ABERRANT HYPERMETHYLATION OF THE CHFR PROPHASE CHECKPOINT GENE

Download The CHFR gene, which was recently cloned by Scolnick and Halazonetis in search for a novel mitotic checkpoint gene with fork-head associati...
0 downloads 413 Views 362KB Size
Download PDF
Love PNG Images
ã

Oncogene (2002) 21, 2328 ± 2333 2002 Nature Publishing Group All rights reserved 0950 ± 9232/02 $25.00 www.nature.com/onc

Aberrant hypermethylation of the CHFR prophase checkpoint gene in human lung cancers Kotaro Mizuno1,4, Hirotaka Osada1, Hiroyuki Konishi1, Yoshio Tatematsu1, Yasushi Yatabe2, Tetsuya Mitsudomi3, Yoshitaka Fujii1,4 and Takashi Takahashi*,1 1

Division of Molecular Oncology, Aichi Cancer Center Research Institute, 1-1 Kanokoden, Chikusa-ku, Nagoya 464-8681, Japan; Department of Anatomic and Molecular Diagnostic Pathology, Aichi Cancer Center Hospital, 1-1 Kanokoden, Chikusa-ku, Nagoya 464-8681, Japan; 3Department of Thoracic Surgery, Aichi Cancer Center Hospital, 1-1 Kanokoden, Chikusa-ku, Nagoya 464-8681, Japan; 4Department of Surgery II, Nagoya City University School of Medicine, Mizuho-cho, Mizuho-ku, Nagoya 4678061, Japan

2

The CHFR gene, which was recently cloned by Scolnick and Halazonetis in search for a novel mitotic checkpoint gene with fork-head association motifs, has been suggested to play a key role in the mitotic prophase checkpoint. In this study, we demonstrated tumor-speci®c aberrant hypermethylation of the promoter region of the CHFR gene in a signi®cant fraction of lung cancers in association with loss of detectable levels of CHFR transcripts. Aberrant hypermethylation was observed in seven of 37 primary lung cancer cases. Treatment with the demethylating agent 5-aza-2'-deoxycytidine restored expression of the CHFR gene in lung cancer cell lines exhibiting aberrant hypermethylation and loss of its expression. In contrast, genetic alterations were found to be infrequent in lung cancers. This is the ®rst description of aberrant hypermethylation of the CHFR gene in any type of human cancer, and provides further evidence of the involvement of multiple checkpoint alterations in lung cancer. Oncogene (2002) 21, 2328 ± 2333. DOI: 10.1038/sj/ onc/1205402 Keywords: lung cancer; checkpoint; prophase

CHFR;

hypermethylation;

Introduction Lung cancer is the leading cause of cancer deaths in many economically developed countries including the United States and Japan. Molecular biological studies have provided clear evidence that multiple alterations are accumulated in a multi-step manner by both genetic and epigenetic mechanisms during the process of lung carcinogenesis. Representative genes, which are a€ected by genetic mechanisms, include p53, Rb and K-ras (Takahashi et al., 1989; Harbour et al., 1988; Rodenhuis et al., 1987), while aberrant promoter-hypermethylation of the p16 and RASSF1A genes as well as loss of

*Correspondence: T Takahashi; E-mail: [email protected] Received 18 October 2001; revised 11 February 2002; accepted 12 February 2002

imprinting of the insulin-like growth factor II gene belong to the epigenetic type (Merlo et al., 1995; Dammann et al., 2000; Burbee et al., 2001; Suzuki et al., 1994). It is now clear that the G1 checkpoint is frequently impaired as a functional consequence of such defects, while evidence is emerging that mitotic checkpoint impairment may also play an important role (Takahashi et al., 1999), although its molecular mechanisms are less clearly understood (Nomoto et al., 1999; Sato et al., 2000; Gemma et al., 2000). In addition, we have reported in vivo CHK2 inactivation, which suggests the possible involvement of G2 checkpoint impairment in the pathogenesis of human lung cancers (Haruki et al., 2000; Matsuoka et al., 2001). Scolnick and Halazonetis (2000) recently identi®ed a gene, named CHFR, which appears to coordinate another important phase of cell cycle progression, i.e., mitotic prophase. Their examination of eight human cancer cell lines showed loss of CHFR expression in one neuroblastoma and two colon cancer cell lines, although the molecular mechanisms underlying the inactivation were not investigated. In addition, a potential missense mutation was reported in the U2OS osteosarcoma cell line. Importantly, cells expressing wild-type CHFR were shown to exhibit normally delayed entry into metaphase (i.e., prophase delay) in response to mitotic stress, whereas the tumor cell lines that had lost CHFR function entered metaphase with no such delay. In this study, both genetic and epigenetic alterations in the CHFR prophase checkpoint gene were searched for in human lung cancers. We report here that the CHFR gene is inactivated in a signi®cant fraction of lung cancers due to aberrant hypermethylation of its promoter region, providing further evidence of multiple checkpoint alterations in this major form of cancer.

