Epigenetic Regulation of the Ras Effector/Tumour Suppressor RASSF2 in Breast and Lung Cancer

Oncogene (2008) 27, 1805–1811 & 2008 Nature Publishing Group All rights reserved 0950-9232/08 $30.00 www.nature.com/onc ONCOGENOMICS Epigenetic regulation of the ras effector/tumour suppressor RASSF2 in breast and lung cancer

WN Cooper1, RE Dickinson1, A Dallol1, EV Grigorieva2,3, TV Pavlova2,4, LB Hesson1, I Bieche5, M Broggini6, ER Maher1, ER Zabarovsky2,4, GJ Clark7 and F Latif1

1Department of Medical and Molecular Genetics, Division of Reproductive and Child Health, Institute of Biomedical Research, University of Birmingham, Edgbaston, Birmingham, UK; 2MTC, Karolinska Institute, Theorells vag, Stockholm, Sweden; 3Institute of Molecular Biology and Biophysics SD RAMS. Timakova 2, Novosibirsk, Russia; 4Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Moscow, Russia; 5Laboratoire d’Oncogenetique-INSERM E0017, Centre Rene Huguenin, St-Cloud, France; 6Laboratory of Molecular Pharmacology, Istituto di Ricerche Farmacologiche ‘Mario Negri’, Milan, Italy and 7Molecular Targets Group, Department Of Medicine, J. G. Brown Cancer Center, University of Louisville, Louisville, KY, USA

RASSF2 is a recentlyidentified member of a class of novel Keywords: RASSF2; breast cancer; lung cancer; epi- tumour suppressor genes, all containing a ras-association genetic inactivation; Nuclear localization signal domain. RASSF2 resides at 20p13, a region frequently lost in human cancers. In this report we investigated methylation status of the RASSF2 promoter CpG island in a series of breast, ovarian and non-small cell lung Introduction cancers (NSCLC). RASSF2 was frequentlymethylatedin breast tumour cell lines (65%, 13/20) and in primary Epigenetic silencing of tumour suppressor genes plays an breast tumours (38%, 15/40). RASSF2 expression could important role in the development of many cancer types. be switched back on in methylated breast tumour cell lines 0 0 Identification of genes that are inactivated by promoter after treatment with 5 -aza-2 deoxycytidine. RASSF2 was region CpG island hypermethylation will lead not only to also frequentlymethylatedin NSCLC tumours (44%, insights into novel disease mechanisms but also to (22/50). The small number of corresponding normal breast development of new diagnostic molecular markers and and lung tissue DNA samples analysed were unmethy- potential reactivation therapies. We have identified mem- lated. We also did not detect RASSF2 methylation in bers of a novel family of genes that contain a ras-association ovarian tumours (0/17). Furthermore no mutations were domain (reviewed in Agathanggelou et al., 2005; Hesson found in the coding region of RASSF2 in these ovarian et al., 2007). So far the Ras-association domain family tumours. We identified a highlyconserved putative bi- (RASSF) of genes contains six members, namely, RASSF1 partite nuclear localization signal (NLS) and demon- located at 3p21.3, RASSF2 located at 20p13, RASSF3 strated that endogenous RASSF2 localized to the nucleus. located at 12q14.1, AD037 (also known as RASSF4)located Mutation of the putative NLS abolished the nuclear at 10q11.21, NORE1 located at 1q32.1 and RASSF6 at localization. RASSF2 suppressed breast tumour cell 4q21.21. RASSF1 was identified from a region of over- growth in vitro and in vivo, while the abilityof NLS- lapping homozygous deletions at 3p21.3 in lung and breast mutant RASSF2 to suppress growth was much diminished. tumour cell lines (Dammann et al., 2000; Lerman and Hence we demonstrate that RASSF2 has a functional Minna, 2000; Agathanggelou et al., 2001; Burbee et al., NLS that is important for its tumour suppressor gene 2001). Expression of isoform A of RASSF1 is frequently function. Our data from this and a previous report lost following methylation of its promoter in various adult indicate that RASSF2 is frequentlymethylated in color- and childhood cancers (reviewed in Agathanggelou et al., ectal, breast and NSCLC tumours. We have identified 2005). So far RASSF1A is thought to play an important RASSF2 as a novel methylation marker for multiple role in microtubule stability, it has also shown to be ONCOGENOMICS malignancies and it has the potential to be developed into involved in cell cycle regulation, apoptosis and cell a valuable marker for screening several cancers in parallel migration (see Agathanggelou et al., 2005; Dallol et al., using promoter hypermethylation profiles. 2005). RASSF1A has also been shown to be part of a Oncogene (2008) 27, 1805–1811; doi:10.1038/sj.onc.1210805; complex similar to the Drosophila Hippo/Salvador/Lats published online 24 September 2007 tumour suppressor pathway (Guo et al., 2007). Rassf1a knock-out mice showed increased susceptibility to sponta- Correspondence: Dr F Latif, Department of Medical and Molecular neous and carcinogen-induced tumour formation (Tommasi Genetics, Division of Reproductive and Child Health, Institute of et al., 2005; van der Weyden et al., 2005). Biomedical Research, University of Birmingham, Edgbaston, Birming- While there are other putative transcripts from the ham B15 2TT, UK. E-mail: [email protected] RASSF2 locus, the longest and most abundant is the Received 25 June 2007; revised 18 August 2007; accepted 20 August isoform that has previously been known as RASSF2A, 2007; published online 24 September 2007 and will be hereafter referred to as RASSF2. Epigenetic regulation in breast and lung cancer WN Cooper et al 1806

