Previous Article | Next Article ![]()
Molecular and Cellular Biology, February 2008, p. 1265-1273, Vol. 28, No. 4
0270-7306/08/$08.00+0 doi:10.1128/MCB.01108-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
,
Heather A. Steinman,
Kathleen Hoover, and
Stephen N. Jones*
Departments of Cell Biology and Cancer Biology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, Massachusetts 01655
Received 21 June 2007/ Returned for modification 31 July 2007/ Accepted 11 November 2007
|
|
|---|
|
|
|---|
The amplification and overexpression of either Mdm2 (31) or MdmX (7, 36) have been observed in a variety of human cancers, including sarcoma, glioma, and, in the case of MdmX, retinoblastoma (24), suggesting that either Mdm2 or MdmX can function as an oncogene to inhibit p53 activity and promote tumorigenesis. Since many of these Mdm-overexpressing tumors retain wild-type p53 alleles, the reactivation of p53 by small-molecule inhibition of the Mdm2-p53 or MdmX-p53 interaction is an attractive strategy for treating these cancers (26, 27).
The results of experiments in vitro or in vivo involving the forced overexpression of Mdm proteins suggest that Mdm2 and MdmX may also have p53-independent roles in promoting cell growth (13, 23, 29, 38, 41). However, molecular targets for Mdm2 or MdmX activity other than p53 have yet to be confirmed. Furthermore, it remains unclear if physiologic levels of either Mdm2 or MdmX expression can exert functions that are distinct from their abilities to downregulate p53 activity. We have previously characterized mice and primary mouse embryonic fibroblasts (MEFs) with p53 or both Mdm2 and p53 deleted. These cells display equivalent rates of cell proliferation and cell transformation in culture, and Mdm2/p53 double null mice and p53-null mice present with the same rate and tissue spectrum of spontaneous tumor formation (22), indicating that normal cellular levels of Mdm2 function primarily to regulate p53 activity. In contrast, MEFs generated from Mdm2-transgenic mice with intact MdmX proliferated slower than MEFs lacking MdmX (42), suggesting that MdmX might have antiproliferative properties when p53 functions are inhibited. To determine if physiologic levels of MdmX can regulate cell proliferation in a p53-independent manner, we have generated and analyzed the growth and transformation of MEFs derived from mice with p53 or both p53 and MdmX deleted. Unlike our previous results with Mdm2, the deletion of MdmX increases the spontaneous transformation and proliferation of immortalized p53-null MEFs. Also, in contrast to what occurs with Mdm2, the loss of MdmX seems to have a profound impact on chromosome stability. Although both Mdm2/p53 double null cells and MdmX/p53 double null cells initially display the chromosomal hyperploidy characteristic of p53-null cells, MdmX-deficient cells undergo a reduction in chromosome number that correlates with increased cell proliferation and transformation during their growth in culture. Furthermore, MdmX/p53 double null mice display increased rates of spontaneous tumorigenesis relative to those of p53-null mice. Similarly to MEFs, tumor cells isolated from MdmX/p53-null mice proliferate faster and have fewer chromosomes than p53-null tumor cells, corroborating the role for MdmX in the maintenance of genome stability. Hyperploid, p53-deficient MEFs and tumor cells that lack MdmX display reduced centrosome clustering and high levels of multipolar mitotic spindle formations, likely accounting for the aberrant chromosome segregation and loss of chromosomes during mitosis. The reintroduction of MdmX into MdmX/p53-null tumor cells increased the ploidy, reduced the incidence of multipolar spindles, and decreased the proliferation rate in these cells. These data reveal that MdmX has a p53-independent role in suppressing cell proliferation, transformation, and tumorigenesis by promoting bipolar mitosis and preventing chromosome loss in polyploid p53-deficient cells.
|
|
|---|
PDL = log(nf/n0)/log2, where n0 is the initial number of cells and nf is the final number of cells]. For the colony formation assay, cells were plated at a low density (200 cells per cm2) and grown for 2 weeks in medium that was changed every 3 days. Cells were fixed with methanol and stained with crystal violet to evaluate colony formation. For the transformation assay (focus formation assay), 2 x 104 cells per cm2 were grown in Dulbecco's modified Eagle medium with 7% serum. The medium was changed every 3 days. After 14 days, the cells were fixed with methanol and stained with Giemsa stain. For the transfection, cells were transfected with an MDMX expression construct and/or with the puromycin expression construct pBABE-Puro at a 5:1 molar ratio by use of Lipofectamine 2000 (Invitrogen) in a six-well plate format. Cells were replated 24 h after transfection and selected for puromycin resistance. MdmX expression was confirmed by reverse transcriptase PCR (data not shown). FACS analysis of DNA content. For fluorescence-activated cell sorting (FACS) analysis, cells were plated at the density of 1 x 104/cm2 and grown asynchronously to 90% confluence. The cells were fixed in ethanol, stained with propidium iodide, and analyzed by a Becton Dickinson FACScan flow cytometer (Mountain View, CA) for DNA content. FACS profiles of wild-type MEFs were generated in parallel as controls for diploid-cell DNA content.
