterozygous animals at the Tufts University animal facility. The genotype of the mice was determined by PCR amplification using a combination of primers specific for Msh2 and for the targeted locus. The wild-type allele was amplified by using the primers 5′-AAAGTGCACGTCATTTGGA-3′ and 5′-GCTCACTTAGACGCCATTGT-3′; the latter primer and a primer specific for the targeted locus, 5′-GCCTTCTTGACGAGTTCTTC-3′, were used to amplify the targeted allele. Each PCR mixture contained 200 ng of DNA, 200 μM deoxynucleoside triphosphate mix (Pharmacia), 1× PCR buffer (Perkin-Elmer Cetus), a 1 μM concentration of each primer in reactions designed to amplify the wild-type allele or a 2 μM concentration of each primer in reactions designed to amplify the targeted allele, and 1 U of Taq polymerase (Perkin-Elmer Cetus). Samples were amplified for 40 cycles of 94°C for 1 min, 58°C for 30 s, and 72°C for 1 min, followed by a 5-min extension at 72°C; the products were fractionated on agarose gels and visualized by ethidium bromide staining.
Apoptosis and cell cycle analysis.Cells were harvested and frozen in supplemented RPMI medium containing 20% fetal calf serum and 10% dimethyl sulfoxide. At the time of analysis, the cells were thawed, washed in phosphate-buffered saline, stained with propidium iodide solution (10 μg/ml in H2O), and immediately analyzed by flow cytometry with a FACScan instrument and Modfit software (7). In some experiments, the frequency of apoptotic cells obtained by using propidium iodide staining was compared to that obtained by staining cells with merocyanin 540, a dye that specifically stains apoptotic cells (43). Similar results were obtained using either method.
RNA and sequence analysis.Total RNA was prepared using the RNeasy kit (Qiagen) according to the manufacturer's instructions. p19Arf cDNA was prepared from total cellular RNA by reverse transcription-PCR (RT-PCR) as described previously (54). The same protocol was followed to amplify the 1.3-kb full-length p53 cDNA, using the primers 5′-GTGTCTCAGCCCTGAAGTCATAAGAC-3′ and 5′-CTAGCATTCAGGCCCTCATCCT-3′ (51). The PCR products were cloned into the TOP10 cloning vector (Invitrogen) and sequenced on an ABI373 stretch machine (Perkin-Elmer) at the DNA Facility, Department of Physiology, Tufts University. The p53 sequence data were compared to the reported human p53 mutations in the IARC p53 Germline Mutation database (www.iarc.fr/p53/GERM.HTM ) and the European Bioinformatics Institute database (ftp://ftp.ebi.ac.uk/pub/databases/p53 ).
Protein analysis.Cells were harvested, washed once in phosphate-buffered saline, and lysed in lysis buffer (10 mM Tris [pH 7.4], 1% sodium dodecyl sulfate, 1 mM phenylmethylsulfonyl fluoride, and 1 mM sodium orthovanadate). The lysates were heated at 95°C for 5 to 10 min, and either 50 μg of the sample or the entire sample was fractionated through a sodium dodecyl sulfate-polyacrylamide gel. The samples were transferred to polyvinylidene difluoride membranes (Millipore), which were probed with anti-p19Arf (NB200-106; Novus Biological), anti-cdk4 (C-21; Santa Cruz Biotechnology), anti-Gag/v-Abl (H548) (8), and anti-p53 (Ab-3; Oncogene Research Products) antibodies. The blots were treated according to the Western-Light kit protocol (Tropix), using alkaline phosphatase-conjugated secondary antibodies and the CSPD substrate (Tropix).
RESULTS
Msh2 status does not influence events in primary pre-B-cell transformation.To determine if the absence ofMsh2 influences the frequency of cells susceptible to Ab-MLV transformation, bone marrow cells from Msh2 −/−and Msh2 +/+ mice were infected with Ab-MLV-P160 and plated in soft agar (48). Macroscopic colonies of primary transformants were counted 10 days later. Analyses of multiple sets of littermates revealed that transformation frequency did not correlate with the presence of a functional Msh2 gene (Table1). Although more transformants were recovered from Msh2 null animals in two experiments, this pattern was reversed in another experiment, and similar numbers of transformants were recovered in a fourth analysis. This variation most likely reflects the mixed background of the mice; susceptibility to Ab-MLV transformation is strongly influenced by genetic background (48), and similar differences have been observed when other animals from mixed backgrounds have been analyzed (reference54 and our unpublished data). Thus, the presence of Msh2 does not alter the frequency of pre-B cells susceptible to primary Ab-MLV transformation.