Figure 1 (a) Frequent methylation of the RASSF2 CpG island is observed in breast tumour cell lines as determined by CoBRA digestion of PCR products. Sodium bisulphite modiﬁcation was performed as described previously (Agathanggelou et al., 2001). The primers and conditions for the PCR have been described previously (Hesson et al., 2005). PCR products were then assayed by digestion with TaqIorBstUI for 2 h at 65 or 60 1C, respectively. (b) Direct bisulphite sequencing of the RASSF2 CpG island in breast tumour cell lines. CG dinucleotides within the ampliﬁcation region are numbered 1–23. The underlined CpGs are contained within BstUI or TaqI recognition sequences. White and black squares represent unmethylated and methylated CpGs, respectively. (c) Expression of RASSF2 in breast tumour cell lines. RASSF2 CpG island hypermethylation corresponds with loss of RASSF2 expression. Expression was restored after treatment with the demethylating agent 50-aza-20deoxycytidine (5-aza-2dC 2.5 mM; Sigma (Poole, UK), media replaced daily for 5 days). Breast tumour line HCC38 is not methylated for RASSF2, hence expression is unchanged after 5-aza-2dC treatment. Primers and conditions have been described previously (Hesson et al., 2005).

As well as RASSF1A, RASSF2 (Hesson et al., 2005), tumour suppressors (Agathanggelou et al., 2005) and AD037/RASSF4 (Eckfeld et al., 2004) and NORE1A demonstrated that it binds directly to K-Ras via the Ras (Hesson et al., 2003; Irimia et al., 2004) are also effector domain (Vos et al., 2003; Hesson et al., 2005). epigenetically inactivated in tumours, while RASSF3 K-Ras is a member of a family of oncoproteins that play (Tommasi et al., 2002; Hesson et al., 2004) and RASSF6 an important role in many biological processes (Rom- (Allen et al., 2007) are not methylated. We have recently mel and Hafen, 1998; Malumbres and Barbacid, 2003; demonstrated that RASSF2 is frequently methylated in Schubbert et al., 2007). These abilities are mediated by colorectal tumours and adenomas and showed that the ability of RAS proteins to interact with a wide range methylation was cancer specific (Hesson et al., 2005). of effector proteins (Shields et al., 2000; Cox and Der, Hence, RASSF2 methylation is a potential molecular 2003); for example via signalling through phosphatidy- marker for early detection of colorectal cancer. RASSF2 linositol 3-kinase and AKT, RAS proteins can induce protein expression is frequently downregulated in human cell division and oncogenesis, whereas interactions with lung tumour cell lines including non-small cell lung cancer RASSF2 can mediate apoptosis and cell cycle arrest (NSCLC) (Vos et al., 2003). Using array-based compara- (Vos et al., 2003). We had previously demonstrated that tive genomic hybridization in ovarian tumour cell lines RASSF2 was frequently methylated in colorectal tu- Lambros et al. (2005) identified regions of deletions mours and adenomas (Hesson et al., 2005) and that including homozygous deletions on several chromosomes. methylation was tumour specific. Furthermore, using Candidate tumour suppressor genes localized to these the same sets of methylation primers as in Hesson et al. regions included CDKN2C and MADHIP (1p32.3), WW (2005), two recent reports (Park et al., 2007; Zhang domain protein WWOX (16q23.1) and RASSF2 (20p13). et al., 2007) have demonstrated that methylation of In this report we determined the methylation status of RASSF2 corresponds to loss of expression in primary RASSF2 in breast, NSCLC and ovarian tumours. We also colorectal and nasopharyngeal tumours as well as undertook experiments to determine tumour suppres- tumour cell lines. sion properties of RASSF2 and characterized a putative NLS in the RASSF2 sequence. RASSF2 methylation analysis in breast tumour cell lines In this report we used combined bisulphite restriction analysis (CoBRA) (Hesson et al., 2005) and bisulphite Results and discussion sequencing to demonstrate for the first time the methylation status of RASSF2 promoter region CpG Previously we identified RASSF2 as a member of the island in a series of breast tumour cell lines. RASSF2 recently discovered RASSF family of Ras effectors/ was methylated in 13 of 20 (65%) of these breast tumour

