Msh2 deficiency leads to chromosomal abnormalities ... - Nature

Oncogene (2006) 25, 2531–2536

& 2006 Nature Publishing Group All rights reserved 0950-9232/06 $30.00 www.nature.com/onc

SHORT COMMUNICATION

Msh2 deficiency leads to chromosomal abnormalities, centrosome amplification, and telomere capping defect MR Campbell1, Y Wang2, SE Andrew1 and Y Liu2 1 Department of Medical Genetics, University of Alberta, Edmonton, Alberta, Canada; 2Life Sciences Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA

Msh2 is a key mammalian DNA mismatch repair (MMR) gene and mutations or deficiencies in mammalian Msh2 gene result in microsatellite instability (MSI þ ) and the development of cancer. Here, we report that primary mouse embryonic fibroblasts (MEFs) deficient in the murine MMR gene Msh2 (Msh2/) showed a significant increase in chromosome aneuploidy, centrosome amplification, and defective mitotic spindle organization and unequal chromosome segregation. Although Msh2/ mouse tissues or primary MEFs had no apparent change in telomerase activity, telomere length, or recombination at telomeres, Msh2/ MEFs showed an increase in chromosome end-to-end fusions or chromosome ends without detectable telomeric DNA. These data suggest that MSH2 helps to maintain genomic stability through the regulation of the centrosome and normal telomere capping in vivo and that defects in MMR can contribute to oncogenesis through multiple pathways. Oncogene (2006) 25, 2531–2536. doi:10.1038/sj.onc.1209277; published online 5 December 2005 Keywords: mismatch repair; centrosome amplification; telomere capping; chromosomal abnormalities

Mutations or deficiency of MMR genes have been linked to the development of cancers. Human patients inherit a heterozygous mutation in one of the MMR genes, most commonly hMSH2 or hMLH1, develop the human cancer syndrome Hereditary NonPolyposis Colorectal Cancer (HNPCC) (reviewed in Buermeyer et al. (1999)). Mice completely deficient in one of the MMR genes, namely Msh2, Mlh1, or Msh6, most commonly develop early onset thymic lymphomas (de Wind et al., 1995; Reitmair et al., 1995; Edelmann et al., 1997; Prolla et al., 1998). MSH2 is the key MMR protein that initiates the recognition of a base mispair and the subsequent recruitment of additional MMR proteins to complete the repair process (reviewed in Buermeyer et al. (1999)). MMR proteins also have roles in cell cycle arrest, cell signaling, apoptosis and the Correspondence: Dr Y Liu, Life sciences Division, Oak Ridge Natiional Laboratory, Oak Ridge, TN, USA. E-mail: [email protected] Received 5 July 2005; revised 19 October 2005; accepted 25 October 2005; published online 5 December 2005

prevention of homeologous recombination (reviewed in Bellacosa (2001)). In this study, we have used isogenic wild type and Msh2/ mouse tissues and cells to assess possible roles of MMR in mammalian telomere maintenance, centrosome fidelity, and thus chromosomal stability. We examined the possible role of MMR in maintaining chromosomal stability by performing standard Giemsa-staining of wild type and Msh2/ primary MEFs and scoring for chromosome copy number changes (both chromosomal gains and losses). In all, 12/40 (30%) wild-type metaphases, in contrast to 35/44 (79.5%) of Msh2/ metaphases, showed chromosomal number changes (Figure 1, Table 1). The majority of wild-type MEFs retained normal chromosome counts (40, or 80 if cell division has not yet occurred), however, Msh2/ MEFs had chromosomal counts that range from 27 to 83 chromosomes (Figure 1d). While most chromosomal aberrations seen were in the form of numerical changes, at least one Msh2/ metaphase showed extensive levels of chromatid type breaks and aberrations (Figure 1c). This observation was also confirmed by telomeric FISH of Msh2/ metaphase spreads (see details below). Our results demonstrate that MMR is critical in maintaining chromosomal stability and loss of MMR function leads to chromosomal abnormalities, mainly chromosome aneuploidy, in primary MEFs. The observation of a high incidence of chromosome aneuploidy in primary Msh2/ MEFs led us to examine centrosome numbers in these cells, because aberrations in the number of centrosomes almost inevitably cause chromosome mis-segregation and thus aneuploidy (reviewed in Doxsey et al. (2005), Nigg (2002), Saunders (2005)). To examine centrosome numbers in primary Msh2/ MEFs, we stained wild-type and Msh2/ MEFs with an antibody against the g-tubulin component of the centrosome. The majority (88%) of wild-type cells displayed a normal distribution of 1–2 centrosomes per cell; in contrast, only 66% of Msh2/ cells had normal numbers of centrosomes (Figure 2a, Table 2). We observed frequent amplification of centrosomes in Msh2/ cells. In all, 34% of Msh2/ cells showed centrosome amplification (>2 centrosomes), with 11% of cells having multiple centrosomes (nX5). This is significantly different than wild-type cells in which only 12% of cells had more than two centrosomes (Po0.0001).

