Main text

It is now widely accepted that tumorigenesis and cancer progression result from the progressive accumulation of genetic alterations conferring a growth advantage to the cells (Loeb and Christians, 1996; Jackson and Loeb, 1998). Numerous observations strongly suggest that the generation of multiple mutations is facilitated by a genetic instability acquired in the early steps of malignant transformation. Two major types of instability have been described in solid cancers, based on the model of colorectal cancer (Lengauer et al., 1998). In a small subset of tumors, microsatellite instability (MIN), which is tightly associated with mismatch repair (MMR) deficiency, results in substitutions, deletions or insertions of few nucleotides (Duval and Hamelin, 2002). Most other sporadic cancers are aneuploid and exhibit an abnormal rate of gross chromosomal abnormalities such as loss of heterozygosity. This latter type of defect is named chromosomal instability (CIN). Despite its relatively common phenotype, the molecular mechanisms giving rise to CIN remain widely undefined. The tumor suppressor p53 acts as the 'guardian of the genome', its inactivation was first proposed to be a major cause of CIN. However, two observations disprove this hypothesis. First, a number of cancers with mutant p53 have been shown to be diploid with no obvious chromosomal defects. Second, Bunz et al. (2002) recently showed that the targeted inactivation of p53 in human cells was not able to cause measurable changes in the structural or numerical stability of chromosomes. Another mechanistic hypothesis relies on the disruption of mitotic checkpoints. Indeed, in vitro, an abnormal partitioning of chromosomes during cell division leads to losses and gains of whole or part of chromosomes (Sen, 2000). So far, most of the research on mitotic checkpoints has focused on the spindle checkpoint that controls the transition from metaphase to anaphase. However, alterations in the expression or sequence of key factors of this checkpoint have been detected in only a few cases, and their causal effect in the acquisition of CIN remains to be demonstrated (Imai et al., 1999; Sato et al., 2000; Wang et al., 2003). Scolnick and Halazonetis (2000) recently described a novel mitotic checkpoint, involving the Chfr gene, which delays chromosomal condensation and entry into metaphase when centrosome separation is inhibited by mitotic stress. The Chfr protein has ubiquitin ligase activity, and negatively regulates the activation of Cdc2 kinase at the G2–M transition, by targeting PLK1 for ubiquitination (Kang et al., 2002). Interestingly, the Chfr is frequently inactivated by an epigenetic mechanism in several human cancers, including colon, lung and esophageal cancers (Mizuno et al., 2002; Shibata et al., 2002; Corn et al., 2003). Colon and breast cancers frequently harbor a high CIN (Kallioniemi et al., 1994; Tsushimi et al., 2001). Given the biological role of Chfr in response to mitotic stress, we analysed the genetic and epigenetic status of the Chfr gene in colon and breast cancer cells displaying either MIN or CIN.

To test whether the loss of expression of Chfr is a frequent event in colon and breast carcinogenesis, Chfr mRNA levels were first evaluated in a matched tumor/normal expression array (Clontech Laboratories, Palo Alto, USA) allowing the parallel analysis of cDNA pairs from tumor and corresponding normal tissues from individual patients (Figure 1a) (Luker et al., 2001). Whereas Chfr expression was easily detected in all breast tumor samples tested, eight of 11 (72%) colon cancers demonstrated a significant decrease of expression as compared to their normal counterparts. This preliminary observation led us to hypothesize that the loss of expression of the Chfr gene is a frequent event during colon tumor progression, but not during breast tumor progression. Chfr expression was then evaluated in 46 primary breast cancers and 25 primary colon cancers by real-time RT–PCR. In support of our hypothesis, Chfr gene was expressed in all breast cancer samples, whereas 20 colon tumors displayed a loss of Chfr expression (data not shown).

Figure 1.

Expression analysis of the Chfr gene. (a) Expression analysis of Chfr using a matched tumor/normal expression array (Clontech, CA, USA), consisting of cDNAs pairs from tumor (T) and corresponding normal (N) tissues from individual patients. Hybridization results obtained with colon and breast samples are presented. A Chfr-specific cDNA fragment was radiolabeled, hybridized overnight at 68°C, washed, then exposed to Biomax MS X-ray film. Signal intensities were then evaluated with a phosphoimager apparatus. (b) Expression analysis of 21 colorectal cancer cell lines (eight MIN and 13 CIN) by Northern blot. Northern blot analysis was performed using 10 g of total RNA according to the standard protocol. To verify equal loading of RNA in each lane, the gel was stained with ethidium bromide before transfer of RNA to nylon membranes. The chfr-specific probe (623 pb) was generated by PCR using 5'-AGT CCA GGA TTA CGT GTG CCC-3' (sense) and 5'-CCT CCA GTG CTG CTG AAA GCT GCT C-3' (antisense) primers

