Epigenetic inactivation of the candidate 3p21. 3 suppressor gene BLU ...

Oncogene (2003) 22, 1580–1588

& 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00 www.nature.com/onc

Epigenetic inactivation of the candidate 3p21.3 suppressor gene BLU in human cancers Angelo Agathanggelou1, Ashraf Dallol1, Sabine Zo¨chbauer-Mu¨ller2,8, Catherine Morrissey2, Sofia Honorio1, Luke Hesson1, Tommy Martinsson3, Kwun M Fong4, Michael J Kuo5, Po Wing Yuen6, Eamonn R Maher1,7, John D Minna2 and Farida Latif*,1,7

ONCOGENOMICS

1

Section of Medical and Molecular Genetics, Department of Paediatrics and Child Health, University of Birmingham, The Medical School, Edgbaston, Birmingham B15 2TT, UK; 2Hamon Center for Therapeutic Oncology Research, University of Texas Southwestern Medical Center at Dallas, TX, USA; 3Department of Clinical Genetics, Gothenburg University, Sahlgrenska University, Hospital/Ostra, S-416 85 Gothenburg, Sweden; 4The Prince Charles Hospital, Rode Road, Chermside, Brisbane 4032, Australia; 5 Birmingham Children’s Hospital, Steelhouse Lane, Birmingham B4 6NH, UK; 6Department of Surgery, University of Hong Kong, Hong Kong, China; 7Cancer Research UK Renal Molecular Oncology Research Group, University of Birmingham, The Medical School, Birmingham B15 2TT, UK

Many distinct regions of 3p show frequent allelic losses in a wide range of tumour types. Previously, the BLU candidate tumour suppressor gene (TSG) encoded by a gene-rich critical deleted region in 3p21.3 was found to be inactivated rarely in lung cancer, although expression was downregulated in a subset of lung tumour cell lines. To elucidate the role of BLU in tumorigenesis, we analysed BLU promoter methylation status in tumour cell lines and detected promoter region hypermethylation in 39% lung, 42% breast, 50% kidney, 86% neuroblastoma and 80% nasopharyngeal (NPC) tumour cell lines. Methylation of the BLU promoter region correlated with the downregulation of BLU transcript expression in tumour cell lines. Expression was recovered in tumour cell lines treated with 5-aza 2-deoxycytidine. Exogenous expression of BLU in neuroblastoma (SK-N-SH) and NSCLC (NCIH1299) resulted in reduced colony formation efficiency, in vitro. Furthermore, methylation of the BLU promoter region was detected in primary sporadic SCLC (14%), NSCLC (19%) and neuroblastoma (41%). As frequent methylation of the RASSF1A 3p21.3 TSG has also been reported in these tumour types, we investigated whether BLU and RASSF1A methylation were independent or related events. No correlation was found between hypermethylation of RASSF1A and BLU promoter region CpG islands in SCLC or neuroblastoma. However, there was association between RASSF1A and BLU methylation in NSCLC (P ¼ 0.0031). Our data suggest that in SCLC and neuroblastoma, RASSF1A and BLU methylations are unrelated events and not a manifestation of a regional alteration in epigenetic status, while in NSCLC there may be a regional methylation effect. Together, these data suggest a significant role for epigenetic inactivation of

*Correspondence: Dr F Latif; E-mail: fl[email protected] 8 Current address: Department of Medicine, Clinical Division of Oncology, University Hospital, Vienna, Austria Received 18 September 2002; revised 12 November 2002; accepted 13 November 2002

BLU in the pathogenesis of common human cancers and that methylation inactivation of BLU occurs independent of RASSF1A in SCLC and neuroblastoma tumours. Oncogene (2003) 22, 1580–1588. doi:10.1038/sj.onc. 1206243 Keywords: epigenetic inactivation; 3p21.3 region; human cancers