Msh2 does not affect Ab-MLV transformation frequency a
Absence of MMR increases the frequency of p53mutations.The absence of MMR increases the frequency with which mutations accumulate, and elevated frequencies of mutations in genes that may affect the oncogenic process have been identified in some MMR-deficient tumor cells (23, 32, 42, 46). Thep53 tumor suppressor gene is mutated in approximately 50% of Ab-MLV-transformed pre-B cells (51). To determine if the absence of Msh2 increases the frequency of p53 mutations, established transformants from Msh2 null and wild-type mice were screened for loss of p53 function by using susceptibility to gamma irradiation-induced apoptosis. Ab-MLV-transformed pre-B cells expressing mutant forms of p53 are resistant to apoptosis by ionizing radiation (51). Analyses of more than 100 transformants derived from Msh2−/− mice revealed that 93% survived irradiation, implying the presence of mutant p53; among 25 transformants derived from wild-type littermates, only 28% survived irradiation.
Nearly all Ab-MLV-transformed pre-B cells that harbor p53mutations express readily detectable p53 protein in the absence of irradiation because the stability of the protein is altered as a consequence of the mutation (52). To confirm that the transformants screened by gamma irradiation displayed the expected pattern of p53 expression, 40 transformants, including the representatives shown in Fig. 1, were analyzed by Western blotting with anti-p53 antibodies. Consistent with expression of wild-type p53, cells that underwent rapid apoptosis following irradiation expressed very low levels of p53 under control conditions, and these levels increased following irradiation. Most cells that did not undergo rapid apoptosis following irradiation expressed abundant p53 in the absence of treatment. Three transformants derived from Msh2 null animals failed to express detectable levels of p53 (Fig. 1 and data not shown). Loss of p53 expression has been observed in erythroleukemias induced by Friend leukemia virus (2) but is not common in Ab-MLV-transformed cells (51, 56).
Some Msh2 null transformants do not express p53. Established transformants from Msh2 null animals and wild-type mice were treated with approximately 1,000 rads of gamma irradiation, and lysates were prepared from these cells (+) and mock-treated cells (−) 4 h later. The proteins were examined by using Western blotting. The blots were cut and one portion was probed with the anti-p53 monoclonal antibody Ab-3 (Oncogene Research Products). The other portion of the blot was probed with the anti-Gag/v-Abl monoclonal antibody H548 (8) to control for protein loading. Null, lysate from L1-2, a p53-negative pre-B-cell line transformed with the Ab-MLV-P120 strain (56); WT, lysate from 38B9, a pre-B-cell line transformed with Ab-MLV-P160 that expresses wild-type p53 (51). Cell lines 10-18, 10-14, and 10-11 express mutant p53; these cell lines and 10-31, 10-2, and the null control do not undergo rapid apoptosis following high-dose gamma irradiation. The WT control, 10-5, and 10-12 cell lines express wild-type p53 and undergo rapid apoptosis following high-dose gamma irradiation (data not shown).
p53 mutations in the absence of MMR.The observation that three transformants derived from the Msh2null background failed to express p53 suggested that these cell lines might carry mutations that are distinct from those observed previously. The only reported Ab-MLV transformant that lacks p53 expression lost one copy of the gene and contains a proviral integration in the other copy (56); however, Southern analyses of DNA from theMsh2 null cells suggested that at least one copy of thep53 gene remained intact in the nonexpressing cells (data not shown). To explore the changes present in these cells more fully, RT-PCR was used to amplify p53 sequences from twoMsh2 null transformants that expressed mutant forms of p53 and from the three transformants that appeared to lack p53expression. An amplification product of the expected size was recovered from all samples. The PCR products were cloned and sequenced from at least two independent RT-PCRs, and only sequence changes present in both amplifications were considered significant. Similar to the pattern seen in other Ab-MLV transformants (51), a single mutation that affected the DNA binding domain was detected in the two cell lines which expressed abundant p53 (Fig. 2). A single nucleotide deletion or insertion that caused a frameshift affecting the carboxyl terminus was detected in the cDNAs recovered from the three transformants which failed to express detectable p53 protein. Insertion and deletion are common in cells lacking MMR, particularly in regions with stretches of A and G residues (32, 42); the mutation found in the 10-2 cell line occurred within a 7-base stretch of A residues, suggesting that this change may reflect the MMR deficiency. However, more clones would have to be analyzed to determine if the absence of MMR influences the types of mutations found in Msh2 null transformants.