Oncogene Epigenetic regulation in breast and lung cancer WN Cooper et al 1807

Figure 2 Methylation status of the RASSF2 CpG island in primary tumours (T) and corresponding normal (N) tissue was determined by MSP where M represents methylation-specific PCR and U represents unmethylated-specific PCR. The primers and conditions have been described previously (Hesson et al., 2005). (a) The RASSF2 CpG island promoter region was hypermethylated in 38% (15/40) primary breast tumours. The MCF7 cell line was used as a positive control for the M PCR, CT represents water control. (b) Methylation was found in 22/50 (44%) primary NSCLC tumours. (c) In NSCLC RASSF2 CpG island, methylation was tumour specific. (d) The RASSF2 CpG island promoter region was unmethylated in primary ovarian tumours. OVCA28, an ovarian tumour cell line partially methylated for RASSF2, acts as a positive control for both M and U sets of primers. RASSF2 CoBRA was also carried out on these samples to confirm the result (data not shown). cell lines (Figure 1a). Direct sequencing confirmed these region in the methylated primary breast tumours, and CoBRA results and demonstrated dense methylation revealed no methylation in samples deemed unmethylated across the region (Figure 1b). Treatment with 50-aza- by digestion. There was no correlation between RASSF2 20deoxycytidine (5-aza-2dC) of tumour cell lines where methylation and breast tumour grade. We also analysed RASSF2 was methylated and expression silenced methylation status of RASSF1A inthesameseriesof reactivated RASSF2 expression, while no change in breast tumours. RASSF1A wasmethylatedin18of32 RASSF2 expression was seen in unmethylated tumour (56%) of these breast tumours. There was no correlation lines (Figure 1c). between RASSF1A and RASSF2 methylation status. RASSF2 was methylated in 44% (22/50) of NSCLC tumours (Figure 2b). Where corresponding normal RASSF2 methylation analysis in breast, NSCLC and DNA was available it was not methylated, suggesting ovarian tumours that the methylation was likely to be a tumour-specific We had previously developed a methylation-specific PCR event (Figure 2c). Methylation was found at an equal (MSP) assay to analyse RASSF2 methylation in colorectal frequency in all grades of tumour (I/II ¼ 7/16 (44%); tumours and adenomas (Hesson et al., 2005). The same IIIA ¼ 4/9 (44%); IV ¼ 10/23 (44%) and 1of 2 NSCLC assay was utilized for analysing a series of breast, NSCLC of unknown grade was also methylated) suggesting that and ovarian tumours for RASSF2 methylation. RASSF2 it is an early event in lung tumourigenesis. The was found to be methylated in 38% (15/40) of breast methylation status of RASSF1A in this series of tumours tumours, and the corresponding normal DNA was had previously been determined (Agathanggelou et al., unmethylated (n ¼ 10) (Figure 2a). To determine the extent 2001). There was no association between RASSF1A and of methylation, CoBRA PCR was performed on three RASSF2 methylation status. methylated and one unmethylated breast tumour and the We also analysed ovarian tumours for RASSF2 methyla- PCR products were cloned and sequenced. Sequencing tion using both MSP and CoBRA assays. None of the 17 confirmed the extent of methylation of the promoter ovarian tumours analysed demonstrated methylation of the

Oncogene Epigenetic regulation in breast and lung cancer WN Cooper et al 1808 120 140 140 100 120 120 100 80 100 80 80 60 60 60 40 40 40 Relative colony Relative