Msh2 deficiency leads to chromosomal abnormalities MR Campbell et al

2532

Figure 1 Chromosomal instability in primary Msh2/ MEFs. (a–c) Representative photomicrographs of Giemsa-stained metaphase spreads of primary wild type and Msh2/ MEFs. (a) A wild-type MEF showing normal chromosomal numbers (n ¼ 40). (b) An Msh2/ MEF showing aneuploidy with n ¼ 83. (c) An Msh2/ MEF showing extensive chromatid type breaks and aberrations. (d) Chromosomal number distribution in primary wild type and Msh2/ MEFs. A total of 40 wild-type and 44 Msh2/ metaphase spreads were scored. The graph shows that the majority of wild-type MEFs contain a normal number of chromosomes n ¼ 40. Although Msh2/ MEFs show a clustering of chromosome counts around the expected 40 chromosomes, they illustrate extensive deviation from the expected normal number of chromosomes (n ¼ 40). Aneuploidy changes in Msh2/ cells ranged from 23 to 87. For experimental detail, see supplemental material.

Table 1 Increased chromosome aneuploidy detected by Giemsa staining in primary Msh2-/- MEFs Cell type

# Metaphases

# Aneuploid

# Diploid

Percentage

Wild type Msh2/

40 44

12 35

28 9

30% 79.5%

Three individual embryos for each genotype (wild type and Msh2/) were analysed. A two-sided fisher’s exact test was used to compare normal and abnormal numbers of chromosomes per cell (Po0.0001).

We next examined if centrosome amplification in Msh2/ MEFs can lead to defective mitotic spindle organization and unequal chromosome segregation in mitosis. Using double immunostaining with g- and atubulin antibodies, we observed multiple centrosomes associated with multiple spindle poles and misaligned chromosomes in primary Msh2/ MEFs (Figure 2b). This result demonstrates that multiple centrosomes persist through interphase to mitosis and lead to multiple spindle poles, causing mis-segregation of chromosomes and thus chromosomal aneuploidy in Msh2/ MEFs. Centrosomes are critical in maintaining proper chromosome segregation during mitosis and centrosome aberrations have been implicated in the development of tumor aneuploidy (reviewed in Doxsey et al. (2005), Nigg (2002), Saunders (2005)). Ghadimi et al. (2000) reported the existence of centrosome amplification in aneuploid but not in diploid human MMR deficient Oncogene

colorectal cancer cell lines, however, the molecular mechanism that led to this phenotype is not clear. Here, we demonstrate that loss of MMR function, specifically the MSH2 function, can lead to abnormal numbers of centrosomes accompanied by a high incidence of chromosomal aneuploidy in mice. Our data suggest that defects in MSH2 function could lead to centrosome amplification, likely one of the major mechanisms leading to the chromosomal instability observed in the Msh2 deficient MEFs and possibly in human cancers. Our finding implies a novel role for the MMR proteins in the fidelity of the centrosome and thus genomic stability. The observation from primary isogenic MEFs suggests that the role of MMR proteins in centrosome instability may be an early event in genome instability leading to tumorigenesis. Telomeres are the chromosome end structures essential in protecting chromosome ends from rearrangement. Loss of telomere protection at chromosome ends leads to chromosome end-to-end fusions, subsequently chromosome fusion-breakage cycles in cell divisions and eventually genomic instability (reviewed in Artandi and DePinho (2000), Greider (1998)). In the event that increased chromosome abnormalities in MMR deficient mice may be due to telomere dysfunction, we examined the status of telomeres in MMR deficient mice. Chromosome end-to-end fusions and ends without detectable telomeric DNA (also referred as telomere signal free ends or SFEs) are signatures of telomere dysfunction. We examined frequencies of these events in primary Msh2/ MEFs. When compared to wild-type