Introduction Allelic losses on the short arm of chromosome 3 are the most frequent genetic alterations detected in many sporadic human cancers, including lung, kidney, breast, ovary and neural crest-derived tumours (Ejeskar et al., 1998; Fullwood et al., 1999; Martinez et al., 2000, 2001; Wistuba et al., 2000; Maitra et al., 2001). Detailed studies on loss of heterozygosity (LOH) have identified several distinct 3p loci likely to contain tumour suppressor genes (TSGs), 3pl2, 3pl4, 3p21.3 and 3p25– 26. The VHL tumour suppressor gene (TSG) maps to 3p25 and is inactivated in a majority of sporadic renal cell carcinomas, but inactivation of VHL TSG is rare in other tumour types showing frequent 3p LOH (Latif et al., 1993; Gnarra et al., 1994; Herman et al., 1994; Sekido et al., 1994; Foster et al., 1995; Clifford and Maher, 2001). This suggests that 3p may contain multiple TSGs important in one or more tumour types. 3p21.3 is one of two main lung cancer TSG regions (LCTSGR1), the other being 3pl2 (LCTSGR2) (Sundaresan et al., 1998). Overlapping homozygous deletions in lung and breast tumour cell lines have narrowed the minimal critical region in 3p21.3 to 120 kb (Wei et al., 1996; Sekido et al., 1998; Lerman and Minna, 2000). This region was cloned and sequenced as part of a 600 kb cosmid contig coding for 19 genes. Eight genes were located within the 120 kb minimal region including alpha-2 delta-2 calcium channel subunit, PL6, 101F6,

BLU methylation in human cancers A Agathanggelou et al

1581

NPRL2/G21, BLU, RASSF1, FUS1 and LUCA2 (Lerman and Minna, 2000). Several of these genes, including isoform A of RASSF1 and BLU, are downregulated in lung tumour cell lines. However, extensive mutation analysis in a large panel of sporadic lung and breast tumours and tumour cell lines did not identify any of these candidate genes as classical TSGs (Angeloni et al., 2000, 2001; Lerman and Minna, 2000; Honorio et al., 2001). Recently, a new paradigm was set by the identification of RASSF1A as a TSG based on the frequent methylation inactivation of its promoter region in a wide range of tumour types and inhibition of the malignant phenotype in both in vitro and in vivo assays (Dammann et al., 2000; Agathanggelou et al., 2001; Burbee et al., 2001). This evidence is supported by the reactivation of RASSF1A in tumour cell lines treated with demethylating agents. Many other known TSGs are inactivated by methylation of their promoter regions such as VHL, pl6INK4a and RB1 but these genes, unlike RASSF1A, are also frequently inactivated by somatic mutations. Epigenetic inactivation of RASSF1A occurs in 72% of SCLC and this correlates well with the frequency of LOH in this region (Agathanggelou et al., 2001). However, RASSF1A methylation is detected less often in other tumour types with frequent 3p LOH (e.g. NSCLC), suggesting that one or more of the other 3p21.3 genes may be important in the pathogenesis of non-SCLC tumours. BLU is a candidate 3p21.3 TSG immediately upstream of RASSF1. The BLU gene was isolated using PCR primers to screen a cosmid contig (containing the critical region for a lung TSG) for the presence of b-catenin gene, which had been assigned to 3p21.3. The sequence analysis of the PCR product obtained using b-catenin gene showed no sequence homology to b-catenin. The PCR product was used to screen a cDNA library identifying the BLU gene (Lerman and Minna, 2000). Two isoforms of BLU are generated by alternative splicing and are referred to as the lung (12 exons) and testis (11 exons) isoforms after the cDNA libraries from which they were cloned (Figure 1a). BLU shares 30–32% identity with proteins of the MTG/ETO family of transcription factors and the suppressins, which are thought to regulate entry into the cell cycle. Compared with the lung isoform (440 amino acids) of BLU, the testis isoform (435 amino acids) has a different sequence of amino acids at positions 199–234 resulting in the loss of one of the three PKC phosphorylation sites. The mouse ortholog has 87% identity at the cDNA level and 89% at the protein level. The CG11253 gene product is the fly ortholog and is 49% identical in amino-acid sequence (Figure 1b). However, no true orthologs are present in yeast and worm proteomes. The promoter region of BLU is associated with a predicted CpG island and BLU is abundantly expressed in normal lung tissue. However, expression is drastically reduced in a subset of lung tumour cell lines. Missense mutations have been detected in only 3/61 (4.9%) lung cancer cell lines. Therefore, we have investigated the

possibility that hypermethylation of the BLU promoter region is involved in the downregulation of BLU expression in tumour cell lines and we have used a functional assay to examine the candidacy of BLU as a 3p21.3 TSG.