20 % colonies in agar % colonies in agar

formation ability (%) formation 20 20 0 0 0 VectorRASSF2A Vector only CI.3 Vector only CI.4

d 60 Vector 1 50 )

3 Vector 2 RASSF2A Cl. 3 CI.3 CI.4 Vector 1 Vector 2 Vector 40 RASSF2A Cl. 4 RASSF2A 30

Tumour size (mm size Tumour 10

0 7 1014172225 28 31 35 Days Figure 3 (a) Colony formation. The number of Geneticin-resistant colonies in cells transfected with empty vector was set at 100%. Values are the mean 7s.d. of three separate experiments each calculated from triplicate plates. The breast tumour cell line MCF7 was transfected with 2 mg of pFLAG or pFLAG-RASSF2 using 6 ml of FuGENE 6 (Roche, Burgess Hill, UK). Cells were selected with 600 mgmlÀ1 Geneticin (Gibco, Paisley, UK) in DMEM. Twenty-one days after transfection-surviving colonies were stained with 0.4% crystal violet. (b) Immunoblot stained with anti-FLAG (Sigma; clone M5) to detect RASSF2 in protein extracts from MCF7 cells stably transfected with pFLAG-RASSF2 (stable clones, Cl.3 and Cl.4). (c) Anchorage-independent growth studies. Growth of MCF7- RASSF2 stable clones (Cl.3 and Cl.4) and MCF7 cells stably transfected with an empty vector were compared using a soft agar assay. In a six-well plate a base layer of 2 ml growth medium containing 0.7% Noble agar was covered with 1 ml of growth medium containing 0.35% Noble agar and 5 Â 104 cells. Finally 2 ml growth medium containing 0.7% Noble agar was layered on top of the cells. After 35 days colonies comprised of at least 32 cells were counted on an inverted microscope. The number of colonies in cells transfected with empty vector was set at 100%. Values are the mean 7s.d. calculated from triplicate plates. (d) Tumour formation in SCID mice. Two different MCF7-RASSF2 stable clones were used for inoculation into SCID mice (Cl.3 and Cl.4). MCF7 cells stably transfected with an empty vector were used as a control (vectors 1 and 2). Tumourigenicity was assayed by s.c. injection as described by Protopopov et al. (2002) and Kashuba et al. (2004). In brief 3 Â 106 cells in combination with Matrigel were inoculated into 4- to 8- week-old SCID mice. Three mice were used per construct and tumour growth in animals was checked twice a week, if tumour formation was observed, tumours were measured using calipers. Animals were maintained according to institutional guidelines.

RASSF2 promoter CpG island (Figure 2d). Since no independent MCF7 clones stably expressing RASSF2 methylation was found in ovarian tumours, we looked for were generated (Cl.3 and Cl.4). These clones inhibited mutations in the coding region of RASSF2 in these samples. anchorage-independent growth in a soft agar assay (Cl.3 No somatic inactivating mutations were found but one and Cl.4 vs vector Po0.05; Figures 3b and c). Cells from silent substitution was detected in exon 8 in two tumours. these RASSF2-expressing clones were also inoculated in to We also analysed methylation status of RASSF1A in the SCID mice to study tumour growth in vivo. Each MCF7- same series of ovarian tumours. RASSF1A was methylated RASSF2 stable clone (Cl.3 and Cl.4) was mixed with in 10 of 17 (59%) of ovarian tumours respectively. This Matrigel and inoculated into three mice. As a control, frequency of RASSF1A methylation in ovarian tumours is MCF7 cells stably transfected with an empty vector were similar to that which has been reported previously (Yoon used (also with Matrigel; vector clones 1 or 2). Solid et al., 2001). tumour nodes were only formed at the inoculation point and their size was measured with calipers. Over the course RASSF2 expression inhibits growth of breast cancer cells of 31–35 days, solid tumours were detected for all mice in vitro and in vivo inoculated with the control MCF7 cells (vector clones 1 or Previous studies have suggested that RASSF2 may 2) but no tumours were observed for mice inoculated with function as a tumour suppressor gene in vitro in cell MCF7-RASSF2 cells (Cl.3 and Cl.4; Figure 3d). This migration, cell cycle progression and colony formation experiment was also done without Matrigel, when after assays (Vos et al., 2003; Akino et al., 2005; Zhang et al., 140 days no tumours were observed in mice inoculated 2007), we have conﬁrmed and extended this work by with RASSF2-expressing clones 3 and 4 but tumours were demonstrating that transfection of Flag-tagged wild-type formed within 120 days with vector control cells. RASSF2 into the MCF7 breast tumour cell line (where Thus we have demonstrated that RASSF2 can endogenous RASSF2 expression is lost by promoter function as a tumour suppressor gene in vitro inhibiting methylation) reduces colony formation ability by 66% the growth of breast cancer cell lines in colony formation compared to cells transfected with the empty vector and soft agar growth assays, and in vivo inhibiting (P ¼ 0.0005, Figure 3a). This effect was consistent through tumour formation when cells expressing RASSF2 were three independent experiments (similar results were subcutaneously injected into SCID mice. This is the ﬁrst obtained in breast tumour cell line HTB19). Two report demonstrating that RASSF2 can suppress tumour