Full figure and legend (210K)

To further study the potential involvement of Chfr alteration in these two frequent types of human cancers, we next studied the Chfr gene status in 21 breast cancer cell lines (BT 20, BT 474, CAL 51, CAMA-1, HBL 100, HCC 1937, HME-1, Hs 578 T, MCF 7, MCF 12A, MDA-MB 157, MDA-MB 231, MDA-MB 361, MDA-MB 436, MDA-MB 453, MDA-MB 459, MDA-MB 549, SK-Br3, T47D, UACC 812 and ZR 75-1) and 21 colon cancer cell lines (CACO 2, CO 115, COLO 320, FET, FRI, HCT 15, HCT 116, HT 29, ISRECO1, ISRECO2, ISRECO3, LOVO, LS 174 T, LS 1034, SW 48, SW 480, SW 620, SW 837, SW 1116 and TC 7, TC 71). In all, 13 colon cancer cell lines displayed a CIN whereas eight were diploid, with a known inactivation of the MMR process by either hMSH2 or hMLH1 mutation leading to MIN (Wheeler et al., 1999; Gayet et al., 2001). All breast cancer cell lines were aneuploid with the exception of HME-1 and Cal-51 (Gioanni et al., 1990; Forozan et al., 2000; Kytola et al., 2000; Rummukainen et al., 2001).

In the initial study by Scolnick and Halazonetis (2000), a missense mutation was identified in the cystein-rich region of Chfr resulting in a loss of function, in the U2OS osteosarcoma cell line. Consequently, in a first approach to evaluate the status of Chfr, we studied the entire coding sequence (17 exons) of the gene in all breast and colon cancer cell lines. No chfr mutation was detected, confirming that gene alteration is not a common mechanism for Chfr inactivation.

Then, we studied Chfr expression in all 42 cancer cell lines using Northern blot and real-time RT–PCR. Results obtained with both methodologies were highly correlated. All breast cancer cell lines exhibited a significant, although variable, expression of Chfr, confirming that the loss of Chfr expression is a rare event in breast carcinogenesis (data not shown). In contrast, undetectable or very low levels of Chfr mRNA were found in 12 of the 21 (57%) colorectal cell lines (Figure 1b, Table 1). More specifically, five of the 13 (38%) CIN colon cancer cell lines, and four of the eight (50%) MIN colon cancer cell lines showed a complete loss of Chfr expression. The latter observation demonstrates that Chfr inactivation is not sufficient to cause CIN.

Table 1 - Summary of Chfr gene analysis in colon cancer cells.

Full table

We next studied the molecular mechanisms underlying the loss of expression of the Chfr gene in colon cancer cells. Numerous studies have demonstrated that the hypermethylation of promoter CpG islands represents an important mechanism in the inactivation of specific gene expression during tumorigenesis; besides, recent data showed an abnormal methylation of the Chfr gene in a variety of human cancer cells (Mizuno et al., 2002; Shibata et al., 2002; Corn et al., 2003). Thus, hypermethylation was plausibly responsible for the loss of expression in our samples. To analyse this hypothesis, demethylation studies were first carried out. Treatment with 5-aza-dC restored Chfr gene expression in the cells with a loss of Chfr transcription, as assessed by real-time RT–PCR (data not shown). These results strongly suggested that methylation plays a causal role in Chfr gene silencing in colorectal cancer cells. Hypermethylation in CpG islands of Chfr promoter was then further determined by a PCR-based methylation assay. In this assay, all unmethylated cytosines are converted to uracil by treatment with sodium bisulfite, followed by methylation-specific PCR. In order to determine the extension and the involvement of specific CpG sequences with abnormal methylation, a 498 bp fragment (+20 -478) encompassing more than 50 CpG sequences within the promoter was amplified from six cell lines with or without endogenous Chfr expression after sodium bisulfite treatment. Subsequent PCR products were then cloned and more than 10 clones of each cell line were sequenced. As shown in Figure 2a, CpG sites in cell lines with no endogenous Chfr expression were frequently methylated. On the basis of this observation, PCR primers were designed from the sequences encompassing the CpG sites sensitive to methylation in order to amplify specifically methylated (M primer set) or unmethylated (U primer set) CpG sequences. Once validated in sequenced clones, the methylation-specific PCR was carried out on a larger panel of cancer cell lines. In support of our previous observations, none of the 19 (0%) breast cancer cell lines showed a methylation of the Chfr DNA sequences, whereas nine of the 21 (43%) colon cancer cell lines displayed a complete methylation, indicating that the epigenetic silencing of Chfr is a tissue-specific mechanism (Figure 2b and Table 1). Finally, to investigate whether the aberrant methylation of the Chfr promoter was a relevant process in vivo, we further analysed the methylation status of 22 primary colon cancers matched with adjacent normal tissues. Aberrant methylation was observed in eight samples (36%) (data not shown).