Results Tissue distribution of BLU PCR amplification was used to determine the relative abundance of BLU transcript in a multiple tissue cDNA panel (Clontech panels I and II). BLU is well represented in cDNA from brain, lung, liver, kidney, pancreas, testis and ovary (Figure 2). Low-to-moderate levels of BLU expression was detected in cDNA from heart, placenta, spleen, thymus, prostate, small intestine, colon and peripheral blood; control GAPDH data are not shown. Detection of methylation of the BLU promoter region in tumour cell lines A 12 kb region upstream of the BLU genomic locus was analysed using promoter prediction programs. Using the PromoterInspector software (Genomatix), a promoter region 311 to 67 bp of the BLU translational start was predicted. Furthermore, the transcriptional start of BLU is predicted to begin 178 bp by the TSSW program (Transcription Start Site, Wingender database, Nix software, UK HGMP Resource Centre). A CpG island (fulfilled the criteria of a GC content >0.5 and an observed : expected CpG ratio of 0.81) overlapping these sites (309 to +126 bp) was predicted using GRAIL/CpG (Nix software, UK HGMP Resource Centre). The methylation status of the above CpG island was analysed in tumour cell lines and primary tumours. DNA from a panel of 54 lung, six breast, six kidney, seven neuroblastoma and five NPC tumour cell lines was modified with sodium bisulphite. The methylation status of the BLU promoter region was determined by either combined bisulphite restriction analysis (COBRA) and/ or methylation-specific PCR (MSP) analysis as described in Materials and methods. The modified methylated sequence of the BLU promoter region and COBRA/MSP primer locations are shown in Figure 3a. Methylation of the BLU promoter region CpG island was detected in 21/54 (39%) of lung, 3/7 (42%) breast, 3/6 (50%) kidney, 6/7 (86%) neuroblastoma and 4/5 (80%) NPC cell lines (Figure 3b). In all cases undigested (for COBRA) or unmethylated (for MSP) PCR product remained, indicating that cell lines were partially methylated. COBRA PCR products were cloned and sequenced (for each tumour cell line four to six clones were sequenced) to verify the methylation status of the cell lines and to determine the pattern of CpG island methylation (Figure 3c). In all TaqI digestion positive cell lines, the majority of the CpGs were methylated. Oncogene


1582

Figure 1 (a) Schematic representation of the transcription pattern for the lung (12 exons) and testis (11 exons) isoforms of BLU showing the relative position of the predicted CpG island in the promoter region. (b) Alignment of the amino-acid sequence of the lung and testis isoforms of BLU with the mouse and drosophila homologues. Residues shaded in dark grey are conserved between the four proteins. The predicted MYND domain is shaded in light grey and shares 91 and 45% identity with Ms BLU and Dm CG11253, respectively

Figure 2 Tissue expression profile of BLU. In total, 2 mg of cDNA from each of the tissues indicated (MTC panel I and II, Clontech) was PCR amplified using BLU-specific primers to determine the relative abundance of BLU transcripts in various cell types

Effect of promoter region methylation on the expression of BLU in tumour cell lines RT–PCR was used to compare the expression levels of BLU in tumour cell lines with and without BLU promoter methylation. Figure 4a shows that expression Oncogene

of BLU is reduced in cell lines where methylation of the BLU promoter region is detected. To determine the relation between BLU promoter methylation and BLU expression, three neuroblastoma (SK-N-SH, SK-N-AS and Kelly) and one lung cancer cell lines (NCI-H187), all partially methylated, were treated with


1583

Figure 3 (a) Schematic representation of the forward sequence of the sodium bisulphite-treated BLU promoter region CpG island fragment. The predicted CpG island (NIX program) is in bold text and the positions of the CpG residues are underlined (in this sequence they are shown as methylated) and the restriction site for Taql is shown. Arrows indicate the position of the COBRA and MSP primers. The dotted arrow shows the transcriptional start of BLU. (b) The methylation status of the BLU promoter region in tumour cell lines was determined using either COBRA (left panel) or MSP primers (right panel). Unmodified DNA (UM) was included to show primer specificity for bisulphate-modified DNA. In vitro methylated DNA (IVD) was used as a positive control for BLU promoter methylation, and normal lymphocyte DNA (NL) was used as a negative control for methylation. COBRA: Neuroblastoma cell lines SK-N-DZ and SK-N-SH; breast tumour cell lines MCF-7 and 435, and kidney tumour cell line KTCC2, all showing partial methylation and the kidney tumour cell line SKRC54 not methylated for BLU. MSP: Neuroblastoma cell lines CHP212 and KELLY partially methylated while SK-N-MC is not methylated. (c) COBRA PCR products were cloned and sequenced from neuroblastoma cell lines (Kelly and CHP212), neuroblastomas (Stl96 and St208), SCLC cell line (107) and NSCLC tumours (76 and 83). For each tumour cell line and primary tumour four to six clones were sequenced. The methylation status of CpG residues 1–20 is indicated by shading, black (methylated), white (unmethylated) and grey (partially methylated). All COBRA PCR products showed both methylated and unmethylated alleles, figure only shows the methylated allele

the demethylating agent 5-aza-2-deoxycytidine for 6 days. Demethylation increased the expression of BLU in the partially methylated cell lines, but had no effect in the unmethylated control cell line SK-N-MC (Figure 4b,c). BLU promoter region methylation in primary tumours In view of the effect of promoter region methylation on the expression of BLU in tumour cell lines, we proceeded to determine the methylation status of BLU in sporadic lung and neuroblastoma tumours. BLU methylation was detected in 4/29 (13.7%) SCLC, 26/145