Results Search for genetic alterations in the CHFR gene Forty-four lung cancer cases were examined for the presence of genetic alterations in the CHFR gene by

Hypermethylation of CHFR in lung cancers K Mizuno et al

reverse transcription (RT)-polymerase chain reaction (PCR)-single strand conformation polymorphism (SSCP) and subsequent sequencing analyses. We observed distinct mobility shifts in fragments S6-AS6, S10-AS10 and S11-AS11, which turned out to be newly identi®ed polymorphisms, including those at codons 270 and 497 with amino acid substitutions, and silent polymorphisms at codons 295 and 569 (Table 1). In addition, we found two heterozygotes for the G to A transition at codon 580, which had been reported to be a potential missense mutation with functional impairment of CHFR in the U2OS osteosarcoma cell line (Scolnick and Halazonetis, 2000). It was noted that while both cases showed loss of heterozygosity in the lung tumors, the A allele, which has been reported to confer reduced function, was lost in one case but retained in the other (Figure 1). These ®ndings suggested that genetic alterations of the CHFR gene are very infrequent, if at all, in lung cancers, although it remains possible that mutations might preferentially occur in the regulatory or intronic regions of the CHFR gene.

line with reduced CHFR expression, was also found to carry signi®cant, but mostly partial, methylation of the CpG dinucleotides. In contrast, aberrant hypermethylation was not detected in cell lines with normal CHFR expression such as VMRC-LCD. Association between the aberrant hypermethylation and loss of expression of CHFR was further studied by treatment with a demethylating agent, 5-aza-2'-deoxycytidine (5-aza-dC). RT ± PCR analysis clearly showed restoration of CHFR expression after 96 h of treatment with 5-aza-dC in all three lines with undetectable CHFR expression, supporting the notion of inactivation of CHFR by aberrant hypermethylation (Figure 2d, data not shown for QG90 and Calu6). We then investigated whether aberrant promoter hypermethylation of the CHFR gene was also present

2329

Identification of aberrant hypermethylation of the CHFR gene There is accumulating evidence that epigenetic alterations are an important alternative mechanism of inactivation of genes involved in carcinogenesis. Since the putative promoter region of the CHFR gene appeared to harbor a CpG island, we ®rst investigated the potential involvement of aberrant hypermethylation of CHFR by examining 16 lung cancer cell lines, which were devoid of the contamination by normal cells such as stromal cells and lymphocytes inevitable in primary tumors. Genomic DNAs were treated with sodium bisul®te according to the standard procedures, followed by PCR ampli®cation using oligonucleotide primers that do not correspond to CpG sites and thus can be used for PCR ampli®cation irrespective of the methylation status. Direct sequencing of the resultant PCR products spanning a part of the CpG island showed that heavily methylated CpGs were present in three lung cancer cell lines (QG90, small cell carcinoma; QG56, squamous cell carcinoma; and Calu6, large cell carcinoma) in contrast to lack of hypermethylation in all of the six normal lung tissues examined (Figures 2a ± c). Northern blot analysis revealed a signi®cant association between the aberrant hypermethylation and lack of detectable CHFR expression (Figure 3). PC10, a squamous cell carcinoma cell Table 1

Codon 270 295 497 569 580

Sequence variations identi®ed within the coding region of CHFR Nucleotide GGG/AGG CCA/CCC GCG/GTG TTG/CTG GTG/ATG