2533

Figure 2 Centrosome amplifications and defective mitotic spindle and chromosomal separation in primary Msh2/ MEFs. (a) Immunofluorescence analysis of the centrosomes by antibody against g-tubulin. Representative normal staining for one or two centrosomes (red) (upper panel) and abnormal staining for more than two centrosomes and centrosomes in cluster (middle and lower panels) in primary Msh2/ MEFs. Enlarged versions of multiple centrosomes are shown in black and white at the right bottom. (b) Double immunofluorescence analysis of the centrosomes by antibody against g-tubulin and mitotic spindles by antibody against atubulin. Upper panel: chromosomes (blue). Middle panel: overlap images of mitotic spindles (green) and centrosomes (red). Lower panel: overlap images of chromosomes, mitotic spindles, and centrosomes. Representative normal staining for two centrosomes and spindle poles (n ¼ 2) and abnormal staining for more than two centrosomes and spindle poles (n>2). Notice that chromosomes are misaligned in mitotic cells with more than two centrosomes and spindle poles. For experimental detail, see supplemental material.

Table 2 Centrosome amplification in primary Msh2/ MEFs Cell type

Wild type Msh2-/-

Number of centrosome

Summary

n ¼ 1(%)

n ¼ 2(%)

n ¼ 3 or 4(%)

n ¼ 5–6(%)

nX7(%)

n ¼ 1 or 2(%)

nX3(%)

21.9 11.3

66 54.8

9.9 23.7

1.8 6.6

0.4 3.5

87.9 66.1

12.1 33.9

n indicates the number of centrosomes. n>7 (n rages from 7 to 14). For each genotype, three embryos were examined. Approximately 600 cells per genotype were scored. The difference in the numbers of centrosomes between wild type and Msh2/ MEFs is significant (Po0.0001) by Student’s ttest. Table 3

The frequencies of chromosome abnormalities detected by telomeric FISH in primary Msh2/ MEFs

Cell type

# Metaphases

# End-to-end fusions

# Chromosome breakages

#Telomere signal free ends

Wild type Msh2/

126 170

4 (3%) 11 (6.5%)

6 (4.7%) 10 (5.9%)

0 (0%) 14 (8.2%)

For each genotype, two primary MEF lines were examined. % represents average numbers of chromosome abnormalities/metaphase.

MEFs, Msh2/ MEFs showed a slight increase in chromosome end-to-end fusions (6.5% Msh2/ metaphases vs 3% wild-type metaphases) and telomere signal-free-ends in Msh2/ MEFs (8.2% Msh2/ metaphases vs 0% wild-type metaphases) (Figure 3a, Table 3). These results demonstrate that MMR deficient cells have slight telomere capping defects. Telomerase and telomere length is essential in maintaining telomere integrity (reviewed in Greider (1998)). To determine whether loss of MMR function could lead to abnormal telomere length and telomerase activity thus telomere capping defect, we measured telomere length and telomerase activity of wild type, Msh2 þ /, and Msh2/ mouse tissues and primary MEFs by Flow- and Q-FISH (Zijlmans et al., 1997; Rufer et al., 1998) and TRAP assays (Kim and Wu, 1997). No

difference in relative telomere signal intensities and telomerase activity was observed in these samples (Figure 3b–d). Our data suggest that MSH2 does not play a major role in telomere length and telomerase regulation in vivo and end-to-end fusions in Msh2/ mouse cells are unlikely the result of apparently normal telomere length or telomerase activity. All the chromosome end-to-end fusions in Msh2/ MEFs showed detectable telomeric DNA signals at the fusion points. Therefore, end-to-end fusions in Msh2/ MEFs are unlikely cause by telomere signal free ends but likely through other mechanisms that may affect telomere capping. For example, a loss of MMR may reduce the inhibition of recombination to a level that permits fusion events. Alternatively, an increased mutation frequency in telomere sequence may occur in Msh2 Oncogene