Metaphase spreads and chromosome counting. Proliferating cells were incubated with colcemid (0.02 µg/ml medium) for 90 min, harvested and incubated in hypotonic solution (0.075 M KCl) for 15 min at 37°C, and fixed with methanol-acetic acid. Drops of cell suspension were placed on microscope slides, dried, DAPI (4'-6'-diamidino-2-phenylindole) stained, and analyzed by fluorescence microscopy.
Immunofluorescence microscopy.
Cells were grown on glass coverslips, fixed in methanol, and stained with mouse monoclonal
-tubulin (Sigma-Aldrich, St. Louis, MO) and rabbit polyclonal anti-
-tubulin (Sigma-Aldrich) primary antibodies, as well as secondary antibodies coupled to Alexa 488 or Alexa 594 (Molecular Probes). DNA was stained with DAPI. Immunofluorescence images were recorded using a charge-coupled-device camera attached to an epifluorescence Zeiss Axioplan 2 microscope (Zeiss, Thorwood, NY). Single images were acquired using Metamorph imaging software (Universal Imaging, Downingtown, PA).
Animal studies. MdmX+/– p53+/– mice were intercrossed to generate p53-deficient mice that were MdmX wild type, MdmX heterozygous, or MdmX null. Cohorts of these mice were monitored for tumor formation as previously described (43). Necropsies were performed on mice bearing obvious tumors, and tumors were harvested and fixed in 10% phosphate-buffered formalin for histological analysis. The pathological analysis of tumors arising in the mice was performed by IDEXX (Grafton, MA). Statistical analyses of the rate of tumor formation was performed using Student's t test, with a P value of <0.05 considered significant. Animals were used in accordance with regulations established by the University of Massachusetts Medical School's Institute for Animal Care and Use Committee and with federal guidelines.
|
|
|---|
![]() View larger version (58K): [in a new window] |
FIG. 1. The loss of MdmX increases the growth potential of p53-null cells in culture. (A) At passage 4 (Pass 4), growth rates of p53-null and MdmX/p53-null cells are equivalent and much higher than that of wild-type (WT) cells. At passage 9, MdmX/p53-null cells proliferate faster than p53-null cells. Error bars represent standard errors of the mean from two independent cell lines. (B) The loss of MdmX confers a growth advantage to p53-deficient cells during long-term growth in culture. A 3T3 immortalization assay was performed with two or three cell lines of each genotype. The actual cell number at each passage (left panel) and accumulated PDLs (right panel) are shown. By passage 17, MdmX/p53-null populations underwent almost 10 more doublings than the p53-null population. PD, number of population doublings. (C) The loss of MdmX allows immortalized p53-deficient cells to proliferate much faster and to reach much higher saturation densities than the wild-type cells. A proliferation assay was performed with passage 17 immortalized cells. (D) The loss of MdmX stimulates growth within the individual clones at a low plating density. Colony formation was evaluated for crystal violet-stained cells at passage 18. (E) The loss of MdmX triggers the spontaneous transformation of p53-deficient cells. In contrast to p53-null cells, which failed to form foci, MdmX/p53-null cells formed numerous dark-blue foci on Giemsa-stained plates. Counts of triplicate 10-cm plates revealed 188 ± 53 foci larger than 1 mm in diameter per plate.
|
The increased growth capacity of p53-null cells relative to that of wild-type cells is also reflected in their ability to grow when plated at a low cell density (15). However, p53-null cells with MdmX deleted not only retain the same plating efficiency and ability to grow at a low density (see Table S1 in the supplemental material) but also form much larger colonies than p53-null cells with intact MdmX (Fig. 1D). Eighty percent of all MdmX/p53 double null colonies exceeded 200 cells per colony at day 5 postplating, whereas only 30% of the p53-null colonies reached this size. These data further demonstrate an increase in the rate of p53 colony proliferation when MdmX is deleted. The ability of the MdmX/p53-null cells to reach a much higher saturation density than the p53-null cells (Fig. 1C) suggested that the loss of MdmX further compromises the regulation of cell growth and might induce the transformation of immortalized p53-deficient cells. Therefore, we performed a cell transformation assay by allowing the MEFs to grow to saturation density and assaying for the formation of transformed foci at 2 weeks postplating. Primary cells deficient for p53 do not form foci under these conditions, unless they are transduced with an activated oncogene such as ras (10). However, the results of the transformation assay (Fig. 1E) revealed that MdmX/p53-null cells formed numerous foci in culture. Counts of triplicate 10-cm plates of MdmX/p53-null cells documented 188 (±53 standard deviations) foci larger than 1 mm in diameter per plate. In contrast, p53-null MEFs did not form any foci in culture. These results confirm that MdmX inhibits the spontaneous transformation of immortalized p53-deficient cells.