(18.6%) NSCLC and 20/49 (40.8%) neuroblastomas, Figures 3c and 5. While no methylation was detected in corresponding blood or fibroblast DNA from neuroblastoma patients, BLU promoter region methylation was detected in 1/10 of corresponding normal lung tissues from lung cancer patients, the corresponding tumour sample was also methylated for BLU. The unmethylated allele was amplified in all c ases either as a consequence of normal cell contamination as none of the tumour samples were microdissected or because of partial methylation as seen in tumour cell lines. Oncogene


1584

Figure 4 (a) RT–PCR showing the expression of BLU (expression primers were lung isoform specific) in a panel of tumour cell lines of known BLU methylation status, NCI-H249 and NCI-H290 are lung tumour cell lines; SKRC47 and KTCL26 are kidney tumour cell lines; HCC712; T47D and MCF-7 are breast tumour cell lines; SK-N-SH and SK-N-MC are neuroblastoma lines: , not methylated; +, methylated for BLU. (b) Neuroblastoma cell line SK-N-SH over a range of PCR cycles before and after treatment with demethylating agent. (c) Top panel: 30 cycles of BLU RT–PCR in a range of neuroblastoma cell lines showing that BLU expression is recovered by treatment with 5-aza-2-deoxycytidine. Bottom panel: RT–PCR in a lung tumour cell line, NCI-H187, showing that BLU expression is recovered after treatment with 5-aza-2-deoxycytidine. Transcripts for BLU and GAPDH were amplified using genespecific primers and Ready-To-Go RT–PCR beads (Amersham Pharmacia). Cells were seeded at 30% confluency and incubated for 6 days in 10% FCS/DMEM with and without 10 mm 5-aza-2-deoxycytidine. RNA was extracted using RNeasy kit (Qiagen) and 2 mg of RNA was used in each RT–PCR: , before 5-aza treatment; +, after 5-aza treatment

Inhibition of colony formation by BLU in tumour cell lines Colony formation assays were used to test for the growth-inhibitory effect of exogenously expressed BLU in lung and NB tumour cell lines. Empty plasmid pcDNA3.1 or pcDNA3.l containing the full open reading frame of BLU (lung isoform Accession number U 70824) was transfected into the lung NCI-H1299 (homozygous for multiple 3p21.3 region markers) and neuroblastoma SK-N-SH (partially methylated for BLU) tumour cell lines. NCI-H1299 was also transfected with a vector containing wtp53 as a positive control for the inhibition of colony formation. Cells were selected with geneticin (G418), 48 h after transfection, and resistant colonies developing 12–21 days later were stained with 0.4% crystal violet. The number of G418-resistant colonies after transfection with wtp53 was reduced by 84% in NCI-H1299 (n ¼ 9). Colony formation efficiency after transfection with BLU was reduced by 75% in SK-N-SH (n ¼ 5) cells and 49% in Oncogene

Figure 5 Methylation-specific PCR (MSP) for the detection of methylated (M) and unmethylated (U) BLU alleles in bisulphitemodified DNA from primary neuroblastoma and NSCLC. N ¼ normal; T ¼ tumour. In the upper panel, neuroblastomas T1, T2, T4 and T5 are methylated, while T3 and T6 are unmethylated. In the lower panel NSCLC T1, T2 and T3 are methylated, while T4 is unmethylated


1585

Figure 7 Comparative analysis of methylation in neuroblastoma. Shown for 49 NB cases are comparative analysis of BLU methylation, RASSF1A methylation and CASP8 methylation. Dark grey bars represent a methylated case, white bars represent nonmethylated cases. Dashed grey bars indicate unavailable data Figure 6 Growth-suppressive effect of BLU in lung (NCI-H1299) and NB (SK-N-SH) tumour cell lines. In each case, the number of G418-resistant colonies from vector-transfected cell lines was set at 100%. Error bars represent the standard deviation of at least three separate experiments, each calculated from triplicate plates