Oncogene Epigenetic regulation in breast and lung cancer WN Cooper et al 1809

Figure 4 (a) RASSF2 protein sequences were aligned using the ClustalW program (www.ebi.ac.uk). The * illustrates the location of amino acids identical in all proteins; : denotes amino acids that undergo conservative substitution; .denotes amino acids that undergo semiconservative substitution. The putative bipartite nuclear localization domain is underlined. The protein schematic of human RASSF2 is coloured black to represent the Ras-association (RA) domain and hatched to illustrate the SARAHdomain. ( b) Cos7 cells transfected with green fluorescent protein (GFP)-RASSF2. (c) Cos7 cells transfected with GFP-RASSF2 MtNLS (The QuikChange site-directed mutagenesis kit (Stratagene, Amsterdam, The Netherlands) was used to mutate the nuclear localization signal of RASSF2. The mutagenic primers were AGTGATGTTGGGGTGGCTGCCGCTGGCAATGTGAGGACG and CGTCCTCACATTGC- CAGCGGCAGCCACCCCAACATCACT). (d) Cos7 cells were transfected with GFP-RASSF2 or GFP-RASSF2 MtNLS and the localization scored as more intense in the nucleus, more intense in the cytoplasm or equal intensity. For each vector at each time point at least 65 cells were analysed. (e) Protein from H720 human lung cancer cell line was fractionated by standard CLB buffer protocol, and the blot immunoprobed with anti-RASSF2, (Vos et al., 2003) then reprobed with a-tubulin and TFIIH(Santa Cruz, Santa Cruz, CA, USA) to confirm that the nuclear and cytoplasmic fractions were free of contaminating cytoplasmic and nuclear proteins, respectively. (f) Growth in soft agar of HTB19 cells stably transfected with an empty vector was compared with HTB19 cells stably expressing RASSF2 or RASSF2 with a mutated NLS as described above except colonies were counted after 21 days. Asterisk (*) signifies Po0.05, evaluated by Student’s t-test. cell growth in vivo hence providing further evidence for alteration between vertebrates and in Xenopus converts its role as a tumour suppressor gene. an R to K thereby retaining the positive charge. Therefore we generated a green fluorescent protein (GFP)-tagged RASSF2 and a version with basic amino RASSF2 contains a functional nuclear acids within the bipartite NLS-mutated (R150A/R151A/ localization signal R152A) transfection of these constructs into Cos7 cells Bioinformatic analysis of RASSF2 was performed using demonstrated predominantly nuclear localization of wild PSORT (www.psort.org) and two putative nuclear type (Figure 4b) and predominantly cytoplasmic locali- localization signals (NLS) were predicted; a monopartite zation of the mutant NLS (Figure 4c). More than 65 cells NLS of the pat 4 subcategory (RRHR between amino from triplicate experiments were analysed for localiza- acids 165 and 168) and a bipartite NLS (Robbins et al., tion of the GFP signal, and the results are presented in 1991) between amino acids 151 and 167. The two motifs Figure 4d. Similar results were obtained in HeLa and 293 of basic amino acids RR(N)10RIRRHare perfectly cells, and the breast tumour cell lines MCF7 and HTB19 conserved among vertebrates Figure 4a, and the only and also with FLAG-tagged RASSF2 constructs.