p53 mutations found in Msh2 null transformants. p53 sequences were amplified from total cellular RNA isolated from five independently derived Msh2null transformants, and the cDNA was cloned and sequenced. The effect of each mutation is illustrated and reflects the following changes: A135V, GCG to GTG; R246Q, CGA to CAA; frameshift 298, deletion of a C at 1046; frameshift 314, insertion of a C at 1093; and frameshift 379, deletion of A at 1289. Conserved regions of the protein are shaded.
Lack of Msh2 increases the rate and frequency of establishment.Ab-MLV-induced transformants acquire p53mutations during an apoptotic crisis period that characterizes the multistep process of in vitro transformation (41). Only a fraction of the primary transformants survive this phase. The high percentage of p53 mutations in Msh2 null cells suggested that this process might be altered in the absence of MMR. To test this idea, the rate and frequency of outgrowth were examined for primary transformants derived from Msh2 −/−mice and normal littermates. Cells were monitored daily, and once a culture maintained high viability and grew predictably it was considered established (41, 54). Analyses of 170 primary transformants derived from normal mice revealed that about 15% of the cells became established (Fig. 3). This value is similar to that obtained with other strains of mice on similar mixed backgrounds (reference 54 and our unpublished data). In contrast to this pattern, about 70% of the 180Msh2 −/− primary transformants became established. The length of the crisis period was also affected by the presence of Msh2; primary transformants from null animals became established 15 to 21 days earlier than those from wild-type littermates (Fig. 3). Therefore, the absence of MMR allows for an enhanced rate and frequency of transition from a primary Ab-MLV-transformant to a fully established cell line.
Primary transformants from Msh2 null mice are established at a high frequency. Primary transformants fromMsh2 −/− (open symbols) and wild-type (filled symbols) mice were plated in liquid medium, and their ability to develop into established transformants was monitored. A transformant is considered established when levels of apoptosis are less than 10% and the cells can be subcultured at regular intervals (41, 54). Each point represents the frequency of primary transformants that were established at the day shown. Each line represents data from an independent experiment; a total of 180 transformants fromMsh2 null animals and 170 transformants from wild-type animals were examined.
Absence of MMR does not alter p19Arf expression.It is thought that elevated expression of the p19Arf protein during Ab-MLV transformation allows stabilization of p53 and leads to apoptotic crisis, thereby providing a selective advantage to cells that harborp53 mutations (41). To determine if the shortened crisis and increased frequency of p53 mutations inMsh2 null cells reflect an altered p19Arf response, the pattern of p19Arf expression in representative transformants was examined by Western blotting (Fig.4). The blots were stripped and probed with an anti-cdk4 antibody to control for protein loading. Readily detectable p19Arf protein was present in bothMsh2 −/− and Msh2 +/+transformants prior to the onset of crisis and during the crisis period. Thus, the pattern of p19Arf expression is unaffected by MMR status, suggesting that the phenotype of theMsh2 −/− cells is not controlled by altered p19Arf expression. Consistent with this idea, sequence analysis of p19Arf cDNAs from seven Msh2 −/−, oneMsh2 +/−, and two Msh2 +/+transformants revealed that no mutations had occurred in the p19Arf-coding sequence.
p19Arf expression is similar in primary transformants from Msh2 null and wild-type mice. Primary transformants from Msh2 null and wild-type mice were explanted from agar, and populations were expanded in liquid culture. Lysates were prepared from representative cells 4 and 8 days postexplantation and analyzed by Western blotting with anti-p19Arf antibodies (Novus Biological). The blot was reprobed with anti-cdk4 antibody to control for protein loading. C, controls with lysates from the Ink4a/Arf null cell line IA-17 (−) and the p19Arf-positive cell line L1-2 (+) (41).