HI299 (n ¼ 8) compared with transfection with empty vector control (n ¼ 12), Figure 6. Correlation between BLU and RASSF1A CpG island methylation In previous studies, we had determined the methylation status of RASSF1A in the same series of SCLC and NSCLC primary tumours (Agathanggelou et al., 2001; Burbee et al., 2001), and RASSF1A and CASP8 in neuroblastoma (Astuti et al., 2001). Therefore, we looked for any correlation in the methylation of these genes with methylation of BLU. There was no correlation between the methylation of BLU and RASSF1A in SCLC or neuroblastoma or between BLU and CASP8 in neuroblastoma (Figure 7). However, there was significant association between RASSF1A and BLU methylation in our series of NSCLC (P ¼ 0.0031). In earlier studies, we had analysed methylation status of RARB, TIMP-3, pl6, MGMT, DAPK, ECAD, pl4 and GSTP1 (Zochbauer-Muller et al., 2001) in the same series of primary NSCLC. Methylation status of BLU was independent when compared with the methylation status of the above eight genes. In neuroblastoma samples, there was no correlation detected between BLU methylation status and N-myc amplification, 1 p deletion, 17q gain or survival (Table 1) (Ejeskar et al., 1998). However, lower stage neuroblastomas (stages I, II, IVS) were more frequently methylated for BLU compared to higher stage tumours (stages III, IV), P ¼ 0.025.

Discussion We have investigated tumour cell lines for the epigenetic inactivation of BLU. Methylation of the BLU promoter region was detected in 39% lung, 42% breast, 50% kidney, 86% neuroblastoma and 80% NPC tumour cell lines, and this correlated with the downregulation of BLU expression. Conversely, the expression of BLU was reconstituted in tumour cell lines treated with 5-aza-2deoxycytidine supporting the role of DNA methylation in the transcriptional repression of BLU. Although we did not find a tumour cell line that had complete absence of BLU expression, a recent article by Yan et al. (2002) demonstrates that small changes in the expression of the APC TSG can lead to the development of familial adenomatous polyposis. Methylation of the BLU promoter region was also detected in primary tumours 14% SCLC, 19% NSCLC and 41% neuroblastoma. It was interesting that BLU methylation in neuroblastoma was inversely correlated with stage, but was not correlated with survival; it is very likely that larger studies are required to determine the effect on survival. The BLU promoter region is within 6 kb of the RASSF1A promoter. However, we found that there was no correlation between BLU and RASSF1A methylation in SCLC and neuroblastoma primary tumours, suggesting that inactivation of BLU is an independent event and not the result of methylation of the 3p21.3 region. Gene NPRL2/G21 (Lerman and Minna, 2000) is located just centromeric to BLU (while RASSF1 is located telomeric to BLU) and also has a CpG island. Our preliminary data indicate that in neuroblastomas where BLU is methylated, NPRL2/G21 is unmethylated. Hence, another indication that BLU promoter region CpG island methylation in neuroblastoma is a specific Oncogene


1586

event and is not because of methylation of the 3p21.3 region. Also, there was no correlation between BLU and CASP8 methylation in neuroblastoma. However, we did find a significant association between RASSF1A and BLU methylation status in NSCLC (P ¼ 0.0031), suggesting that in NSCLC there may be a regional alternation in epigenetic status. It is interesting to note that in the two neural crest-derived tumours (SCLC and NB), there was no association between RASSF1A and BLU methylation status. In two tumour cell lines (an NB and an NSCLC) with partial endogenous BLU expression, exogenous overexpression of BLU reduced colony formation efficiency by 49–75%. These results indicate that overexpressed BLU has growth suppressor activity in neuroblastoma and lung cancer cells. BLU contains a predicted MYND domain at its Cterminus that shares 45% identity with the MYND domains found in important transcription regulatory proteins such as ETO(MTG8) (Wolford and Prochazka, 1998), suppressins (LeBoeuf et al., 1998), BS69 (Masselink and Bernards, 2000) and the Drosophila developmental proteins DEAF-1 (Michelson et al., 1999) and nervy (Feinstein et al., 1995). The MYND motif is essential to the function of many of these proteins. The AML/ETO fusion protein functions as a transcriptional repressor of AML-1B-dependent transactivation of TCRB, interleukin 3, GM-CSF, neutrophil protein 3 and MDR-1 promoters. Mutation of the AML/ETO MYND domain is sufficient to impair repression of the basal transcription of MDR-1 (Lutterbach et al., 1998). Furthermore, the AML/ETO-MYND domain has been shown to interact with the corepressor N-CoR that represses transcription by recruiting histone deacetylases (Amann et al., 2001). The potential TSG and transcriptional repressor, BS69, has a MYND domain that has been shown to bind oncoviral proteins such as El A and EBNA2 (Ansieau and Leutz, 2002). Deletions in E1A and EBNA2, which block interaction with BS69, are nontransforming. This suggests that abrogation of BS69 function through interaction with the MYND domain is important in tumorigenesis. Hence, BLU as an MYND domain containing protein is likely to be involved in important transcriptional regulation pathways. Coupled with our findings that BLU is epigenetically inactivated in a significant proportion of tumour cell lines and in a subset of primary sporadic tumours, and has an inhibitory growth effect in vitro, we postulate that BLU is a likely candidate 3p21.3 TSG involved in the development of several common sporadic tumour types.