Oncogene Epigenetic regulation in breast and lung cancer WN Cooper et al 1810 To conﬁrm that the locilization was not due to an methylation of the RASSF2 promoter is an early and a artefact of using the overexpressed tagged protein, the frequent event in a range of epithelial-derived tumours H720 lung cancer cell line was fractionated into nuclear (including NSCLC, breast and colorectal cancers). Thus and cytoplasmic portions, and the lysates subjected to RASSF2 promoter methylation shows potential as a immunoblotting using an antibody against RASSF2. biologically relevant biomarker for the early detection of This conﬁrmed the predominant nuclear localization of these cancers and for developing novel therapeutic endogenous RASSF2 (Figure 4e). intervention regimes. We also demonstrate that RASSF2 HTB19 clones stably expressing RASSF2 or MtNLS can function as a TSG both in vitro and in vivo and has a were generated and while expression of RASSF2 was able functional NLS that is important for its tumour to suppress anchorage-independent growth by 56% suppressor gene function(s). compared to cells transfected with the empty vector (Po0.05), cells expressing MtNLS showed a much smaller (17%) reduction in ability to form colonies in soft agar Acknowledgements compared with the empty vector. When compared with cells expressing RASSF2 with the MtNLS, cells expressing Research in FL laboratory funded in part by Breast Cancer wild-type RASSF2 suppressed growth by 47% (P 0.05; Campaign, Cancer Research UK, Association for International o Cancer Research and Birmingham Children’s Hospital Re- Figure 4f). This suggests that the nuclear localization of search Foundation. ERZ was supported by research grants RASSF2 is important for its function as a tumour from the Swedish Cancer Society, Swedish Research Council, suppressor. Swedish Foundation for International Cooperation in Research The data presented in this study and our previously and Higher Education (STINT), Swedish Institute, Royal published work (Hesson et al., 2005) demonstrate that Swedish Academy of Sciences, INTAS and Karolinska Institute.

References

Agathanggelou A, Cooper WN, Latif F. (2005). Role of the RASSF1A and BLU genes located within the critical 3p21.3 ras-association domain family 1 tumor suppressor gene in region in gliomas. Oncogene 23: 2408–2419. human cancers. Cancer Res 65: 3497–3508. Hesson LB, Cooper WN, Latif F. (2007). Evaluation of the Agathanggelou A, Honorio S, Macartney DP, Martinez A, 3p21.3 tumour-suppressor gene cluster. Oncogene; e-pub Dallol A, Radar J et al. (2001). Methylation associated ahead of print: 28 May 2007. inactivation of RASSF1A from region 3p21.3 in lung, breast Hesson L, Dallol A, Minna JD, Maher ER, Latif F. (2003). and ovarian tumours. Oncogene 20: 1509–1518. NORE1A, a homologue of RASSF1A tumour suppressor Akino K, Toyota M, Suzuki H, Mita H, Sasaki Y, gene is inactivated in human cancers. Oncogene 22: 947–954. Ohe-Toyota M et al. (2005). The Ras effector RASSF2 is Hesson LB, Wilson R, Morton D, Adams C, Walker M, a novel tumor-suppressor gene in human colorectal cancer. Maher ER et al. (2005). CpG island promoter hypermethy- Gastroenterology 129: 156–169. lation of a novel Ras-effector gene RASSF2A is an early Allen NP, Donninger H, Vos MD, Eckfeld K, Hesson L, event in colon carcinogenesis and correlates inversely with Gordon L et al. (2007). RASSF6 is a novel member of the K-ras mutations. Oncogene 24: 3987–3994. RASSF family of tumor suppressors. Oncogene; e-pub Irimia M, Fraga MF, Sanchez-Cespedes M, Esteller M. (2004). ahead of print: 2 April 2007. CpG island promoter hypermethylation of the Ras-effector Burbee DG, Forgacs E, Zochbauer-Muller S, Shivakumar L, gene NORE1A occurs in the context of a wild-type K-ras in Fong K, Gao BN et al. (2001). Epigenetic inactivation lung cancer. Oncogene 23: 8695–8699. of RASSF1A in lung and breast cancers and Kashuba VI, Li JF, Wang FL, Senchenko VN, Protopopov A, malignant phenotype suppression. J Natl Cancer Inst 93: Malyukova A et al. (2004). RBSP3 (HYA22) is a tumor 691–699. suppressor gene implicated in major epithelial malignancies. Cox AD, Der CJ. (2003). The dark side of Ras: regulation of Proc Natl Acad Sci USA 101: 4906–4911. apoptosis. Oncogene 22: 8999–9006. Lambros MB, Fiegler H, Jones A, Gorman P, Roylance RR, Dallol A, Agathanggelou A, Tommasi S, Pfeifer GP, Maher ER, Carter NP et al. (2005). Analysis of ovarian cancer cell lines Latif F. (2005). Involvement of the RASSF1A tumor using array-based comparative genomic hybridization. suppressor gene in controlling cell migration. Cancer Res 65: J Pathol 205: 29–40. 7653–7659. Lerman MI, Minna JD. (2000). The 630-kb lung cancer Dammann R, Li C, Yoon JH, Chin PL, Bates S, Pfeifer GP. homozygous deletion region on human chromosome 3p21.3: (2000). Epigenetic inactivation of a RAS association domain identiﬁcation and evaluation of the resident candidate tumor family protein from the lung tumour suppressor locus suppressor genes. Cancer Res 60: 6116–6133. 3p21.3. Nat Genet 25: 315–319. Malumbres M, Barbacid M. (2003). RAS oncogenes: the ﬁrst Eckfeld K, Hesson L, Vos MD, Bieche I, Latif F, Clark GJ. 30 years. Nat Rev Cancer 3: 459–465. (2004). RASSF4/AD037 is a potential ras effector/ Park HW, Kang HC, Kim IJ, Jang SG, Kim K, Yoon HJ et al. tumor suppressor of the RASSF family. Cancer Res 64: (2007). Correlation between hypermethylation of the 8688–8693. RASSF2A promoter and K-ras/BRAF mutations in micro- Guo C, Tommasi S, Liu L, Yee JK, Dammann R, Pfeifer GP. satellite-stable colorectal cancers. Int J Cancer 120: 7–12. (2007). RASSF1A is part of a complex similar to the Protopopov AI, Li J, Winberg G, Gizatullin RZ, Kashuba VI, Drosophila Hippo/Salvador/Lats tumor-suppressor network. Klein G et al. (2002). Human cell lines engineered for Curr Biol 17: 700–705. tetracycline-regulated expression of tumor suppressor can- Hesson L, Bieche I, Krex D, Criniere E, Hoang-Xuan K, didate genes from a frequently affected chromosomal Maher ER et al. (2004). Frequent epigenetic inactivation of region, 3p21. J Gene Med 4: 397–406.