2534

Figure 3 Chromosome abnormalities, telomere length, and telomerase activity in primary Msh2/ MEFs. (a) Telomeric-FISH analysis of metaphase spreads of primary Msh2/ MEFs showing DAPI staining (blue) and telomere fluorescence signals (red). Arrows indicate chromosome end-to-end fusion events with detectable telomeric DNA signal at the fusion sites (left panel, top, and middle), SFEs (left panel, bottom), chromosome fragments (right panel). (b) Quantitative measurement of telomeric DNA signal intensity at individual chromosome end by Q-FISH analysis of metaphase spreads of primary wild type and Msh2/ MEFs. Data was accumulated using approximately 10 metaphases for each histogram. The number of telomeres within a given range of telomeric DNA intensities was plotted against the telomere DNA signal intensity using arbitrary units. (c) Flow-FISH analysis of average telomere signal intensity of thymocytes and splenocytes derived from wild type ( þ / þ ), Msh2 þ /, and Msh2/ mice. In each set, data were pooled from at least five individual mice (error bars represent the s.d.). (d) TRAP was performed for 25 PCR cycles on 10, 5, and 2.5 mg of mouse tissue extracts prepared from wild type, Msh2 þ /, and Msh2/ mouse testes. An internal PCR standard for the telomerase repeat amplification protocol is shown at bottom left with an arrow. Nonspecific products are as indicated. A titration of cell extract was used to demonstrate that the TRAP products were in the near-linear range. For experimental detail, see supplemental material.

deficiency and thus compromise telomere function. Pickett et al. (2004) report a correlation between MSI þ and telomere sequence mutations, suggesting that an increased mutation frequency from loss of MMR function leads to telomere mutations. Increased mutational load at telomeres may alter telomere sequence to a degree that telomerase and telomere-binding or associated proteins may have altered binding affinities to telomeres (de Lange, 2002), leading to consequent change in telomere integrity. It is worthy to point out that low frequencies of fusion events can lead to the fusion-breakage-fusion cycles during cell proliferation Oncogene

and thus contribute to genomic instability (Artandi and DePinho, 2000) to a certain degree, but are unlikely the cause of the high incidence of chromosome aneuploidy observed in the Msh2/ MEFs. MMR proteins play a role in the prevention of homeologous recombination and loss of antirecombination activity of MMR proteins may thus promote telomere recombination, which, in turn, could affect telomere integrity. To examine the frequencies of telomeric recombination through telomere sister-chromatid exchanges (T-SCEs), we performed Strandspecific Chromosome Orientation-FISH (CO-FISH)


2535 Table 4

The frequencies of telomere sister chromatid exchanges (T-SCEs) in primary Msh2 deficient MEFs

Cell type

Metaphases

# T-SCEs/ chromosomes (%)

Ratio relative to control (wild type)

Wild type

200

1.00

Msh2/

200

15/8505 (0.18%) 18/8541 (0.21%)

1.17

A chromosome with either three or four telomeric DNA signals was considered as T-SCE-positive. For each genotype, two primary MEF lines at passage two were examined.

Figure 4 Normal telomere sister chromatid exchange in primary Msh2/ MEFs.CO-FISH analysis of metaphase spreads of primary Msh2/ MEFs showing DAPI staining (blue) and telomere fluorescence signals (red). Arrows indicate a chromosome fragment (a, left), a telomere signal free end (a, right and b), and a chromosome with T-SCE (b). A chromosome with more than two telomeric DNA signals is regarded as T-SCE positive. For experimental detail, see supplemental material.