Absence of Msh2 accelerates recovery from apoptotic crisis.Recovery from the crisis that characterizes Ab-MLV-induced pre-B-cell transformation involves both suppression of apoptosis and acquisition of a stable growth phenotype (41, 54). Msh2 has been reported to affect both apoptosis and cell cycling under certain conditions (6, 10, 11, 17, 20, 52). To determine if alteration in either or both of these parameters correlated to the accelerated establishment of the Msh2 null transformants, propidium iodide staining and fluorescence-activated cell sorter analysis were used to assess cell cycling and apoptosis. Similar to other primary transformants (41, 54), cells from both Msh2 null mice and their normal littermates displayed high levels of apoptosis 4 to 5 days postexplantation (data not shown), which was still evident after 2 weeks (Fig.5; Table2). Thus, the absence of MMR does not decrease the severity of apoptosis during crisis.
Primary transformants from Msh2 null and wild-type backgrounds undergo crisis. Primary transformants were plated in liquid medium, and populations were expanded. The plots shown are representative of an established cell line from theMsh2 −/− background (A), an established cell line from the Msh2 +/+ background (B), anMsh2 −/− transformant (18-6) emerging from crisis at 14 days postexplantation (C), anMsh2 −/− transformant (18-46) in crisis at day 14 (D), and an Msh2 +/+ transformant (18-137) in crisis at day 14 (E). Additional cell cycle information can be found in Table 2.
Msh2 null and wild-type transformants undergo crisis a
Although cell cycle analysis of transformants in crisis has not been reported, the unpredictable population doubling during this period (41, 54) suggests that altered cell cycle progression may contribute to the crisis phenomenon. Analysis of the cell cycle profile of primary transformants confirmed this idea; compared to established cell lines, in which about 40% of the population resides in G1, 60 to 70% of the cells in populations derived from both wild-type and MMR-deficient animals were in the G1 compartment 2 weeks after explantation (Fig. 5; Table 2). The increased frequency of cells in G1 was mirrored by a decreased frequency of cells in the S and G2phases of the cell cycle. These data demonstrate that the absence of Msh2 influences both the apoptotic program and the ability of cells to transit the G1 phase of the cell cycle during the establishment process.
Loss of p53 function is the critical step during the transformation process.The accelerated crisis recovery ofMsh2 null transformants suggests that Msh2 affects apoptotic and cell cycle regulatory pathways in a direct fashion. However, this phenotype also correlates with an increased frequency of p53mutation. Because a functional p53 is required for crisis (54), the lack of MMR may facilitate the accumulation ofp53 mutations. Thus, loss of p53 function may account for the ease with which Msh2 null transformants become established and for the accompanying effects on apoptosis and cell cycling. To test this idea, the relationship between loss of p53 function and emergence from crisis was compared in sevenMsh2 null and four wild-type cell lines. Cells were seeded at 106 per ml and growth and viability were monitored by counting the trypan blue-stained populations. When the population doubled, cells were reseeded at a concentration of 106 per ml. To monitor the frequency of cells expressing mutant forms of p53, samples of each culture were also treated with gamma irradiation, and the frequency of cells that did not undergo rapid apoptosis was monitored 10 to 12 h later. As expected, population doubling was erratic and cell viability was low early in the transformation process (Fig. 6). As the transformants emerged from crisis, viability increased and the cultures could be reseeded in a more predictable fashion. This phase is between days 28 and 38 in the example shown in Fig. 6. During this time, population doubling stabilizes, the viability of the cells steadily increases, and the frequency of p53 mutant cells within the culture increases. Throughout this phase, population doubling time is only slightly influenced by the presence of p53 mutant cells, because these cells do not cycle fast enough to have a noticeable impact on population doubling until they represent a very high percentage of the population. This growth profile was similar in Msh2 null and wild-type cells (data not shown). However, crisis was extended for several weeks in the wild-type transformants.