Materials and methods Patients and samples DNA from a total of 174 lung tumours (29 SCLC and 145 NSCLC with corresponding normal tissue from 10 cases) and 54 lung tumour cell lines was used, these samples have been described in earlier publications (Agathanggelou et al., 2001; Burbee et al., 2001). The primary tumours and corresponding normal tissue were either obtained at autopsy or by operation. DNA from primary tumour tissue from 40 neuroblastomas Oncogene

and corresponding normal tissue (fibroblast or blood sample) were used, this material has been described previously (Ejeskar et al., 1998). DNA from five NPC cell lines (CNE1, CNE2, Ml, SUNE1 and NPC 391) were obtained from Dr Po Wing Yuen.

Cell lines A total of 34 lung tumour cell lines including (A549, H460, H1299, H82, H1092, H1105, H1184, H1284, H1304, H1436, H1522, H1618, H1628, H1672, H60, H69, H128, H187, H209, H250, H345, H378, H417, H592, H711, H719, H740, H748, H841, H845, H889, H290, H838, H2052), seven breast tumour cell lines (MCF7, HCC38, MDA MB 435, MDA MB 436, MDA MB 231, HCC712, HCCHB2), six renal cell carcinoma cell lines (KTCT26, KTCC2, SKRC54, SKRC39, SKRC47, SKRC45) and seven neuroblastoma cell lines (SK-N-SH, SKN-DZ, SK-N-AS, SK-N-MC, SK-N-FI, Kelly, CHP212) were used. Tissue culture media supplemented with 10mm 5-aza-2deoxycytidine was used to treat cell lines for 6 days.

Bisulphite modification Bisulphite modification of DNA was performed as described (Agathanggelou et al., 2001). Briefly, 0.5–1.0 mg of genomic DNA was denatured in 0.3 m NaOH for 15 min at 371C and then unmethylated cytosine residues were sulphonated by incubation in 3.12 m sodium bisulphite (pH 5.0) (Sigma)/5 mm hydroquinone (Sigma) in a thermocycler (Hybaid) at 991C/30 s and 501C/15 min for 20 cycles. The sulphonated DNA was recovered using the Wizard DNA cleanup system (Promega) in accordance with the manufacturer’s instructions. The conversion reaction was completed by desulphonating in 0.3 m NaOH for 10 min at room temperature. DNA was ethanol precipitated and resuspended in water.

COBRA and MSP DNA sequences specific for the BLU promoter region were amplified using primer sets for COBRA and MSP as described in Table 1. The conditions for PCR amplification with the COBRA primer set were 1 PCR buffer (Invitrogen), 0.25 mm dNTPs, 1.5 mm MgCl2, 0.4 mm forward/reverse primers, 100 ng bisulphite-modified DNA in a 50 ml reaction volume. The reaction mix was overlaid with mineral oil, and incubated at 951C/5 min for one cycle during which 5 units of DNA polymerase (Invitrogen) were added and then incubated at 951C/min, 551C/min, 721C/2 min for 40 cycles. Methylated cytosine residues were identified by restriction enzyme digestion and by sequencing of cloned PCR products. Briefly, 16 ml of PCR product was incubated with 20 units of TaqI (Roche) (New England BioLabs) for 2 h at 651C. The restriction enzyme digestion products were then visualized by separation in a 2% agarose gel stained with ethidium bromide. PCR products were cloned using an AT-cloning kit (Promega) in accordance with the manufacturer’s instructions and sequenced using Big Dye (versionII) sequence cycling kit (Perkin–Elmer). BLU-specific MSP was performed using a 25 ml reaction volume containing 150 ng bisulphite modified DNA, 1 PCR buffer, 0.16 mm dNTPs, 1.5 mm MgCl2, 0.6 mm each primer and 1 unit HotStart Taq (Invitrogen). The reaction mixture was incubated at 951C for 12 min, followed by six touchdown cycles of 951C/20 s, X1C/40 s, 721C/20 s (Umsp X ¼ 661C, msp X ¼ 721C), X11C per cycle and then 30 cycles of 951C/20 s, X61C/30 s, 721C/30 s.