Change

Amino acid Gly/Arg Pro/Pro Ala/Val Leu/Leu Val/Met

Figure 1 Identi®cation of two heterozygotes for the G to A transition (GTG to ATG) at codon 580. Note that while both cases show loss of heterozygosity in each lung tumor, the A allele reported to have reduced function was lost in one case but retained in the other Oncogene

Hypermethylation of CHFR in lung cancers K Mizuno et al

2330

Figure 2 Detection of aberrant hypermethylation of the promoter region of the CHFR gene. (a) Schematic diagram of locations of the CpG sites and potential binding sites of transcription factors within the PCR products ampli®ed with primers indicated by arrows. (b) Direct sequencing analysis of PCR products using genomic DNAs treated with sodium bisul®te. Complete methylation is seen at the 16th to 19th CpGs in Calu6 without CHFR expression, whereas such aberrant methylation is absent in VMRC-LCD and a representative normal lung tissue (6N), both of which express CHFR. PC10 shows partial methylation. (c) Summary of the results of methylation analysis. Filled circles, complete methylation; shaded circles, partial methylation; open circles, no methylation. (d) Induction of CHFR in QG56 with promoter hypermethylation by treatment with a demethylating agent, 5-aza-2'-deoxycytidine

in primary lung cancer specimens in vivo using oligonucleotide primers designed to speci®cally amplify Oncogene

either methylated or unmethylated genomic DNA when coupled with sodium bisul®te conversion.

Hypermethylation of CHFR in lung cancers K Mizuno et al

Thirty-seven of the 44 lung cancer cases for which genomic DNAs were available were examined by methylation-speci®c PCR analysis. Seven of 37 (19%) cases were shown to carry apparent hypermethylation of the CHFR promoter region (Figure 4a). Concurrent ampli®cation with both methylated DNA-speci®c and unmethylated DNA-speci®c primers in some cases may have been due to either contamination by normal cells or the presence of hemizygous methylation. Aberrant hypermethylation of the CHFR promoter region was further con®rmed in cases 6, 14, 25, 26 and 34 by direct sequencing of the sodium bisul®te-treated genomic DNAs (data not shown). Taken together, these ®ndings suggested that aberrant hypermethylation of the CHFR promoter region may lead to loss of its expression and consequential prophase checkpoint impairment in lung cancers both in vitro and in vivo. Although the number studied was too small to draw any de®nitive conclu-

sions, our preliminary analysis did not show any signi®cant associations between aberrant hypermethylation of CHFR and clinical parameters including histological type, tumor size, the presence and extent of metastasis, disease stage, smoking history and survival after surgical treatment (data not shown).

2331

Discussion There is accumulating evidence that impairment of various checkpoint functions plays an important role in the pathogenesis of cancers (Levine, 1997; Elledge, 1996). We and others have shown that inactivation of the p53 gene, which disrupts the G1 checkpoint, plays a major role in the pathogenesis of this fatal disease (Takahashi et al., 1989). In addition, our previous report on the possible involvement of mitotic

Figure 3 Loss of CHFR expression in lung cancers. Northern blot analysis shows that CHFR expression is undetectable in QG56, QG90 and Calu6, while PC10 shows reduced CHFR expression. Note that immortalized human bronchial (BEAS2B) and peripheral lung (HPL1D) epithelial cell lines express abundant CHFR transcripts. Two colon cancer cell lines (HT29 and HCT116), which were used by Scolnick and Halazonetis (2000), are also included as positive and negative controls

Figure 4 Representative results of aberrant hypermethylation of CHFR in primary lung cancer specimens in vivo. Methylationspeci®c PCR ampli®cation shows the presence of aberrant hypermethylation in cases 6, 14, 25, 26 and 34. Reverse images of ethidium bromide-stained gels are shown. M, methylated DNA-speci®c ampli®cation; U, un-methylated DNA-speci®c ampli®cation. N, normal lung; T, lung cancer Oncogene

Hypermethylation of CHFR in lung cancers K Mizuno et al

2332

Oncogene

checkpoint impairment in lung cancer development (Takahashi et al., 1999) was supported by the subsequent identi®cation of alterations of the mitotic checkpoint genes such as MAD1 and BUB1 (Nomoto et al., 1999; Sato et al., 2000; Gemma et al., 2000). Inactivation of the CHK2 gene, which is expected to result in the impairment of another important checkpoint, i.e., the G2 checkpoint, has also been identi®ed in lung cancers (Haruki et al., 2000; Matsuoka et al., 2001). Recent molecular cloning of the CHFR gene by Scolnick and Halazonetis (2000), which has checkpoint function in another important cell cycle phase, i.e., prophase, thus prompted us to investigate whether this gene may also be altered in lung cancers. Consequently, we identi®ed loss of CHFR expression in a signi®cant fraction of lung cancers. This study further provides the ®rst description of tumor-speci®c aberrant promoter hypermethylation of CHFR in any type of human cancer. Although it remains to be clari®ed how CHFR inactivation contributes to the development of lung cancers in future biological and biochemical studies, our ®ndings expand on the observations of Scolnick and Halazonetis (2000), lending further support to a connection between CHFR inactivation and cancer. It will be interesting to investigate which types of checkpoint impairment alone or in combination may be associated with speci®c clinical features such as biological aggressiveness and sensitivity to chemotherapeutic drugs and radiation. Both genetic and epigenetic mechanisms have been implicated as mechanisms responsible for the inactivation of various checkpoint-related genes in lung cancer. Although epigenetic silencing is a frequent mechanism of loss of gene function (Baylin and Herman, 2000), it is interesting to note that either type of alteration seems to be predominant in di€erent genes a€ected in lung cancers. In this regard, our ®ndings suggested that the CHFR gene may be preferentially inactivated by the epigenetic mechanism, i.e., aberrant hypermethylation of the promoter region. We noted that the G to A missense substitution, which was previously suggested as a mutation conferring signi®cantly reduced function (Scolnick and Halazonetis, 2000), may represent a functional polymorphism. Although we observed loss of this less ecient allele in an overt lung cancer specimen, this does not preclude the possibility that this functional polymorphism plays a role in the initiation step, showing an association with an individual's susceptibility to lung cancer. Further molecular epidemiological, case-control surveys will therefore be interesting. In conclusion, there is increasing evidence that various checkpoints are perturbed in lung cancers, suggesting that such defects may play an important role in carcinogenesis. By the same token, it has also become clear that additional genes responsible for such impairment need to be identi®ed for a better understanding of this fatal disease, which should ultimately lead to breakthrough in its prevention, diagnosis and treatment.

Materials and methods Lung cancer specimens and cell lines Tumor samples, along with uninvolved lung tissue where available, were collected from 44 lung cancer patients (nine small cell carcinomas, 16 adenocarcinomas, 11 squamous cell carcinomas and eight large cell carcinomas). All tissues were quickly frozen in liquid nitrogen and stored at 7808C until analysis. Sixteen (®ve small cell carcinomas, ®ve squamous cell carcinomas, four adenocarcinomas and two large cell carcinomas) lung cancer cell lines were analysed in this study. Two immortalized human lung epithelial cell lines representing proximal and peripheral airway cells (BEAS2B and HPL1D, respectively) as well as two colon cancer cell lines used by Scolnick and Halazonetis were also included as controls (Scolnick and Halazonetis, 2000; Reddel et al., 1988; Masuda et al., 1997). RT ± PCR ± SSCP analysis PCR ampli®cation using random-primed ®rst-strand cDNAs was performed with the aid of the following oligonucleotide primers in the presence of [32P]dCTP, followed by electrophoretic separation on 6% nondenaturing polyacrylamide gels in both the presence of 5% glycerol at room temperature and in the absence of glycerol at 48C. The primer pairs used for ampli®cation of CHFR were as follows: S1 (sense; 5'GAGGCCGCAATGTCTCTT) and AS1 (antisense; 5'AAGGTCGCAACCTCGTCT); S2 (sense; 5'-AAGCGGGAGTGGACCATG) and AS2 (antisense; 5'-GATGACATCCCCAGTCTG); S3 (sense; 5'-AAGAAGCAGACATGCCCT) and AS3 (antisense; 5'-ACCTGCACCTGAGGTATC); S4 (sense; 5'-ATGTGTTCCATGGGACCA) and AS4 (antisense; 5'-TTAGGGGAGATGCCACCA); S5 (sense; 5'-TCCTCCAGTTGTGGGTCT) and AS5 (antisense; 5'-TGTGCGACCAACAACTGC); S6 (sense; 5'-GATGGGGACCTTGACCTG) and AS6 (antisense; 5'-CATCCAGCCCGAGTAGCA); S7 (sense; 5'-TGCATGCACACGTTCTGC) and AS7 (antisense; 5'-GGCATCCATACTTTGCAC); S8 (sense; 5'-CCAGACAAGAGTCGCAGT) and AS8 (antisense; 5'CCTTCTGTACTCAGGACA); S9 (sense; 5'-CAGCCATACGTCGTGTGC) and AS9 (antisense; 5'-CTGCTCGCGCTCCGCTCT); S10 (sense; 5'-TGCTGCTTCCAGCCCATG) and AS10 (antisense; 5'-GGATGTCTGACTCGTAGC); S11 (sense; 5'-CTGGACGGCGTGCTGAAC) and AS11 (antisense; 5'-GGTCAGCTCACGGAAGCT); and S12 (sense; 5'GTTACTGCTGTGGCCTGC) and AS12 (antisense; 5'TGCTCAGGGCCTCTGGAT). PCR ampli®cation was carried out in the presence of 10% glycerol and consisted of 35 cycles of 948C for 30 s, 588C for 30 s and 728C for 30 s (S1AS1, S2-AS2, S3-AS3, S9-AS9, S10-AS10, S11-AS11); 948C for 30 s, 538C for 30 s and 728C for 30 s (S4-AS4, S5-AS5, S6-AS6, S7-AS7, S8-AS8); or 948C for 30 s, 638C for 30 s and 728C for 30 s (S12-AS12). Sequencing analysis RT ± PCR products of lung cancer specimens showing distinct PCR ± SSCP patterns were sequenced directly using an ABI3100 (Perkin-Elmer, Foster City, CA, USA) DNA sequencer and a Dye Terminator Cycle Sequencing Kit (Perkin-Elmer). Northern blot analysis Northern blot analysis was performed using 5 mg of total RNA according to the standard protocol. A 557 bp cDNA