Oncogene Epigenetic regulation in breast and lung cancer WN Cooper et al 1811 Robbins J, Dilworth SM, Laskey RA, Dingwall C. (1991). Tommasi S, Dammann R, Zhang ZQ, Wang Y, Liu LM, Two interdependent basic domains in nucleoplasmin nuclear Tsark WM et al. (2005). Tumor susceptibility of Rassf1a targeting sequence: identification of a class of bipartite knockout mice. Cancer Res 65: 92–98. nuclear targeting sequence. Cell 64: 615–623. van der Weyden L, Tachibana KK, Gonzalez MA, Adams DJ, Rommel C, Hafen E. (1998). Ras—a versatile cellular switch. Ng BL, Petty R et al. (2005). The RASSF1A isoform of Curr Opin Genet Dev 8: 412–418. RASSF1 promotes microtubule stability and suppresses Schubbert S, Bollag G, Shannon K. (2007). Deregulated Ras tumorigenesis. Mol Cell Biol 25: 8356–8367. signaling in developmental disorders: new tricks for an old Vos MD, Ellis CA, Elam C, Ulku AS, Taylor BJ, Clark GJ. dog. Curr Opin Genet Dev 17: 15–22. (2003). RASSF2 is a novel K-Ras-specific effector and Shields JM, Pruitt K, McFall A, Shaub A, Der CJ. (2000). potential tumor suppressor. J Biol Chem 278: 28045–28051. Understanding Ras: ‘it ain’t over ‘til it’s over’. Trends Cell Yoon JH, Damman R, Pfeifer GP. (2001). Hypermethylation Biol 10: 147–154. of the CpG island of the RASSF1A gene in ovarian and Tommasi S, Dammann R, Jin SG, Zhang XF, Avruch J, renal cell carcinomas. Int J Cancer 94: 212–217. Pfeifer GP. (2002). RASSF3 and NORE1: identifica- Zhang Z, Sun D, Van do N, Tang A, Hu L, Huang G. (2007). tion and cloning of two human homologues of the puta- Inactivation of RASSF2A by promoter methylation corre- tive tumor suppressor gene RASSF1. Oncogene 21: lates with lymph node metastasis in nasopharyngeal 2713–2720. carcinoma. Int J Cancer 120: 32–38.

Oncogene