Loss of p53 function correlates with increased growth and viability. Primary transformants were explanted from agar, and populations were expanded. After several weeks, the cells were counted and 106 viable cells were seeded into 35-mm dishes. The cells were counted daily after staining with trypan blue, and viability was assessed. When the population had doubled, the cells were reseeded at 106 cells per ml. The percentage of viable cells was also calculated. At regular intervals, the frequency of cells lacking functional p53 was assessed by using gamma irradiation; the percentage of cells expressing mutant p53 is indicated in the boxes. The data shown are representative of analyses with seven transformants derived from Msh2 null mice and four transformants derived from wild-type mice.
The relationship between p53 mutation and emergence from crisis can be revealed by analyzing the kinetics with which loss of p53 function occurs. If this event provides the dominant selective force for emergence from crisis, then cells expressing wild-type p53 should be lost from independent populations at a rapid and similar rate once a cell carrying a loss-of-function mutation arises. Alternatively, if changes in addition to the loss of p53 function are required, the rate with which cells expressing mutant p53 accumulate should vary among independently derived cell lines. To distinguish between these possibilities, the kinetics with which cells expressing wild-type p53 were lost from the population was examined (Fig.7). Cells that had lost p53 function were detected in primary transformants from Msh2 null animals 1 to 2 weeks before they were observed in those derived from normal mice. Thus, the absence of MMR facilitates the occurrence of mutations affecting p53, thereby accelerating the emergence of cells which have lost p53 function. Although the rates of loss were not identical for all of the samples, comparison of the slopes of the curves, using a two-tailed t test, during the phase when p53 was rapidly lost revealed that these differences were not statistically significant (P = 0.1). Small differences in the curves could reflect the presence of additional changes in other genes which contribute to the full-blown transformed phenotype. However, the reproducibility with which functional p53 was lost in a group of independently derived primary transformants strongly suggests that this event is the rate-limiting step controlling the emergence of fully transformed cells.
Loss of p53 function is strongly selected in primary transformants from both Msh2 null and wild-type backgrounds. Independently derived primary transformants from Msh2 null mice (open circles) and normal mice (filled circles) were expanded and monitored periodically for the appearance of p53 mutations by using a gamma irradiation assay. The fraction of cells resistant to rapid apoptosis following gamma irradiation, a feature that strongly correlates with loss of p53 function in Ab-MLV transformants (51), was calculated at each time point.
DISCUSSION
Our results demonstrate that the absence of MMR alters the outcome of Ab-MLV-mediated transformation by influencing the way in which primary transformants evolve to fully established cell lines. This process is marked by a profound apoptotic crisis which is dependent on functional p53 and Ink4a/Arf genes (41, 54). As many as half of the primary transformants derived from normal mice lose p53 function coincident with emergence from crisis; most of the others down-modulate expression of the p19Arf protein (41, 51), a molecule that regulates p53 levels (50). In the absence of Msh2, almost all of the survivors express mutant forms of p53. Thus, the absence of MMR accelerates and skews the molecular events by which cells evolve from primary transformants to fully established cell lines.
The p53 gene may be an ideal target for mutation in the Ab-MLV system because cells expressing mutant p53 have a strong selective advantage and because a wide range of mutations can abolish p53 function (26). The inherently high mutation frequency found in the absence of Msh2 (1, 12, 44) likely increases the chances that a mutation facilitating escape from crisis will occur. Although loss of p19Arf expression is another mechanism by which primary transformants recover from crisis (41), mutations affecting the p19Arf coding sequence have not been observed in any Ab-MLV transformants, including those studied here (35, 54). The mechanism by which p19Arf is regulated in Ab-MLV transformants has not been elucidated. However, theInk4a/Arf locus is methylated in other lymphoid malignancies (21, 22, 31, 36, 39, 47), and this type of modification may be important in the Ab-MLV transformants. MMR status would not be expected to affect this sort of epigenetic change.