1587 Table 1

COBRA MSP

RT–PCR

Primer

Sequence

COBRA F COBRA R UmspF UmspR msp F msp R BLU ex 6.2F BLU ex 8R GAPDH F GAPDH R

atagagtttagattataaggatttggagtt ataccaaaatctaaaacaaaacaattac tttgtgggttatagtttgagaaagtg aacaaattaaccacacctacac ttcgtgggttatagttcgagaaagcg aacgaattaaccgcgcctacgc gcttagcacacacaacctgc ccttggcaaaacttgtgagg tgaaggtcggagtcaacggatttggt catgtgggccatgaggtccaccac

DNA and RNA extraction DNA and RNA were prepared using the DNeasy and RNeasy kits (Qiagen), respectively, in accordance with the manufacturer’s instructions. RT–PCR RT–PCR amplification of BLU- and GAPDH-specific transcripts was carried out using primers described in Table 1, 0.5 mg of RNA and Ready-To-Go RT PCR beads (Amersham) in accordance with the manufacturer’s instructions. The semiquantitative RT–PCR approach used, compared the expression levels of BLU with that of the housekeeping gene GAPDH. A range of cycles was employed to determine the number needed to detect specific RT–PCR product while still in the exponential phase of the amplification of both genes. The ethidium bromide signal obtained during this phase of the RT–PCR was then compared to give a ratio of BLU: GAPDH representative of the levels of these transcripts in the RNA

isolate. By comparing the ratios obtained before and after 5aza-2-deoxycytidine treatment, the effect of promoter demethylation on the expression of BLU was determined. Figure 4b clearly shows that the BLU signal from 5-aza-2deoxycytidine treated SK-N-SH cells has almost reached a plateau at 30 cycles, whereas the signal from the untreated cells is much weaker. Transfection and colony formation assays Transfection with Fugene 6 reagent (Roche) used l mg of each plasmid in a 3 : 1 ratio per 35 cm2 dish containing 5 l04 cells seeded 24 h before transfection. At 48 h after transfection, cells were supplemented with 600–800 mg/ml G418 BME basal medium or DMEM medium with 10% fetal bovine serum (Life Technologies GIBCO). Surviving colonies were counted 12–21 days later after staining with 0.4% crystal violet. Statistical analysis Comparisons were made using Fisher’s exact test or w2 test as appropriate. P values o0.05 were taken as statistically significant.

Acknowledgements This work is supported in part by SPARKS (Sports Aiding Medical Research for Kids), Association for International Cancer Research, Cancer Research UK. SH is supported by the Portuguese Foundation for Science and Technology. TM is supported by the Swedish Cancer Society and the Swedish Children’s Cancer Foundation. JDM supported by Grants NCI P50 CA70907 and CA71618.

References Agathanggelou A, Honorio S, Macartney DP, Martinez A, Dallol A, Rader J, Fullwood P, Chauhan A, Walker R, Shaw JA, Hosoe S, Lerman MI, Minna JD, Maher ER and Latif F. (2001). Oncogene, 20, 1509–1518. Amann JM, Nip J, Strom DK, Lutterbach B, Harada H, Lenny N, Downing JR, Meyers S and Hiebert SW. (2001). Mol. Cell Biol., 21, 6470–6483. Angeloni D, Duh FM, Wei MF, Johnson BE and Lerman MI. (2001). Mol. Cell Probes, 15, 125–127. Angeloni D, Wei MH, Duh FM, Johnson BE and Lerman MI. (2000). Mol. Cell Probes, 14, 53–54. Ansieau S and Leutz A. (2002). J. Biol. Chem., 277, 4906– 4910. Astuti D, Agathanggelou A, Honorio S, Dallol A, Martinsson T, Kogner P, Cummins C, Neumann HP, Voutilainen R, Dahia P, Eng C, Maher ER and Latif F. (2001). Oncogene, 20, 7573–7577. 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. Clifford SC and Maher ER. (2001). Adv. Cancer Res., 82, 85– 105. Dammann R, Li C, Yoon JH, Chin PL, Bates S and Pfeifer GP. (2000). Nat. Genet., 25, 315–319. Ejeskar K, Aburatani H, Abrahamsson J, Kogner P and Martinsson T. (1998). Br. J. Cancer., 77, 1787–1791.

Feinstein PG, Kornfeld K, Hogness DS and Mann RS. (1995). Genetics, 140, 573–586. Foster K, Osborne RJ, Huddart RA, Affara NA, FergusonSmith MA and Maher ER. (1995) Eur. J. Cancer, 31A, 2392–2395. Fullwood P, Marchini S, Rader JS, Martinez A, Macartney D, Broggini M, Morelli C, Barbanti-Brodano G, Maher ER and Latif F. (1999). Cancer Res., 59, 4662–4667. Gnarra JR, Tory K, Weng Y, Schmidt L, Wei MH, Latif F, Liu S, Chen F, Duh FM, Lubensky L, Duan DR, Florence C, Pozzati R, Walther MM, Bander NH, Grossman BH, Drauch H, Pomer S, Brooks D, Isaacs WB, Lerman MI, Zbar B and Linehan WM. (1994). Nat. Genet., 7, 85–90. Herman JG, Latif F, Weng Y, Lerman MI, Zbar B, Liu S, Samid D, Duan DR, Gnarra JR, Linehan WM and Baylin SB. (1994). Proc. Natl. Acad. Sci. USA, 91, 9700– 9704. Honorio S, Gordon K, MacCartney D, Agathanggelou A and Latif F. (2001). Mol. Cell Probes, 15, 391–393. Latif F, Tory K, Gnarra J, Yao M, Duh FM, Orcutt ML, Stackhouse T, Kuzmin I, Modi W, Geil L, Schmidt L, Zhou F, Li H, Wei MH, Chen F, Glenn G, Choyke P, Walther MM, Weng Y, Duan DR, Dean M, Glavac D, Richards FM, Crossey PA, Ferguson-Smith MA, Paslier DL, Chumakov I, Cohen D, Chinault AC, Maher ER, Linehan WM, Zbar B and Lerman MI. (1993). Science, 260, 1317– 1320. Oncogene


1588 LeBoeuf RD, Ban EM, Green MM, Stone AS, Propst SM, Blalock JE and Tauber JD. (1998). J. Biol. Chem., 273, 361– 368. Lerman MI and Minna JD. (2000). Cancer Res., 60, 6116– 6133. Lutterbach B, Sun D, Schuetz J and Hiebert SW. (1998). Mol. Cell Biol., 18, 3604–3611. Maitra A, Wistuba II, Washington C, Virmani AK, Ashfaq R, Milchgrub S, Gazdar AF and Minna JD. (2001). Am. J. Pathol., 159, 119–130. Martinez A, Fullwood P, Kondo K, Kishida T, Yao M, Maher ER and Latif F. (2000). Mol. Pathol., 53, 137–144. Martinez A, Walker RA, Shaw JA, Dearing SJ, Maher ER and Latif F. (2001). Mol. Pathol., 54, 300–306. Masselink H and Bernards R. (2000). Oncogene, 19, 1538– 1546. Michelson RJ, Collard MW, Ziemba AJ, Persinger J, Bartholomew B, and Huggenvik JI. (1999). J. Biol. Chem., 274, 30510–30519. Sekido Y, Ahmadian M, Wistuba II, Latif F, Bader S, Wei MH, Duh FM, Gazdar AF, Lerman MI and Minna JD. (1998). Oncogene, 16, 3151–3157.

Oncogene

Sekido Y, Bader S, Latif F, Gnarra JR, Gazdar AF, Linehan WM, Zbar B, Lerman MI and Minna JD. (1994). Oncogene, 9, 1599–1604. Sundaresan V, Chung G, Heppell-Parton A, Xiong J, Grundy C, Roberts I, James L, Cahn A, Bench A, Douglas J, Minna J, Sekido Y, Lerman M, Latif F, Bergh J, Li H, Lowe N, Ogilvie D and Rabbitts P. (1998). Oncogene, 17, 1723–1729. Wei MH, Latif F, Bader S, Kashuba V, Chen JY, Duh FM, Sekido Y, Lee CC, Geil L, Kuzmin I, Zabarovsky E, Klein G, Zbar B, Minna JD and Lerman MI. (1996). Cancer Res., 56, 1487–1492. 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. Wolford JK and Prochazka M. (1998). Gene, 212, 103–109. Yan H, Dobbie Z, Gruber SB, Markowitz S, Romans K, Giardiello FM, Kinzler KW and Vogelstein B. (2002). Nat. Genet., 30, 25–26. Zochbauer-Muller S, Fong KM, Virmani AK, Geradts J, Gazdar AF and Minna JD. (2001). Cancer Res., 61, 249– 255.