Figure 2.

Methylation status of 27 CpG sites within the Chfr promoter. (a) Sequencing of 27 CpG dinucleotides spanning the Chfr promoter in six colon cancer cell lines after bisulfite treatment. Each circle was filled according to the degree of methylation on each CpG site. The Chfr mRNA expression level for each cell line is indicated at the left of the diagram (+ or -). Bisulfite reactions were carried out using reagents provided in the CpGenome DNA Modification Kit (Serologicals, Corporation, Norcross, CA, USA). Briefly, 1 g of genomic DNA was denatured with NaOH and treated 16 h with sodium bisulfite following the manufacturer's recommendations. After purification, the bisulfite-converted DNA was PCR amplified using a primer set that spanned a CpG-rich region within the Chfr promoter. Subsequent PCR products were cloned and sequenced. More than 10 clones from each cell line were analysed to obtain the percentage of methylation. (b) Analysis by methylation-specific PCR (MSP) of the promoter region of the Chfr gene. PCR was performed with methylation-specific primers, 5'-ATA TAA TAT GGC GTC GAT C-3' (sense) and 5'-TCA ACT AAT CCG CGA AAC G-3' (antisense); unmethylation-specific primers, 5'-ATA TAA TAT GGT GTT GAT T-3' (sense) and 5'-TCA ACT AAT CCA CAA AAC A-3' (antisense), using 150 ng of the bisulfite-modified genomic DNA as template under the following PCR conditions: one cycle at 95°C for 15 min; 40 cycles at 94°C for 1 min, 58 or 50°C for 1 min, 72°C for 1 min; and one cycle at 72°C for 5 min. PCR products were analysed by agarose-gel electrophoresis and ethidium bromide staining. Lanes U, amplification with primers recognizing unmethylated Chfr alleles. Lanes M, amplification with primers recognizing methylated Chfr. (+) and (-) indicate the presence or absence of Chfr mRNA, as evaluated by real-time RT–PCR and Northern blot analysis

Full figure and legend (194K)

Numerous observations suggest that CIN is caused by mitotic abnormalities such as errors in the partitioning of chromosomes (Paulovich et al., 1997; Cahill et al., 1998). Chfr was recently defined as a central component of a new mitotic checkpoint that delays chromosome condensation in response to mitotic stress; besides, in vitro, Chfr-deficient cells subjected to mitotic stress condense their chromosomes despite failing to separate their centrosomes (Scolnick and Halazonetis, 2000). The role of Chfr as a monitor of centrosome function led us to hypothesize that the loss of Chfr may participate in the acquisition of genetic instability. Indeed, it is now accepted that mitotic fidelity is directly dependent on normal centrosome function. As the centrosome is involved in the control of chromosome segregation, it has been hypothesized that the deregulation of centrosome duplication affects centrosome number and promotes aneuploidy, both common features of human breast and colon cancer cells (Doxsey, 2002). In support of this hypothesis, centrosome amplification is detected in early cancer lesions and is highly associated to aneuploidy and chromosomal instability (Lingle et al., 2002). Indeed, a comparison of MMR-deficient diploid versus MMR-proficient aneuploid colorectal carcinoma cell lines revealed the exclusive occurrence of centrosome amplification and instability in all aneuploid tumor cell lines analysed (Ghadimi et al., 2000). Furthermore, the overexpression of the STK15 kinase (Aurora-A) and the POLO kinase (PLK1), which are among the kinases that colocalize with the centrosome, can induce centrosome amplification and aneuploidy in cell lines (Sen et al., 1997; Zhou et al., 1998). It is worth noting that PLK1 is a known target of Chfr for ubiquitin-dependent proteolysis (Kang et al., 2002). All these observations are consistent with a role of Chfr inactivation in the acquisition of a CIN. Nevertheless, our data clearly demonstrate that Chfr inactivation is not associated with CIN in cancer cells. First, the loss of expression was not observed in breast cancers, whereas centrosomal abnormalities are clearly associated to genomic instability in this tumor type. In contrast, the epigenetic inactivation of Chfr is a frequent event in colorectal cancer cells, as confirmed in a recent report (Corn et al., 2003). However, this inactivation is present in aneuploid cancer cells as well as in diploid cancer cells with MMR deficiency, clearly indicating that Chfr disruption alone does not cause CIN. Conversely, the observation of normal Chfr levels, in the absence of genetic alteration, in CIN cancer cells, suggests that Chfr inactivation is not necessary for CIN. However, it is important to note that our observations do not discard the hypothesis of a role of Chfr alteration in the maintenance of genomic instability. Indeed, it remains possible that, in the presence of other genetic or epigenetic alterations, the loss of Chfr could exacerbate a pre-existing tendency toward such instability.