(Bailey et al., 2004) analysis on metaphase spreads of wild type and Msh2/ primary MEF. 0.18% versus 0.21% of chromosomes from wild type or Msh2/ MEFs were positive for T-SCEs, respectively (Figure 4, Table 4). Thus Msh2/ MEFs has a similar low rate of T-SCEs as wild type MEFs. This result suggests that MMR deficiency does not affect telomeric recombination through telomere sister-chromatid exchanges. Previous reports have indicated that MMR deficiency may help cells to overcome cellular crisis caused by dysfunctional telomeres through telomeric recombination in telomerase deficient yeast or mammalian cells (Rizki and Lundblad, 2001; Bechter et al., 2004). The primary Msh2/ MEFs have normal telomere length

and are telomerase proficient. It is thus possible that MMR might affect T-SCE or telomere length maintenance only when telomeres become critically short and/or during cellular immortalization in telomerase deficient cells. Future analysis of MMR deficiency in mice with both telomerase and telomere dysfunction may help us to determine the role of MMR in telomerase-independent telomere length maintenance in mammals. The majority of tumors are characterized by genomic instability; either chromosomal instability or microsatellite instability. Both centrosomes and telomeres are essential in preserving genomic stability. Msh2 is a key mammalian MMR gene and mutations at Msh2 gene contributes to MSI þ and is associated with the human cancer syndrome HNPCC. Here, we represent the first characterization of a novel role of Msh2 gene in maintaining chromosome stability, through regulating centrosome fidelity and telomere function in the Msh2 deficient mice. These findings suggest that defects in MMR can contribute to oncogenesis through multiple pathways including centrosome amplification, telomere capping defects and chromosomal abnormalities. Moreover, the use of isogenic nontumor cells for these experiments demonstrates the importance of MMR in early instabilities that ultimately lead to tumorigenesis. Acknowledgements We acknowledge the support of the office of Biological and Environmental Research, US Department of Energy under Contract DE-AC056-960R22464 with UT-Battelle, LLC. MRC is supported by Alberta Heritage Foundation for Medical Research (AHFMR) and Alberta Cancer Board studentships. SEA is a scholar of the AHFMR and the Canadian Genetic Diseases Network. We thank Ms. Cecilia Wang and Ms. Marla Gomez for assisting the image analysis of FISH.

References Artandi SE, DePinho RA. (2000). Curr Opin Genet Dev 10: 39–46. Bailey SM, Brenneman MA, Goodwin EH. (2004). Nucleic Acids Res 32: 3743–3751. Bechter OE, Zou Y, Walker W, Wright WE, Shay JW. (2004). Cancer Res 64: 3444–3451.

Bellacosa A. (2001). Cell Death Differ 8: 1076–1092. Buermeyer AB, Deschenes SM, Baker SM, Liskay RM. (1999). Annu Rev Genet 33: 533–564. de Lange T. (2002). Oncogene 21: 532–540. de Wind N, Dekker M, Berns A, Radman M, te Riele H. (1995). Cell 82: 321–330. Oncogene


2536 Doxsey S, Zimmerman W, Mikule K. (2005). Trends Cell Biol 15: 303–311. Edelmann W, Yang K, Umar A, Heyer J, Lau K, Fan K et al. (1997). Cell 91: 467–477. Ghadimi BM, Sackett DL, Difilippantonio MJ, Schrock E, Neumann T, Jauho A, Auer G et al. (2000). Genes Chromosomes Cancer 27: 183–190. Greider CW. (1998). Proc Natl Acad Sci USA 95: 90–92. Kim NW, Wu F. (1997). Nucleic Acids Res 25: 2595–2597. Nigg EA. (2002). Nat Rev Cancer 2: 815–825. Pickett HA, Baird DM, Hoff-Olsen P, Meling GI, Rognum TO, Shaw J et al. (2004). Oncogene 23: 3434–3443.

Prolla TA, Baker SM, Harris AC, Tsao JL, Yao X, Bronner CE et al. (1998). Nat Genet 18: 276–279. Reitmair A, Schmits R, Ewel A, Bapat B, Redston M, Mitri A et al. (1995). Nat Genet 11: 64–70. Rizki A, Lundblad V. (2001). Nature 411: 713–716. Rufer N, Dragowska W, Thornbury G, Roosnek E, Lansdorp PM. (1998). Nat Biotechnol 16: 743–747. Saunders W. (2005). Semin Cancer Biol 15: 25–32. Zijlmans JM, Martens UM, Poon SS, Raap AK, Tanke HJ, Ward RK, Lansdorp PM. (1997). Proc Natl Acad Sci USA 94: 7423–7428.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene