The CpG island of the novel tumor suppressor gene ... - Nature

26 downloads 0 Views 326KB Size Report
The CpG island of the novel tumor suppressor gene RASSF1A is intensely methylated in primary small cell lung carcinomas. Reinhard Dammann1, Takashi ...
Oncogene (2001) 20, 3563 ± 3567 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc

SHORT REPORTS

The CpG island of the novel tumor suppressor gene RASSF1A is intensely methylated in primary small cell lung carcinomas Reinhard Dammann1, Takashi Takahashi2 and Gerd P Pfeifer*,1 1

Department of Biology, Beckman Research Institute, City of Hope Cancer Center, Duarte, California 91010, USA; 2Division of Molecular Oncology, Aichi Cancer Center Research Institute, 1-1 Kanokoden, Chikusa-ku, Nagoya 464-8681, Japan

Loss of heterozygosity at 3p21.3 occurs in more than 90% of small cell lung carcinomas (SCLCs). The Ras association domain family 1 (RASSF1) gene cloned from the lung tumor suppressor locus 3p21.3 consists of two major alternative transcripts, RASSF1A and RASSF1C. Epigenetic inactivation of isoform A (RASSF1A) was observed in 40% of primary non-small cell lung carcinomas and in several tumor cell lines. Transfection of RASSF1A suppressed the growth of lung cancer cells in vitro and in nude mice. Here we have analysed the methylation status of the CpG island promoters of RASSF1A and RASSF1C in primary SCLCs. In 22 of 28 SCLCs (=79%) the promoter of RASSF1A was highly methylated at all CpG sites analysed. None of the SCLCs showed evidence for methylation of the CpG island of RASSF1C. The results suggest that hypermethylation of the CpG island promoter of the RASSF1A gene is associated with SCLC pathogenesis. Oncogene (2001) 20, 3563 ± 3567. Keywords: RASSF1; methylation; CpG island; tumor suppressor gene; small cell lung cancer; 3p21 Introduction Lung cancer currently is the leading cause of cancer death in the United States with more than 150 000 deaths annually attributed to this disease (Wingo et al., 1999). Lung cancers are broadly classi®ed into two groups: small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC), the latter being divided into the major histological subtypes adenocarcinoma, squamous cell carcinoma, and large cell carcinoma. Losses of heterozygosity (LOH) on chromosome arms 3p, 9p, 13q and 17p are among the most frequent alterations in lung tumors (Whang-Peng et al., 1982; Yokota et al., 1987; Naylor et al., 1987; Hibi et al., 1992; Kohno and Yokota, 1999; Girard et al., 2000). However, only a limited number of tumor suppressor genes have been identi®ed as potential targets for chromosomal deletion events in these tumors. These

*Correspondence: GP Pfeifer Received 5 January 2001; revised 27 February 2001; accepted 15 March 2001

genes include the RB gene (13q), the p53 gene (17p), and the p16 CDK inhibitor gene (9p). Allelic loss at 3p21 is an early event in lung tumor pathogenesis, occurs at the stage of hyperplasia/ metaplasia (Sundaresan et al., 1992; Hung et al., 1995; Thiberville et al., 1995; Wistuba et al., 1999), and as such may be critically involved in tumor initiation. Cross-sectional examination of individual NSCLC tumors showed that all portions of the tumor shared concordant LOH at 3p despite morphological diversity (Yatabe et al., 2000). Homozygous deletions at 3p21.3 have been described in several cancer cell lines and in primary lung tumors (Killary et al., 1992; Yamakawa et al., 1993; Wei et al., 1996; Kok et al., 1997; Todd et al., 1997). Recently, a region of minimum homozygous deletion was narrowed to 120 kb using several lung cancer and a breast cancer cell line (Sekido et al., 1998). Until now, several genes located in the 3p21.3-deleted chromosomal region have been isolated as candidates for tumor suppressor genes. However, none of the genes located within the 3p21.3 homozygous deletion area was found to be mutated in more than 5 ± 10% of lung tumors (Lerman and Minna, 2000). This supports the assumption that the putative 3p21.3 tumor suppressor gene is inactivated by mechanisms other than mutation of the coding sequence. Transcriptional silencing by hypermethylation of CpG islands containing the promoter regions of tumor suppressor genes is becoming recognized as a common mechanism (Baylin et al., 1998; Jones and Laird, 1999; Baylin and Herman, 2000). Recent studies have demonstrated that the CpG islands in the RB, p16, VHL, APC, MLH1, and BRCA1 genes are frequently methylated in a variety of human cancers, but are usually methylation-free in the corresponding normal tissues (Baylin et al., 1998; Jones and Laird, 1999; Baylin and Herman, 2000). CpG islands that are hypermethylated in lung cancer include those of the p16 gene (Shapiro et al., 1995; Otterson et al., 1995; Merlo et al., 1995; Belinsky et al., 1998), the DNA repair gene MGMT (Palmisano et al., 2000), the deathassociated protein kinase (Tang et al., 2000), and DMBT1, a candidate tumor suppressor gene located at 10q25.3 ± 26.1 (Wu et al., 1999). Inactivation of tumor suppressor genes due to hypermethylation may play an important role in carcinogenesis.

RASSF1A methylation in lung cancer R Dammann et al

3564

Figure 1 Methylation analysis of the RASSF1A gene in primary small cell lung carcinomas. Sequences of PCR products from bisul®te-treated DNA for the CpG island spanning the RASSF1A promoter were obtained from di€erent cancer and normal samples. Methylated cytosines appear as a G signal in the complementary strand (bold Gs). The sequence is shown before and after bisul®te conversion at the top. The position of a TaqI restriction site is underlined. Two Sp1 consensus sites are boxed. The data shown were obtained by direct sequencing of PCR products from bisul®te-treated DNA or from PCR products subcloned into plasmids (two primary tumors and matched normal tissue controls). Small cell lung carcinomas and the corresponding normal lung tissues were obtained from the Aichi Cancer Center Research Institute and the City of Hope National Medical Center. DNA was isolated from tumors, and the methylation status of the RASSF1A promoter region was determined by a bisul®te genomic sequencing protocol (Clark et al., 1994; Dammann et al., 2000). DNA sequences were ampli®ed by mixing 100 ng of bisul®te-treated DNA with primers MU379 (5'-GTTTTGGTAGTTTAATGAGTTTAGGTTTTTT) and ML730 (5'-ACCCTCTTCCTCTAACACAATAAAACTAACC) in 100 mMl reaction bu€er containing 200 mM of each dNTP and Taq polymerase (Roche Diagnostics Corp.; Indianapolis, IN, USA) and incubating at 958C for 15 s, 558C for 15 s and 748C for 30 s, for 20 cycles. A semi-nested PCR was performed using 1/50 of the initially ampli®ed products and an internal primer ML561 (5'-CCCCACAATCCCTACACCCAAAT) and primer MU379 with similar conditions as described for the preceding PCR ampli®cation, but for 30 cycles. PCR products were initially sequenced directly to obtain average methylation levels and were then subcloned for con®rmation Oncogene

In previous work we have cloned and characterized the Ras association domain family 1A gene (RASSF1A) located within the 120 kb 3p21.3 minimal homozygous deletion region and found that this gene is epigenetically inactivated in 40% of primary nonsmall cell lung cancers (Dammann et al., 2000). The RASSF1 protein is 55% homologous to the mammalian Ras e€ector Nore1 (Vavvas et al., 1998). Reexpression of RASSF1A in A549 lung cancer cells reduced colony formation, suppressed anchorage-independent growth and inhibited tumor formation in nude mice (Dammann et al., 2000). In the present study we have investigated the methylation status of the RASSF1A gene in primary SCLC samples. Previously, we have observed that all 17 of the 17 analysed SCLC cell lines had lost the expression of RASSF1A and silencing of RASSF1A correlated with hypermethylation of the CpG island promoter sequence (Dammann et al., 2000). However, cell lines are known to have unusually high frequencies of CpG island methylation (Antequera et al., 1990; Jones et al., 1990), which makes an analysis of primary tumors a necessity. To analyse the RASSF1 gene in primary SCLCs, we determined the methylation status of the two CpG islands, in which the alternative RASSF1A and RASSF1C transcripts initiate. We used bisul®te sequencing of genomic DNA (Clark et al., 1994) to determine the methylation status of CpG dinucleotides. We analysed 16 CpGs in a 205 bp fragment containing three Sp1 consensus binding sites and the putative transcription and translation initiation sites of RASSF1A. All guanines present after bisul®te sequencing in Figure 1 are derived from methylated cytosines on the complementary strand. In Figure 1, one SCLC cell line (H345) shows complete methylation of the sequence analysed. Further, we analysed the methylation status of the RASSF1A promoter in 28 primary non-microdissected SCLC samples. Of the 28 samples analysed, 22 (=79%) were methylated in the promoter region of RASSF1A. Five matching normal tissue samples were available. Several examples, including two matching tumor and normal tissues, are shown in Figure 1. One of the normal tissue samples was methylated, but to a lower degree than the corresponding tumor (25% versus 80%). The other four normal matching tissue samples were unmethylated. In Figure 1, ®ve methylated primary small cell lung carcinomas are shown. We estimated from the sequence scans of the uncloned PCR products that the tumor samples SC-1, SC-3, and SC-7 showed more than 80% bisul®te modi®cationresistant cytosines at all CpGs analysed. In total, more than 75% of the SCLC samples that were methylated contained more than 60% bisul®te-resistant cytosines at all CpG sequences. This high level of methylation indicates that a large fraction of the tumor cell population has the promoter of RASSF1A methylated. Since LOH at 3p21.3 occurs in almost 100% of all SCLC tumors (Whang-Peng et al., 1982; Naylor et al., 1987; Hibi et al., 1992; Kok et al., 1997; Wistuba et

RASSF1A methylation in lung cancer R Dammann et al

al., 2000; Girard et al., 2000; Lindblad-Toh et al., 2000), the remaining allele is silenced by methylation. The high methylation frequency of the RASSF1A promoter correlates well with the high LOH frequencies at 3p21.3 generally found in SCLC. Another methodology to analyse the PCR fragments obtained from bisul®te-modi®ed DNA is by further digestion with a restriction enzyme that has a CpG in its recognition sequence (Xiong and Laird, 1997). TaqI (5'TCGA-3') or BstUI (5'CGCG-3') will cut only previously methylated DNA after bisul®te treatment and PCR. The consensus sequences will be lost by cytosine deamination in unmethylated samples. The analysed 205 bp fragment of RASSF1A contains two TaqI restriction sites after bisul®te conversion of CpG methylated DNA. Restriction digestion of PCR products obtained from DNA methylated at both TaqI sites results in three bands (90, 81 and 34 bp; the two larger ones migrating together). PCR products obtained from DNA from HeLa cells were unmethylated (consistent with expression of RASSF1A in this cell line; Dammann et al., 2000), whereas in vitro methylated HeLa DNA (Methyl.) was digested at these TaqI sites (Figure 2). In Figure 2, we analysed seven primary small cell lung carcinomas by TaqI restriction (cases SC-1 ± 7). The tumor samples showed approximately 60 ± 95% methylation of the TaqI sites except for sample number SC-2, which was methylated at a level of 30%. For all the analysed SCLCs the results obtained by restriction digestion and direct sequencing were practically identical. For almost half of the SCLC cases the methylation levels of these CpG sites were higher than 80%. For comparison, we analysed the CpG island promoter of RASSF1C. Expression analysis of RASSF1C in SCLC cell lines showed no reduction of transcription (Dammann et al., 2000). In order to determine the methylation status of the CpG island promoter of RASSF1C in primary SCLCs, we performed restriction analysis of PCR ampli®ed bisul®te-modi®ed DNA. The analysed 311 bp promoter fragment contains 38 CpGs, one Sp1 consensus binding site and the putative transcription and translation initiation sites of RASSF1C. This methylated fragment has ®ve BstUI sites and digestion results in bands of 140, 89, 31, 21, 16 and 14 bp. The seven SCLCs with high methylation levels at the CpG island of RASSF1A showed no evidence for methylation in the CpG island of RASSF1C (Figure 2). All analysed SCLCs were unmethylated at all CpG sites in the promoter of RASSF1C and this was con®rmed by direct sequencing (data not shown). This shows that methylation of the RASSF1A CpG island is speci®c for this locus. Methylation does not occur on the RASSF1C CpG island, which is downstream of the RASSF1A island, nor does it occur in the CpG island of the BLU gene, which is upstream (proximal) of RASSF1A. We did not ®nd any methylation in 20 primary NSCLCs and in 10 primary SCLCs in the BLU gene (C-L Chiang, R Dammann, T Takahashi, GP Pfeifer, unpublished results).

3565

Figure 2 Methylation analysis of RASSF1A and RASSF1C by restriction digestion. PCR products from bisul®te-treated DNA obtained from HeLa cells, primary SCLCs (1 ± 7) and in vitro methylated HeLa DNA (Methyl.) were digested (+) or mockdigested (7). The sizes of molecular weight markers (M) are shown on the right in base pairs. For the restriction enzyme analysis of PCR products from bisul®te-treated DNA (Xiong and Laird, 1997), 50 ng of the PCR products were digested with 10 units of TaqI (for RASSF1A) or 10 units of BstUI (for RASSF1C) according to conditions speci®ed by the manufacturer of the enzymes (New England Biolabs; Beverly, MA, USA) and analysed on 2.2% TBE agarose gels. To determine the methylation status of the RASSF1C promoter region, PCR reactions similar to those described in Figure 1 were performed. For the ®rst PCR reaction primers M1305 (5'GTTTTTTGTGGTAGGTGGGGTTTG) and M1627 (5'-AATCCRAATCCTCTTAACTACAATAACCAC) were used. For the nested PCR, the internal primers M1318 (GGTGGGGTTTGTGAGTGGAGTTT) and M1599 (5'-ACTACTCRTCRTACTACTCCAAATCATTTC) were used

Methylation and LOH are the major loss of function pathways for the RASSF1A gene since somatic mutations appear to be rare in this gene (Dammann et al., 2000). We have found no mutations in 16 SCLC cell lines. Epigenetic silencing is a common mechanism for loss of tumor suppressor gene function in cancer. Promoter methylation of tumor suppressor genes, such as p16, and DNA repair genes, such as MGMT, have been detected in DNA from sputum in patients with squamous cell lung cancer several years before clinical diagnosis (Palmisano et al., 2000). A similar methylation analysis of the RASSF1A promoter from sputum of smokers could serve as a sensitive early detection method for small cell lung cancer. This gene may be particularly relevant in this type of analysis, since RASSF1A promoter hypermethylation is a very frequent event, and LOH analysis of the same region predicts that it might be an early change in the pathogenesis of lung cancer. In a recent study, Vos et al. (2000) have shown that RASSF1 binds RAS in a GTP-dependent manner. Over-expression of RASSF1C induced apoptosis (Vos et al., 2000). This pro-apoptotic e€ect of RASSF1C is enhanced by activated RAS and inhibited by dominant negative RAS. No epigenetic inactivation of RASSF1C has so far been shown in lung cancer (Dammann et al., 2000; this study). Vos et al. (2000) reported a reduction of RASSF1C expression in ovarian tumor cell lines. Since isoforms A and C encode for the identical RAS association domain, it is possible that RASSF1A binds to RAS in a similar same manner as RASSF1C. Activated RAS proteins are usually associated with growth enhancement and transformation. However, Oncogene

RASSF1A methylation in lung cancer R Dammann et al

3566

RAS also induces growth inhibitory e€ects manifested by senescence (Serrano et al., 1997), terminal di€erentiation (Bar-Sagi and Feramisco, 1985) or apoptosis (Mayo et al., 1997; Chen et al., 1998; Downward, 1998; Shao et al., 2000). RASSF1 might be responsible for the RAS-dependent growth inhibition through its proapoptotic function (Vos et al., 2000). Loss of RASSF1 expression by methylation in human cancer may shift the balance of RAS activities towards a growth promoting e€ect without the necessity of RAS activating mutations. Indeed, RAS mutations are found in less than 1% of SCLCs (Mitsudomi et al., 1991; Wagner et al., 1993), whereas inactivation of

RASSF1A is close to 80 or 100% by hypermethylation and LOH, respectively. This is consistent with the possibility that RASSF1A might be the tumor suppressor gene, which is associated with the early and frequent loss of the 3p21.3 locus in lung cancer development.