Hypermethylation of CHFR in lung cancers K Mizuno et al

probe for the CHFR gene was generated by PCR ampli®cation with 5'-TGCTGCTTCCAGCCCATG and 5'TGCTCAGGGCCTCTGGAT oligonucleotide primers. Methylation analysis Sodium bisul®te conversion of genomic DNA of lung cancers was performed essentially as described by Ferguson et al. (2000), followed by PCR ampli®cation using F1 (sense; 5'GTTTTTTTTGTTTTAATA) and R1 (antisense; 5'-CCYCTATCAAAAAACATTAC) oligonucleotide primers, which generated a 248 bp PCR product. PCR ampli®cation consisted of 35 cycles of 948C for 1 min, 488C for 1 min and 728C for 1 min. Genomic organization of the CHFR gene was obtained through GenBank (accession no. NT009455) and potential binding sites of various transcription factors were predicted using TFSEARCH program. The resultant products were puri®ed using a Qiagen PCR puri®cation kit (Qiagen, Inc., Chatsworth, CA, USA) sequenced using the sense primer with an ABI3700 DNA sequencer (Perkin-Elmer) and a Big Dye Terminator Cycle Sequencing Kit (Perkin-Elmer). Bisul®te conversion was con®rmed to be complete by the presence of substituted thymine for all cytosine residues at non-CpG sites. Methylation-speci®c PCR ampli®cation was carried out with the following oligonucleotide primers, which were designed to be speci®c to either methylated or unmethylated DNA after sodium bisul®te conversion as described above. Methylated DNA-speci®c primers were MF1 (sense; 5'ATATAATATGGCGTCGATC) and MR1 (anti-sense; 5'TCAACTAATCCGCGAAACG). Unmethylated DNA-spe-

2333

ci®c primers were UF1 (sense; 5'-ATATAATATGGTGTTGATT) and UR1 (5'-TCAACTAATCCACAAAACA). PCR ampli®cation consisted of 35 cycles of 948C for 1 min, 588C for 1 min and 728C for 1 min (MF1 and MR1); 948C for 1 min, 508C for 1 min and 728C for 1 min. The resultant PCR products were separated on 2% agarose gels. 5-aza-dC treatment Cell lines were treated with 1.25 mM 5-aza-dC for 96 h. RNAs were extracted according to standard procedures, and RT ± PCR analysis was carried out with the S6 and AS6 oligonucleotide primers as described above. b-actin was used as loading control, and the primers used were: 5'-GACTACCTCATGAAGATC and 5'-GATCCACATCTGCTGGAA.