The process of a hypermethylation of the Chfr promoter sequence is relevant in vivo, since levels of methylated DNA are significantly higher in primary colon cancer cells as compared to their normal counterparts from the same patients, as evaluated by MSP analysis. As demonstrated in cancer cell lines, there is no association with the type of genetic instability (MIN versus CIN). The biological significance of the high frequency of Chfr inactivation in MIN colorectal cancers remains unknown. It may either indicate that the loss of Chfr is a central event in colorectal carcinogenesis, or that it reflects a more general anomaly such as epigenetic instability. Indeed, it has been proposed that a CpG islands methylator phenotype, strongly associated with MMR deficiency, results in unusually widespread hypermethylation and simultaneous silencing of multiple genes (Ahuja et al., 1997; Toyota et al., 1999). Additional studies should help to clarify this issue.

References

1. Ahuja N, Mohan AL, Li Q, Stolker JM, Herman JG, Hamilton SR, Baylin SB and Issa JP. (1997). Cancer Res., 57, 3370–3374. | PubMed | ISI | ChemPort |
2. Bunz F, Fauth C, Speicher MR, Dutriaux A, Sedivy JM, Kinzler KW, Vogelstein B and Lengauer C. (2002). Cancer Res., 62, 1129–1133. | PubMed | ISI | ChemPort |
3. Cahill DP, Lengauer C, Yu J, Riggins GJ, Willson JK, Markowitz SD, Kinzler KW and Vogelstein B. (1998). Nature, 392, 300–303. | Article | PubMed | ISI | ChemPort |
4. Corn PG, Summers MK, Fogt F, Virmani AK, Gazdar AF, Halazonetis TD and El-Deiry WS. (2003). Carcinogenesis, 24, 47–51. | Article | PubMed | ISI | ChemPort |
5. Doxsey S. (2002). Mol. Cell, 10, 439–440. | Article | PubMed | ISI | ChemPort |
6. Duval A and Hamelin R. (2002). Ann. Genet., 45, 71–75. | Article | PubMed | ISI |
7. Forozan F, Mahlamaki EH, Monni O, Chen Y, Veldman R, Jiang Y, Gooden GC, Ethier SP, Kallioniemi A and Kallioniemi OP. (2000). Cancer Res., 60, 4519–4525. | PubMed | ISI | ChemPort |
8. Gayet J, Zhou XP, Duval A, Rolland S, Hoang JM, Cottu P and Hamelin R. (2001). Oncogene, 20, 5025–5032. | Article | PubMed | ISI | ChemPort |
9. Ghadimi BM, Sackett DL, Difilippantonio MJ, Schrock E, Neumann T, Jauho A, Auer G and Ried T. (2000). Genes Chromosomes Cancer, 27, 183–190. | Article | PubMed | ISI | ChemPort |
10. Gioanni J, Le Francois D, Zanghellini E, Mazeau C, Ettore F, Lambert JC, Schneider M and Dutrillaux B. (1990). Br. J. Cancer, 62, 8–13. | PubMed | ISI | ChemPort |
11. Imai Y, Shiratori Y, Kato N, Inoue T and Omata M. (1999). Jpn. J. Cancer Res., 90, 837–840. | PubMed | ChemPort |
12. Jackson AL and Loeb LA. (1998). Semin. Cancer Biol., 8, 421–429. | PubMed |
13. Kallioniemi A, Kallioniemi OP, Piper J, Tanner M, Stokke T, Chen L, Smith HS, Pinkel D, Gray JW and Waldman FM. (1994). Proc. Natl. Acad. Sci. USA, 91, 2156–2160. | Article | PubMed | ChemPort |
14. Kang D, Chen J, Wong J and Fang G. (2002). J. Cell Biol., 156, 249–259. | Article | PubMed | ISI | ChemPort |