Acknowledgments This work was supported by a grant from the University of California Tobacco Related Disease Research Program (9RT-0175) to GP Pfeifer.

References Antequera F, Boyes J and Bird A. (1990). Cell, 62, 503 ± 514. Bar-Sagi D and Feramisco JR. (1985). Cell, 42, 841 ± 848. Baylin SB and Herman JG. (2000). Trends Genet., 16, 168 ± 174. Baylin SB, Herman JG, Gra€ JR, Vertino PM. and Issa JP. (1998). Adv. Cancer Res., 72, 141 ± 196. Belinsky SA, Nikula KJ, Palmisano WA, Michels R, Saccomanno G, Gabrielson E, Baylin SB and Herman JG. (1998). Proc. Natl. Acad. Sci. USA, 95, 11891 ± 11896. Chen CY, Liou J, Forman LW and Faller DV. (1998). J. Biol. Chem., 273, 16700 ± 16709. Clark SJ, Harrison J, Paul CL and Frommer M. (1994). Nucleic Acids Res., 22, 2990 ± 2997. Dammann R, Li C, Yoon JH, Chin PL, Bates S and Pfeifer GP. (2000). Nature Genet., 25, 315 ± 319. Downward J. (1998). Curr. Opin. Genet. Dev., 8, 49 ± 54. Girard L, Zochbauer-Muller S, Virmani AK, Gazdar AF and Minna JD. (2000). Cancer Res., 60, 4894 ± 4906. Hibi K, Takahashi T, Yamakawa K, Ueda R, Sekido Y, Ariyoshi Y, Suyama M, Takagi H and Nakamura Y. (1992). Oncogene, 7, 445449. Hung J, Kishimoto Y, Sugio K., Virmani A, McIntire DD, Minna JD and Gazdar AF. (1995). JAMA, 273, 558 ± 563. Jones PA, Wolkowicz MJ, Rideout WM, Gonzales FA, Marziasz CM, Coetzee GA and Tapscott SJ. (1990). Proc. Natl. Acad. Sci. USA, 87, 6117 ± 6121. Jones PA and Laird PW. (1999). Nature Genet., 21, 163 ± 167. Killary AM, Wolf ME, Giambernardi TA and Naylor SL. (1992). Proc. Natl. Acad. Sci. USA, 89, 10877 ± 10881. Kohno T and Yokota J. (1999). Carcinogenesis, 20, 1403 ± 1410. Kok K, Naylor SL and Buys CHCM. (1997). Adv. Cancer Res., 71, 27 ± 92. Lerman MI and Minna JD. (2000). Cancer Res., 60, 6116 ± 6133. Lindblad-Toh K, Tanenbaum DM, Daly MJ, Winchester E, Lui WO, Villapakkam A, Stanton SE, Larsson C, Hudson TJ, Johnson BE, Lander ES and Meyerson M. (2000). Nature Biotechnol., 18, 1001 ± 1005. Mayo MW, Wang CY, Cogswell PC, Rogers-Graham KS, Lowe SW, Der CJ and Baldwin Jr. AS, (1997). Science, 278, 1812 ± 1815. Merlo A, Herman JG, Mao L, Lee DJ, Gabrielson E, Burger PC, Baylin SB and Sidransky D. (1995). Nature Med., 1, 686 ± 692. Mitsudomi T, Viallet J, Mulshine JL, Linnoila RI, Minna JD. and Gazdar AF. (1991). Oncogene, 6, 1353 ± 1362. Naylor SL, Johnson BE, Minna JD and Sakaguchi AY. (1987). Nature, 329, 451 ± 454. Oncogene