Acknowledgments We would like to thank Drs Curtis C Harris (National Cancer Institute), LJ Old (Memorial Sloan Kettering Cancer Center), M Akiyama (Radiation E€ect Research Foundation) and Y Hayata (Tokyo Medical University) for their generous gifts of cell lines. This work was supported in part by a Grant-in-Aid for Scienti®c Research on Priority Areas from the Ministry of Education, Science and Culture of Japan and a Grant-in-Aid for the Second Term Comprehensive Ten-Year Strategy for Cancer Control from the Ministry of Health and Welfare, Japan.

References Baylin SB and Herman JG. (2000). Trends Genet., 16, 168 ± 174. Burbee DG, Forgacs E, Zochbauer-Muller S, Shivakumar L, Fong K, Gao B, Randle D, Kondo M, Virmani A, Bader S, Sekido Y, Latif F, Milchgrub S, Toyooka S, Gazdar AF, Lerman MI, Zabarovsky E, White M and Minna JD. (2001). J. Natl. Cancer Inst., 93, 691 ± 699. Dammann R, Li C, Yoon JH, Clin PL, Bates S and Pfeifer GP. (2000). Nat. Genet., 25, 315 ± 319. Elledge SJ. (1996). Science, 274, 1664 ± 1672. Ferguson AT, Evron E, Umbricht CB, Pandita TK, Chan TA, Hermeking H, Marks JR, Lambers AR, Futreal PA, Stampfer MR and Sukumar S. (2000). Proc. Natl. Acad. Sci. USA, 97, 6049 ± 6054. Gemma A, Seike M, Seike Y, Uematsu K, Hibino S, Kurimoto F, Yoshimura A, Shibuya M, Harris CC and Kudoh S. (2000). Genes Chromosomes Cancer, 29, 213 ± 218. Harbour JW, Lai SL, Whang-Peng J, Gazdar AF, Minna JD and Kaye FJ. (1988). Science, 241, 353 ± 357. Haruki N, Saito H, Tatematsu Y, Konishi H, Harano T, Masuda A, Osada H, Fujii Y and Takahashi T. (2000). Cancer Res., 60, 4689 ± 4692. Levine AJ. (1997). Cell, 88, 323 ± 331. Masuda A, Kondo M, Saito T, Yatabe Y, Kobayashi T, Okamoto M, Suyama M, Takahashi T and Takahashi T. (1997). Cancer Res., 57, 4898 ± 4904.

Matsuoka S, Nakagawa T, Masuda A, Haruki N, Elledge SJ and Takahashi T. (2001). Cancer Res., 61, 5362 ± 5365. Merlo A, Herman JG, Mao L, Lee DJ, Gabrielson E, Burger PC, Baylin SB and Sidransky D. (1995). Nat. Med., 1, 686 ± 692. Nomoto S, Haruki N, Takahashi T, Masuda A, Koshikawa T, Takahashi T, Fujii Y, Osada H and Takahashi T. (1999). Oncogene, 18, 7180 ± 7183. Reddel RR, Ke Y, Gerwin BI, McMenamin MG, Lechner JF, Su RT, Brash DE, Park JB, Rhim JS and Harris CC. (1988). Cancer Res., 48, 1904 ± 1909. Rodenhuis S, van de Wetering ML, Mooi WJ, Evers SG, van Zandwijk N and Bos JL. (1987). N. Engl. J. Med., 317, 929 ± 935. 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. Scolnick DM and Halazonetis TD. (2000). Nature, 406, 430 ± 435. Suzuki H, Veda R, Takahashi T and Takahashi T. (1994). Nat. Genet., 6, 332 ± 333. Takahashi T, Haruki N, Nomoto S, Masuda A, Saji S, Osada H and Takahashi T. (1999). Oncogene, 18, 4295 ± 4300. Takahashi T, Nau MM, Chiba I, Birrer MJ, Rosenberg RK, Vinocour M, Levitt M, Pass H, Gazdar AF and Minna JD. (1989). Science, 246, 491 ± 494.

Oncogene