MdmX suppresses chromosome loss in p53-null cells. To investigate the possible effects of MdmX on genome stability, we compared the cell ploidy, measured as the DNA content per cell, of propidium iodide-stained p53-null cells with that of MdmX/p53-null cells at different passages using wild-type MEFs at passage 2 as a diploid control (Fig. 2A). Very low passage number (passage 2) MEFs with either p53 or both MdmX and p53 deleted displayed normal, diploid FACS profiles similar to that of the wild-type MEFs. It was reported previously (6, 12, 15) that MEFs with inactivated p53 genes are genetically unstable and contain supernumerary chromosome numbers after repeated passages in culture. Indeed, at passage 9, both MdmX/p53-null MEFs and p53-null MEFs displayed very small 2N cell subpopulations, and the majority of the cells had 4N or >4N DNA content. However, while the p53-null cells maintained a significant hyperploid subpopulation for more than 20 additional passages, MEFs with both MdmX and p53 deleted underwent a drastic reduction in ploidy at higher passages.
![]() View larger version (37K): [in a new window] |
FIG. 2. MdmX suppresses the reduction in ploidy and loss of chromosomes in immortalized p53-null cells. (A) Flow cytometry profiles of PI-stained MEFs. Histograms of DNA content in MdmX/p53-null cells (second row, red) and in p53-null cells (third row, red) are overlayered with a histogram of passage 2 (p2) wild-type (WT) MEFs (blue) shown as a diploid control. At passage 9, the majority of the cells of both genotypes show 4N or >4N DNA content. While p53-null cells maintain hyperploid DNA contents at high passage numbers, MdmX/p53 double null cells display drastic reductions in hyperploid subpopulations at high passage numbers. (B) Chromosome number distribution in low- and high-passage-number MEFs. For chromosome number determination, chromosomes in at least 50 DAPI-stained metaphase spreads for each cell line and each passage shown were counted. The deletion of MdmX, but not of Mdm2, caused a drastic reduction in the chromosome number of high-passage-number p53-null cells. (C) Percentages of cells with greater-than-triploid chromosome contents (more than 60 chromosomes per cell) in populations of MdmX/p53-null (three cell lines) and p53-null (four cell lines) MEFs. (D) The deletion of MdmX also leads to a reduction in chromosome numbers in immortalized p21-null cells.
|
In contrast to MdmX, the presence or absence of Mdm2 does not seem to affect chromosome numbers of p53-null cells (Fig. 2B). In order to examine whether the loss of MdmX affects chromosome number in polyploid cells of a different background, we examined MEFs previously generated from MdmX-null mice with the p53 target gene, p21WAF1, deleted (43). The loss of p21 function in MEFs is known to induce endoreduplication, resulting in polyploid cells (44). The results in Fig. 2C show that MdmX/p21 double null MEFs have much lower chromosome numbers per cell than p21-null MEFs at similar passage numbers, suggesting that the MdmX-mediated maintenance of the chromosome number is not unique to p53-deficient cells.
MdmX suppresses spontaneous tumorigenesis in p53-deficient mice. Spontaneous transformation of p53-null cells in the absence of MdmX coupled with the increased rate of chromosomal loss in the MdmX/p53-null cells prompted us to investigate whether the antiproliferative growth properties of MdmX could alter spontaneous tumorigenesis in p53-null mice. Cohorts of p53-deficient mice that were wild type, heterozygous, or null for MdmX were generated, and a tumor assay was performed. We have previously demonstrated that Mdm2 status does not affect the onset of tumorigenesis in p53-deficient mice, and mice with both Mdm2 and p53 deleted present with the same tumor types as p53-null mice (22). In contrast, the deletion of one or both MdmX alleles significantly altered tumor formation in p53-heterozygous or p53-null mice (Fig. 3A and B). The tumor spectrum and tumor burden of p53-heterozygous mice or p53-null mice in the presence or absence of MdmX remained unchanged, with thymic lymphomas and sarcomas being the predominant tumor types observed (data not shown); however, the rate of tumorigenesis was accelerated in the absence of MdmX. Fifty percent of the p53-null mice present with spontaneous tumors by 25 weeks of age. On average, mice lacking both p53 and MdmX develop tumors 5 weeks earlier. These results reveal that MdmX suppresses oncogenesis when p53 is deleted and confirm in vivo that MdmX has a p53-independent role in regulating cell transformation.