Otterson GA, Khleif SN, Chen W, Coxon AB and Kaye FJ. (1995). Oncogene, 11, 1211 ± 1216. Palmisano WA, Divine KK, Saccomanno G, Gilliland FD, Baylin SB, Herman JG and Belinsky SA. (2000). Cancer Res., 60, 5954 ± 5958. Sekido Y, Ahmadian M, Wistuba II, Latif F, Bader S, Wei M-H, Duh F-M, Gazdar AF, Lerman MI and Minna JD. (1998). Oncogene, 16, 3151 ± 3157. Serrano, M, Lin AW, McCurrach ME, Beach D and Lowe SW (1997). Cell, 88, 593 ± 602. Shao J, Sheng H, DuBois RN and Beauchamp RD. (2000). J. Biol. Chem., 275, 22916 ± 22924. Shapiro GI, Park JE, Edwards CD, Mao L, Merlo A, Sidransky D, Ewen ME and Rollins BJ. (1995). Cancer Res., 55, 6200 ± 6209. Sundaresan V, Ganly P, Hasleton P, Rudd R, Sinha G, Bleehen N and Rabbitts P. (1992). Oncogene, 7, 1989 ± 1997. Tang X, Khuri FR, Lee JJ, Kemp BL, Liu D, Hong WK and Mao L. (2000). J. Natl. Cancer Inst., 92, 1511 ± 1516. Thiberville, L, Payne P, Vielkinds J, LeRiche J, Horsman D, Nouvet G, Palcic B and Lam S. (1995). Cancer Res., 55, 5133 ± 5139. Todd S, Franklin WA, Varella-Garcia M, Kennedy T, Hilliker Jr. CE, Hahner L, Anderson M, Wiest JS, Drabkin HA and Gemmill RM. (1997). Cancer Res., 57, 1344 ± 1352. Vavvas D, Li X, Avruch J and Zhang X-F. (1998). J. Biol. Chem., 273, 5439 ± 5442. Vos MD, Ellis CA, Bell A, Birrer MJ and Clark GJ. (2000). J. Biol. Chem., 275, 35669 ± 35672. Wagner SN, Muller R, Boehm J, Putz B, Wunsch PH and Ho¯er H. (1993). Virchows Arch. B., 63, 325 ± 329. Wei M-H, Latif F, Bader S, Kashuba V, Chen J-Y, Duh FM, Sekido Y, Lee C-C, Geil L, Kuzmin I, Zabarovsky E, Klein G, Zbar B, Minna JD and Lerman MI. (1996). Cancer Res., 56, 1487 ± 1492. Whang-Peng J, Kao-Shan CS, Lee EC, Bunn PA, Carney DN, Gazdar AF and Minna JD. (1982). Science, 215, 181 ± 182. Wingo PA, Ries LA, Giovino GA, Miller DS, Rosenberg HM, Shopland DR, Thun MJ and Edwards BK. (1999). J. Natl. Cancer Inst.,91, 675 ± 690. Wistuba II., Behrens C, Milchgrub S, Bryant D, Hung J, Minna JD and Gazdar AF. (1999). Oncogene, 18, 643 ± 650.

RASSF1A methylation in lung cancer R Dammann et al

Wistuba II, Behrens C, Virmani AK, Mele G, Milchgrub S, Girard L, Fondon III, JW, Garner HR, McKay B, Latif F, Lerman MI, Lam S, Gazdar AF and Minna JD. (2000). Cancer Res., 60, 1949 ± 1960. Wu W, Kemp BL, Proctor ML, Gazdar AF, Minna JD, Hong WK and Mao L. (1999). Cancer Res., 59, 1846 ± 1851. Xiong Z and Laird PW. (1997). Nucleic Acids Res., 25, 2532 ± 2534.

Yamakawa K, Takahashi T, Horio Y, Murata Y, Takahashi E, Hibi K, Yokoyama S, Ueda R and Nakamura Y. (1993). Oncogene, 8, 327 ± 330. Yatabe Y, Konishi H, Mitsudomi T, Nakamura S and Takahashi T. (2000). Am. J. Pathol., 157, 985 ± 993. Yokota J, Wada M, Shimosato Y, Terada M and Sugimura T. (1987). Proc. Natl. Acad. Sci. USA, 84, 9252 ± 9256.

3567

Oncogene