15. Kytola S, Rummukainen J, Nordgren A, Karhu R, Farnebo F, Isola J and Larsson C. (2000). Genes Chromosomes Cancer, 28, 308–317. | Article | PubMed | ISI | ChemPort |
16. Lengauer C, Kinzler KW and Vogelstein B. (1998). Nature, 396, 643–649. | Article | PubMed | ISI | ChemPort |
17. Lingle WL, Barrett SL, Negron VC, D'Assoro AB, Boeneman K, Liu W, Whitehead CM, Reynolds C and Salisbury JL. (2002). Proc. Natl. Acad. Sci. USA, 99, 1978–1983. | Article | PubMed | ChemPort |
18. Loeb LA and Christians FC. (1996). Mutat. Res., 350, 279–286. | PubMed | ISI | ChemPort |
19. Luker KE, Pica CM, Schreiber RD and Piwnica-Worms D. (2001). Overexpression of IRF9 confers resistance to antimicrotubule agents in breast cancer cells. Cancer Res., 61, 6540–6547. | PubMed | ChemPort |
20. Mizuno K, Osada H, Konishi H, Tatematsu Y, Yatabe Y, Mitsudomi T, Fujii Y and Takahashi T. (2002). Oncogene, 21, 2328–2333. | Article | PubMed | ISI | ChemPort |
21. Paulovich AG, Toczyski DP and Hartwell LH. (1997). Cell, 88, 315–321. | Article | PubMed | ISI | ChemPort |
22. Rummukainen J, Kytola S, Karhu R, Farnebo F, Larsson C and Isola JJ. (2001). Cancer Genet. Cytogenet., 126, 1–7. | Article | PubMed | ISI | ChemPort |
23. Sato M, Sekido Y, Horio Y, Takahashi M, Saito H, Minna JD, Shimokata K and Hasegawa Y. (2000). Jpn. J. Cancer Res., 91, 504–509. | PubMed | ChemPort |
24. Scolnick DM and Halazonetis TD. (2000). Nature, 406, 430–435. | Article | PubMed | ISI | ChemPort |
25. Sen S. (2000). Curr. Opin. Oncol., 12, 82–88. | Article | PubMed | ISI | ChemPort |
26. Sen S, Zhou H and White RA. (1997). Oncogene, 14, 2195–2200. | Article | PubMed | ISI | ChemPort |
27. Shibata Y, Haruki N, Kuwabara Y, Ishiguro H, Shinoda N, Sato A, Kimura M, Koyama H, Toyama T, Nishiwaki T, Kudo J, Terashita Y, Konishi S, Sugiura H and Fujii Y. (2002). Carcinogenesis, 23, 1695–1699. | Article | PubMed | ISI | ChemPort |
28. Toyota M, Ahuja N, Ohe-Toyota M, Herman JG, Baylin SB and Issa JP. (1999). Proc. Natl. Acad. Sci. USA, 96, 8681–8686. | Article | PubMed | ChemPort |
29. Tsushimi T, Noshima S, Oga A, Esato K and Sasaki K. (2001). Cancer Genet. Cytogenet., 126, 34–38. | Article | PubMed | ISI | ChemPort |
30. Wang Q, Moyret-Lalle C, Couzon F, Surbiguet-Clippe C, Saurin JC, Lorca T, Navarro C and Puisieux A. (2003). Oncogene, 22, 1486–1490. | Article | PubMed | ISI | ChemPort |
31. Wheeler JM, Beck NE, Kim HC, Tomlinson IP, Mortensen NJ and Bodmer WF. (1999). Proc. Natl. Acad. Sci. USA, 96, 10296–10301. | Article | PubMed | ChemPort |
32. Zhou H, Kuang J, Zhong L, Kuo WL, Gray JW, Sahin A, Brinkley BR and Sen S. (1998). Nat. Genet., 20, 189–193. | Article | PubMed | ISI | ChemPort |

Acknowledgements

We thank T Halazonetis for Chfr constructs. This work was supported by grants from the Comité Départemental du Rhône and the Comité Départemental de l'Ain de la Ligue de Lutte contre le Cancer, and from the Association pour la Recherche sur le Cancer.