![]() View larger version (22K): [in a new window] |
FIG. 3. MdmX suppresses tumorigenesis in p53-deficient mice. Cohorts of mice from the indicated genotypes were monitored for spontaneous tumor formation. (A) Kaplan-Meier plot of tumorigenesis in p53-heterozygous mice that are either wild type or heterozygous for MdmX. The loss of one functional allele of MdmX accelerates tumorigenesis in mice heterozygous for p53. (B) Kaplan-Meier plot of tumorigenesis in p53-null mice that are either wild type or null for MdmX. The loss of MdmX function accelerates tumorigenesis in the absence of p53. P values are shown for each curve.
|
![]() View larger version (36K): [in a new window] |
FIG. 4. MdmX suppresses cell proliferation and the loss of chromosomes in p53-null tumor cells. Cells were isolated from the thymic tumors of p53-null, MdmX/p53-null, and Mdm2/MdmX/p53-null animals. (A) MdmX/p53-null tumor cells proliferate much faster and reach higher saturation densities than p53-null tumor cells. Error bars represent standard errors of the means from two independent experiments. (B) Double and triple null tumor cells have reduced ploidy and chromosome numbers compared to those of the p53-null tumor cells. Chromosome number ranges are the same as in Fig. 2. The results represent the average chromosome numbers from two independent experiments for null and double null cells and from one experiment for triple-null cells. (C) DNA content of MdmX/p53-null tumor cells transfected with an MdmX-expressing plasmid. Parental, nontransfected cells (Control), cells transfected with the puromycin vector only (Mock), and stable MdmX transfectants (clones c1 and c2) were analyzed for DNA content. MdmX-transfected clones showed higher DNA contents than those of both controls. (D) The reintroduction of MdmX into the double null tumor cells (clones c1 and c2) significantly reduced the rate of cell proliferation.
|
-tubulin and
-tubulin were used to stain microtubules and centrosomes, respectively, and DAPI was used to counterstain DNA. MEFs or tumor cells with p53 deleted clearly showed increased numbers of centrosomes, as previously reported (4, 12). Centrosomes were often clustered around the opposite poles, giving rise to bipolar spindles and facilitating equal or semiequal segregations of the supernumerary chromosomes during mitosis (Fig. 5A). In sharp contrast, centrosome clustering was very rare in cells with both MdmX and p53 deleted. Instead, almost one-quarter of the mitotic MdmX/p53 double null cells formed symmetrical and asymmetrical multipolar spindles (Fig. 5B). Interestingly, these multipolar spindles permit chromosome segregation, as telophase figures with more than two groups of separated chromosomes are frequently observed in these cells (Fig. 5C). While it is doubtful that these groups contain the correct diploid chromosome set, some may generate viable cells, albeit with lower chromosome numbers than parental cells. Furthermore, when transfected with exogenous MdmX, MdmX/p53-null tumor cells also displayed increased centrosome clustering (Fig. 5D) accompanied by a drastic reduction in the formation of multipolar spindles. Overall, only about 10% of the mitotic cells lacking p53 were found to have multipolar spindles, whereas on average 23% of the p53-null cells displayed multipolar spindles when MdmX was absent (Fig. 5E). These data indicate that the presence of MdmX correlates with centrosomal clustering and reduced multipolar spindle formation in p53-null MEFs and in cells derived from p53-null tumors.
![]() View larger version (38K): [in a new window] |
FIG. 5. MdmX suppresses spindle multipolarity in p53-null MEFs and tumor cells. Cells were stained with mouse monoclonal anti- -tubulin (green) antibodies, rabbit polyclonal anti- -tubulin (red) antibodies, and DAPI (blue). (A) Bipolar spindles in p53-null cells with amplified centrosomes. More than 30% of the mitotic p53-null MEFs and tumor cells with amplified centrosomes show bipolar spindle assembly. In those cells, chromosomes are aligned on one metaphase plate and are likely to be distributed equally or almost equally between the two daughter cells, as evident in the telophase cell (two right-hand sets of images). (B) Multipolar spindles in MdmX/p53-null cells with amplified centrosomes. Symmetrical (left and middle sets of images) and asymmetrical (right set of images) multipolar spindles are common in MdmX/p53-null cells. (C) In cells with multipolar spindles, chromosomes can separate and generate telophase figures with more than two groups of chromosomes. (D) Bipolar spindles in MdmX/p53-null tumor cells transfected with an MdmX-expressing plasmid. (E) The incidence of mitotic cells with multipolar spindles was evaluated by scoring from 100 to 200 mitotic figures for MEFs (M) and tumor cells (T). The histograms show the averages (from left to right) for two p53-null MEF lines, two experiments with p53-null tumor cells, three double null MEF lines, two experiments with double null tumor cells, and two clones of Mdm4-transfected double null tumor cells. Error bars represent standard errors of the means. The loss of MdmX function causes a more-than-twofold increase in spindle multipolarity. Scale bars = 10 µm.
|
|
|
|---|
MEFs deficient for p53 or for both MdmX and p53 display supernumerary centrosomes and hyperploidy at low passage numbers. However, MdmX/p53 double null cells undergo a loss of chromosomes during growth in culture, and reintroduction of MdmX into MdmX/p53-null cells increases their ploidy. This appears to be a unique feature of MdmX, as ploidy in p53-null cells is not affected by the presence or absence of Mdm2.
Deletion of MdmX in p53-null cells correlates with an increase in the subpopulation of cells displaying multipolar spindles during mitosis, and the addition of exogenous MdmX to MdmX/p53 double null cells increases the coalescence of centrosomes and promotes bipolar cell division.
Since Hansemann's (14) and Boveri's (2) experiments more than a century ago, multipolar mitosis and aneuploidy have been tightly linked to tumorigenesis (35, 45, 48). Centrosomal clustering and bipolar spindle formation have been suggested to be critical for preserving the genomic stability of noncancer cells (34, 37) and for the maintenance and propagation of hyperploid cells with amplified centrosomes (3, 40). The reduction of chromosome numbers observed in polyploid p53-deficient cells after prolonged culturing in vitro has been proposed to give rise to a more stable, yet aneuploid, karyotype that confers a growth advantage to a cell (5). This reduction in ploidy is thought to be analogous to the previously described genomic convergence during tumor progression in vivo (17), and the relatively stable karyotypes found in many advanced tumors and established tumor cell lines may reflect the evolvement of a genotype optimized for growth under specific conditions (1).
The increased spindle multipolarity, chromosomal loss, and increased proliferation of p53-null cells lacking MdmX indicate a p53-independent role for MdmX in the suppression of tumorigenesis through the maintenance of genome stability. As the effect of MdmX loss on the proliferation and ploidy of p53-null cells is observed only at late passages or in immortalized cells, it is likely that MdmX plays a p53-independent role in regulating cell transformation and subsequent tumorigenesis. Although the presence or absence of MdmX does not seem to impact the amplification of centrosomes in p53-null cells, the increase in bipolar spindle polarity and chromosome ploidy following the reintroduction of MdmX into MdmX/p53-null cells confirms that MdmX alters the organization of these supernumerary centrosomes. Therefore, we propose that the MdmX-mediated clustering of amplified centrosomes facilitates bipolar mitosis and the propagation of hyperploid genomes. The loss of MdmX increases spindle multipolarity and the rate of chromosomal loss, and the resulting reduction in ploidy should yield a faster-replicating population of double null cells. Conversely, chromosomal clustering and the antiproliferative effect of exogenous MdmX in p53-null cells suggest that cells with supernumerary chromosomes that undergo bipolar mitosis proliferate slower than cells with multipolar spindles and further support a p53-independent role for MdmX in tumor suppression.
MdmX has recently been recommended as a suitable target for small-molecule inhibition in the treatment of human cancers (24, 46), as the inhibition of the MdmX-p53 interaction in tumors with wild-type p53 likely releases tumor suppressor p53 activity. However, it is unclear what effect this inhibition has on p53-independent MdmX activities, especially in the approximately half of all human tumors that lack functional p53. Therefore, further work is needed to define the precise molecular mechanism of the MdmX-mediated suppression of proliferation and tumorigenesis.
Core facilities were supported by program project grant 5P30DK32520 from the National Institute of Diabetes and Digestive and Kidney Diseases. This work was supported by a grant from the National Institutes of Health to S.N.J. (RO1CA77735).
Published ahead of print on 26 November 2007. ![]()
Supplemental material for this article may be found at http://mcb.asm.org/. ![]()
These authors contributed equally to this work. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»