Although a high frequency of Ab-MLV transformants isolated fromMsh2 null mice lose p53 function, the loss of MMR does not always provide a strong selection for such a change. Indeed, the changes that confer a selective advantage are not always predictable. For example, although loss of p53 function in colon cancer occurs in more than 50% of sporadic cases (14, 28), this change occurs in less than 20% of the colon cancers in hereditary nonpolyposis colon cancer patients, individuals that lack MMR, and appears less commonly in tumor cells displaying error-prone replication (9, 28). In contrast, mutations affecting the TGF-β RII gene, which is targeted in about 2% of sporadic colon cancers, are found in 60 to 70% of colon cancers in hereditary nonpolyposis colon cancer patients (28). Mutation of TGF-β RII has also been detected in lymphomas arising in Msh2-deficient mice (30). In contrast, loss of p53 function occurs in about 35% of patients with MMR-deficient acute myelogenous leukemia (3, 59), a frequency that is higher than that observed when MMR is intact. Thus, the type of mutational event that confers a selective advantage to a cell in the absence of MMR function differs depending on the cell type and the environmental cues to which the cell is exposed. However, very little is known about the nature of these features and how they exert their influence.
Previous studies have shown that most Ab-MLV-transformed pre-B cells that carry p53 mutations express readily detectable p53 (51). The majority of the transformants derived fromMsh2 null mice display a similar pattern, and consistent with this finding, mutations affecting the DNA binding domain were identified in two of these cell lines. However, three cell lines failed to express detectable p53 protein and had frameshift mutations affecting the carboxyl-terminal portion of the molecule. Only 5 to 15% of all p53 mutations found in human tumors affect the carboxyl terminus (www.iarc.fr/p53/GERM.HTM :ftp://ftp.ebi.ac.uk/pub/databases/p53 ), and many alterations affecting these sequences result in synthesis of a truncated p53 protein (24, 38, 58). p53 proteins lacking portions of the carboxyl terminus should have been detected by the antibodies we used. However, all three of the cell lines which failed to express detectable p53 protein also contained much lower levels of p53 RNA as assessed by Northern blotting (data not shown). Thus, while the mutations affecting the carboxyl terminus were the only changes detected in the cDNA, they may not be responsible for the absence of p53 expression.
The crisis that is characteristic of Ab-MLV-induced pre-B-cell transformation is marked by high levels of apoptosis and an irregular growth pattern. The effects of Msh2 on both of these processes have been documented in several systems (6, 13, 17, 20, 52, 57). However, primary Ab-MLV transformants fromMsh2 null and control mice enter crisis with similar kinetics, and the magnitude of apoptosis and the frequency of G1 cells are similar in the two cell types. Thus, Msh2 does not appear to be directly involved in regulating these responses during Ab-MLV transformation, which suggests the MMR system does not respond directly to oncogenic signals. This idea is consistent both with the observation that crisis requires functional p53 and p19Arf pathways (41, 54) and with the normal p19Arf response found in primary transformants from Msh2 null mice.
Analysis of the kinetics with which cells containing a p53mutation become dominant within populations undergoing crisis revealed that loss of p53 function coincides with a predictable pattern of high viability and stable proliferation in both normal and Msh2null cells. In addition, the rates with which cells expressing wild-type p53 were lost from populations of primary transformants were very similar among the group of independent transformants examined. These data suggest that loss of p53 function is the rate-limiting step in Ab-MLV pre-B-cell transformation. The similarity in the rate of loss of p53 wild-type cells in transformants from normal and Msh2null mice further supports the idea that once a selectable mutation has arisen, transformants from both backgrounds behave identically. Thus, Msh2 loss increases the probability that a disabling p53 mutation will occur, thereby accelerating the later phases of the transformation process. This interpretation is consistent with the view that loss of MMR function predisposes cells to tumor development by increasing the chances that mutations in genes regulating cell growth or survival will arise, rather than by affecting these processes directly.
ACKNOWLEDGMENTS
We are grateful to John Coffin and Zohar Sachs for helpful discussions and Anne Halgren for technical assistance.
J.J.-W. was partially supported by grant GM 08448, D.R.-C. was supported by grant HL07785, and A.H.R. was supported by a fellowship from the German Academic Exchange Service (DAAD). The work reported here was supported by grant CA 33771 from the National Cancer Institute.
FOOTNOTES
- Received 6 June 2000.
- Returned for modification 2 August 2000.
- Accepted 14 August 2000.
- Copyright © 2000 American Society for Microbiology
